Handbook of Applied Behavior Analysis (Wayne W. Fisher)
Handbook of Applied Behavior Analysis (Wayne W. Fisher)
HANDBOOK
OF APPLIED
BEHAVIOR ANALYSIS
SECOND EDITION
Edited by
Wayne W. Fisher
Cathleen C. Piazza
Henry S. Roane
No part of this book may be reproduced, translated, stored in a retrieval system, or transmitted,
in any form or by any means, electronic, mechanical, photocopying, microfilming, recording,
or otherwise, without written permission from the publisher.
The authors have checked with sources believed to be reliable in their efforts to provide
information that is complete and generally in accord with the standards of practice that are
accepted at the time of publication. However, in view of the possibility of human error or
changes in behavioral, mental health, or medical sciences, neither the authors, nor the editors
and publisher, nor any other party who has been involved in the preparation or publication
of this work warrants that the information contained herein is in every respect accurate or
complete, and they are not responsible for any errors or omissions or the results obtained from
the use of such information. Readers are encouraged to confirm the information contained in
this book with other sources.
Wayne W. Fisher, PhD, BCBA-D, is the Henry Rutgers Endowed Professor of Pediatrics at Robert
Wood Johnson Medical School and a core faculty member of the Brain Health Institute. He is also
inaugural director of the Rutgers Center for Autism Research, Education, and Services (RUCARES)
at Rutgers, The State University of New Jersey. His influential research has focused on preference
assessment, choice responding, and the assessment and treatment of autism and severe behavior
disorders. Dr. Fisher has published over 200 peer-reviewed journal articles. He is a past editor of the
Journal of Applied Behavior Analysis, past president of the Society for the Experimental Analysis of
Behavior (SEAB), and a Fellow of the Association for Behavior Analysis International (ABAI). Dr.
Fisher is a recipient of the Bush Leadership Award; the Nathan H. Azrin Distinguished Contribu-
tion to Applied Behavior Analysis Award and the SEAB Don Hake Translational Research Award
from Division 25 (Behavior Analysis) of the American Psychological Association; and research and
mentorship awards from the University of Nebraska.
Cathleen C. Piazza, PhD, BCBA-D, is Professor in the Graduate School of Applied and Professional
Psychology at Rutgers, The State University of New Jersey, and founding director of the Pediatric
Feeding Disorders Program at Children’s Specialized Hospital in New Jersey. Dr. Piazza and her col-
leagues have examined various aspects of feeding disorders—among the most common health prob-
lems in children—and have developed a series of interventions to address them. Her research has
established strong empirical support for applied-behavior-analytic interventions for feeding disorders.
Dr. Piazza is a former editor of the Journal of Applied Behavior Analysis, past president of the SEAB,
and a Fellow of the ABAI. She is a recipient of the Nathan H. Azrin Distinguished Contribution to
Applied Behavior Analysis Award from Division 25 (Behavior Analysis) of the American Psycho-
logical Association and the Outstanding Mentor Award from the ABAI.
Henry S. Roane, PhD, BCBA-D, is Vice President of Clinical Quality at Sprout Therapy and holds
the Gregory S. Liptak, MD, Professorship of Child Development in the Department of Pediatrics at
the State University of New York (SUNY) Upstate Medical University. Dr. Roane serves as Chief of
v
vi About the Editors
the Division of Development, Behavior and Genetics; Director of the Golisano Center for Special
Needs; and Chair of the Behavior Analysis Studies Program at Upstate. He is Editor-in-Chief of
Behavioral Development and has served as an associate editor and editorial board member for journals
in the fields of pediatrics, behavior analysis, and school psychology. Dr. Roane has coauthored over
100 research articles and chapters as well as several academic texts on the assessment and treatment
of behavior disorders in children with autism and related disorders. He is a Fellow of the ABAI, a
member of the Society for Pediatric Research, and a previous recipient of the SUNY Chancellor’s
Award for Excellence in Faculty Service.
Contributors
vii
viii Contributors
Addressing the social and behavior problems that humans display is a daunting task; the first and most
important step is the realization that changing behavior inevitably involves learning by doing in an envi-
ronmental context in which the contingencies of reinforcement promote desirable behavior over unde-
sirable behavior. As behavior analysts, we may identify and understand the prevailing contingencies of
reinforcement better than most laypersons, but that knowledge does not inoculate us from the potent
effects those contingencies have on our own behavior. As applied behavior analysts, in particular, we
often find ourselves in contexts involving strong social contingencies in which we are expected to solve
complex and socially important behavior problems. Such powerful contingencies are likely to shape and
hone our behavior-analytic skills—much more so than reading this or any other book. Nevertheless,
an informative text can provide a roadmap that helps us respond to those social contingencies more
effectively and rapidly. We have developed and revised this book specifically for that purpose.
As we conceived of and developed the first edition of this book, it occurred to us that there
was no single source we would consistently go to when faced with a particularly challenging clini-
cal or applied research problem. Rather, we might start by going to any one of a number of different
sources, including (1) consulting with in-house colleagues or those at other institutions, (2) conduct-
ing searches of the Journal of Applied Behavior Analysis or the Journal of the Experimental Analysis
of Behavior, or (3) reading sections of the many books covering behavior-analytic topics relevant to
the specific clinical or research challenge before us. Thus our central goal for the first edition of this
book was to develop a resource for those behavior analysts working in service and applied research
settings—one that, we hoped, would be the first source they would turn to when presented with a
unique or difficult clinical or applied research problem. In fact, in selecting the authors for each of
the chapters, we spent a considerable amount of time asking ourselves whom we would call upon as
a first choice for a consultant on the specific topic, and then we invited that person to be the senior
author for the chapter. It was exceedingly reinforcing when our first choices accepted our invitation
to author the chapters on their specific areas of expertise in almost every case.
In planning the second edition of this book, we relied heavily on the feedback we received
from applied behavior analysts who used the first edition, particularly the feedback we received from
professors who employed the first edition to aid in the teaching of applied behavior analysis. Based
on that feedback, we have worked diligently to update and improve the integration, readability, and
relevance of the book’s contents. Importantly, we have expanded the number of chapters from 30
xi
xii Preface
to 34, in order to cover topics important to our readers that we did not cover sufficiently in the first
edition. We have added two chapters on the quantitative analysis of behavior: one on the matching
law and behavioral persistence by Podlesnik et al. (Chapter 6), and another on behavioral econom-
ics by DeLeon et al. (Chapter 7). These chapters cover highly relevant topics in applied work, since
they address the basic processes that govern how individuals allocate their time to various response
options, such as why individuals chose to emit problem behavior over adaptive behavior in certain
environmental conditions. We have also added a chapter on the assessment and treatment of pedi-
atric feeding disorders by Piazza and Kirkwood (Chapter 25), one on teacher training by DiGennaro
Reed and colleagues (Chapter 27), one on providing in-home behavioral services via telehealth by
Wacker et al. (Chapter 31), and one on organizational behavior management by Wilder and Gravina
(Chapter 32). These chapters provide important and timely information on topics highly relevant to
applied behavior analysts providing services in these areas.
The overall organization of this second edition of the book is similar to that of the first, in that it
provides the reader with the foundations of behavior analysis in the early chapters and then ties these
basic concepts to applications in subsequent chapters. As such, it strikes a balance between emphasis
on research and clinical applications. The book provides a detailed level of analysis for both general
and specialized areas of behavior analysis. Its content reflects the breadth of behavior analysis and the
expansion of applied behavior analysis into mainstream domains such as pediatric care, psychology,
organization management, psychiatry, and drug addiction.
After the book’s Introduction (Part I/Chapter 1), Part II of the book devotes six chapters to a
concise yet detailed review of the history, philosophy, and basic principles that provide the foun-
dational basis for applied behavior analysis. Part III is devoted to two chapters on measurement,
experimental design, and related methodological issues. Part IV consists of four chapters that discuss
stimulus preference assessments and both functional and structural approaches to assessing problem
behavior, as well as specific chapters on indirect, direct, and controlled functional assessments. Part
V of the book describes a variety of concepts and procedures relevant to interventions for increasing
desirable behavior. The four chapters in this section cover topics such as differential-reinforcement
procedures, building complex repertoires and establishing stimulus control, teaching verbal behav-
ior, and staff training and management. Part VI of the book includes five chapters covering issues
related to developing interventions for decreasing problem behavior. Topics in this section include
developing antecedent interventions; designing function-based extinction, reinforcement, and pun-
ishment interventions; and developing token economies. Part VII describes a variety of important
subspecialties within the field of applied behavior analysis, including treatment of autism spectrum
disorder, behavioral pediatrics, treatment of pediatric feeding disorders, behavioral approaches to
education, teacher training, establishing safety skills in children, behavioral treatment of drug addic-
tion, behavioral gerontology, telehealth delivery of behavioral services, and organizational behavior
management. Part VIII, the final section of the book, focuses on professional issues in applied behav-
ior analysis; it includes a chapter on ethics and training, and another on professional certification.
Of interest to the reader is that although the topics of each chapter are specific to that content area,
there are several overlapping themes across chapters. The discussion of specific principles across dif-
ferent content domains is representative of the breadth of the basic tenets of behavior analysis.
This book can be used as a core or primary textbook for courses in psychology, education, or
behavior analysis. The target audiences for the book are practicing behavior analysts and students
in graduate classes in psychology, education, or other related fields, and it could serve as a primary
source for preparing for professional certification. The quality and comprehensiveness of the book
make it a must-have for any behavior analysis library. We hope that our readers will find this text as
informative as it was enjoyable for us to edit.
Wayne W. Fisher
Cathleen C. Piazza
Henry S. Roane
Contents
I. INTRODUCTION 1
4. Schedules of Reinforcement 58
F. Charles Mace, Duncan Pritchard, and Heather Penney
xiii
xiv Contents
Author Index 579
I n Chapter 1, Fisher, Groff, and Roane introduce the reader to the field of
behavior analysis and its three subfields: behaviorism; the experimental analysis
of behavior; and applied behavior analysis, the principal topic of this book. The
chapter introduces many concepts (e.g., stimulus equivalence) that are discussed
in greater depth in subsequent sections of the handbook. Because the history of
applied behavior analysis has not changed appreciably since the first edition of
this handbook appeared, this introductory chapter has undergone only minor
changes for the second edition.
1
CHAP TER 1
Behavior analysis is a discipline with three primary havior of individuals rather than groups; (4) envi-
branches (Morris, Todd, Midgley, Schneider, & ronmental explanations of behavior; and (5) the
Johnson, 1990): (1) behaviorism, which focuses on study of behavior as a natural science. We discuss
the worldview or philosophy of behavior analysis; each of these tenets before turning our attention
(2) the experimental analysis of behavior, which to the dimensions that specifically define ABA.
focuses on identifying and analyzing the basic
principles and processes that explain behavior;
and (3) applied behavior analysis (ABA), which BEHAVIOR AS SUBJECT MATTER
focuses on solving problems of social importance
using the principles and procedures of behavior As behavior analysts, we believe that the ap-
analysis. Although this third branch is the prima- propriate subject matter for our field is behavior.
ry topic of our text, a basic knowledge of the other We define behavior quite broadly to include any-
branches is necessary to appreciate the develop- thing an individual does when interacting with
ment and dimensions of ABA fully. the physical environment (Catania, 2013; Skin-
Behavior analysis began as a school or subfield ner, 1938), including crying, speaking, listening,
within the discipline of psychology. Some still view running, jumping, shifting attention, and even
behavior analysis as a subspecialty in psychology, thinking. This behavioral philosophy contrasts
whereas others believe that the basic tenets of be- with that of mentalists or cognitive psychologists:
havior analysis and traditional psychology are so They view thinking, feeling, and other internal
fundamentally at odds that the two cannot coexist events as activity that occurs within metaphysical
within a single discipline (e.g., Fraley & Vargus, entities such as the self, the psyche, or the mind,
1986). The basic tenets that distinguish behavior and they believe that these entities influence or
analysis from other areas of psychology include its control outward behavior. Mentalists observe be-
emphasis on (1) behavior as the basic datum for havior to draw inferences about these hypotheti-
the field, rather than the psyche, the self, or other cal structures, which they view as the appropriate
internal mental or metaphysical structures or phe- subject matter for the field of psychology. They
nomena; (2) continuity between publicly observ- believe that understanding these inner constructs
able behavior and private events, such as thinking helps to explain observable behavior. Behaviorists
and feeling; (3) prediction and control of the be- believe that behavior itself is their appropriate sub-
3
4 In t r o d u c t i o n
ject matter, and that they should study it directly, ing the existence of ghosts. Modern-day mental-
without references to internal causes. Behaviorists ists (e.g., cognitive psychologists) do not often talk
view the brain as real but the mind as an inven- about the mind per se, but they are much more
tion, something thought up rather than something likely than behaviorists to look to internal vari-
that thinks and controls behavior. ables (e.g., thoughts and feelings) to explain how
Although people in all walks of life talk about behavior and similar logical problems arise. That
the mind as if it were a real entity, when questioned is, they use observable behavior (e.g., preparing a
about its location and its characteristics, they find sandwich) to formulate hypotheses about internal
that the mind is difficult if not impossible to lo- constructs (e.g., the individual is hungry), which
cate or describe in precise terms. Another problem they then use to explain the observed behavior
that arises when one attempts to explain outward, (e.g., the person has prepared a sandwich because
observable behavior by appealing to causation via of the hunger). Skinner (1953, p. 31) pointed out
internal events is that one then must explain what that the two statements “He eats” and “He is hun-
causes the internal events. Two philosophical ar- gry” describe a single set of facts. Thus the observ-
guments illustrate this problem: Ryle’s regress and er cannot use one statement, “He is hungry,” to
the homunculus fallacy. explain the other, “He eats.” Skinner also argued
Ryle (1949) identified a logical flaw in the that appeals to such inner causes impede scientific
traditional (dualist) view of intelligent behavior inquiry, because once we identify a supposed cause
(see Tanney, 2018). The dualist position, view- of behavior (i.e., “He eats because he is hungry”),
ing the mind and body as two distinct entities, there is no reason to continue searching for an ex-
is that when an individual displays an intelligent planation of the behavior.
act (i.e., an observable response), internal, mental By contrast, B. F. Skinner’s approach to ex-
reflection on how to act intelligently must have plaining behavior represents a constantly evolving
preceded and directed it. Ryle pointed out that if one in which experimental findings guide theory
the logic of the dualist view were accurate, then much more than theory guides experimentation.
the internal operation of reflection also would be In fact, revisions and updates of behavior-analytic
an intelligent act (albeit an internal one) that explanations of behavior are often based on new
would need to be preceded and guided by reflec- experimental findings—an approach that some
tion about various alternative ways of reflecting, have referred to as “a work in progress” (e.g., Cata-
thus creating a potentially never-ending succes- nia, 1988, 2013). One notable example of the way
sion of reflecting about reflecting about reflecting. we have updated our conceptualizations of behav-
The endless need for a predecessor and director of ior based on new experimental findings is the way
every intelligent act is called Ryle’s regress. we define our subject matter, behavior.
The homunculus fallacy is analogous to Ryle’s Early definitions of behavior focused on its phys-
regress, except that it focuses on how we inter- ical or topographical characteristics; for example,
pret visual stimulation. A mentalist viewpoint “thought processes are really motor habits in the
is that light projects onto the back of the retina, larynx, improvements, short cuts, changes, etc.”
and that the mind views these images like the way (Watson, 1913, p. 177). Skinner (1938) provided a
an individual views a motion picture. The mind broader definition of behavior and introduced the
is thus akin to a little man or homunculus who concept of the three-term contingency (anteced-
is metaphorically sitting inside the brain viewing ent–behavior–consequence) that defines operant
the movie. The question, then, is how the mind behavior. That is, we define operant behavior by
or the homunculus sees and interprets the motion its topographical features and by its functional
picture playing inside the human brain. In keeping properties—namely, the environmental anteced-
with the mentalist hypothesis, another, smaller ents and consequences functionally related to the
homunculus inside the first would have to see and specific response topography. The topographical
interpret its movie, which would need to have an features of a person running to catch a bus may
even smaller homunculus inside it to interpret its be like those of someone running out of a burn-
movie. We call the endless need for another ho- ing building. The two forms of running, however,
munculus to explain the visual interpretations of are distinctly separate operant responses because
the prior one the homunculus fallacy (Uttal, 2000). they are under the control of different environ-
These arguments illustrate that proving or dis- mental antecedents and consequences, and these
proving the existence of the mind is impossible, environment–behavior relations define operant
much like the impossibility of proving or disprov- behavior (Donahoe, 2004).
History, Philosophy, Principles, and Methods 5
More recent empirical findings have led to ad- driving home alone in another car. The married
ditional refinements regarding what constitutes man looks at his spouse while stopped at a traffic
behavior. For example, research has shown that light and says, “Remind me to take the garbage out
operant behavior is sensitive to both molecular when we get home.” At the same stoplight, the sin-
and molar patterns of reinforcement (e.g., Herrn- gle man thinks silently to himself, “I’ve got to re-
stein, 1969). Based in part on this empirical find- member to take the garbage out when I get home.”
ing, teleological behaviorism attempts to explain Behaviorists would view the talking done by the
complex behavior (e.g., building a house, falling in married man and the thinking done by the single
love) by identifying organized patterns of environ- man as distinct forms of behavior governed by the
ment–behavior relations that involve both proxi- same laws, in which talking is a public behavior
mal and ultimate causes or consequences. Rachlin that others can observe, and thinking is a private
(2012, 2018) explains that hammering a nail is a behavior that only the single man can observe.
function of not only the immediate consequence Behavior analysts almost exclusively study pub-
of fastening two boards together, but also the larg- lic behavior because they can observe, quantify,
er task of constructing a floor, which in turn is and subject it objectively to the scientific method.
a function of the task of building a house; these However, behaviorists believe that the general
nested responses are functions of the ultimate con- principles of behavior derived from the study of
sequence of sheltering and protecting one’s family. public responses (e.g., talking aloud) also apply
Our conception of what constitutes behavior to and explain private responses (e.g., thinking or
also has expanded because of research on stimulus talking silently to oneself). They also believe that
equivalence and relational-frame theory (Hayes, private events (e.g., pain) often correlate with ob-
Barnes-Holmes, & Roche, 2001; Sidman, 2000). servable events, thus allowing an individual to tact
Research in this area has shown consistently that or describe common private events (e.g., skinning
when we train certain stimulus relations (e.g., a knee; cf. Welsh, Najdowski, Strauss, Gallegos, &
“Mike is heavier than Bill; Bill is heavier than Fullen, 2019).
Sam”) with verbally competent human partici- Behaviorists focus on general principles that re-
pants, other stimulus relations emerge without late to the function of behavior—that is, its pur-
specific training (e.g., “Sam is lighter than Mike”). pose or why it occurs. Behaviorists believe that
These emergent or derived relations are important environmental events that occur in close physical
because they may be prerequisites to and form the and temporal proximity to the behavior determine
basis of generative-language acquisition. They also the function of a response. Important environmen-
are potentially important because they require a tal events that influence behavior include (1) the
broader definition of what constitutes operant context in which the response occurs (e.g., teenag-
behavior; that is, equivalence classes or relational ers behave differently at home with parents than at
frames represent broader units of operant behavior a party with peers); (2) motivational factors (e.g.,
that include both trained (i.e., reinforced) and un- searching for a restaurant is more likely if one has
trained stimulus relations. not eaten in a while); (3) antecedents that signal
which responses will be successful (e.g., proceeding
if a traffic light is green because it signals safety,
PRIVATE EVENTS and stopping if it is red because it signals danger),
and (4) the consequences or outcomes of responses
A common misconception of behavior analy- that influence whether they will recur (e.g., an
sis is that it does not acknowledge or attempt to individual is more likely to study for a test in the
explain internal, private events such as thoughts future if he or she gets a better grade as a result).
and dreams. Many behavior analysts believe that Applying these general principles to the previ-
the same laws that govern overt behavior govern ous scenario, a behavior analyst might hypothesize
private events, and they do not explain these pri- that the married man asks his spouse to remind
vate events by using mentalistic processes (Moore, him to take out the trash because (1) stopping at a
2003). The major difference between public and traffic light provides a signal or cue indicating that
private behavior is that others can observe and it is momentarily safe to shift his attention to mat-
verify public behavior, but only the individual per- ters other than driving the car; (2) the man has
forming the behavior can observe private behavior. experienced a negative outcome associated with
Consider the scenario of a married man driving forgetting to take out the trash previously, such as
home with his spouse in one car, and a single man trash piling up because the cans would not hold
6 In t r o d u c t i o n
it all; and (3) asking his spouse to remind him to 1950; Mace, 1996). The experimental methods
take out the garbage increases the likelihood that of mainstream psychology, which studies groups,
he will take out the trash and avoid the negative and behavior analysis, which studies individu-
consequence of the trash piling up. The same als, reflect this fundamental difference between
three reasons would apply to the single man, ex- the two fields. Most psychological researchers
cept that he has no companion in the car to help employ group comparison designs and use infer-
him remember to take out the trash, so he says the ential statistics to identify significant differences
words silently rather than aloud. Thus, although between various groups, whereas behavior analysts
the two responses in this example (talking aloud use single-case designs to study the generality of
and thinking about the trash) are different topo- principles of behavior (e.g., behavioral momen-
graphically because others can observe talking but tum, delay discounting). Behavior analysts find
not thinking, they are similar functionally because the prediction and control of the behavior of indi-
both are occasioned by the same antecedent (sit- viduals rather than groups advantageous, because
ting at the stoplight) and reinforced by the same individuals engage in behavior but groups do not
consequence (avoiding the trash piling up). (Johnston & Pennypacker, 1993, p. 23).
The only way to identify whether a private Researchers using group comparison methods
event has occurred is through self-report, because often present their results in terms of statistical
others cannot observe an individual’s covert be- means to describe how the average individual in
haviors. Self-observation, however, is often unreli- the group behaved, and they use standard devia-
able (Skinner, 1953). In fact, Skinner points out tions to describe how much behavioral variability
the irony in the fact that the verbal community was present in the group. From a behavioral per-
teaches an individual to “know oneself.” That is, spective, these statistics do not describe the be-
the two primary ways in which an individual learns havior of any individual in the group accurately,
to identify and label his or her private events is (1) which is a limitation (Johnston & Pennypacker,
to “find commonalities between private and public 1993, p. 324). Each individual in the group has a
events,” or (2) for others to “identify things that unique genetic makeup and an extensive learning
usually occasion it [the private event] or behavior history. Consequently, environmental manipula-
that usually co-occurs” (p. 259). For example, if a tions may evoke different behavior in one individ-
mother and child both cover their ears as a low- ual than in another. To illustrate, one interven-
flying jet passes them, and then the parent says, tion that may be effective for one individual in a
“That hurt my ears,” the child may subsequently group may not be as effective for another.
learn to use the label hurt to describe or tact a sim- Conversely, in a single-case design experiment,
ilar sensation in the ear caused by an ear infection. an individual serves as his or her own experimen-
Similarly, if a child vomits, refuses to eat food, tal control. Thus the experiment accounts for the
and has a temperature, a parent might tell the individual’s unique genetic makeup and operant-
child, “Your stomach hurts.” Skinner (1953) ex- learning history. Because the individual serves as
plains that if a culture cannot teach an individ- his or her own control (i.e., the researcher com-
ual to discriminate private events, the individual pares his or her behavior in baseline or control
may never develop the skill of identifying private conditions to that in intervention conditions),
events, and the individual may not have an exten- this type of research can determine whether an
sive knowledge of self. intervention is effective for an individual.
and procreate; consequently, those adaptive traits ganisms to outrun predators. These organisms are
are more likely to appear in the next generation more likely to survive than are organisms that run
than traits that do not facilitate survival and pro- more slowly and get eaten by predators. Conse-
creation. Natural selection is typically a gradual quently, the organisms with fast-twitch muscles are
process. The genetic makeup of an individual more likely to reproduce and pass their fast-twitch-
evolves gradually over many generations to a point muscle genes to the next generation. By contrast,
where it is drastically different from the genetic suppose the organism with fast-twitch muscles also
makeup of its ancestors (Skinner, 1969). These ge- has blue eyes. Blue eyes may not aid in the sur-
netic variables in conjunction with an individual’s vival of the organism. The organism’s opportunity
environment contribute to both respondent and to reproduce, however, is increased because of its
operant behavior. In fact, Skinner (1981) postu- fast-twitch muscles; thus the organism is likely to
lated that “operant behavior is an evolved process” pass the trait of blue eyes to the next generation.
(p. 502); that is, operant behavior evolved and per- Therefore, blue eyes are a spandrel or by-product
sisted through the phylogenetic process of natural of natural selection. Similarly, reading a textbook
selection because it provided a means by which in- before taking a test may increase the probability
dividuals could acquire behavior that was adaptive that an individual achieves a good grade on a test;
to novel and changing environments. consequently, textbook-reading behavior may in-
Ontogenetic variables are like phylogenetic vari- crease in the future. The environment directly
ables and natural selection, except that the chang- strengthens this behavior through its consequenc-
es occur in an individual’s lifetime and often from es. If the individual drinks green tea while reading
moment to moment, rather than across multiple a textbook, then the behavior of drinking green
generations (Skinner, 1969). Ontogeny refers to tea may increase as a by-product of its correlation
the natural selection of behaviors resulting from with the reinforcing consequences associated with
their environmental consequences. If an individ- reading. The increase in green tea does not cause
ual emits a response (such as betting on the most the individual to do well on his or her test, but the
muscular-looking horse in a race) that produces behavior increases as a by-product of the reinforce-
a favorable or reinforcing consequence (such as ment associated with reading.
winning the bet), the probability that he or she Knowledge of spandrels plays a role in the ap-
will repeat that response in similar environmental plication of behavior analysis. To illustrate, when
contexts increases. That is, the environment se- a behavior analyst implements an intervention to
lects and shapes the behavior because individuals either decrease or increase a specific target behav-
repeat responses that produce favorable outcomes ior, he or she must consider what other behaviors
or consequences in that environment. Similarly, if the intervention will affect as a by-product of the
an individual emits a behavior (such as reaching intervention for target behavior and must plan
into a hole in the ground) that produces an unfa- accordingly. For example, extinction (i.e., no lon-
vorable or punishing consequence (such as being ger providing reinforcement for a behavior that
bitten by an unseen animal), the probability that is maintained by that reinforcer) of disruptive
he or she will emit a similar response in the future behavior may result in an increase in aggression,
decreases. Thus both natural selection and operant even if this latter response did not produce rein-
selection involve selection by consequences. The forcement in the past. Thus the behavior analyst
environment selects traits correlated with survival should add an intervention component to account
of the species with natural selection, and changes for this (e.g., providing access to the reinforcer,
in such traits evolve slowly over many generations. contingent on an alternative behavior).
The environment selects responses correlated with
favorable consequences (e.g., satiation of hunger,
quenching of thirst, numbing of pain) with oper- STRUCTURAL VERSUS FUNCTIONAL
ant selection, and changes in response patterns CLASSIFICATION OF BEHAVIOR
can occur from one moment to the next or over a
person’s lifetime. Most approaches for classifying and understand-
In both phylogeny and ontogeny, some genetic ing aberrant behavior emphasize (1) the behav-
traits and behaviors are not selected directly; rath- ior’s topographical features and (2) how certain
er, they are spandrels—by-products or “free riders” responses tend to co-occur. For example, a boy
of selection of other traits or behaviors (Skinner, who avoids physical contact and eye contact with
1969). For example, suppose a genetic trait for others and who displays peculiar vocal and motor
fast-twitch muscles aids in survival, allowing or- responses (e.g., referring to himself as “you” and
8 In t r o d u c t i o n
others as “I” and repetitively spinning objects) may physics. The behavior of scientists, like that of any
receive a diagnosis of autism spectrum disorder other organism, is a consequence of their interac-
(ASD). Clinicians then may use this diagnosis as tion with the environment. Consequently, behav-
an explanation of the aberrant behavior that led ior analysts must apply the same behavior-analytic
to the diagnosis (e.g., “He repetitively spins objects principles to themselves as they do to the individu-
because he has ASD”). As discussed earlier in the als with whom they conduct research to improve
example provided by Skinner (1953; i.e., “He eats,” their own behavior (Johnston & Pennypacker,
“He is hungry”), the statements “He has ASD” 1993). Skinner (1953) stated that “science is first
and “He repetitively spins objects” are two ways of all a set of attitudes,” and “science [should] re-
of describing the same set of facts; thus one state- ject even its own authorities when they interfere
ment does not explain the other. with the observation of nature” (p. 12). Skinner
Behavior analysts frequently work with children emphasized that “science is a willingness to accept
with ASD (see, e.g., Kodak, Grow, & Bergmann, facts even when they are opposed to wishes,” and
Chapter 23, this volume), but they view the di- that scientists should “remain without an answer
agnosis as descriptive rather than explanative. until a satisfactory one can be found” (pp. 12–13).
Because behavior analysts work to identify oper- This approach to science and to scientists’ atti-
ant contingencies that are maintaining a behav- tudes is equally relevant to clinicians who apply
ior, they assess and categorize aberrant behavior the natural science of behavior analysis to prob-
according to its function. Other fields of science, lems of social importance.
such as microbiology, have long understood the im-
portance of analyzing both the structure and func-
tion of dynamic entities. Behavior analysts employ DIMENSIONS OF THE EXPERIMENTAL ANALYSIS
a similar practice by categorizing behavior by its OF BEHAVIOR
structural characteristics and its function. For ex-
ample, one child with ASD might slap other people In addition to Skinner’s (1969) general views on
because when he does, others are less likely to ap- scientific attitudes, several specific attitudes form
proach him with schoolwork to complete. In this the basis of the experimental analysis of behav-
case, the function of aggression would be to avoid ior as a natural science. These include (1) deter-
schoolwork. By contrast, another child with ASD minism, (2) experimentation, (3) empiricism, (4)
might slap other people because when she does, her reliability, (5) philosophical doubt, and (6) par-
caregivers are more likely to give her physical atten- simony. Behavior analysts are more likely to con-
tion in the form of tactile stimulation (e.g., sensory duct objective research that aids in furthering the
integration). In this case, the function of aggres- theories and principles of behavior analysis if they
sion would be to gain a specific form of caregiver maintain these attitudes.
attention. Thus, although both cases involve the
aggressive act of slapping others, the function of
Determinism
the behavior differs. Analyzing the function of an
individual’s aberrant behavior allows us to predict Determinism is the belief or attitude that all events
the differential effectiveness of interventions more in the universe including behavioral events are
accurately. For example, a time out from attention orderly, lawful, predictable, and determined by
would be an effective intervention for self-injuri- physical causes (Cooper, Heron, & Heward, 2007;
ous behavior maintained by attention, but it would Mazur, 2006; Neuringer, & Englert, 2017). In gen-
likely worsen self-injurious behavior maintained by eral, this means that behavior does not spontane-
avoidance or escape from social interaction. ously occur (e.g., a child does not hit a sibling out
of the blue); there is always a reason why an indi-
vidual or organism emits a behavior (e.g., hitting
THE STUDY OF BEHAVIOR AS A NATURAL SCIENCE in the past resulted in the sibling’s leaving the play
area and the child’s gaining access to the video
The final tenet that distinguishes behavior analy- game). As behavior analysts, we believe that the
sis from traditional psychology is that it examines phylogenetic and ontogenetic variables described
behavior as a natural science (Baum, 2018); thus previously determine current behavior, and we
professionals in the field of behavior analysis con- focus on current operant contingencies because we
duct research and develop theories in ways similar can alter them in ways that promote socially im-
to those in the natural sciences of chemistry and portant changes (e.g., reducing sibling aggression).
History, Philosophy, Principles, and Methods 9
An individual does not have to accept the Skinner (1953) acknowledged that other non-
premise that all behavior is determined to be a experimental methods play a role in the scientific
behavior analyst and to approach the study of be- analysis of behavior, including casual, clinical,
havior as a natural science. To do so would conflict and controlled observations. He also acknowl-
with philosophical doubt, which maintains that we edged that scientists achieve the rigor and control
should continually question our assumptions, find- in the laboratory with nonhuman species at the
ings, and conclusions, and with empiricism, which price of ecological validity or “unreality in condi-
requires that we empirically demonstrate deter- tions” (p. 37). However, the experimental analysis
minism before we accept it fully (see discussions of behavior counters this limitation by focusing on
below). the identification of the basic behavioral processes
Scientists in the field of physics, which is clearly that underlie both simple animal behavior and
a natural science, have adopted stochastic models complex human behavior.
and quantum mechanics (which are not determin- Skinner (1953) argued that “the commonest
istic) to explain certain phenomena that classical objection to a thoroughgoing functional analysis
Newtonian mechanics (which is deterministic) [of complex human behavior] is simply that it can-
does not explain well. Nevertheless, a general be- not be carried out, but the only evidence for this
lief in determinism at this juncture in the develop- is that it has not yet been carried out” (p. 41). As
ment of behavior analysis is at least useful, because several chapters in this book show, we have made
it focuses our attention on the functional charac- considerable progress in conducting functional
teristics of behavior. After we identify the func- analyses of complex human behavior since Skin-
tional variables maintaining behavior, we then ner’s time, such as self-injurious behavior (Iwata,
can manipulate these variables to either increase Dorsey, Slifer, Bauman, & Richman, 1982/1994).
desirable behavior or decrease aberrant behavior.
Scientists would be unable to identify why an in-
Empiricism
dividual emits a behavior and would be unable to
modify the behavior if the behavior of organisms Empiricism is the attitude or viewpoint that the
was not lawful. information available to science comes from the
senses, and that scientists should base their conclu-
sions primarily on sensory evidence. This means
Experimentation
that scientists should be careful observers and
Adopting experimentation as the principal method believe what they observe the world to be, rather
of studying behavior is the only reasonable option than what others have taught them it should be.
if we accept that natural physical events determine When conducting an experiment, behavior
behavior, and that the primary goals of a natural analysts must maintain the attitude of empiricism,
science of behavior are the prediction and control which is the practice of making objective scien-
of its subject matter. Skinner (1953) speculated tific decisions based on factual data regarding in-
that “perhaps the greatest contribution which a terventions, research, and theory development. A
science of behavior may make to the evaluation of scientist’s behavior is a function of environmental
cultural practices is an insistence upon experimen- variables (Johnston & Pennypacker, 1993); thus
tation” (p. 436). numerous variables are controlling his or her be-
Behavior analysts are interested in experimenta- havior at any given time. These variables may
tion involving the manipulation of environmental include personal experiences, personal advance-
antecedents, consequences, or both as the inde- ment, opinions, or beliefs. As much as possible, a
pendent variables, and behavior as the dependent behavior analyst’s decisions should be a function
variable. The purpose of this type of experimenta- of the available empirical data and not of these
tion is to identify the specific environmental vari- other variables. Conversely, if variables other than
ables of which a behavior is a function. A function- objective data are controlling a scientist’s behav-
al relation exists when a change in an independent ior, then the results of the experiment will not be
behavior reliably produces a defined change in the empirical or valid.
dependent variable. Describing a functional rela-
tion between a response and its reinforcer under
Reliability
a specified environmental context is more precise
than saying that the environmental events caused Conducting a single experiment does not provide
the behavior. sufficient evidence to determine how an indepen-
10 In t r o d u c t i o n
child to glue pieces of construction paper together involved in the experimental analysis of behavior.
would not. At any point, a behavior analyst might Examples of conceptually systematic intervention
target several response classes, and he or she must components are extinction and schedules of rein-
prioritize which behaviors are most important to forcement.
modify. Many experiments that use group designs incor-
Consistent with the other two branches of be- porate inferential statistics to determine whether
havior analysis, a principal dimension of ABA statistically significant differences between groups
is its focus on behavior, exemplified through the exist. Applied behavior analysts rarely use statis-
direct observation, objective measurement, quan- tics to determine whether a behavior change is
tification, prediction, and control of behavior significant. Instead, behavior analysts determine
(Baer et al., 1968). Behavior analysts typically do the effectiveness of their procedures by evaluating
not rely on indirect measures of behavior, such as their data, often through visual inspection (Fish-
self-reports, interviews, or checklists (Baer, Wolf, er, Kelley, & Lomas, 2003)—that is, whether the
& Risley, 1987). In addition, they do not attribute individual whose behavior was changed and the
behavior to characteristics of inner qualities, such family, caregivers, and friends of that individual
as personality traits. Instead, they attempt to iden- find the behavior change significant. The fact that
tify a function of the behavior by manipulating en- a behavior change is statistically significant does
vironmental events as independent variables and not mean that the change is socially important.
observing changes in behavior as the dependent For example, a reduction in a boy’s head banging
variable. from a rate of 12 per minute to 6 per minute may
The third dimension of ABA is that it is ana- be statistically significant. However, the boy is still
lytic, which means that when we treat behavior, hitting his head over 300 times an hour. Conse-
we use objective and controlled single-case de- quently, this is not a socially acceptable reduction
signs that permit a believable demonstration of in head banging. A more significant reduction
the effectiveness of our intervention. We strive needs to occur before the intervention can be clas-
to demonstrate a functional relation between our sified as effective.
intervention and any observed changes in the tar- The last principle of ABA is that the findings
get behavior (Baer et al., 1968), using single-case must be generalizable to other settings, caregivers,
experimental designs, including reversal, multiele- or behaviors (Baer et al., 1968). If we decrease a
ment, and multiple-baseline designs (see DeRosa, child’s aggressive behavior to near-zero levels in
Sullivan, Roane, Craig, & Kadey, Chapter 9, this our clinic, but the child still engages in aggression
volume). Baer et al. (1987) emphasized that behav- at school and at home, then the behavior reduc-
ior analysts should select the design that is best tion has not generalized. Generalization is impor-
suited for the experimental question, rather than tant because a decrease in aberrant behavior is not
adjust the experimental question to fit a specific beneficial if the child only spends a few hours a
experimental design. week in the clinic. The behavioral intervention is
Behavior analysts should also be technological, only beneficial if it decreases the child’s behavior
which means that they should thoroughly and ac- across different settings when different caregivers
curately describe their procedures while conduct- implement it. The most effective way to ensure
ing research and in clinical practice. They should that generalization occurs is to program it into the
document this information, which includes a writ- intervention (Stokes & Baer, 1977).
ten procedure, operational definitions of target
behaviors, and procedural-integrity data, in a way
that allows another reasonably competent behav- SUMMARY
ior analyst to replicate the procedure after reading
these documents (Baer et al., 1968, 1987). To summarize, there are three branches of behav-
The assessments and interventions applied be- ior analysis: behaviorism, experimental behavior
havior analysts implement are applied in nature. analysis, and ABA. Each branch is interested in
However, these interventions and the approaches directly studying, predicting, and controlling be-
used to develop the interventions should be con- havior, rather than observing behavior as a means
ceptually systematic (Baer et al., 1968), which of making inferences about the mind, the psyche,
means that they should be based on the basic the self, or other internal mental or metaphysical
behavior principles validated through empirical structures or phenomena. Behaviorists believe
research conducted over many years by scientists that there is continuity between the behavior of
12 In t r o d u c t i o n
human and nonhuman species and between pub- Keller, F. S., & Schoenfeld, W. N. (1950). Principles of
lic and private behavior (e.g., thinking, feeling). psychology. New York: Appleton-Century-Crofts.
Behaviorists believe that behavior is lawful, that Mace, F. C. (1996). In pursuit of general behavioral
it occurs for a reason, and that they can study it relations. Journal of Applied Behavior Analysis, 29,
557–563.
by using the rigorous scientific methods applied in
Mazur, J. E. (2006). Learning and behavior (6th ed.).
other natural or “hard” sciences. Finally, behavior- Upper Saddle River, NJ: Pearson/Prentice Hall.
ists focus on the function(s) of behavior and be- Moore, J. (2003). Explanation and description in tradi-
lieve that they can explain and control behavior tional neobehaviorism, cognitive psychology, and be-
by observing and manipulating environmental havior analysis. In K. A. Lattal & P. N. Chase (Eds.),
events that occur in relation to the behavior. Behavior theory and philosophy (pp. 13–39). New York:
Kluwer Academic/Plenum Press.
Morris, E. K., Todd, J. T., Midgley, B. D., Schneider, S.
REFERENCES M., & Johnson, L. M. (1990). The history of behav-
ior analysis: Some historiography and a bibliography.
Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some Behavior Analyst, 13, 131–158.
current dimensions of applied behavior analysis. Neuringer, A., & Englert, W. (2017). Epicurus and B.
Journal of Applied Behavior Analysis, 1, 91–97. F. Skinner: In search of the good life. Journal of the
Baer, D. M., Wolf, M. M., & Risley, T. R. (1987). Some Experimental Analysis of Behavior, 107, 21–33.
still-current dimensions of applied behavior analysis. Rachlin, H. (2012). Our overt behavior makes us
Journal of Applied Behavior Analysis, 20, 313–327. human. Behavior Analyst, 35, 49–57.
Baum, W. M. (2018). Multiscale behavior analysis and Rachlin, H. (2018). Is talking to yourself thinking?
molar behaviorism: An overview. Journal of the Ex- Journal of the Experimental Analysis of Behavior, 109,
perimental Analysis of Behavior, 110, 302–322. 48–55.
Catania, A. C. (1988). The behavior of organisms as a Ryle, G. (1949). The concept of mind. Chicago: Univer-
work in progress. Journal of the Experimental Analysis sity of Chicago Press.
of Behavior, 50, 277–281. Sidman, M. (2000). Equivalence relations and the re-
Catania, A. C. (2013). Learning (5th ed.). Cornwall-on- inforcement contingency. Journal of the Experimental
Hudson, NY: Sloan. Analysis of Behavior, 74, 127–146.
Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Skinner, B. F. (1938). The behavior of organisms. Acton,
Applied behavior analysis (2nd ed.). Upper Saddle MA: Copley.
River, NJ: Pearson/Merrill–Prentice Hall. Skinner, B. F. (1953). Science and human behavior. New
Donahoe, J. W. (2004). Ships that pass in the night. York: Free Press.
Journal of the Experimental Analysis of Behavior, 82, Skinner, B. F. (1969). Contingencies of reinforcement: A
85–93. theoretical analysis. Englewood Cliffs, NJ: Prentice-
Fisher, W. W., Kelley, M. E., & Lomas, J. E. (2003). Vi- Hall.
sual aids and structured criteria for improving visual Skinner, B. F. (1981). Selection by consequences. Sci-
inspection and interpretation of single-case designs. ence, 213, 501–504.
Journal of Applied Behavior Analysis, 36, 387–406. Smith, T. L. (2017). The epistemologies of parsimony: A
Fraley, L. E., & Vargas, E. A. (1986). Separate disci- review of Ockham’s razors: A user’s manual by Elliott
plines: The study of behavior and the study of the Sober. Journal of the Experimental Analysis of Behav-
psyche. Behavior Analyst, 9, 47–59. ior, 108, 485–498.
Hayes, S. C., Barnes-Holmes, D., & Roche, B. (2001). Stokes, T. F., & Baer, D. M. (1977). An implicit tech-
Relational frame theory: A precis. In S. C. Hayes, D. nology of generalization. Journal of Applied Behavior
Barnes-Holmes, & B. Roche (Eds.), Relational frame Analysis, 10, 349–367.
theory: A post-Skinnerian account of human language Tanney, J. (2018). Remarks on the “thickness” of ac-
and cognition (pp. 141–154). New York: Kluwer Aca- tion description: With Wittgenstein, Ryle, and Ans-
demic/Plenum Press. combe. Philosophical Explorations, 21, 170–177.
Herrnstein, R. J. (1969). Method and theory in the Uttal, W. R. (2000). The war between mentalism and
study of avoidance. Psychological Review, 76, 49–70. behaviorism: On the accessibility of mental processes.
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Mahwah, NJ: Erlbaum.
Richman, G. S. (1994). Toward a functional analysis Watson, J. B. (1913). Psychology as the behaviorist views
of self-injury. Journal of Applied Behavior Analysis, 27, it. Psychological Review, 20, 158–177.
197–209. (Reprinted from Analysis and Intervention in Welsh, F., Najdowski, A. C., Strauss, D., Gallegos, L.,
Developmental Disabilities, 2, 3–20, 1982) & Fullen, J. A. (2019). Teaching a perspective-taking
Johnston, J. M., & Pennypacker, H. S. (1993). Strategies component skill to children with autism in the natu-
and tactics of behavioral research (2nd ed.). Hillsdale, ral environment. Journal of Applied Behavior Analysis,
NJ: Erlbaum. 52, 439–450.
PAR T II
BASIC PRINCIPLES
AND CONCEPTS
A proverb attributed to the Chinese philosopher Lao Tzu is this: “Give a man a
fish and you will feed him for a day; teach him to fish and you will feed him for
a lifetime.” The same can be said for the importance of teaching behavior analysts
the basic principles and concepts of our field, rather than only giving them spe-
cific procedures for specific problems. Learning the basic principles and concepts
of behavior analysis will not only allow applied behavior analysts to address the
specific socially significant problems they are likely to face upon graduation (e.g.,
teaching verbal behavior skills to a child with autism spectrum disorder), but will
also provide them with the tools they will need to address novel problems for
which no protocols exist (e.g., teaching handwashing skills and social distancing
to schoolchildren during a coronavirus pandemic).
In the first chapter of Part II (Chapter 2), Donahoe and Vegas introduce
the basic principles of respondent (or classical) conditioning, in which a neutral
stimulus that previously had no effect on the target response comes to reliably
evoke the target response after the behavior analyst has consistently paired that
previously neutral stimulus with a stimulus that consistently and unconditionally
evokes the target response. These authors discuss how respondent conditioning
is important to the development and amelioration of various clinical disorders,
such as drug addiction and panic disorder. Next, Catania introduces the reader
in Chapter 3 to basic operant contingencies of reinforcement and punishment,
as well as the antecedent stimuli that occasion and establish the effectiveness of
those contingencies. Understanding and analyzing operant contingencies in rela-
tion to the stimuli that evoke and motivate responding are critical to the develop-
ment of effective behavioral assessments (e.g., analyzing idiosyncratic functions of
problem behavior) and behavioral interventions (e.g., developing novel function-
based treatments). In Chapter 4, Mace, Pritchard, and Penney provide a detailed
discussion of basic and complex schedules of reinforcement, including differential,
compound, and concurrent schedules. They provide a variety of examples illus-
13
14 B a s i c P r in c ip l e s a nd C o n c e p t s
trating how such reinforcement schedules are pertinent to applied work. Spradlin,
Simon, and Fisher introduce the basic processes and principles of stimulus control
in Chapter 5, which describes how and why the behavior of individuals changes
and adapts to the stimulus conditions present in different situations and at differ-
ent times (e.g., why spectators at a baseball game sing during the seventh-inning
stretch but not during other parts of the game). These authors describe how
stimulus control is relevant to a wide variety of clinical applications, ranging from
teaching simple discriminations (e.g., only crossing a street when the traffic light
is green) to teaching in ways that promote the emergence of generative verbal
behavior.
In a new contribution to this edition of the handbook, Chapter 6, Podlesnik,
Jimenez-Gomez, and Kelley discuss two quantitative theories of behavior that
have considerable relevance to applied work—namely, the generalized matching
law and behavioral-momentum theory. The matching law is particularly relevant
to how and why individuals allocate their time and responding to adaptive and
maladaptive responses, and behavioral-momentum theory is pertinent to under-
standing and mitigating treatment relapse. In another new addition, Chapter
7, DeLeon and colleagues describe the subarea of behavioral economics, which
integrates principles from microeconomics and behavior analysis to demonstrate
and analyze how the demand for commodities (i.e., reinforcers) relates to the
work or effort an individual puts forth to obtain those reinforcers. They illustrate
how these concepts are directly applicable to disorders of impulse control, such as
attention-deficit/hyperactivity disorder, and to various health-related behaviors.
CHAP TER 2
John W. Donahoe and Rocío Vegas
At the dawn of the 20th century, two scientists— reinforcement could occur rapidly and be studied
one in St. Petersburg, Russia, and the other in within the laboratory.
Cambridge, Massachusetts—independently began Pavlov’s and Thorndike’s procedures differed
the search to discover how individual experi- in a critically important respect, but they began
ence produced long-lasting changes in behavior. from the same starting point—by presenting the
The first scientist was Ivan Pavlov (1927/1960), learner with a stimulus to which the learner would
a physiologist whose earlier research on digestion already reliably respond. Stated more technically,
would earn a Nobel Prize. The second was Edward both Pavlov and Thorndike presented an eliciting
Thorndike (1903), a psychologist whose published stimulus (say, food) that reliably evoked a behav-
work would eventually exceed in volume that of ior (say, eating). Because of natural selection, the
any other psychologist—past or present (Jonçich, taste and smell of food elicit a variety of respons-
1968). The methods these pioneers used differed, es—including salivation, approach, and eating.
but both described themselves as following in The experimenter could manipulate an eliciting
Darwin’s footsteps: That is, they were trying to stimulus easily and readily measure the elicited
explain complex phenomena as the cumulative response. Pavlov and Thorndike differed about
product of more basic processes. For Darwin, the the type of event that reliably preceded the elicit-
basic process was evolution through natural selec- ing stimulus and its elicited response. In Pavlov’s
tion. For Pavlov and Thorndike, the basic process laboratory, a specified stimulus (such as the tick-
has come to be known as selection by reinforce- ing sound of a metronome) preceded the presenta-
ment or simply reinforcement. Darwin primarily tion of food to a dog. In Thorndike’s procedure, a
studied how changes in structure could arise from specified behavior (such as moving a rod) allowed a
natural selection. Pavlov and Thorndike stud- cat to escape from a cage and gain access to food.
ied how changes in function could arise from se- Figure 2.1 shows the relation between P avlov’s and
lection by reinforcement. Both shared the belief Thorndike’s procedures. As indicated in this fig-
that even the most complex phenomena could be ure, the critical difference between the two pro-
understood as the cumulative result of selection cedures is that a specific stimulus (Si) reliably pre-
processes acting over time. The selection process ceded the eliciting stimulus in Pavlov’s procedure,
discovered by Darwin usually acted over extremely whereas a specific response (Rj) reliably preceded
long periods and could be known largely through the eliciting stimulus in Thorndike’s procedure. B.
naturalistic observation. By contrast, selection by F. Skinner was the first to appreciate fully that this
15
16 B a s i c P r in c ip l e s a nd C o n c e p t s
typical case. In the present example, after the ex- second variable was not identified until the late
perimenter paired the ticking sound with food, the 1960s, beginning with the work of Leon Kamin
ticking sound evoked salivation when presented (1968, 1969). Kamin used a method developed by
alone. The conditioned response (CR) was the re- Skinner’s student William Estes (Estes & Skin-
sponse that the CS evoked after the experimenter ner, 1941). Kamin found that temporal contiguity
repeatedly paired the CS with the US/UR, and is between the CS and US/UR was sometimes not
the behavioral effect scientists usually measure in enough to produce conditioning. The US also had
the Pavlovian procedure. The process whereby the to evoke a change in ongoing behavior in addition
environment acquired its ability to control behav- to contiguity with the CS. That is, the US had
ior was called conditioning because the ability of to evoke behavior that was not already occurring
the CS to evoke the CR was conditional upon (i.e., before the experimenter presented the US. A con-
dependent upon) pairing the CS with the US/UR. tiguous CS–US/UR relation would produce con-
Scientists generally call Pavlov’s procedure clas- ditioning only if such a change occurred. We call
sical conditioning in recognition of his historical this second factor a behavioral discrepancy.
priority. Scientists ultimately used the term con-
ditioning for Thorndike’s procedure—but called
Temporal Contiguity
it operant conditioning to distinguish it from Pav-
lov’s procedure. As already noted, the Thorn- The classical procedure permits an analysis of the
dike–Skinner procedure differed from Pavlov’s effects of the temporal relation between the CS
in a critical respect: The event that reliably pre- and the US/UR on conditioning (Gormezano &
ceded the eliciting stimulus was a response, not a Kehoe, 1981; Kirkpatrick & Balsam, 2016). Experi-
stimulus. Because an increase in the strength of mental analysis is facilitated because the experi-
the response was dependent on presentation of an menter can control the presentation of both stim-
eliciting stimulus in both procedures, we call the uli and can measure the relevant behavior. Figure
eliciting stimulus a reinforcing stimulus or simply a 2.2 shows a representative effect of varying the
reinforcer. An eliciting stimulus functions as a rein- temporal relation between the onset of the CS and
forcer in either conditioning procedure. US (Smith, Coleman, & Gormezano, 1969). Here
the CS was a tone, the US was a mild shock given
Temporal Relation between the CS and the US/UR FIGURE 2.2. Effect of the CS–US interval on the per-
Given an appropriate choice of CS and US, what centage of CRs with the Pavlovian procedure. Differ-
must take place for conditioning to occur? Re- ent groups of rabbits were trained at each of the CS–US
intervals. The CS was a tone, and the US was a mild
search has identified two critical variables: (1) the
shock in the region of the eye. The shock elicited a
temporal relation between the CS and US/UR, “blink” of the nictitating membrane (NM). ms, milli-
and (2) a change in the ongoing behavior evoked seconds. From Donahoe and Palmer (1994/2005), based
by the US. Pavlov demonstrated the first vari- on findings from Smith, Coleman, and Gormezano
able, the temporal relation between the CS and (1969). Copyright © 2005 John W. Donohoe. Reprinted
the US/UR, known as temporal contiguity. The by permission.
18 B a s i c P r in c ip l e s a nd C o n c e p t s
near a rabbit’s eye, and the CR was movement of men (e.g., Kehoe & Macrae, 1994; Van Willigen,
the nictitating membrane that the shock elicited. Emmett, Cote, & Ayres, 1987). Longer-term rela-
(The nictitating membrane is a semitransparent tions between the environment and behavior may
membrane that extends over the eye to protect it. occur only by filling the gap between the CS and
This membrane is present in many animals, such remote US/URs with more immediate events that
as dogs and cats, but is vestigial in humans, where sustain conditioning because the conditioning
only the pink tissue in the nasal corner of each eye process operates over very short time intervals. We
remains.) The nictitating membrane’s response is describe how this occurs in a subsequent section
particularly well suited for experimental analysis, on higher-order conditioning and conditioned re-
because membrane movement is rare except when inforcement.
the eye is “threatened.” Thus any movement of
the nictitating membrane following the CS is very
Critical Temporal Relation: CS–US or CS–UR?
likely to be a CR and not the result of other fac-
tors. The experimenter in the Pavlovian procedure
After CS–US/UR pairings in which the experi- manipulates the relation between stimuli—the
menters trained different animals with different CS and the reinforcing US. By contrast, we have
intervals between the CS and the US/UR, the seen that the experimenter in the operant proce-
major findings shown in Figure 2.2 were these: dure manipulates the relation between a response
(the operant) and the reinforcing stimulus. Note
1. When the CS came after the US/UR, a back- that although the experimenter directly controls
ward-conditioning procedure, conditioning did the temporal relation between the CS and US,
not occur. the relation between the CS and UR necessar-
2. As the interval between the CS and the US/ ily varies as well (see Figure 2.3). Thus whether
UR increased, CR responding became more the CS–US relation or the CS–UR relation is
probable at first and then reached a maximum critical for conditioning is unclear. Teasing apart
at a relatively short value (here, less than a these relations might appear unimportant, except
half-second, or 500 milliseconds). that the difference between the events that the
3. CR responding declined when the CS–US/ experimenter manipulates in the Pavlovian and
UR interval further increased. operant procedures has led many to interpret the
difference as more than a procedural distinction
To summarize, selection by reinforcement oc- (e.g., Rescorla, 1991). Specifically, the subject in
curs over only a relatively brief interval in a well- a classical procedure is sometimes said to learn a
controlled Pavlovian procedure. Stimuli (CSs) stimulus–stimulus relation, whereas the subject in
that reliably precede the US/UR acquire control an operant procedure learns a stimulus–response
over the CR due to reinforcement. Conditioning relation. Figure 2.1 illustrates the stimulus–re-
may occur after only one or a few CS–US/UR sponse relation of the operant procedure. Note
pairings, depending on the specific training regi- that the reinforced operant necessarily occurs in
the presence of some environmental stimulus. Sci-
entists call the stimulus that guides behavior in
(tone) the operant procedure a discriminative stimulus. As
CS Skinner (1937) noted, “It is the nature of [oper-
(shock) ant] behavior that . . . discriminative stimuli are
US practically inevitable” (p. 273; see also Catania &
(nictitating membrane response) Keller, 1981; Dinsmoor, 1995; Donahoe, Palmer, &
UR Burgos, 1997). Thus some environmental stimulus
is likely to acquire control over behavior in the
time operant procedure, even though the experimenter
may not manipulate this relation directly. The
idea that the participant acquires different kinds of
FIGURE 2.3. Schematic diagram of the events in a typi-
cal Pavlovian (respondent) procedure. A specified en- relations in the two procedures rests upon the fact
vironmental stimulus (here, the CS of a tone) precedes that the experimenter manipulates different kinds
an eliciting stimulus (here, the US of a mild shock in of events in the two procedures. Whether the CS–
the region of the eye of a rabbit) that evokes a response US relation, a relation between two stimuli, or the
(here, the UR of a brief nictitating membrane response). CS–UR relation, a relation between a stimulus
Respondent (Pavlovian) Conditioning 19
and a response, is fundamental in the Pavlovian tating membrane UR. The autonomic nervous
procedure becomes important to determine. system governs heart rate, which is important in
Donahoe and Vegas (2004) developed an ex- emotional responding. We may anticipate that
perimental procedure in which the UR occurred the more delayed and longer-lasting emotional
with enough delay after the US that they could responses mediated by the autonomic nervous sys-
separate the effects of the CS–UR relation from tem may be acquired and maintained under cir-
the CS–US relation. The experimenters injected cumstances in which more rapidly occurring and
water into a pigeon’s mouth as a US and identified shorter-duration responses are not.
the throat movements that accompanied swallow-
ing as a UR. The advantage of this procedure is
Behavioral Discrepancy Produced
that throat movements, the UR, begin about 500
by the Reinforcing Stimulus
milliseconds after the injection of water, the US,
and these movements continued for several sec- As noted earlier, experiments conducted by Leon
onds. Thus the CS could be introduced after the Kamin in the late 1960s showed that something
onset of the US, but before or during the UR. This in addition to contiguity between the CS and US/
produced a backward CS–US relation—a relation UR was needed for conditioning. Many subse-
that does not promote conditioning with brief- quent experiments confirmed and extended these
duration URs, such as the nictitating membrane findings by using both Pavlovian and operant pro-
response. Also, experimenters could introduce the cedures (e.g., Fam, Westbrook, & Holmes, 2017;
CS after the onsets of both the US and UR but Rescorla & Wagner, 1972; Vom Saal & Jenkins,
still overlap the UR, because the swallowing UR 1970). Some previous studies had pointed in a sim-
continued for some seconds. The central finding ilar direction, but their significance was not fully
was that the CS, a light, came to evoke the CR, appreciated (e.g., Johnson & Cumming, 1968; Re-
swallowing, independently of the relation of the scorla, 1967).
CS to the US as long as the CS preceded or over- Kamin devised a multiphase Pavlovian proce-
lapped the UR. Thus conditioning in the Pavlov- dure known as the blocking design, which is sum-
ian procedure varied more systematically with the marized in Table 2.1. Kamin conditioned the CRs
temporal relation between the CS and UR than to CS1 in the experimental group of animals dur-
between the CS and US. The inference that the ing Phase 1. He continued to pair CS1 with the
learner acquired a different kind of relation with US in Phase 2, but CS1 was now accompanied by
the Pavlovian procedure—a relation between two CS2, a stimulus that came on and went off at the
environmental events (CS–US) instead of an en- same time as CS1. Note that the temporal relation
vironment–behavior relation (CS–UR)—arose of CS2 to the US/UR should also support condi-
from a misinterpretation of the finding that varia- tioning if CS–US/UR contiguity were the only
tions in the CS–US relation affected condition- requirement: The temporal relation of CS2 to the
ing. Varying the temporal relation between the US was the same as for CS1. To determine wheth-
CS and US also varied the CS-UR relation, and er each stimulus had acquired the CR, Kamin pre-
this is the relation that appears critical. sented CS1 and CS2 separately in a test phase. As
The finding that the relation of the CS to the shown in Table 2.1, the CR occurred to CS1 but
behavior evoked by the US allowed us to under- was attenuated or eliminated to CS2. An other-
stand certain previously puzzling findings. For ex- wise effective temporal relation of CS2 to the US/
ample, conditioning of the nictitating membrane UR did not condition a CR. Prior conditioning to
does not occur if a CS occurs after a shock US pre- CS1 had blocked conditioning to CS2.
sented near the rabbit’s eye, which is a backward- One possible interpretation of the blocking of
conditioning procedure. However, conditioning of conditioning to CS2 is that two CSs cannot be
a change in heart rate does occur with this same conditioned simultaneously to the same US. Vari-
preparation (Schneiderman, 1972). If condition- ous control experiments eliminated this possibil-
ing is dependent on the CS-UR relation, these ity. For example, experimenters first conditioned
different results may be explained as follows: The animals to an unrelated stimulus CS3 during
nictitating membrane UR is very rapid and of very Phase 1 (see Table 2.1). These animals then re-
brief duration, whereas the heart rate UR is more ceived the same training as the experimental
delayed and longer lasting. Thus the backward- group. The experimenters simultaneously present-
conditioning procedure allowed the CS to precede ed CS1 and CS2 followed by the US/UR during
and overlap the heart rate UR but not the nicti- Phase 2. Then, when the experimenters presented
20 B a s i c P r in c ip l e s a nd C o n c e p t s
Test phase CS1 (tone) presented alone—CR CS1 (tone) presented alone—CR
CS2 (light) presented alone—no CR CS2 (light) presented alone—CR
CS1 and CS2 separately to the control animals continue to be an effective reinforcer. The more
during the test phase, each stimulus evoked a CR. deprived the learner is of contact with a stimulus,
Thus two stimuli could be conditioned to the same the more vigorous is the behavior evoked by that
US simultaneously, and the explanation of block- stimulus, and the more effectively it can function
ing had to be sought elsewhere. as a reinforcer (cf. Donahoe, 1997a; Premack, 1959;
Robert Rescorla and Allan Wagner (1972) first Seaver & Bourret, 2020; Timberlake & Allison,
offered a compelling explanation of blocking. 1974). Food for a food-deprived animal evokes vig-
Stated in behavioral terms instead of in the as- orous eating; food for a satiated animal does not.
sociationist language of the original formulation,
a stimulus becomes a CS when the UR that is
A Unified Principle of Reinforcement
evoked by the US differs from the behavior occur-
ring just before the US was presented, given an ap- Our present understanding of the conditions re-
propriate temporal relation to the UR (Donahoe, quired for selection by reinforcement may be sum-
Crowley, Millard, & Stickney, 1982; Mizunami, marized as follows: If a stimulus evokes a change in
Terao, & Alvarez, 2018; Stickney & Donahoe, ongoing behavior (behavioral discrepancy), then
1983). Technically speaking, the US must evoke that stimulus functions as a reinforcer for the en-
a behavioral change or discrepancy for it to function vironment–behavior events that accompany the
as a reinforcer. In the experimental group, the ex- discrepancy (temporal contiguity; Donahoe et al.,
perimenter blocked conditioning to CS2 during 1982, 1993).
Phase 2, because CS1 was already evoking the Figure 2.1 shows that in the Pavlovian proce-
CR (e.g., salivation) before the US evoked the UR dure, the stimulus that reliably precedes the dis-
(also salivation). The UR did not constitute a suf- crepancy is the CS, and the behavior that reliably
ficient change in ongoing behavior to produce new instantiates the discrepancy is the UR. Figure 2.1
conditioning. In the control group, however, when also shows that in the simplest operant proce-
CS2 occurred during Phase 2, it accompanied a dure, no particular stimulus reliably precedes the
stimulus (CS1) that did not already evoke a CR, discrepancy, and the events that accompany the
and both CS1 and CS2 became effective condi- discrepancy are the operant and the UR. Thus
tioned stimuli. both the operant and the CR are acquired in the
The significance of the behavioral-discrepancy operant procedure. The basic conditioning pro-
requirement is that a stimulus must evoke a change cess, selection by reinforcement, appears to be the
in behavior for it to function as a reinforcer. In same in both the classical and operant procedures.
everyday language, the learner must be surprised However, the events that reliably accompany the
to make the elicited response. Natural selection discrepancy in the two procedures are different,
has chosen neural mechanisms of conditioning and the outcomes of the two procedures are there-
that function only when the would-be reinforcer fore different. In the Pavlovian procedure, a spe-
evokes a behavioral change. As a possible everyday cific stimulus, the CS, gains control over a specific
example, parents who lavish praise on a child in- response, the CR, but whatever other responses
dependently of the child’s behavior may find that occur at the time of the discrepancy are uncon-
their praise becomes ineffective as a reinforcer. Fre- trolled. In the operant procedure, two specific re-
quent and indiscriminate praise is not surprising. sponses, the operant and the CR, are acquired, but
Conversely, sparingly delivered parental praise may the antecedent stimuli that accompany the oper-
Respondent (Pavlovian) Conditioning 21
ant are uncontrolled. (Note that a discriminated nonchance relations between events only through
operant procedure does specify the antecedent repeated experience. A unified reinforcement
stimuli.) Because the reinforcement process ap- principle accommodates the different behavioral
pears to be fundamentally the same in the Pavlov- outcomes produced by the respondent and oper-
ian and operant procedures, we call it the unified ant procedures, while also allowing for the occa-
reinforcement principle (Donahoe et al., 1982). sional emergence of superstitious conditioning in
In the Pavlovian procedure, no specified re- either procedure. Natural selection has produced
sponse necessarily precedes the UR-instigated a conditioning process that is sensitive to reliable
discrepancy, and in the simple operant procedure, relations between the environment and behavior,
no specified environmental stimulus necessarily but the process is not infallible.
precedes the discrepancy. However, this does not
mean that a response other than the CR cannot
be acquired in the classical procedure, or that no CONDITIONED REINFORCEMENT
stimulus acquires control of behavior in the oper- AND COMPLEX BEHAVIOR
ant procedure (Donahoe et al., 1997). To the ex-
tent that conditioning is promoted by only a single In the larger world outside the laboratory, many
occurrence of a discrepancy, other responses may stimuli that serve as effective reinforcers do not
become conditioned in the Pavlovian procedure, elicit responses that are readily detectable at the
and some stimuli may acquire control of the op- behavioral level of observation. Talking with an-
erant in the operant procedure. Indeed, Skinner other person continues when the other person
(1948) demonstrated that an operant response engages in that subtle behavior we call paying at-
may be conditioned when reinforcers occur inde- tention. But paying attention is not an elicited be-
pendently of behavior. The responses the partici- havior and includes barely perceptible and highly
pant acquired are those that happen by chance to variable behavior such as maintaining eye contact.
precede the reinforcer. Thus a pigeon that is given In addition, stimuli that serve as reinforcers in the
occasional presentations of food independently real world often seem delayed far beyond the time
of its behavior may nevertheless acquire behavior that experimental analysis indicates is effective:
such as pecking screw heads on the wall of the test The worker who comes to the job on Monday may
chamber or pacing in front of the feeder (Staddon not receive a paycheck until Friday. How, then,
& Simmelhag, 1970; Timberlake & Lucas, 1985). are we to understand the conditioning and main-
Once such a response happens to precede food, it tenance of human behavior when it apparently
can be further strengthened by later presentations does not satisfy the requirements for condition-
of food that also follow that response. Skinner ing revealed through experimental analysis? The
referred to this phenomenon as superstitious con- phenomenon of conditioned reinforcement provides
ditioning. Experimenters discovered an analogous a major part of the answer.
phenomenon in the Pavlovian procedure (Bene-
dict & Ayres, 1972). When an experimenter pres-
Higher‑Order and Conditioned‑Reinforcement
ents a CS and a US independently, chance con-
Procedures
junctions of the CS with the US may allow the CS
to acquire control of the CR, especially when the The experimental analysis of conditioning with
chance pairings occur early in training. the Pavlovian and operant procedures uses re-
The conditioning process cannot distinguish inforcers that elicit a readily measured response,
between a chance and a nonchance contiguity of such as salivating or eating. The ticking sound of
an event with a reinforcer on a single occasion. Per- a metronome was followed by food, and salivation
haps Pavlov’s dog perked up its ears after hearing came to be evoked by the ticking sound. The ex-
the metronome when the experimenter presented perimenter presented food that elicited eating after
food. The procedure might then strengthen the a rat pressed the lever, which conditioned both bar
behavior of raising the ears in the presence of the pressing and salivation (cf. Shapiro, 1962). Stimuli
sound of the metronome and not merely the CR that function as reinforcers when the learner does
of salivating. Similarly, Skinner’s rat likely looked not have specific prior experience with those stim-
at the bar when about to press it, or his pigeon uli are called unconditioned reinforcers. These stim-
looked at the disk when about to peck it before uli evoke behavior that has benefited survival of
the response produced the food. The condition- the species over evolutionary time. Stimuli such as
ing process can discriminate between chance and sweet-tasting substances, which are generally rich
22 B a s i c P r in c ip l e s a nd C o n c e p t s
in calories, and sexual stimuli, which are linked to experimental chamber. The click occurred imme-
reproductive behavior, are prime examples. diately after the rat pressed the lever and served as
However, other stimuli acquire the ability to a conditioned reinforcer for lever pressing.
function as reinforcers only after they have been Temporal contiguity and behavioral discrepan-
paired with existing reinforcers. These are con- cy are required for both conditioned and uncondi-
ditioned reinforcers, whose reinforcing function is tioned reinforcement. In support of the contiguity
dependent on individual experience. A stimulus requirement for conditioned reinforcement, after
functions as a conditioned reinforcer in the Pav- a stimulus (CS1) has become a CS by pairing it
lovian procedure after it has been paired with an with an unconditioned reinforcer, another stimu-
unconditioned reinforcer. A CS can then func- lus (CS2) will become a CS only if it precedes CS1
tion as a reinforcer for another stimulus if the new by no more than a few seconds (Kehoe, Gibbs,
stimulus is followed by a CS that already evokes Garcia, & Gormezano, 1979). Findings obtained
behavior (the CR) due to prior conditioning. We with the blocking design show that behavioral
can observe such effects in humans as when the discrepancy is also required for conditioned rein-
thought of a stimulus that has been paired with forcement: The experimenter first paired a CS1,
food, such as the sight of food, may function as such as a tone, with food for a pigeon. Then a
an effective CS. Imagine seeing food that you in- CS2, such as a light, accompanied the tone, and
tend to eat at your next meal, particularly if you the experimenter continued to pair the compound
are hungry. Can you detect an increase in saliva- light–tone stimulus with food. When the experi-
tion? Imagining a favored food evokes conditioned menter later presented the pigeon with two disks,
salivation through previous experience with that pecking the disk that produced the tone increased,
food. but pecking the disk that produced the light did
A higher-order conditioning procedure is a Pavlov- not. The light was blocked from becoming a con-
ian experimental arrangement in which the ex- ditioned reinforcer (Palmer, 1987). To summarize,
perimenter pairs a previously neutral stimulus with conditioned and unconditioned reinforcement
an established CS. As a laboratory example, the both require temporal contiguity and behavioral
experimenter first pairs a CS1, such as a tone, with discrepancy.
food. After CS1 evokes a salivary CR, the experi-
menter then introduces a CS2, such as a light, that
Role of Conditioned Reinforcement
precedes CS1 alone. As a result, CS2 also acquires
in Complex Behavior
the ability to evoke a salivary CR, even though
CS2 was never paired with food. CS1 has func- Human behavior is increasingly acquired and
tioned as a conditioned reinforcer. If the experi- maintained by conditioned reinforcers. Consider
menter continues the higher-order procedure, but the following example: Assume that a child has
the CS2–CS1 sequence is never followed by food, already learned the alphabet. Stated more techni-
then CS1 will cease to function as a conditioned cally, the visual stimuli provided by letters have
reinforcer. In laboratory experiments on higher- become discriminative stimuli that control say-
order conditioning, responding to CS2 is usually ing the names of the letters; B controls /bee/, C
maintained by occasionally presenting CS1 alone, controls /see/, D controls /dee/, and so on. Keller
followed by food (cf. Rescorla, 1980). and Schoenfeld (1950) proposed that discrimina-
We also can study conditioned reinforcers using tive stimuli can function as conditioned reinforc-
operant procedures (Williams, 1994a, 1994b). For ers when they follow other responses. Dinsmoor
example, if the experimenter has previously fol- (1950) demonstrated experimentally that this
lowed the sound of operation of the feeder with conjecture was correct. (For reviews, see Williams,
access to food, then rats may acquire lever pressing 1994a, 1994b.) Given these findings, suppose that
when pressing is followed by that sound (Skinner, an experimenter asks children who have learned
1938). This procedure is called a conditioned or sec- to say letter names when they see a written letter
ondary reinforcement procedure, because the sound are then asked to write letters when the experi-
of feeder operation, not food, has strengthened menter says each letter’s name. The closer their
the operant. Skinner appreciated that a reinforcer writing approaches the appearance of the letter
must follow almost immediately after an operant if the experimenter has instructed them to write, the
the operant is to be acquired. Therefore, the click- greater the conditioned reinforcement provided by
ing sound of the feeder mechanism was paired with the visual stimuli produced by their own writing.
food before introducing the response lever into the In short, writing a letter produces immediate con-
Respondent (Pavlovian) Conditioning 23
ditioned reinforcement, the magnitude of which tems involved in unconditioned and conditioned
increases as the writing more closely approximates reinforcement indicates that both classes of rein-
the correct appearance of the letter. Learning to forcers engage the same basic neural system.
say letters has allowed the children to write letters
with little or no reliance on external reinforcers
Neural Systems of Unconditioned
from others. Conditioned reinforcers that shape
and Conditioned Reinforcement
behavior in this manner illustrate the process of
automatic conditioned reinforcement and play an Figure 2.4 displays a side view of the human cere-
especially important role in the acquisition of bral cortex. The shapes outlined by dashed lines
verbal behavior (Donahoe & Palmer, 1994/2005; indicate the subcortical structures and pathways
Petursdottir & Lepper, 2015; Sundberg, Michael, beneath the cortex that are critical for reinforce-
Partington, & Sundberg, 1996; Vaughan & ment. All unconditioned reinforcers, including
Michael, 1982). drugs of abuse, food, and sexual stimuli, activate
Behavior analysts appreciate the central role brain cells (neurons) that are in a subcortical
of reinforcement in understanding behavior, and area called the ventral tegmental area (see Figure
automatic conditioned reinforcement provides 2.4). Neurons whose cell bodies are in the ventral
a persuasive account of the acquisition of some tegmental area send widely distributed projec-
aspects of complex behavior. However, for many tions (axons) to the prefrontal cortex. The curved
psychologists, the concept of conditioned rein- arrow between the ventral tegmental area and the
forcement seems to appeal to something akin to prefrontal cortex indicates these pathways in Fig-
magic. The reluctance to accept conditioned rein- ure 2.4. The prefrontal cortex receives projections
forcement arises in part because the conditioned from sensory areas of the brain that converge on
reinforcer often evokes no measurable behavior. neurons that lead ultimately to behavior via their
We can observe only an increase in the behav- connections to the motor cortex. The prefrontal
ior that precedes the conditioned reinforcer. We cortex is thus the area of the brain in which inputs
can document the occurrence of conditioned re- from the environment converge on neurons that
inforcement independently, however, by moving lead to behavior. Because of this, connections be-
from the behavioral to the neural level of observa- tween these neurons are prime targets for selection
tion. The following brief sketch of the neural sys- by reinforcement.
n. Acc
VTA
FIGURE 2.4. Side view of the left side of the human cerebral cortex. The front of the brain is toward the left portion
of the figure. The cortical regions are labeled for purposes of this chapter and are not intended to be complete. For
example, the region designated as primary sensory cortex is concerned with vision and does not include other sensory
areas such as temporal cortex, which concerns audition. The forms outlined with dashed lines represent subcortical
structures and pathways involved in the reinforcement of behavior. VTA, ventral tegmental area; n. Acc, nucleus ac-
cumbens. See the text for further information.
24 B a s i c P r in c ip l e s a nd C o n c e p t s
ers. Although stimuli that function as conditioned havior has been reinforced. For a learner with an
reinforcers can also serve other functions (Bullock appropriate history of reinforcement, stimuli that
& Hackenberg, 2014; Russell, Ingvarsson, Haggar, engage the neural mechanisms of conditioned re-
& Jessel, 2018; Shahan, 2010), their reinforcing inforcement continuously and immediately rein-
function is clear. force temporally extended sequences of behavior.
We should mention one additional aspect of the
neural mechanisms of reinforcement. The lower
panel (C) of Figure 2.5 shows the response of do- FACTORS THAT AFFECT BEHAVIORAL CHANGE
pamine neurons when the experimenter presented IN THE CLASSICAL PROCEDURE
the CS but did not follow it by the US. The CS
produced a burst of firing of dopamine neurons, The classical procedure is best suited for the exper-
but then a decrease in firing at the time when the imental analysis of the effects of varying the char-
US would normally occur. At that time, the CS acteristics of the CS and the reinforcer (the US)
briefly inhibited the dopamine neurons in the and of the temporal relation between them. By
VTA. This had the effect of preventing the US contrast, the operant procedure is best suited for
from serving as a reinforcer at that moment. The the experimental analysis of the effects of varying
inhibition of dopamine activity by a CS in the the characteristics of the response, the reinforcer,
Pavlovian procedure, or by a discriminative stimu- and the temporal relation between them. Discrim-
lus in the operant procedure during the time when inated operant conditioning, which is considered
the US occurs, is the neural basis of blocking (Do- by Catania in Chapter 3 of this volume, permits
nahoe et al., 1993; cf. Schultz, 1997, 2001; Waelti, the experimental analysis of all three events—the
Dickinson, & Schultz, 2001). Thus behavioral and environmental stimulus, the behavior that occurs
neuroscientific research agree that both temporal in the presence of the stimulus, and the reinforcer.
contiguity and discrepancy are required for con-
ditioning. Neuroscience complements behavior
Characteristics of the CS
analysis (Donahoe. 2017). As Skinner (1988) rec-
ognized, “The . . . gap between behavior and the Experimenters have used many types of stimuli as
variables of which it is a function can be filled only CSs in the classical procedure. They include the
by neuroscience, and the sooner . . . the better” usual exteroceptive stimuli—visual, auditory, and
(p. 460). tactile stimuli—but also interoceptive stimuli, or
Finally, consider how the neural mechanisms of those produced by stimulation of internal recep-
conditioned reinforcement help us understand the tors. Indeed, Pavlovian conditioning influences
acquisition of complex behavior by experienced the regulation of many intraorganismic responses,
learners. Let us return to the example of children such as blood pressure, glucose levels, and other
learning to write their letters after having learned behavior mediated by the autonomic nervous sys-
to speak their letters. When the children’s writing tem (Dworkin, 1993). Emotional behavior is espe-
behavior produces a visual stimulus that approxi- cially affected by the variables manipulated in the
mates the appearance of a well-formed letter, that Pavlovian procedure because of their pervasive ef-
stimulus then presumably initiates a brief burst of fects on autonomic responses (Skinner, 1938). As
dopamine firing proportional to how closely the one example of interoceptive conditioning, stimuli
visual stimulus corresponds to the letter that con- from inserting a hypodermic needle into the skin
trols saying its name. The cumulative effect of this precede the effects of an injected drug, and these
process is that the children produce a progressively stimuli then become CSs for drug-related respons-
well-formed letter. The same process occurs with es. The effects of such CSs can be complex. When
the behavior of an office worker who receives a internal receptors on neurons sense the increased
paycheck only at the end of the week. Behavior concentration of the injected compound, the level
is acquired and maintained not by the delayed of that compound normally produced by neurons
paycheck, but by immediate conditioned reinforc- decreases. For example, cocaine raises the level of
ers. The sources of these conditioned reinforcers circulating dopamine, and receptors on neurons
vary and include concurrent social stimuli (such that release dopamine in the brain detect this in-
as praise from colleagues) and stimuli produced by crease. These neurons then reduce their produc-
the worker’s own behavior (such as seeing that he tion of dopamine. Thus the true UR is not an in-
or she has done a job well). A well-done job is one crease in dopamine from the injection of cocaine,
that produces stimuli in whose presence past be- but a decrease in the production of dopamine by
26 B a s i c P r in c ip l e s a nd C o n c e p t s
neurons whose receptors detect the increased lev- ences in which the object of the phobia has been
els of dopamine. As a result, for someone addicted paired with an aversive stimulus (Merckelbach &
to cocaine, neurons show a conditioned decrease Muris, 1997).
in the production of dopamine when the person Instances of differential associability also arise
is given a placebo (i.e., an injection CS that is not from the history of the species of which the in-
followed by cocaine) or presented with stimuli that dividual is a member. Taste (gustatory) or smell
have previously accompanied drug intake. De- (olfactory) stimuli more readily become CSs when
creases in dopamine induce withdrawal symptoms, paired with food and the consequences of inges-
because the stimulus of the injection produces a tion than do visual or auditory stimuli (Garcia,
conditioned reduction in the internal production Erwin, & Koelling, 1966). If nausea is a conse-
of dopamine (Eikelboom & Stewart, 1982; see also quence of ingestion, as occurs with poisons, then
Sokolowska, Siegel, & Kim, 2002). Classical con- an aversion to that food is conditioned. This
ditioning plays an important role in dysfunctional phenomenon is called taste aversion and undoubt-
behavior, such as drug addiction. edly owes its occurrence to the special status that
Classical conditioning can also affect panic dis- olfactory and gustatory stimuli have with respect
order (e.g., Bouton, Mineka, & Barlow, 2001). The to the ingestion of food. Over evolutionary time,
life histories of those afflicted with panic disorder such stimuli have come came immediately before
often include pairings of the feared stimulus with the ingestion of food, thus providing the relative
an aversive US (Acierno, Hersen, & Van Hasselt, constancy of environmental conditions required
1993). for natural selection to operate. Under constant
Although many stimuli can function as CSs, conditions, special neural circuitry is selected be-
all stimuli are not equally effective with all USs. tween these sensory modalities and the behavior
As a laboratory example, if the experimenter pres- associated with food intake. Taste aversions are
ents food to a pigeon following a localized visual generally affected by the same variables as other
stimulus, the pigeon will begin to peck the visual CRs, although conditioning may take place over
stimulus (Brown & Jenkins, 1968). This proce- longer intervals between the nominal CS and the
dure, known as autoshaping, meets the definition US, because the UR (e.g., nausea) is delayed (LoL-
of a classical procedure. Pecking, which was ini- ordo & Droungas, 1989). Conditioned aversions to
tially elicited by the sight of food, is now directed food eaten before chemotherapy often occur be-
to a stimulus—the localized light—that reliably cause of the nausea-inducing effects of the treat-
precedes the food. However, if the experimenter ment. Appropriate conditioning regimens, such as
pairs food with a stimulus that is not spatially lo- pairing the treatment with a nonpreferred food,
calized, such as a sound, pecking is not observed, reduce aversions to other foods. In this way, taste
although other measures indicate that condition- aversions develop toward the nonpreferred food
ing has occurred (Leyland & Mackintosh, 1978). and not the food normally eaten (Bernstein, 1991;
The expression of the CR depends in part on the Wang, Lee, He, & Huang, 2017).
CS with which the US is paired. Some instances
of this phenomenon—called differential associabil-
Characteristics of the US/UR
ity—arise from the history of the individual. For
example, presenting the textual stimulus DON’T The stimuli used as USs vary almost as widely as
BLINK to a human as a CS before a puff of air those used as CSs. Generally, we may divide USs
to the eye impairs conditioning of the eye blink into two classes: those that are appetitive, or stimuli
relative to a neutral stimulus, such as the presenta- that elicit approach behavior; and those that are
tion of a geometric form. Conversely, condition- aversive, or stimuli that elicit escape behavior. Ap-
ing is facilitated if the CS is BLINK (Grant, 1972). petitive USs, such as food or water, evoke a range of
Research has also shown interactions between behaviors, including approaching the stimulus and
the CS and US in the conditioning of phobias. a variety of consummatory responses, when they
Stimuli that are often the objects of phobias, such are presented to an appropriately deprived animal.
as spiders, become CSs more rapidly when they Similarly, aversive stimuli elicit a range of behavior,
are paired with an aversive US, such as a moder- including escaping or attacking and freezing when
ate electric shock (Ohman, Fredrikson, Hugdahl, the organism cannot escape the stimulus. The CRs
& Rimmo, 1976; Lindström, Golkar, & Olsson, conditioned to environmental stimuli can either
2015). When we examine the histories of persons facilitate or interfere with operants when the rein-
with phobic behavior, they often contain experi- forcers occur in an operant procedure. To interpret
Respondent (Pavlovian) Conditioning 27
possible interactions of respondents with operants, rabbits. During acquisition, every presentation of
we should remember that the total CR is not nec- the CS was followed by the US/UR. The three
essarily restricted to the CRs that we measure. USs groups of animals then received different percent-
generally elicit many URs, some of which are de- ages of CS–US/UR pairings. One group received
tected less easily at the behavioral level of measure- reinforcers after 100% of CS presentations, and
ment, such as heart rate changes mediated by the responding was maintained at the same high level
autonomic nervous system. as during acquisition. The remaining two groups
received a gradually reduced percentage of rein-
forcement. The US/UR ultimately followed the
SOME PHENOMENA ASSOCIATED CS on only 50% of the trials in one group and
WITH THE PAVLOVIAN (RESPONDENT) only 25% of the trials in the other group. As the
CONDITIONING PROCEDURE middle panel of Figure 2.6 shows, performance was
relatively unchanged, even though the percentage
Thus far, we have been concerned with the acqui- of reinforced CSs was very substantially reduced.
sition of environment–behavior relations via the The procedure is called continuous reinforcement
Pavlovian procedure, and with the process of re- when the US/UR follows every CS presentation,
inforcement that produces acquisition. In this sec- and intermittent (or partial) reinforcement when
tion, we examine several phenomena encountered the US/UR follows only some CSs. In these terms,
during the acquisition of CS–CR relations. efficient acquisition of CRs requires continuous re-
inforcement, but the gradual introduction of inter-
mittent reinforcement can maintain responding.
Maintenance of Conditioning
The acquisition of conditioning proceeds most
Stimulus Generalization
rapidly when every presentation of the CS is fol-
lowed by a reinforcer—whether an unconditioned The stimulus that reliably precedes the reinforcer
or conditioned reinforcer. However, behavior can during acquisition is the CS. However, the CS is
be maintained at high levels with less frequent re- not the only stimulus whose control of the CR is
inforcement once it has been acquired. The left affected by conditioning. Other stimuli that share
panel of Figure 2.6 shows the acquisition of CRs properties in common with the CS also come to
with the nictitating membrane preparation in evoke the CR, although with less strength. For
FIGURE 2.6. Acquisition, maintenance, and extinction of a Pavlovian nictitating membrane CR in the rabbit. During
acquisition, 100% of the CSs were followed by the US. During maintenance, different groups of animals received
either 100%, 50%, or 25% CS–US pairings. During extinction, CS presentations were not followed by the US. From
Donahoe and Palmer (1994/2005), based on findings from Gibbs, Latham, and Gormezano (1978). Copyright © 2005
John W. Donahoe. Reprinted by permission.
28 B a s i c P r in c ip l e s a nd C o n c e p t s
example, if the CS is a tone with a frequency of lessened the value of the food US. An interpreta-
1,000 Hertz (Hz), then tones of 800 Hz are likely tion of this finding is that pairing food with shock
to evoke CRs, although to a lesser degree. Similar- changed the interoceptive stimuli that the CR
ly, tones of 600 Hz may also evoke CRs, but to an and the CS jointly controlled during condition-
even lesser degree. Other stimuli acquire the abil- ing, and that this change weakened food-related
ity to evoke CRs in proportion to their physical CRs. Clearly, a complex array of stimuli, including
similarity to the training CS. Many classical pro- the effects of stimulus generalization and control
cedures with human and nonhuman participants by contextual and interoceptive stimuli, may af-
have documented this phenomenon, known as fect the CR.
stimulus generalization (e.g., Gynther, 1957; Hupka,
Liu, & Moore, 1969). The experimental analysis
Extinction
of neuroscience is consistent with the behavioral
analysis. Responding to a generalization stimu- After a CS has acquired control of a CR, present-
lus occurs to the extent that the generalization ing the CS but omitting the US weakens control,
stimulus activates the same sensory neurons as the which is an extinction procedure. The right panel
training stimulus (Thompson, 1965). of Figure 2.6 shows the effect of an extinction pro-
A second source of stimulus generalization aris- cedure on a CR. The percentage of CS presenta-
es from the other stimuli that accompany the CS. tions that evoked a CR decreased more slowly after
These stimuli define the stimulus context. The intermittent reinforcement than after continuous
stimulus context seldom evokes the CR by itself, reinforcement. The responding of animals that re-
because the more reliably present CS blocks con- ceived 100% reinforcement declined most rapidly,
trol by contextual stimuli. However, the CS and followed by animals receiving 50% and then 25%
the contextual stimuli furnish the full stimulus reinforcement.
compound with which the US/UR is paired, and
the context can also affect responding (Burns,
Punishment
Burgos, & Donahoe, 2011; Donahoe et al., 1997).
Contextual stimuli are sometimes said to function Punishment is a term that, strictly speaking, applies
as occasion setters (Grahame, Hallam, & Geier, only to the operant procedure. In a punishment
1990). procedure, the operant response produces a stimu-
lus that decreases the strength of the operant. As
a laboratory example, food-reinforced lever press-
Control by CR‑Related Interoceptive Stimuli
ing can be punished by the occasional presenta-
As conditioning proceeds, the CR begins to occur tion of a moderate electric shock. Food-reinforced
during the CS before the presentation of the US/ lever pressing declines under this procedure, and
UR (Shapiro, 1962). Thus CR-produced stimuli we say that shock functions as a punisher. By con-
may begin to appear before acquisition is complete. trast, conditioning with the Pavlovian procedure
As a result, these interoceptive events bear a tem- always produces an increase in responding, which
poral relation to the behavioral discrepancy that is an increase in the behavior that the US elicits.
permits them also to acquire control of the CR. Although punishment occurs only in operant pro-
In an illustrative study, the experimenter paired cedures, conditioning in the classical procedure is
an appetitive US, food, with an aversive stimulus, relevant because CRs contribute to punishment.
a moderate electric shock, after the experimenter Specifically, participants acquire CRs and oper-
had previously paired the CS with food alone. ants together in the operant procedure. These
(Pairing eliciting stimuli that evoke competing CRs can decrease the operant if the operant and
URs is called a counterconditioning procedure; cf. the CR are incompatible (Donahoe & Palmer,
Richardson & Donahoe, 1967.) Food-related CRs 1994/2005). In the preceding example, food condi-
were weakened when the experimenter presented tions lever pressing, whereas shock conditions es-
the CS after the experimenter had paired the food cape from the lever as well as autonomic responses
with shock (Colwill & Rescorla, 1985; Donahoe (Borgealt, Donahoe, & Weinstein, 1972). Because
& Burgos, 2000; Holland & Rescorla, 1975). Note the organism cannot press the lever while simul-
that food-related CRs were weakened even though taneously escaping from the region with the lever,
the experimenter had never paired the CS itself lever pressing declines. The recovery of lever press-
with shock. This phenomenon is known as revalu- ing from punishment depends on the extinction of
ation, so called because pairing food with shock escape responses (Estes & Skinner, 1941).
Respondent (Pavlovian) Conditioning 29
We can understand certain paradoxical effects criminated operant procedures), and the behavior
of punishment procedures as the product of in- is the operant plus the CR. The remainder of this
teractions between operants and respondents. In chapter describes some implications of the rein-
one line of research, monkeys restrained in a chair forcement process for understanding dysfunctional
were first trained to bite a rubber hose for food. behaviors that reflect both Pavlovian and operant
This was an operant task, with biting as the oper- contingencies.
ant and food as the reinforcer. The experimenter
then altered the procedure so that biting the hose
What Is the Role of Conditioned Reinforcement
continued to produce food, but also an occasional
in Dysfunctional Behavior?
electric shock to the tail. Electric shock to the tail
of a monkey elicited hose biting. Biting is a com- Conditioned reinforcement plays a critical role in
ponent of aggressive behavior that is often elicited the acquisition and maintenance of temporally
by aversive stimuli. Instead of reducing the rate of extended sequences of behavior. Skinner (1938)
hose biting, the addition of shock increased it, par- illustrated this with rats whose bar presses imme-
ticularly at the times when shock was most likely. diately produced a clicking sound that had been
In some cases, the experimenter could eliminate previously paired with food, thereby bridging the
food altogether, and the monkey would continue time before food was actually ingested. The ef-
to bite the hose, the only consequence of which fect of conditioned reinforcement has also been
was now the occasional delivery of shock (Branch illustrated with the example of children whose at-
& Dworkin, 1981; Morse & Kelleher, 1977). This tempts to write their letters produce stimuli that
masochistic behavior is understandable, at least in immediately approximate the appearance of letters
part, as a case in which the operant that produced they have previously learned to speak. The click
food and the respondent evoked by shock were produced by bar pressing and the visual stimuli
similar—biting. produced by writing provide immediate condi-
tioned reinforcers for the behavior. Conditioned
reinforcement is mediated by the neural pathways
SOME IMPLICATIONS OF RESPONDENT between the prefrontal cortex and the nucleus ac-
CONDITIONING FOR APPLIED BEHAVIOR ANALYSIS cumbens and from there to the ventral tegmental
area. (See Figure 2.4.) In this section, we consider
Most human behavior involves operant as well as some of the possible effects on conditioned rein-
respondent contingencies—that is, contiguity of a forcement produced by decreases in the function-
response with a reinforcer, and also contiguity of a ality of connections from the prefrontal cortex
stimulus with a reinforcer. As a result, techniques to the nucleus accumbens. A stimulus acquires
that researchers use to modify dysfunctional be- its ability to function as a conditioned reinforcer
havior implement both operant and respondent through the stimulus–reinforcer relations manipu-
procedures. An understanding of the conditioning lated in Pavlovian procedures.
process by the Pavlovian procedure is important
for two principal reasons. First, operant contingen-
Attention‑Deficit/Hyperactivity Disorder
cies necessarily include stimulus–reinforcer conti-
guities. Some environmental stimulus always pre- One of the behavioral characteristics of attention-
cedes the reinforcing stimulus (or US; see Figure deficit/hyperactivity disorder (ADHD) is difficulty
2.1). Thus reinforcer-evoked responses (CRs) are sustaining activity on a task in which the rein-
acquired inevitably in operant procedures. Sec- forcer is delayed. The conditioned reinforcers that
ond, current accounts of operant and Pavlovian ordinarily maintain temporally extended behavior
procedures indicate that both procedures engage are relatively ineffective, which allows other con-
the same fundamental conditioning process. That currently available stimuli to control behavior that
is, whatever stimuli precede the behavioral dis- competes with the task at hand. In the vernacular,
crepancy acquire control over whatever responses a person with ADHD is easily distracted. Because
accompany the discrepancy. In the Pavlovian pro- conditioned reinforcers play an important role in
cedure, these stimuli are the CS and the context maintaining temporally extended environment–
in which the CS occurs, and the behavior is the behavior relations, a deficit in conditioned rein-
CR (generally components of the UR). In the op- forcement is a likely contributor to ADHD. In sup-
erant procedure, the stimuli are those that precede port of this conjecture, studies have found that the
the discrepancy (discriminative stimuli in dis- introduction of a conditioned reinforcer, especially
30 B a s i c P r in c ip l e s a nd C o n c e p t s
if it occurs immediately after the behavior, facili- ing properties for persons with ASD immediately
tates choice of a larger but delayed unconditioned followed a correct response. The effective images
reinforcer (Williams, 1994c). Gradually increasing were mostly of machines, automobiles, computers,
the time between the response and the reinforcer and other nonsocial objects that the researchers
increases the choice of delayed reinforcement by used because they evoked eye movement fixations
children (Schweitzer & Sulzer-Azaroff, 1988). A from participants with ASD during a pretest. In
gradual introduction of the delay permits stimuli summary, the $ served as an effective conditioned
in the delay period to become conditioned rein- reinforcer for controls but not for participants with
forcers. Studies of the neural basis of ADHD are ASD, as measured by choice behavior and by the
consistent with the hypothesis that a deficit in the activation level of neurons in the nucleus accum-
neural mechanisms of conditioned reinforcement bens. No deficits were observed in the activation
is involved (Donahoe & Burgos, 2005; Johansen of prefrontal neurons by their multisynaptic inputs
et al., 2009). from sensory areas. Thus the nucleus accumbens
neurons would have been activated if the normal
prefrontal connections to the nucleus accumbens
Autism Spectrum Disorder
had been present. Possible causes of such neuro-
Autism spectrum disorder (ASD; see Kodak, Grow, developmental deficits in ASD are under current
& Bergmann, Chapter 23, this volume) is a focus investigation (e.g., Choi et al., 2016; Donahoe,
of applied behavior analysis because its methods 2018).
are the only empirically supported procedures that A deficit in the responsiveness of nucleus ac-
potentially remediate this often debilitating range cumbens neurons to neurons in the prefrontal
of behavioral dysfunctions. Experimental evidence cortex provides a possible mechanism whereby we
suggests that deficits in conditioned reinforcement may understand the variability of ASD symptoms.
and its neural mechanisms play an important role If the number and origins of connections from
in ASD (Donahoe, 2018). The behavioral phe- regions in the prefrontal neurons to nucleus ac-
nomena seen in ASD may include repetitive be- cumbens neurons differ for those with ASD, then
haviors such as hand flapping, verbal deficits such the potential for the stimuli that activate these
as delayed or absent speech, and social challenges neurons to serve as conditioned reinforcers will
such as lack of eye contact when interacting with differ. For example, suppose that during neural de-
others. How might malfunctioning of conditioned velopment connections are missing from some of
reinforcement provide insights into this diverse set the neurons in the premotor cortex to neurons in
of behavioral deficits? the nucleus accumbens. Instead, projections form
To study the effect of conditioned reinforcers from neurons in adjacent premotor areas that gov-
on the behavior of persons with ASD, the experi- ern arm and hand movements to the nucleus ac-
menter placed participants in an apparatus that cumbens. Under these circumstances, hand flap-
monitored the activation levels of various brain ping would activate the ventral tegmental area via
regions during a choice task (Dichter et al., 2012; the nucleus accumbens and would be reinforced.
Knutson, Adams, Fong, & Hammer, 2001). When As another example, suppose that projections
neurons are activated, their demand for oxygen from the region of the prefrontal cortex activated
increases, which increases the blood flow in that by speech movements that normally go to the nu-
region. Functional magnetic resonance imaging cleus accumbens are absent or diminished. In this
detects the increase in blood flow. In the experi- scenario, the articulatory movements that produce
ment, a correct choice was immediately followed by speech sounds could not benefit from the auto-
either a dollar sign ($) or a brief view of a picture. matic conditioned reinforcement that normally
The $ indicated that the participant would receive aids the acquisition of verbal behavior. Deficits in
$1.00 at the end of the experimental session, and verbal behavior would result. Finally, suppose that
was intended to serve as a conditioned reinforcer. neurons in the region of the prefrontal cortex that
For participants with ASD, the $ activated neu- receive inputs from the sensory areas involved in
rons in the nucleus accumbens much less than for face perception do not have connections to the
control participants, where it served as an effec- nucleus accumbens. Under these circumstances,
tive conditioned reinforcer. By contrast, activity seeing a human face could not serve as a source
in the nucleus accumbens did not differ between of conditioned reinforcement. Instead, individu-
participants with ASD and controls when a pic- als might perceive direct eye contact with another
ture of an object with known attention-demand- person as a threat gesture, which is the typical re-
Respondent (Pavlovian) Conditioning 31
action in other primates (Emery, Lorincz, Perrett, natural environment that reliably occurred before
Oram, & Baker, 1997). reinforcers in the learner’s past, we can identify
If alterations in the usual connectivity of pre- these stimuli by noting the situations in which the
frontal regions to the nucleus accumbens occur behavior now habitually occurs. We must identify
with ASD, then the neural mechanisms that these stimuli if the behavioral changes produced
mediate conditioned reinforcement would be af- by a therapeutic environment are to persist. Three
fected. Experimental analysis suggests some po- guidelines are useful:
tentially remedial actions if this is the case. For
individuals with especially severe deficits, we 1. To the extent possible, the remedial environ-
could use stimuli that function as unconditioned ment should include some stimuli that control
reinforcers because they access the dopaminergic the target behavior in the natural environ-
reinforcement system directly through the ven- ment. In this way, we maximize stimulus gen-
tral tegmental area. These stimuli include food eralization from the remedial to the natural
and drink (Lovaas, 1987). Other stimuli that are environment (Stokes & Baer, 1977). We must
attention-demanding, such as certain visual dis- identify the conditions in the natural environ-
plays, could also be effective. Baseline preferences ment that precede dysfunctional behavior to
for various displays would need to be determined identify these controlling stimuli.
to identify stimuli that could function as condi- 2. We must introduce stimuli that control be-
tioned reinforcers (e.g., Mitchell & Stoffelmayr, havior in the remedial environment into the
1973; Premack, 1959). natural environment if those stimuli do not
The establishment of conditioned reinforcers occur in the natural environment. This applies
requires Pavlovian higher-order and compound- whether the intervention seeks to establish
conditioning procedures or operant conditioned- appropriate behavior or to establish behavior
reinforcement procedures. Research indicates that that competes with the dysfunctional behav-
if we pair two stimuli with a reinforcer, then some ior. Reinforcers do not select responses alone;
neurons in sensory association areas of the cortex they select environment–behavior relations
become activated equally by either stimulus. These (Donahoe et al., 1997).
neurons are called pair-coded neurons (Sakai & 3. We must supplement or remove the reinforce-
Miyashita, 1991). If one of the stimuli in the pair ment contingencies that maintain dysfunc-
becomes a conditioned reinforcer, then the other tional behavior in the natural environment
stimulus can also become a conditioned reinforcer, with alternative contingencies that maintain
even though it might not become a conditioned the behavior acquired in the remedial envi-
reinforcer if it were separately paired with the ronment. To do this, the conditions in the
unconditioned reinforcer. Ideally, the two stimuli natural environment that follow dysfunctional
become equivalent in their ability to engage the behavior and serve as reinforcers must be de-
conditioned-reinforcement system. termined. Such reinforcers must be removed,
Matching-to-sample is an operant procedure or competing behavior must be acquired that
that promotes the formation of pair-coded neurons. minimizes contact with them. No behavioral
Behavioral evidence demonstrates that matching- intervention can inoculate a person against the
to-sample procedures produce stimuli that form effect of reinforcers of dysfunctional behavior
equivalence classes for at least some participants encountered in the natural environment.
with ASD (McLay et al., 2013; cf. Sidman, 2000).
In this way, a stimulus that could not function as We can introduce controlling stimuli for alter-
a conditioned reinforcer when separately paired native behavior from the remedial environment
with a reinforcer may become a conditioned rein- into the natural environment in several ways.
forcer through its equivalence with another stimu- These stimuli may be explicitly introduced into
lus that has acquired that function. the natural environment. For example, we might
pair a red card with an aversive stimulus (US) in
the remedial environment as a first step to con-
What Stimuli Control Behavior
trol profligate spending. We then would insert the
in the Natural Environment?
card into the profligate spender’s wallet so that the
The stimuli that guide behavior are those that card is visible before money is accessible. We also
have reliably occurred prior to reinforcers. Al- can condition verbal responses to stimuli present
though we cannot know all of the stimuli in the in the natural environment, and these responses
32 B a s i c P r in c ip l e s a nd C o n c e p t s
may in turn generate verbal stimuli that control dysfunctional behavior in the natural environ-
alternative behavior. Continuing with the exam- ment.
ple of profligate spending, every time the wallet is To be effective, the extinction procedure should
opened, reinforcers are provided in the remedial be continued substantially beyond the time when
environment when the spender verbalizes, “Do I the environment first ceases to occasion the be-
really need to buy this?” havior. So-called “extinction below zero” increases
Verbal stimuli are potentially among the most the likelihood that all of the stimuli controlling
effective stimuli to control behavior in the natu- the behavior have lost their control (Welker &
ral environment, because verbal responses are not McAuley, 1978). If effective stimuli remain, they
dependent on external support in the same way foster resurgence of the maladaptive behavior, and
as most nonverbal responses. Verbal responses— it may be reinforced again (Epstein & Skinner,
and hence the stimuli they produce—are poten- 1980). The recurrence of behavior after extinction
tially in the behavioral repertoire of a person in is called spontaneous recovery (Estes, 1955; Pavlov,
any environment. A second advantage of verbal 1927/1960; Skinner, 1938). Again, the remedial
stimuli is that they can be produced by subvocal environment cannot inoculate behavior against
verbal behavior, and subvocal behavior is not sub- the effects of reinforcers for dysfunctional behav-
ject to contingencies of reinforcement instituted ior provided by the natural environment.
by others (Palmer, 2012). Others can ask why the Addiction provides a particularly striking ex-
red card is in the wallet, because they too can see ample of the recurrence of dysfunctional behavior.
the red card. But others cannot ask why a thought Research with the classical procedure has shown
occurred, because a thought is a subvocal verbal that CRs evoked early in the conditioning pro-
response. Subvocal behavior is private behav- cess give rise to stimuli that acquire control over
ior—that is, behavior whose stimulus properties the CR jointly with the CS. The phenomenon
are detectable only by the person emitting them of revaluation documents the existence of control
(Donahoe & Palmer, 1994/2005; Skinner, 1957). If by CR-related stimuli (Holland & Rescorla, 1975;
the goal is to maintain verbal vocal or subvocal re- Wyvell & Berridge, 2000). In the treatment of ad-
sponses, they too must be followed by reinforcers. diction, we may eliminate physical dependence by
Private behavior, sometimes called cognitive behav- withholding the substance in the remedial envi-
ior, is not immune to the conditioning processes ronment. However, drug-related CRs will recur to
that affect all behavior. the extent that the remedial environment differs
from the environment in which the person ac-
quired the addiction. Moreover, drug-related oper-
What Behavior Is Maintained
ant behavior will recur if it is controlled by intero-
by the Natural Environment?
ceptive stimuli from drug-related CRs. To reduce
As we have seen, behavior is maintained in an resurgence of drug-related CRs and the untoward
environment to the extent that the environment effects of the stimuli they produce, the remedial
contains stimuli in whose presence the behavior environment must gradually introduce stimuli
has been previously reinforced. In the absence of that are CSs for these CRs, possibly including the
reinforcement, an extinction procedure is imple- sight of drug paraphernalia, and withhold rein-
mented, and responding decreases. Intermittent forcement in their presence.
reinforcement during conditioning increases resis- Environment–behavior relations that we select
tance to the effects of extinction, but responding in the remedial environment will endure if the re-
will not continue indefinitely. Thus the natural inforcers that previously maintained the dysfunc-
environment must contain reinforcers for the be- tional behavior are no longer encountered and, in
havior established in the remedial environment. addition, newly established immediate reinforcers
If dysfunctional environment–behavior relations are available for alternative behavior. Eliminating
continue to be reinforced in the natural environ- previously encountered reinforcers requires chang-
ment, then the dysfunctional behavior will recur ing the natural environment—often a daunting
and persist, and more immediate reinforcers will task—or establishing behavior in the remedial
maintain it even if its long-term consequences are environment that reduces contact of the dysfunc-
maladaptive. Behavior that has undergone extinc- tional behavior with these reinforcers. For some-
tion in the remedial environment will reappear in one with an alcohol addiction, a simplistic example
the natural environment if the remedial environ- of the latter would be taking a route that does not
ment does not contain all the stimuli that control pass the local pub and being greeted by an adoring
Respondent (Pavlovian) Conditioning 33
partner upon arrival at home. Important sources Bullock, C. E., & Hackenberg, T. D. (2014). The several
of immediate reinforcement for behavior that was roles of stimuli in token reinforcement. Journal of the
established in a remedial environment are the Experimental Analysis of Behavior, 99, 1–19.
stimuli such behavior produces. For example, be- Burns, R., Burgos, J. E., & Donahoe, J. W. (2011). Pav-
lovian conditioning: Pigeon nictitating membrane.
havior such as fluently reading or facilely writing
Behavioural Processes, 86, 102–108.
a passage produces stimuli that are discriminated Catania, A. C. (1975). The myth of self-reinforcement.
as characteristic of a job well done. The stimuli Behaviorism, 3, 192–199.
produced by such behavior have occurred previ- Catania, A. C., & Keller, K. J. (1981). Contingency, con-
ously in the remedial environment and have been tiguity, correlation, and the concept of causality. In P.
the occasion for praise (a reinforcer) from oth- Harzem & M. D. Zeiler (Eds.), Predictability, correla-
ers. Because these stimuli have been paired with tion, and contiguity (pp. 125–167). New York: Wiley.
praise, they become CSs and can function as con- Choi, G. B., Yim, Y. S., Wong, H., Kim, S., Kim, H., Kim,
ditioned reinforcers (Catania, 1975). However, we S. V., et al. (2016). The maternal interleukin-17a
must continue to pair them with other reinforcers pathway in mice promotes autism-like phenotypes in
offspring. Science, 351, 933–937.
to maintain their status as conditioned reinforc-
Colwill, R. M., & Rescorla, R. A. (1985). Postcondition-
ers. Being literate may enhance one’s ability to ing devaluation of a reinforcer affects instrumental
get a job, but the environment must provide jobs responding. Journal of Experimental Psychology: Ani-
if such stimuli are to endure as conditioned rein- mal Behavior Processes, 11, 120–132.
forcers. Environment–behavior relations track the Dichter, G. S., Felder, J. N., Green, S. R., Ritenberg, A.
momentary contingencies of reinforcement, not M., Sasson, N. J., & Bodfish, J. W. (2012). Reward
remote consequences. In the long run, remedial circuitry function in autism spectrum disorders. So-
interventions cannot overcome the contingencies cial Cognitive and Affective Neuroscience, 7, 160–172.
repeatedly encountered in the natural environ- Dinsmoor, J. A. (1950). A quantitative comparison of
the discriminative and reinforcing function of a stim-
ment. To be otherwise would contradict all that
ulus. Journal of Experimental Psychology, 40, 458–472.
we know from the experimental analysis of re- Dinsmoor, J. A. (1995). Stimulus control: Part I. Behav-
spondent and operant procedures—efforts begun ior Analyst, 18, 51–68.
by Pavlov and Thorndike over 100 years ago. Donahoe, J. W. (1997a). Positive reinforcement: The se-
lection of behavior. In J. R. O’Donohue (Ed.), Learn-
ing and behavior therapy (pp. 169–187). Boston: Allyn
REFERENCES & Bacon.
Donahoe, J. W. (1997b). Selection networks: Simula-
Acierno, R. E., Hersen, M., & Van Hasselt, V. B. (1993). tion of plasticity through reinforcement learning.
Interventions for panic disorder: A critical review of In J. W. Donahoe & V. P. Dorsel (Eds.), Neural-
the literature. Clinical Psychology Review, 6, 561–578. network models of cognition: Biobehavioral foundations
Benedict, J. O., & Ayres, J. J. (1972). Factors affecting (pp. 336–357). Amsterdam: Elsevier Science.
conditioning in the truly random control procedure Donahoe, J. W. (1999). Edward L. Thorndike: The se-
in the rat. Journal of Comparative and Physiological lectionist connectionist. Journal of the Experimental
Psychology, 78, 323–330. Analysis of Behavior, 72, 451–454.
Bernstein, I. L. (1991). Aversion conditioning in re- Donahoe, J. W. (2017). Behavior analysis and neurosci-
sponse to cancer and cancer treatment. Clinical Psy- ence: Complementary disciplines. Journal of the Ex-
chology Review, 2, 185–191. perimental Analysis of Behavior, 107, 301–320.
Borgealt, A. J., Donahoe, J. W., & Weinstein, A. (1972). Donahoe, J. W. (2018). Autism spectrum disorder: Biobe-
Effects of delayed and trace components of a com- havioral mechanisms and remediation procedures. Pre-
pound CS on conditioned suppression and heart rate. sentation at the Berkshire Association for Behavior
Psychonomic Science, 26, 13–15. Analysis and Therapy, Amherst, MA. Retrieved
Bouton, M. E., Mineka, S., & Barlow, D. H. (2001). A from http://lcb-online.org/Autism-Spectrum_Disorder_
modern learning theory perspective on the etiology Biobehavioral_Mechanisms.wmv.
of panic disorder. Psychological Review, 108, 4–32. Donahoe, J. W., & Burgos, J. E. (2000). Behavior analy-
Branch, M. N., & Dworkin, S. I. (1981). Effects of ratio sis and revaluation. Journal of the Experimental Analy-
contingencies on responding maintained by sched- sis of Behavior, 74, 331–346.
ules of electric-shock presentation (response-pro- Donahoe, J. W., & Burgos, J. E. (2005). Selectionism:
duced shock). Journal of the Experimental Analysis of Complex outcomes from simple processes. Behav-
Behavior, 36, 191–205. ioural and Brain Sciences, 28, 429–430.
Brown, P. L., & Jenkins, H. M. (1968). Auto-shaping Donahoe, J. W., Burgos, J. E., & Palmer, D. C. (1993).
of the pigeon’s key-peck. Journal of the Experimental Selectionist approach to reinforcement. Journal of the
Analysis of Behavior, 11, 1–8. Experimental Analysis of Behavior, 60, 17–40.
34 B a s i c P r in c ip l e s a nd C o n c e p t s
Donahoe, J. W., Crowley, M. A., Millard, W. J., & Stick- Grant, D. A. (1972). A preliminary model for processing
ney, K. A. (1982). A unified principle of reinforce- information conveyed by verbal conditioned stimuli
ment. In M. L. Commons, R. J. Herrnstein, & H. in classical conditioning. In A. H. Black & W. F.
Rachlin (Eds.), Quantitative analyses of behavior: Vol. Prokasy (Eds.), Classical conditioning: II. Current re-
2. Matching and maximizing accounts (pp. 493–521). search and theory (pp. 64–99). Englewood Cliffs, NJ:
Cambridge, MA: Ballinger. Prentice-Hall.
Donahoe, J. W., & Palmer, D. C. (2005). Learning and Gynther, M. D. (1957). Differential eyelid conditioning
complex behavior. Richmond, MA: Ledgetop. (Origi- as a function of stimulus similarity and strength of
nal work published 1994) response to the CS. Journal of Experimental Psychol-
Donahoe, J. W., Palmer, D. C., & Burgos, J. E. (1997). ogy, 53, 408–416.
The S-R issue: Its status in behavior analysis and in Hilgard, E. R. (1937). The relationship between the
Donahoe and Palmer’s Learning and complex behavior. conditioned response and conventional learning ex-
Journal of the Experimental Analysis of Behavior, 67, periments. Psychological Bulletin, 34, 61–102.
193–211. Holland, P. C., & Rescorla, R. A. (1975). The effect of
Donahoe, J. W., & Vegas, R. (2004). Pavlovian condi- two ways of devaluing the unconditioned stimulus
tioning: The CS–UR relation. Journal of Experimen- after first- and second-order appetitive conditioning.
tal Psychology: Animal Behavior Processes, 30, 17–33. Journal of Experimental Psychology: Animal Behavior
Dworkin, B. R. (1993). Learning and physiological regula- Processes, 1, 355–363.
tion. Chicago: University of Chicago Press. Hupka, R. B., Liu, S. S., & Moore, J. W. (1969). Audito-
Eikelboom, R., & Stewart, J. (1982). Conditioning of ry differential conditioning of the rabbit nictitating
drug-induced physiological responses. Psychological membrane response: V. Stimulus generalization as a
Review, 89, 507–528. function of the position of CS+ and CS– on the fre-
Emery, N. J., Lorincz, E. N., Perrett, D. I., Oram, M. W., quency dimension. Psychonomic Science, 15, 129–131.
& Baker, C. I. (1997). Gaze following and joint atten- Johansen, E. B., Killeen, P. R., Russell, V. A., Tripp, G.,
tion in rhesus monkeys (Macaca mulatta). Journal of Wickens, J. R., Tannock, R., et al. (2009). Origins
Comparative Psychology, 111, 286–293. of altered reinforcement effects in ADHD. Behavioral
Epstein, R., & Skinner, B. F. (1980). Resurgence of re- and Brain Functions, 5, 1–15.
sponding after the cessation of response-independent Johnson, D. F., & Cumming, W. W. (1968). Some de-
reinforcement. Proceedings of the National Academy of terminers of attention. Journal of the Experimental
Sciences of the USA, 77, 6251–6253. Analysis of Behavior, 11, 157–166.
Estes, W. K. (1955). Statistical theory of spontaneous Jonçich, G. (1968). The sane positivist: A biography of
recovery and regression. Psychological Review, 62, Edward L. Thorndike. Middleton, CT: Wesleyan Uni-
145–154. versity Press.
Estes, W. K., & Skinner, B. F. (1941). Some quantitative Kamin, L. J. (1968). Attention-like processes in classical
properties of anxiety. Journal of Experimental Psychol- conditioning. In M. R. Jones (Ed.), Miami symposium
ogy, 29, 390–400. on the prediction of behavior (pp. 9–31). Miami, FL:
Fam, J., Westbrook, R. F., & Holmes, N. M. (2017). University of Miami Press.
An examination of changes in behavioral control Kamin, L. J. (1969). Predictability, surprise, atten-
when stimuli with different associative histories are tion and conditioning. In B. A. Campbell & R. M.
conditioned in compound. Journal of Experimental Church (Eds.), Punishment and aversive behavior
Psychology: Animal Learning and Cognition, 43(3), (pp. 279–296). New York: Appleton-Century-Crofts.
205–218. Kehoe, E. J., Gibbs, C. M., Garcia, E., & Gormezano,
Garcia, J., Erwin, F. R., & Koelling, R. A. (1966). Learn- I. (1979). Associative transfer and stimulus selection
ing with prolonged delay in reinforcement. Psycho- in classical conditioning of the rabbit’s nictitating
nomic Science, 5, 121–122. membrane response to serial compound CSs. Journal
Gibbs, C. M., Latham, S. B., & Gormezano, I. (1978). of Experimental Psychology: Animal Behavior Process-
Classical conditioning of the rabbit nictitating mem- es, 5, 1–18.
brane response: Effects of reinforcement schedule on Kehoe, E. J., & Macrae, M. (1994). Classical condition-
response maintenance and resistance to extinction. ing of the rabbit nictitating membrane response can
Animal Learning and Behavior, 6, 209–215. be fast or slow: Implications of Lennartz and Wein-
Gormezano, I., & Kehoe, E. J. (1981). Classical condi- berger’s (1992) two-factor theory. Psychobiology, 22,
tioning and the law of contiguity. In P. Harzem & M. 1–4.
D. Zeiler (Eds.), Predictability, correlation, and conti- Keller, F. S., & Schoenfeld, W. N. (1950). Principles of
guity (pp. 1–45). New York: Wiley. psychology. New York: Appleton-Century-Crofts.
Grahame, N. J., Hallam, S. C., & Geier, L. (1990). Con- Kirkpatrick, K., & Balsam, P. D. (2016). Associative
text as an occasion setter following either CS acquisi- learning and timing. Current Opinion in Behavioral
tion and extinction or CS acquisition alone. Learning Sciences, 8, 181–185.
and Motivation, 21, 237–265. Knutson, B., Adams, C. M., Fong, G. W., & Hammer, D.
Respondent (Pavlovian) Conditioning 35
G. (2001). Anticipation of increasing monetary re- Premack, D. (1959). Toward empirical behavioral laws:
ward selectively recruits nucleus accumbens. Journal I. Positive reinforcement. Psychological Review, 66,
of Neuroscience, 21(16), RC159. 219–233.
Leyland, C. M., & Mackintosh, N. J. (1978). Blocking of Rescorla, R. A. (1967). Pavlovian conditioning and its
first- and second-order autoshaping in pigeons. Ani- proper control group. Psychological Review, 74, 71–80.
mal Learning and Behavior, 6, 392–394. Rescorla, R. A. (1980). Pavlovian second-order condi-
Lindström, B., Golkar, A., & Olsson, A. (2015). A clash tioning: Studies in associative learning. Hillsdale, NJ:
of values: Fear-relevant stimuli can enhance or cor- Erlbaum.
rupt adaptive behavior through competition between Rescorla, R. A. (1991). Associative relations in instru-
Pavlovian and instrumental valuation systems. Emo- mental learning: The 18th Bartlett Memorial Lec-
tion, 15, 668–676. ture. Quarterly Journal of Experimental Psychology,
LoLordo, V. M., & Droungas, A. (1989). Selective asso- 43B, 1–23.
ciations and adaptive specializations: Taste aversions Rescorla, R. A., & Wagner, A. R. (1972). A theory of
and phobias. In S. B. Klein & R. R. Mowrer (Eds.), Pavlovian conditioning: Variations in the effective-
Contemporary learning theories: Instrumental condi- ness of reinforcement and nonreinforcement. In A.
tioning theory and the impact of biological constraints H. Black & W. F. Prokasy (Eds.), Classical condition-
on learning (pp. 145–179). Hillsdale, NJ: Erlbaum. ing: II. Current research and theory (pp. 64–99). New
Lovaas, I. (1987). Behavioral treatment and normal edu- York: Appleton–Century–Crofts.
cational and intellectual functioning in young chil- Richardson, W. K., & Donahoe, J. W. (1967). A test of
dren with autism. Journal of Consulting and Clinical the independence of the approach and avoidance
Psychology, 55, 3–9. gradients. Psychonomic Science, 9, 569–570.
McLay, L. K., Sutherland, D., Church, J., & Tyler-Mer- Russell, D., Ingvarsson, E. T., Haggar, J. L., & Jessel, J.
rick, G. (2013). The formation of equivalence classes (2018). Using progressive-ratio schedules to evaluate
in individuals with autism spectrum disorder: A re- tokens as generalized conditioned reinforcers. Journal
view of the literature. Research in Autism Spectrum of Applied Behavior Analysis, 51, 40–52.
Disorders, 7, 418–431. Sakai, K., & Miyashita, Y. (1991). Neural organization
Merckelbach, H., & Muris, P. (1997). The etiology of for the long-term memory of paired associates. Na-
childhood spider phobia. Behaviour Research and ture, 254, 152–159.
Therapy, 35, 1031–1034. Schneiderman, N. (1972). Response system divergencies
Mitchell, W. S., & Stoffelmayr, B. E. (1973). Applica- in aversive classical conditioning. In A. H. Black &
tion of the Premack principle to the behavioral con- W. F. Prokasy (Eds.), Classical conditioning: II. Cur-
trol of extremely inactive schizrophrenics. Journal of rent theory and research (pp. 313–376). New York:
Applied Behavior Analysis, 6, 419–423. Appleton-Century-Crofts.
Mizunami, M., Terao, K., & Alvarez, B. (2018). Ap- Schultz, W. (1997). Adaptive dopaminergic neurons
plication of a prediction error theory to Pavlovian report value of environmental stimuli. In J. W. Do-
conditioning in an insect. Frontiers in Psychology, 9, nahoe & V. P. Dorsel (Eds.), Neural-network models
1272. of cognition: Biobehavioral foundations (pp. 317–335).
Morse, W. H., & Kelleher, R. T. (1977). Determinants Amsterdam: Elsevier Science.
of reinforcement and punishment. In W. K. Honig & Schultz, W. (2001). Reward signaling by dopamine neu-
J. E. R. Staddon (Eds.), Handbook of operant behavior rons. Neuroscientist, 7, 293–302.
(pp. 174–200). Englewood Cliffs, NJ: Prentice-Hall. Schweitzer, J. B., & Sulzer-Azaroff, B. (1988). Self-con-
Ohman, A., Fredrikson, M., Hugdahl, K., & Rimmo, P. trol: Teaching tolerance for delay in impulsive chil-
A. (1976). The premise of equipotentiality in human dren. Journal of the Experimental Analysis of Behavior,
classical conditioning: Conditioned electrodermal 50, 173–186.
responses to potentially phobic stimuli. Journal of Ex- Seaver, J. P., & Bourret, J. C. (2020). Producing mands
perimental Psychology: General, 105, 313–337. in concurrent operant environments. Journal of Ap-
Palmer, D. C. (1987). The blocking of conditioned rein- plied Behavior Analysis, 53, 366–384.
forcement. Unpublished doctoral dissertation, Uni- Shahan, T. A. (2010). Conditioned reinforcement and
versity of Massachusetts, Amherst, MA. response strength. Journal of the Experimental Analy-
Palmer, D. C. (2012). The role of atomic repertoires sis of Behavior, 93, 269–289.
in the interpretation of complex behavior. Behavior Shapiro, M. M. (1962). Temporal relationship between
Analyst, 36, 59–73. salivation and lever pressing with differential rein-
Pavlov, I. P. (1960). Conditioned reflexes. New York: forcement of low rates. Journal of Comparative and
Dover. (Original work published 1927) Physiological Psychology, 55, 567–571.
Petursdottir, A. I., & Lepper, T. L. (2015). Inducing Sidman, M. (2000). Equivalence relations and the re-
novel vocalizations by conditioning speech sounds inforcement contingency. Journal of the Experimental
as reinforcers. Behavior Analysis in Practice, 8, 223– Analysis of Behavior, 74, 127–140.
232. Skinner, B. F. (1935). Two types of conditioned reflex
36 B a s i c P r in c ip l e s a nd C o n c e p t s
and a pseudo type. Journal of General Psychology, 12, Timberlake, W., & Lucas, G. A. (1985). The basis of
66–77. superstitious behavior: Chance contingency, stimu-
Skinner, B. F. (1937). Two types of conditioned reflex: A lus substitution, or appetitive behavior? Journal of the
reply to Konorski and Miller. Journal of General Psy- Experimental Analysis of Behavior, 44, 279–299.
chology, 16, 272–279. Van Willigen, F., Emmett, J., Cote, D., & Ayres, J. J. B.
Skinner, B. F. (1938). The behavior of organisms. New (1987). CS modality effects in one-trial backward
York: Appleton-Century-Crofts. and forward excitatory conditioning as assessed by
Skinner, B. F. (1948). “Superstition” in the pigeon. Jour- conditioned suppression of licking in rats. Animal
nal of Experimental Psychology, 38, 168–172. Learning and Behavior, 15, 201–211.
Skinner, B. F. (1957). Verbal behavior. New York: Apple- Vaughan, M. E., & Michael, J. (1982). Automatic rein-
ton-Century-Crofts. forcement: An important but ignored concept. Be-
Skinner, B. F. (1988). Comments and consequences. haviorism, 10, 217–227.
In A. C. Catania & S. Harnad (Eds.), The selection Vom Saal, W., & Jenkins, H. M. (1970). Blocking the
of behavior: The operant behaviorism of B. F. Skinner development of stimulus control. Learning and Moti-
(pp. 382–461). New York: Cambridge University vation, 1, 52–64.
Press. Waelti, P., Dickinson, A., & Schultz, W. (2001). Do-
Smith, M. C., Coleman, S. R., & Gormezano, I. (1969). pamine responses comply with basic assumptions of
Classical conditioning of the rabbit’s nictitating formal learning theory. Nature, 412, 43–48.
membrane response at backward, simultaneous, and Wang, Y., Lee, H., He, A. B., & Huang, A. C. W. (2017).
forward CS–US intervals. Journal of Comparative and Examinations of CS and US preexposure and postex-
Physiological Psychology, 69, 226–231. posure in conditioned taste aversion: Applications in
Sokolowska, M., Siegel, S., & Kim, J. A. (2002). Intra- behavioral interventions for chemotherapy anticipa-
administration associations: Conditional hyper- tory nausea and vomiting. Learning and Motivation,
algesia elicited by morphine onset cues. Journal of 59, 1–10.
Experimental Psychology: Animal Behavior Processes, Welker, R. L., & McAuley, K. (1978). Reductions in re-
28, 309–320. sistance to extinction and spontaneous recovery as a
Staddon, J. E., & Simmelhag, V. L. (1970). The “super- function of changes in transportational and contex-
stition” experiment: A reexamination of its implica- tual stimuli. Learning and Behavior, 6, 451–457.
tions for the principles of adaptive behavior. Psycho- Williams, B. A. (1994a). Conditioned reinforcement:
logical Review, 78, 3–43. Neglected or outmoded explanatory construct? Psy-
Stickney, K., & Donahoe, J. W. (1983). Attenuation of chonomic Bulletin and Review, 1, 457–475.
blocking by a change in US locus. Animal Learning Williams, B. A. (1994b). Conditioned reinforcement:
and Behavior, 11, 60–66. Experimental and theoretical issues. Journal of the
Stokes, T. F., & Baer, D. M. (1977). An implicit tech- Experimental Analysis of Behavior, 17, 261–285.
nology of generalization. Journal of Applied Behavior Williams, B. A. (1994c). Context specificity of condi-
Analysis, 10, 349–367. tioned-reinforcement effects on discrimination ac-
Sundberg, M. L., Michael, J., Partington, J. W., & Sun- quisition. Journal of the Experimental Analysis of Be-
dberg, C. A. (1996). The role of automatic reinforce- havior, 62, 157–167.
ment in early language acquisition. Analysis of Verbal Wyvell, C. L., & Berridge, K. C. (2000). Intra-accum-
Behavior, 13, 21–37. bens amphetamine increases the conditioned in-
Thompson, R. F. (1965). The neural basis of stimulus centive salience of sucrose reward: Enhancement
generalization. In D. I. Mostofsky (Ed.), Stimulus of reward “wanting” without enhanced “liking” or
generalization (pp. 154–178). Stanford, CA: Stanford response reinforcement. Journal of Neuroscience, 20,
University Press. 8122–8130.
Thorndike, E. L. (1903). Elements of psychology. New Yagishita, S., Hayashi-Takagi, A., Ellis-Davies, G. C.
York: Seiler. R., Urakubo, H., Ishii, S., & Kasai, H. (2014). A
Timberlake, W., & Allison, J. (1974). Response depriva- critical time window for dopamine actions on the
tion: An empirical approach to instrumental perfor- structural plasticity of dendritic spines. Science 345,
mance. Psychological Review, 81, 146–164. 1616–1620.
CHAP TER 3
A. Charles Catania
Handbooks are often consulted as resources for response has produced the event. An organism is
information about specific topics, so this chapter said to come into contact with a contingency when its
is organized as a set of somewhat independent sec- behavior produces some consequences of the con-
tions. It opens with a discussion of operant contin- tingency. Unless otherwise stated, for convenience
gencies, then considers some aspects of the basic the term contingency here implies a response–con-
contingencies known as reinforcement and punish- sequence contingency, rather than contingencies
ment and their positive and negative variants, and more broadly conceived (e.g., stimulus–stimulus
closes with some implications of these contingen- contingencies).
cies and brief surveys of a few related issues. For When responses produce stimuli, the contin-
more detailed treatments, see Skinner (1938, 1953, gent relation is defined by two conditional prob-
1999); Iversen and Lattal (1991a, 1991b); Catania abilities: probability of the stimulus (1) given a
(2013); various volumes of the Journal of the Exper- response and (2) given no response. Without both
imental Analysis of Behavior and The Behavior Ana- probabilities specified, the contingent relations
lyst; and two special issues of the European Journal cannot be distinguished from incidental temporal
of Behavior Analysis, one devoted to contingencies contiguities of responses and stimuli that are occur-
(Arntzen, Brekstad, & Holth, 2006) and the other ring independently over time.
devoted to noncontingent reinforcement (Arnt- Response–reinforcer relations involve two
zen, Brekstad, & Holth, 2004). terms (the response and the reinforcer), but when
correlated with discriminative stimuli (stimuli that
set the occasion on which responses have conse-
RESPONSE–CONSEQUENCE CONTINGENCIES quences), they produce a three-term contingency,
which involves antecedents, behavior, and con-
Contingencies relating responses to their conse- sequences. For example, a child’s touch of a card
quences are properties of environments. They are might be reinforced with an edible if the card is
probability relations among events. When a re- green, but not if it is any other color. In this case,
sponse changes the probability of some event, we green, as the discriminative stimulus, is the first
say that the change is contingent on the response; term; the touch, as the response, is the second
when the change is from a relatively low probabil- term; and the edible, as the reinforcer, is the third
ity to a probability of 1.0, we usually say that the term. Antecedents typically include establishing
37
38 B a s i c P r in c ip l e s a nd C o n c e p t s
both roles simultaneously. In chaining, for exam- But misdiagnosis can go either way. For example,
ple, the stimulus produced by a response early in if such behavior has its source in eliciting stimuli,
a sequence both reinforces that response and sets perhaps for neurological reasons, interpreting it as
the occasion for the next one, as when the open- shaped by reinforcement contingencies could simi-
ing of a door both reinforces the turn of the door- larly lead to ineffective treatment. And it can get
knob and allows the behavior of stepping through even more difficult. In multiple causation, elicit-
to the next room. Stimuli that reinforce or punish ing stimuli and reinforcement contingencies may
some responses can also elicit or occasion others, operate at the same time, so identifying the role of
so choices among such consequences in reinforce- one should not rule out assessments of the other.
ment applications must take into account both the
main reinforcing or punishing effects and their
eliciting or occasioning side effects. REINFORCEMENT
It may be necessary to determine whether be-
havior is maintained by its consequences or is A reinforcer is a type of stimulus, but reinforce-
produced more directly by stimuli. Imprinting ment is neither stimulus nor response. The term
provides a case in point. A newly hatched duck- reinforcement names a relation between behavior
ling ordinarily follows the first moving thing it and environment. The relation includes at least
sees; this imprinted stimulus is usually its mother. three components: (1) Responses must have con-
The duckling’s following is sometimes said to be sequences; (2) their probability must increase (i.e.,
elicited by the imprinted stimulus, but to speak of they must become more probable than when they
elicitation is misleading. A natural consequence of do not have those consequences); and (3) the in-
walking is changing the duckling’s distance from crease must occur because they have those con-
its mother. If closeness is important and requires sequences and not for some other reason. For ex-
behavior other than walking, that other behavior ample, if we knew only that responding increased,
should replace the walking. we could not say that the response must have been
In one early experiment, when a dark compart- reinforced; maybe it was elicited. It would not even
ment containing a moving imprinted stimulus be enough to know that the response was now
was on one side of a one-way window and a re- producing some stimulus it had not been produc-
sponse was available on the other side that lit up ing before. We would still have to know whether
the dark side so the duckling could see it, behav- responding increased because the stimulus was its
ior incompatible with following (such as pecking consequence.
a stationary disk on the wall or standing still on Assume that an abusive parent gets annoyed
a platform) was readily shaped (Peterson, 1960). whenever an infant cries and tries to suppress the
In imprinting, therefore, presentations of the to- crying by hitting the child. The infant cries and
be-imprinted stimulus are establishing events, not then gets hit, which produces even more crying.
eliciting stimuli. Imprinted stimuli, which acquire Here the consequence of crying is getting hit, and
their significance by being presented under ap- getting hit produces more crying, but we cannot
propriate circumstances, begin as stimuli toward argue that the hitting reinforces the crying. Two
which the duckling is relatively indifferent but criteria for reinforcement are satisfied, but not the
end as ones that function as reinforcers. Imprinted third. Stimuli may have other effects along with
stimuli do not elicit following; rather, they become or instead of their effects as consequences of re-
important enough that they can reinforce a vari- sponding. Crying does not increase here because
ety of responses, including following, pecking, and getting hit is a consequence; getting hit brings on
standing still. The point should have been obvious crying even if the infant is not crying at the outset.
to early researchers on imprinting. In natural en- Probably the infant will eventually learn to sup-
vironments, swimming replaces walking when the press the crying. At that point, we will know that
duckling follows its mother into a body of water. the crying was punished rather than reinforced.
If walking were mere elicited behavior, it should In earlier days, scientists would have discussed
not do so. the relation between a response and its reinforcer
Analogous relations can have profound impli- in terms of associations—a principle with substan-
cations in clinical settings. For example, inter- tial precedent in psychology and philosophy—
preting a hospitalized child’s problem behavior rather than in terms of contingencies. In this
as elicited behavior when it has its source in re- discussion, learning was said to occur through the
inforcement contingencies might prevent appro- association of ideas, and the conditional reflexes
priate treatment options from being considered. of Pavlov (1927) seemed to be cases of such associ-
40 B a s i c P r in c ip l e s a nd C o n c e p t s
ations. In one example of Pavlov’s work, the sound phies. Common contingencies select the members
of a metronome consistently preceded the delivery of operant classes, and they do so even if the rela-
of food to a dog, and the dog came to salivate at tions among members are arbitrary. A lever press
the sound of the metronome. Scientists assumed is a lever press, whether the rat presses with right
that learning occurred through the temporal con- paw, left paw, chin, or rump.
tiguity of events, or their occurrence together in The distinction between function and topog-
time. However, we cannot interpret this learn- raphy is particularly crucial when it enters into
ing as an association or as simply substituting one diagnostic categories. The self-injurious behavior
stimulus for another. Pairings in time or contigui- of two children may be similar in topography, but
ties are not equivalent to contingencies, and pair- if one child’s behavior is reinforced socially by at-
ings alone are not sufficient to produce respondent tention and the other’s is reinforced by avoidance
conditioning. Two stimuli can occur together not of compliance with simple requests, effective treat-
only when one never occurs without the other, but ment programs designed for the two children will
also when either can sometimes occur indepen- have to be radically different (Iwata, Pace, Kalsher,
dent of the other. We must specify a contingency, Cowdery, & Cataldo, 1990). The first child must
or the probability of one given the other. An ac- be taught more effective ways of engaging the at-
count solely in terms of association or contiguity tention of others and must be brought into situ-
is inadequate (e.g., Catania, 1971; Donahoe & ations where attention is more readily available.
Vegas, 2004). Requests must be selected for the second child that
are appropriate to the child’s competence, and the
child’s compliance with those requests must be re-
Specificity of Reinforcers
inforced (perhaps in the past, such behavior has
By definition, reinforcement always increases re- instead been punished). What behavior does is
sponding relative to what it would have been like more important than what it looks like.
without reinforcement. Also by definition, that
increase must be specific to the response that pro-
Assessing Reinforcers
duces the consequence. For example, if a rat’s lever
presses produce shock and only the rat’s jumping Events that are effective as reinforcers are often
increases, it would be inappropriate to speak of ei- described in terms of positive feelings or strong
ther pressing or jumping as reinforced. preferences. Such descriptions are subject to the
As an operation, reinforcement is presenting a inconsistent practices of verbal communities, so
reinforcer when a response occurs; it is carried out we must be wary of using them to predict wheth-
on responses, so we speak of reinforcing respons- er particular events will serve as reinforcers. It is
es rather than of reinforcing organisms. We may tempting to equate reinforcers with events col-
say that a pigeon’s key peck was reinforced with loquially called “rewards.” But reinforcers do not
food, but not that food reinforced the pigeon or work because they make the organism “feel good,”
that the pigeon was reinforced for pecking. The or because the organism “likes” them. Our every-
main reason for this restriction is that it is too easy day language does not capture what is important
to be ambiguous by omitting the response or the about reinforcers. For example, staff predictions of
reinforcer, or both, when we speak of reinforcing the reinforcers that might be effective in manag-
organisms. The restriction forces us to be explicit ing the behavior of people with profound disabili-
about what is reinforced by what. For example, ties were inconsistent with reinforcers identified
if we have been told only that a child has been by systematically assessing each individual’s non-
reinforced, we do not know much about actual verbal preferences among those events (Fisher et
contingencies. Although this grammatical restric- al., 1992; Green et al., 1988).
tion forces us to be explicit about which response We sometimes make good guesses about what
has been reinforced, it does not prevent us from will be effective reinforcers, because reinforcers
identifying the organism whose behavior had con- often involve events of obvious biological sig-
sequences. nificance. But reinforcers are not limited to such
events. For example, sensory stimuli, such as flash-
ing lights, can powerfully reinforce the behavior
Function and Topography of Reinforced Responses
of children along the autism spectrum (Ferrari &
Reinforcement creates response classes defined by Harris, 1981). Restraint also seems an unlikely re-
their functions and not by their forms or topogra- inforcer, but in an analysis of self-injurious behav-
Basic Operant Contingencies 41
ior, restraints that prevented children with severe we target for reinforcement. When such behavior
developmental disabilities from poking or biting shares in the effect of the reinforcer, we may mis-
themselves were effective in reinforcing arbitrary takenly conclude that the reinforcer is not doing
responses, such as putting marbles in a box (Favell, its job very well. But if the reinforced behavior in-
McGimsey, & Jones, 1978). cludes response classes that we did not intend to
In the final analysis, the primary criterion for reinforce, it may simply be doing very well a job
reinforcement remains whether the consequences other than the one we wanted it to do. When one
of behavior have raised the likelihood of that be- response is followed by a different reinforced re-
havior. Reinforcers may be correlated with other sponse, the reinforcer may strengthen both, so we
properties, such as reported feelings or preferences, should keep behavior that we do not want to rein-
but they are defined solely by their behavioral ef- force from getting consistently close to reinforcers
fects. produced by other responses.
An intervention for children with autism spec-
trum disorder who displayed persistent errors il-
Delay of Reinforcement
lustrates this principle (Fisher, Pawich, Dickes,
The effects of a reinforcer depend on other re- Paden, & Toussaint, 2014). The children repeated
sponses that preceded it besides the one, usually errors during baseline, even though each correct
most recent, that produced it. Thus, when one response produced an edible reinforcer (FR 1). To
response is followed by a different reinforced re- test whether intermittent errors persisted because
sponse, the reinforcer may strengthen both. Cli- the errors frequently preceded a correct response
nicians and teachers need to take this effect into followed immediately by a reinforcer, the inves-
account, because it is important to recognize that tigators implemented a second-order schedule in
reinforcing a single correct response after a long which they placed an edible reinforcer into one
string of errors may strengthen errors along with of three small glass containers in front of a child.
the correct response. After three consecutive correct responses, the
Assume that a task involves a child’s correct therapist delivered the three accumulated rein-
responses and errors over trials. Reinforcing every forcers to the child, but after an error, the thera-
correct response and repeating any trial with an pist emptied any accumulated reinforcers from the
error until the child gets it right guarantees that containers and the child had to start over. This
any sequence of errors will eventually be followed procedure reduced errors by ensuring that rein-
by a reinforced correct response. Correct responses forcers followed only sequences of three correct
will probably dominate eventually, because the re- responses and thus were not presented soon after
inforcer most closely follows them. But errors may errors.
diminish only slowly and perhaps even continue
indefinitely at a modest level, though they never
Relativity of Reinforcement
actually produce the reinforcer, because they are
reliably followed after a delay by a reinforced cor- Reinforcement is relative in the sense that it de-
rect response. Thus always reinforcing a single cor- pends on relations between the reinforced re-
rect response after a sequence of errors will prob- sponse and the reinforcer. A less probable response
ably maintain errors. may be reinforced by an opportunity to engage in a
Teachers and clinicians must be alert for such more probable response. The inverse relation does
situations. A reinforcer that follows a sequence of not hold. For example, food is not always a rein-
correct responses will probably do a lot more good forcer. When a parent allows a child to go out and
than a reinforcer that follows a single correct re- play with friends only after the child has eaten, the
sponse after several errors. Thus a teacher must opportunity to play may reinforce the eating.
judge whether correct responses are so infrequent The reversibility of the reinforcement relation
that they should be reinforced even though pre- has been amply demonstrated (Premack, 1962).
ceded by errors, or so frequent that the reinforcer For example, levels of food and water deprivation
can wait until the student has made several cor- can be selected so that drinking is reinforced by
rect responses in a row. Another way to reduce the an opportunity to eat at one time, and eating is
strengthening of errors is to extend the time to the reinforced by an opportunity to drink at anoth-
next trial after every error. er. In providing an a priori means for predicting
Many practical applications of reinforcement whether an opportunity to engage in one response
include other behavior that precedes the behavior will reinforce some other response, the relativity of
42 B a s i c P r in c ip l e s a nd C o n c e p t s
reinforcement also avoids the problems of circular If the effects of reinforcement are temporary,
definition inherent in some earlier definitions of then once we have created new behavior with rein-
reinforcement. forcers, we cannot count on its maintenance after
The significance of reinforcers is based on the our intervention ends. Consider children learning
opportunities for behavior that they allow. For ex- to read. Only long after they have learned to name
ample, when time spent in isolation was used in letters of the alphabet and to read whole words are
an attempt to punish the tantrums of a 6-year-old they perhaps ready to read stories, so that reading
girl with autism spectrum disorder, her tantrums can become “its own reward.” Until that happens,
increased substantially instead of decreasing. teachers have no choice but to arrange artificial
This child often engaged in self-stimulation, such contingencies, using extrinsic consequences such
as waving her fingers over her eyes to create vi- as praise to shape the components of reading. Re-
sual flicker, but that behavior was frequently in- sponsible teaching adds extrinsic reinforcers only
terrupted by the staff. Time in the isolation room when there are no effective intrinsic consequenc-
reinforced rather than punished her tantrums, es. It is important to build momentum, but if we
because the isolation room allowed her to engage want to maintain behavior after we terminate arti-
in self-stimulation without interruption (Solnick, ficial consequences, we should do so only if natural
Rincover, & Peterson, 1977). Similarly, a number consequences are in place that will take over that
of investigators have shown that contingent access maintenance.
to stereotypic and repetitive behavior can be used
to reinforce socially appropriate responses (Char-
Side Effects of Reinforcement and Extinction
lop, Kurtz, & Casey, 1990; Fisher, Rodriguez, &
Owen, 2013). Discontinuing reinforcement in extinction has
The relativity of reinforcement reminds us that two components: (1) It terminates a contingency
we should not expect the effectiveness of rein- between responses and reinforcers, and (2) rein-
forcers to be constant across different reinforced forcers are no longer delivered. Because of the for-
responses, different individuals, or even different mer, the previously reinforced responding decreas-
time samples of the behavior of a single individual. es. Because of the latter, unwelcome side effects
When a reinforcer is effective on some behavior in of extinction may appear. For example, aggressive
some context, we must not assume that it will be responding is sometimes a major side effect of ex-
effective on other behavior or even on the same tinction (e.g., Lerman, Iwata, & Wallace, 1999). If
behavior in other contexts. food is suddenly taken away from a food-deprived
rat that has been eating, the rat may become more
active and perhaps urinate or defecate. If the food
Reinforcement and Extinction: Response Rate
was produced by lever presses, the rat may bite
versus Momentum
the lever. If other organisms are in the chamber,
The effects of reinforcers are not permanent. the rat may attack them (Azrin, Hutchinson, &
Reinforcers have temporary effects; when rein- Hake, 1966). These effects and others, though
forcement stops, responding returns to its earlier, observed in extinction, are not produced by the
lower levels. The decrease in responding during termination of the reinforcement contingency
extinction does not require a separate treatment; because they also occur upon the termination of
rather, it is simply one property of reinforcement. response-independent food deliveries, where there
But reinforcement does not merely maintain rates had been no reinforcement contingency. In either
of responding, which are reduced when it is dis- case, a rat that had been eating stops getting food.
continued. It also produces resistance to change, The termination of a reinforcement contingency
or momentum (Nevin, 1992). Two different re- in extinction necessarily entails the termination
sponses with two different histories may be main- of reinforcer deliveries, and the effects of the lat-
tained at similar rates, but one may decrease more ter are necessarily superimposed on the decrease
rapidly in extinction than the other. Responding in previously reinforced responding.
that is more resistant to change is said to have Even if reinforcers have produced problem be-
greater momentum. Extinction is just one type of havior, taking them away may still produce un-
change, and other sources that may be used to as- desired side effects. That is why extinction is not
sess momentum include reinforcement of compet- the method of choice for getting rid of behavior
ing responses, establishing conditions, and delay that has been created by reinforcement. Suppose
of reinforcement. a developmentally delayed boy engages in severe
Basic Operant Contingencies 43
self-injurious behavior such as head banging or gent reinforcement, then responding may decrease
eye poking, and we discover that his behavior is more slowly than in extinction. But such effects
in large part maintained by staff attention as a are usually transient, so this is not a big enough
reinforcer. Because of the harm he might do to concern to rule noncontingent reinforcement out
himself if the self-injurious behavior is ignored, of consideration in either experimental or applied
extinction may be ill advised. Giving him atten- settings. If higher or lower rates of noncontingent
tion independently of the self-injurious behavior reinforcement are available as options, this con-
is one possibility (noncontingent reinforcement, cern favors the lower rates. If behavior persists for
sometimes also called free reinforcement) (Cata- long periods of time under such arrangements, it is
nia, 2005; Lattal, 1974; Sizemore & Lattal, 1977); more appropriate to look for other sources of the
another is to use attention to reinforce alternative behavior than to attribute it to adventitious cor-
responses, and especially ones incompatible with relations of responses and reinforcers.
the self-injury. The self-injury will decrease as al-
ternative responses increase.
Positive Reinforcement and Positive Psychology
These side effects are one reason why extinc-
tion has fallen out of favor in applied settings, Positive reinforcement can be used to change a
compared to procedures such as noncontingent developmentally delayed child who engages exten-
reinforcement. The Journal of Applied Behavior sively in self-injurious behavior into one who has
Analysis has published relatively few examples learned communicative skills and has therefore
of extinction with humans. In general, the solu- been empowered to deal in more constructive ways
tion is not to take the reinforcers away. The bet- with his or her caregivers. If reinforcers are impli-
ter way to reduce misbehavior is to reinforce good cated in the development and maintenance of the
behavior, but sometimes we inadvertently encour- self-injurious behavior, then taking them away is
age the use of extinction (the less effective alter- not the solution. Reinforcement isn’t everything,
native)—especially when we present just a few but extinction isn’t anything. If the reinforc-
basic facts about behavior, as in an introductory ers are already there, they should not be wasted;
psychology course. Generations of students seem they should instead be used constructively. We
to have taken from cursory accounts of behavioral all shape each other’s behavior, and the more we
methods in introductory textbooks the message know about how positive reinforcement works, the
that if one sees a child doing something one does more likely we will be to use it productively and
not approve of, then one should not reinforce that avoid pitfalls such as the coercive practices that
behavior. Instead, one should just ignore it. Left can occur if the control over reinforcers remains
unanswered are the inevitable subsequent ques- one-sided. For these reasons, it might be thought
tions, such as how parents should handle things that positive reinforcement would be especially
when other problematic behavior maintained by important to the practitioners of the approach
the same reinforcer emerges. Rather than teaching called positive psychology. Unfortunately, they es-
parents to ignore the behavior of their children, chew it, along with the establishing events that
we should teach them how to use reinforcers more make it effective; their rhetoric implies that con-
productively, but that alternative is more difficult. tingent acts of kindness should always be replaced
Free noncontingent reinforcement coupled with by random ones (cf. Catania, 2001; Seligman &
the shaping of other behavior should be recom- Csikszentmihalyi, 2000, 2001).
mended to parents or other caregivers, but doing
so poses problems of both communication and
Self‑Reinforcement as Misnomer
implementation (Hagopian, Crockett, van Stone,
DeLeon, & Bowman, 2000). An organism’s delivery of a reinforcer to itself
Why has extinction for so long remained the based on its own behavior has been called self-re-
primary way to study the effects of terminating inforcement, but any effect such an activity might
contingencies? One concern is that accidental have cannot be attributed to the action of the
contiguities of responses and noncontingent re- specific reinforcers delivered by the organism to
inforcers may have effects similar to those of the itself. In so-called “self-reinforcement,” the con-
contiguities that are scheduled when reinforcers tingencies and establishing events that modify
are contingent on responding. If noncontingent the behavior purportedly to be reinforced cannot
and contingent reinforcers have similar effects be separated from those that modify the behav-
on behavior early in the transition to noncontin- ior of self-reinforcing. For example, a student who
44 B a s i c P r in c ip l e s a nd C o n c e p t s
has made a commitment to watch television only stimulus it had not produced before. We would still
after completing a study assignment might think have to know whether responding decreased be-
that this arrangement will reinforce studying. But cause that stimulus was its consequence.
any increase in studying that follows cannot be at- As defined, punishment is the inverse of rein-
tributed to the student’s contingent watching of forcement; it is defined by decreases in consequen-
television: The student has made the commitment tial responding, whereas reinforcement is defined
to deal with studying this way because studying by increases. The vocabulary of punishment paral-
has already become important for other reasons. lels that of reinforcement in its object: Responses,
Whatever has brought the student to commit to not organisms, are said to be punished. If a rat’s
“self-reinforce” studying in the first place has prob- lever pressing produces shock and lever pressing
ably by itself made studying more likely. It is im- decreases, it is appropriate to say that the rat was
possible to pull these variables apart. shocked and that the lever press was punished; it
What was once called self-reinforcement is now goes against colloquial usage, but it is not appro-
more properly called self-regulation (Bandura, 1976, priate to say that the rat was punished. As with
1995; Catania, 1975, 1995; Mahoney & Bandura, reinforcement, this grammatical distinction dis-
1972). To the extent that the activity has effects, courages ambiguities in the observation and de-
it must do so because the individual who appears scription of behavior.
to “self-reinforce” can discriminate behavior that
qualifies for the reinforcer from behavior that does
Parameters of Punishment
not.
This usage also finesses the problem that the As with reinforcement, the effectiveness of pun-
language of self-reinforcement implies reinforce- ishment varies with parameters such as magnitude
ment of the organism rather than reinforcement and delay (Azrin & Holz, 1966). For example, the
of behavior. For example, the commitment to rein- more intense and immediate the punisher, the
force one’s own studying involves setting standards more effectively it reduces behavior. A punisher
for the discrimination between adequate and inad- introduced at maximum intensity reduces respond-
equate studying, so students who try to deal with ing more effectively than one introduced at low
their study habits in this way are discriminating intensity and gradually increased to maximum
properties of their own behavior that have become intensity. The effectiveness of the punisher may
important to them. The contingencies that gener- change over time, such as when a punisher of low
ate these discriminations are complex and prob- intensity gradually becomes ineffective after many
ably involve verbal behavior. The language of self- presentations. As with extinction, it is easier to re-
reinforcement obscures rather than clarifies these duce the likelihood of a response when some other
phenomena. response that produces the same reinforcer is avail-
able than when no alternative responses produce
that reinforcer. And in a parametric relation espe-
PUNISHMENT cially relevant to human applications, punishers
delivered after short delays are more effective than
Paralleling the vocabulary of reinforcement, a those delivered after long ones; with either pets or
punisher is a type of stimulus, but punishment is children, aversive consequences delivered at some
neither stimulus nor response. The term punish- point long after unwanted behavior occurs are not
ment names a relation between behavior and envi- likely to be very effective. If verbal specification of
ronment. The relation includes at least three com- the behavior on which the punisher is contingent
ponents. First, responses must have consequences. matters at all, it can do so only given an extensive
Second, their probability must decrease (i.e., they and sophisticated verbal history on the part of the
must become less probable than when they do not individual at the receiving end (Skinner, 1957).
have those consequences). Third, the decrease A reduction in responding can be studied only
must occur because they have those consequences if some responding already exists. A response that
and not for some other reason. For example, if we is never emitted cannot be punished. Experiments
knew only that responding decreased, we could not on punishment therefore usually superimpose
say that it must have been punished; maybe it was punishment on reinforced responding. But the
previously reinforced responding that had since effects of punishment then also depend on what
been extinguished. It would not even be enough to maintains responding. For example, punishment
know that the response was now producing some by shock probably will reduce food-reinforced lever
Basic Operant Contingencies 45
pressing less if a rat is severely food-deprived than receives shocks during a tone, but during a buzzer
if it is only mildly food-deprived. nothing happens. If chain pulls turn off the tone
and turn on the buzzer, the onset of the buzzer will
reinforce chain pulls; by pulling the chain, the
Recovery from Punishment
rat escapes from the tone and its accompanying
There are ethical constraints on using punish- shock deliveries. Suppose, however, that the rat in-
ment to change behavior. The use of punishment stead receives food during the tone, but during the
as a component of clinical interventions has de- buzzer nothing happens. Now if chain pulls turn
creased in recent years (cf. Greer, Fisher, Saini, off the tone and turn on the buzzer, the onset of
Owen, & Jones, 2016), but punishment cannot the buzzer will punish chain pulls; by pulling the
be eliminated from natural environments (Per- chain, the rat produces a time out from the tone
one, 2003). Without punishment, a child who has and its accompanying food deliveries. In other
been burned upon touching a hot stove or bitten words, the buzzer serves as reinforcer or as punish-
upon approaching an unfamiliar barking dog will er, depending on its context, even though nothing
remain undeterred from doing so again later on. happens during the buzzer in either context. Simi-
Artificial punishment contingencies, however, are larly, as gauged by absenteeism, whether a school
also constrained by practical considerations. Like environment is punishing or reinforcing may de-
reinforcement, the effects of punishment are or- pend on the conditions that prevail at home—as
dinarily temporary; responding usually recovers when going to school is punished for one child
to earlier levels after punishment is discontinued. because it means having to deal with an abusive
This means that just as behavior analysts must school bully, but is reinforced for another because
plan for what will maintain the behavior when it is a convenient way to escape from an even more
reinforcement ends, they also must plan for envi- abusive parent.
ronments in which the relevant contingencies may
be absent. It may do little long-term good to elimi-
Side Effects of Punishment: Eliciting
nate a child’s self-injurious behavior with pun-
and Discriminative Effects
ishment in a hospital setting if the punishment
contingency does not exist when the child returns Aversive stimuli are likely to have other effects
home. The reinforcement of alternative behavior besides those that depend on the punishment con-
might be easier to maintain. tingency (e.g., Azrin, Hutchinson, & McLaughlin,
1965). As with reinforcement, punishment nec-
essarily includes both stimulus presentations and
Relativity of Punishment
a contingency between responses and stimuli, so
The effectiveness of punishers, like that of rein- the effects of the stimulus presentations must be
forcers, is determined by the relative probabilities distinguished from those of the contingency. If
of the punished response and the responses occa- an organism is shocked or pinched, some of its re-
sioned by the punisher; punishment occurs when sponses to those stimuli may have little to do with
a more probable response forces the organism to whether they have been brought on by the organ-
engage in a less probable response. Even stimuli ism’s own behavior. To qualify as punishment, the
that ordinarily serve as reinforcers can become reduction in responding must depend on the con-
punishers under appropriate conditions. For ex- tingent relation between responses and punishers,
ample, food that is reinforcing at the beginning of and not simply on the delivery of punishers.
a holiday feast may become aversive by the time A comparison of the effects of response-pro-
the meal has ended. On the other hand, events duced and response-independent shock on food-
that superficially seem aversive, such as falling reinforced lever pressing in rats (Camp, Raymond,
from a height, may be reinforcing under some cir- & Church, 1967) showed that both procedures
cumstances (consider skydiving). Like reinforcers, reduced lever pressing relative to no-shock condi-
punishers cannot be defined in absolute terms or tions, but that response-produced shock had sub-
in terms of common physical properties. Rather, stantially larger effects than response-independent
they must be assessed in terms of the relation be- shock. Given that both response-produced and
tween punished responses and the responses oc- response-independent shocks reduced responding,
casioned by the punisher. it would not have been possible to assess the ef-
Any given state of affairs may be reinforcing or fect of the punishment contingency without the
aversive, depending on its context. Suppose a rat comparison. The difference made it appropriate to
46 B a s i c P r in c ip l e s a nd C o n c e p t s
call the response-produced shock a punisher. For as punishment is a robust phenomenon that can
example, had response-produced shock instead occur within a short time course (the abrupt in-
produced only the same reduction as response- troduction of a strong punisher reduces responding
independent shock, the appropriate conclusion quickly), the literature on avoidance shows that
would have been that the reduction depended though avoidance is robust once in place, it is dif-
wholly on the eliciting effects of shock, and that ficult and time-consuming to get it started. Passive
the punishment contingency was irrelevant. Just avoidance is therefore best regarded as a misnomer
as we must distinguish between effects of reinforc- for punishment. It is implausible to say that hypo-
er deliveries and effects of the contingent relation thetical behavior presumed to come from relations
between responses and reinforcers, so also we must that are difficult to establish can explain behavior
distinguish between effects of punisher deliveries that is easy to establish. Even if punishment did
and effects of the contingent relation between re- work this way, so that we learn not to stick our
sponses and punishers. hands into a fire because by so doing we avoid the
Punishers can also acquire discriminative prop- aversive proprioceptive stimuli occasioned by ap-
erties, as when a response is reinforced only when proaching the fire, it would make little practical
it is also punished. For example, one experiment difference. For those who have to make decisions
alternated a condition in which a pigeon’s key about whether or when to use punishers, punish-
pecks had no consequences with another condi- ment works pretty much the same way whichever
tion in which every peck produced shock and some theoretical position one assumes.
produced food reinforcers (Holz & Azrin, 1961). A
low rate of pecking was maintained when pecks
produced no shock, because then they never pro- NEGATIVE REINFORCEMENT:
duced food either; pecking increased once pecks ESCAPE AND AVOIDANCE
began to produce shock, however, because only
then did they occasionally produce food. Organisms not only produce stimuli; they also
We can ask whether these shocks should really get rid of them. Without negative reinforcement,
be called punishers. In fact, here we must conclude a child would not learn to escape from the cold
that the shock had become a conditional rein- by coming indoors or to avoid others who might
forcer. The main difference between the shock cheat at games, bully, or lie. A rat does not ordinar-
and other, more familiar reinforcers was that it ily expose itself to shock, and if shock does occur,
acquired its power to reinforce through its relation the rat escapes from it given the opportunity. If
to food; if that relation had been discontinued, it presenting a contingent aversive stimulus punishes
would have lost that power. As an example of a a response, removing or preventing that stimulus
possible source of masochism, these procedures may reinforce a response. When a response ter-
may be relevant to human behavior. For example, minates or prevents an aversive stimulus and be-
a battered child might provoke a parent to the comes more probable for that reason, the stimulus
point of a beating because the beatings are often is called a negative reinforcer, and the operation is
followed by more attention from the then-remorse- called negative reinforcement.
ful parent than ever follows less traumatic parent– In traditional usage, positive and negative,
child interactions. A parent’s attention can be a as modifiers of the term reinforcement, refer to
potent reinforcer and may sometimes override the whether the consequence produced by responding
effects of consequences that would otherwise serve adds something to the environment or takes some-
as punishers. thing away—but we will see later that there are
other, better criteria for the distinction. Negative
reinforcer refers to the stimulus itself and not to its
Passive Avoidance as Misnomer
removal; if removal of shock reinforces a rat’s lever
It has been argued that punishment is reducible press, then shock, not the shock-free period that
to avoidance, in the sense that all behavior other follows the response, is the negative reinforcer.
than the punished response avoids the punisher. Negative reinforcement involving the removal of
For example, if a rat is placed on a platform above a stimulus that is already present is called escape.
an electrified grid, then not stepping down onto When it involves the postponement or prevention
the grid might be called passive avoidance of of a stimulus that has not yet been delivered, it
shock; by not responding, the rat passively avoids is called avoidance. This vocabulary is consistent
what would otherwise be a punisher. But where- with everyday usage: We escape from aversive cir-
Basic Operant Contingencies 47
cumstances that already exist, but we avoid poten- reinforced response is emitted; it is removed only
tial aversive circumstances that have not yet hap- after the response. For example, if the negative
pened. In clinical situations, escape (e.g., from a reinforcer is bright light from which the rat can
medical unit) is often called elopement. escape by pressing a lever, the rat may reduce the
Stimuli that can reinforce by their presentation effects of the light by closing its eyes and hiding its
can punish by their removal, and vice versa. If we head in a corner. Any movement from that posi-
know that a stimulus is effective as a punisher, tion is punished by greater exposure to the light,
then we can reasonably expect it to be effective so the rat is not likely to come out of the corner
as a negative reinforcer, and vice versa; this con- and press the lever. Getting a rat to escape from
sistency is part of our justification for calling the light by lever pressing requires procedures that re-
stimulus aversive. Consistencies are to be expected duce the likelihood of such competing responses
because these categories have their origins in rela- (Keller, 1941).
tions among the probabilities of different response
classes. But we must not take too much for granted.
Avoidance: Hard to Initiate but Easy to Maintain
The fact that we may easily reinforce jumping with
shock removal, whereas we may not so effectively Avoidance involves the prevention of an aversive
punish it with shock presentation, shows that the stimulus by a response; the aversive stimulus is not
symmetry of reinforcement and punishment has present when the reinforced response occurs. The
limits. Reinforcement is most effective if the rein- two major varieties of avoidance procedures are
forced response is compatible with the responding deletion and postponement. Deletion procedures
occasioned by the reinforcer. Inversely, punish- are analogous to swatting a mosquito before it gets
ment is most effective if the punished response is to where it can bite: Once swatted, that mosquito
incompatible with, or at least independent of, the is permanently prevented from biting. Postpone-
responding occasioned by the punisher. Thus it ment procedures are analogous to putting coins in
may be easy to reinforce jumping with shock re- a parking meter: One postpones the violation flag
moval (escape), but hard to punish it with shock as long as one puts coins in the meter and resets it,
presentation. but without additional coins the meter eventually
runs out.
In discriminated or signaled avoidance, a stimu-
Escape: Competition between Contingencies
lus (sometimes called a warning stimulus) precedes
and Elicited Behavior
the aversive stimulus; a response in the presence
In escape, an organism’s response terminates an of this stimulus prevents the aversive stimulus
aversive stimulus. In institutional settings, devel- on that trial. In continuous or Sidman avoidance,
opmentally delayed children sometimes behave no exteroceptive stimulus is arranged. Each re-
aggressively, in that way escaping from simple de- sponse postpones the aversive stimulus (usually,
mands placed upon them, such as tasks designed brief shock) for a fixed time period called the re-
to teach them how to fasten and unfasten clothing sponse–shock (R–S) interval; in the absence of re-
buttons. For two such children, aggression dropped sponses, shocks are delivered regularly according
to near-zero levels when they could escape from to a shock–shock (S–S) interval (Sidman, 1953).
demand situations by engaging in other behavior Shock can be postponed indefinitely, provided
incompatible with aggression (Carr, Newsom, & that no R–S interval ends before a response has
Binkoff, 1980). But such cases of escape might, of been emitted.
course, imply that typical demand situations in Success with avoidance procedures sometimes
such settings do not provide enough reinforcers. depends on whether the experimenter chooses
In positive reinforcement, the reinforcer is a response that the organism is likely to emit in
absent when the reinforced response is emitted. aversive situations. With rats, for example, re-
After the response, the reinforcer is presented and sponses such as jumping a hurdle or running from
occasions other responses. For example, if a rat’s one side of the chamber to the other are likely to
lever press is the reinforced response and food is be elicited by aversive stimuli even in the absence
the reinforcer, food is absent while the rat presses; of a response–shock contingency. Once respond-
eating does not occur until food is presented after ing has been produced by shock, it may contin-
the press. Thus lever pressing and eating do not ue when shock is absent. Thus the rat’s first few
directly compete with each other. In escape, how- avoidance responses may occur mainly because of
ever, the negative reinforcer is present before the their earlier elicitation by shock.
48 B a s i c P r in c ip l e s a nd C o n c e p t s
Avoidance behavior may be persistent after a huddles in a corner and shivers. These responses
long history of avoidance; it can be slow to extin- reduce the likelihood that it will press the lever.
guish. But the consequence of effective avoidance Once its lever press turns on the heat lamp, these
is that nothing happens: The aversive event is competing responses become less likely, but a rat
successfully avoided. Given that an avoidance re- that is no longer cold cannot escape from cold.
sponse is not closely followed by shock, avoidance Responses that competed with the reinforced re-
contingencies implicitly involve delays between sponse occurred before rather than after reinforce-
responses and their consequences. Thus, despite ment, so this example is more like escape from
the persistence of avoidance behavior once it is ad- shock or bright light than like production of food
equately in place, it is often hard to get it started. or water. In general, the language of negative rein-
This may explain why safety measures and other forcement is appropriate when establishing events
preventive procedures are not often shaped by nat- produce behavior that is likely to compete with
ural contingencies. Someone who has never had a the responding to be reinforced.
bad experience with fire may be less likely to in- Another possible criterion is whether one re-
stall a smoke detector than someone who has. One inforcement situation is preferred over another.
significant problem in medicine is the compliance In applied areas, such as management, it is use-
of patients with regimens such as taking prescribed ful to distinguish between two management task
medications. Many patients stop taking their med- contingencies: Managers get employees to carry
ications once their symptoms have disappeared, out tasks either by threatening or criticizing until
even though further doses may have continued tasks are completed, or, more rarely, by providing
benefits. And with preventive medication, such as praise and recognition after tasks are completed.
vaccination, taking a dose is followed by nothing Given a choice, employees are likely to move from
happening right from the start. This problem ex- settings in which they receive criticism to those
ists over a wide range of preventive measures, from in which they receive praise. Assuming that re-
immunizations to safe sex, and from using sterile ductions in threats maintain task completion, we
surgical equipment to purifying drinking water. may call that contingency negative reinforcement.
Given what we know about avoidance contingen- Assuming that recognition for completing tasks
cies, it is no surprise that such measures are some- maintains task completion, we may call that con-
times difficult to shape up and maintain. tingency positive reinforcement. The preferences
of employees for task contingencies justify this
classification.
Behavioral Criteria for Distinguishing Positive
from Negative Reinforcement
The Reinforcer in Negative Reinforcement
Whether stimuli are presented or removed may be
a less important criterion for distinguishing posi- When a successful avoidance response occurs, its
tive from negative reinforcement than whether important consequence is that nothing happens.
responses generated by the reinforcer occur at How can the absence of an event affect behavior?
times when they can compete with the reinforced According to one view, avoidance responding is
response. Consider escape from cold (Weiss & Lat- maintained because the organism escapes from
ies, 1961). In a cold chamber, a rat’s lever presses some properties of the situation that accompanied
turn on a heat lamp. Because presses add energy past aversive stimuli. This view evolved from ear-
in the form of heat, this procedure could be called lier procedures in which a warning stimulus pre-
positive reinforcement. But cold stimulates tem- ceded shock, and the organism prevented shock by
perature receptors in the rat’s skin, and turning on responding in the presence of the warning stimu-
the heat lamp terminates this effect of cold. Cold lus. Avoidance was most easily acquired when the
is a potent aversive event, so by this interpretation avoidance response both terminated the warning
the procedure should be called negative reinforce- stimulus and prevented the shock.
ment. In the context of occasional shocks, a shock-
The justification for choosing the vocabulary free period can serve as a reinforcer. Avoidance
of negative reinforcement lies not with questions contingencies can be arranged in which the or-
of physics, such as whether something is presented ganism can either reduce the total number of
or removed, but with the behavioral effects of the shocks in a session or postpone individual shocks,
stimuli presented before and after emission of the even though the same number of shocks is eventu-
reinforced response. Consider the behavior of the ally delivered in both cases. Either condition can
rat in the cold. Before the reinforced lever press, it maintain avoidance. Situations can be created in
Basic Operant Contingencies 49
which a rat postpones shocks within trials even not be so were shock never present when the light
though it does not reduce the overall shock rate, was off. In all of these cases, contingencies are
or in which it reduces the overall shock rate even about the consequences of responding, whereas es-
though responding shortens the time to the next tablishing or motivating events are about whether
shock (Herrnstein & Hineline, 1966; Hineline, those consequences are important enough to serve
1970). as reinforcers.
NEGATIVE PUNISHMENT: TIME OUT responses falling within the class. Differential re-
inforcement makes responding correspond more
The distinction between positive and negative and more closely to the defining properties of an
reinforcement is easily extended to positive and operant class. It can be based on any dimension of
negative punishment (though here too, ambiguous responding, though the dimension cannot always
cases are possible). Responses can be punished by be unambiguously defined. For example, original-
some events, such as shock or forced running in ity and other complex dimensions of behavior may
a running wheel. Responses also can be punished define operant classes, even though we sometimes
by the termination of events. For example, remov- have difficulty measuring them.
ing food contingent on a food-deprived rat’s lever Key pecks by pigeons and lever presses by rats
presses is likely to reduce the rate of pressing. The are examples of operant classes, but so are our
problem is that it may be hard to demonstrate neg- presses of elevator buttons or our asking someone’s
ative punishment. If the rat is food-deprived and name. Sometimes such classes are created natu-
food is available, it will probably eat rather than rally by contingencies, but sometimes we create
press, so we will have few opportunities to punish them artificially. If we put a pigeon in an exper-
lever pressing by removing food. For this reason, imental chamber, it may or may not peck a key.
studies of negative punishment usually have not Instead of waiting for a peck, we can generate one
removed the positive reinforcer itself; paralleling by shaping (i.e., by successively reinforcing other
the emphasis on avoidance rather than escape in responses that more and more closely approximate
studies of negative reinforcement, the stimulus in it). We start the shaping process by operating a
the presence of which responses are reinforced has feeder only when the pigeon turns toward the key.
been removed instead. After two or three movements toward the key, we
For example, suppose that two levers are avail- then reinforce not just any movement toward the
able to a monkey, and that presses on one lever key, but only those including forward motions of
produce food whenever a light is on. We can ex- the beak. Soon the pigeon’s beak movements are
pect presses on the other lever, but we can punish full-fledged pecks, and one strikes the key. At this
them by making each one produce a time period point, our apparatus takes over, and further pecks
during which the light turns off and presses on operate the feeder automatically.
the first lever do nothing. Such periods are called No set of rules for shaping can substitute for
time out, and the procedure is punishment by time what one learns by actually doing it. It is crucial
out from positive reinforcement (e.g., Ferster, 1958). to be sensitive to the moment-to-moment interac-
Time out originated in experiments like these tions of the organism’s behavior and the delivery
with pigeons, rats, and monkeys, but now is prob- of reinforcers. Shaping is an art applicable to many
ably best known in its human applications (e.g., skills: gymnastics; playing a musical instrument;
Wolf, Risley, & Mees, 1964). For example, time seduction; handwriting; and setting someone up
in an isolation room has sometimes been used to as a victim of a con game. In other words, it can
punish the problem behavior of institutionalized be put to either good use or bad, and many use it
children. In the casual use of time out as a punish- without even knowing they are doing so. As with
er by parents and teachers, contingencies are often reinforcement and punishment, when shaping is
inconsistently applied, and behavior that occurs put to good use, it might as well be done effec-
during the time out is too often neglected. The tively; when it is put to bad use, the best defense
term is occasionally extended to other cases (e.g., against it is knowing how it works.
time out from avoidance, during which no shocks An experienced experimenter can shape a pi-
are delivered). geon’s key peck with just 8–10 reinforcer deliveries.
A shaper who works close to the limits of extinc-
tion—thereby not reinforcing lots of responses
REINFORCEMENT AS SELECTION: now that will have to be extinguished later, when
THE SHAPING OF OPERANT CLASSES behavior has moved closer to the target response—
will probably finish shaping more quickly than one
A class of responses created by its consequences who is generous with reinforcers but risks satiation
is called an operant. It is defined by its function before reaching the target.
and not by its physical or physiological properties Shaping works because behavior is variable. In
(Skinner, 1935). The class depends on differential this respect, shaping is analogous to selection in
reinforcement, or the reinforcement of only those biological systems: Evolution occurs when envi-
Basic Operant Contingencies 51
ronments select individual organisms from popula- change, but 9 years is a very long time over which
tions that vary genetically (e.g., Donahoe, Burgos, contingencies can operate. This does not mean
& Palmer, 1993). Shaping is most obvious when that all such behavior is solely a product of contin-
used by a human trainer, as in the teaching of skills gencies. In the face of such possibilities, however,
to a service dog, but it can also occur as a result of it is certainly more appropriate to be alert for the
natural contingencies. For example, male cowbirds effects of such contingencies than to assume that
in different parts of the United States sing differ- they do not exist.
ent dialects of birdsong (as is usual among song-
birds, the female cowbird doesn’t sing). A female
is most likely to respond with mating-pattern pos- HIGHER‑ORDER CLASSES
tures to songs that sound most like the ones she AND OPERANT CONTINGENCIES
heard in her youth, which were in the dialect of
local males. When a foreign male is introduced, Contingencies can operate in combination and
he begins singing in his own dialect. But he sings present particular challenges when some contin-
with variations, and the more time he spends in gencies are nested in others, in higher-order classes.
her presence, the more his song takes the form of Sometimes when a response class appears insensi-
the local dialect. Her differential reactions are re- tive to its consequences, it is part of a larger class
inforcers, and they shape his song (e.g., King & whose other members continue to have the conse-
West, 1985). quences it once shared with them. In such cases,
Shaping can be hard to see if one does not know the contingencies operating on the higher-order
what to look for; someone who has actually done class may override those arranged for the original
shaping is more apt to notice it in natural environ- class. For example, once generalized imitation has
ments than someone who has only read about it. been established, a child may continue to imitate
Thus parents who always wait a while before at- some instance even though that particular imita-
tending to a crying child may not notice that they tion is never reinforced. That imitation may seem
have gradually shaped louder and more annoying insensitive to operant contingencies, but it will be
cries. The attention reinforces the crying, and an- maintained by the contingencies that operate on
noying cries are, by definition, the ones most likely the higher-order class as long as the higher-order
to get attention. If one watches what a parent does class maintains its integrity.
when a child throws tantrums, it is often easy to We would ordinarily expect subclasses for
guess where the tantrums have come from. The which reinforcement has been discontinued to be
contingencies that produce such problem behav- differentiated from their higher-order classes, but
ior seldom occur in isolation, so other behavior or this may not happen if the integrity of the higher-
other reinforcers may eventually displace the prob- order class depends on its membership in other,
lem behavior, but the spontaneous disappearance interlocking higher-order classes that still include
of problem behavior must not be taken as evidence the subclass (e.g., playing the game “Simon Says”
that the behavior’s source has been independent on the playground may help to maintain general-
of shaping. ized imitation in the classroom, even if imitative
Shaping can be demonstrated over minutes responses in the classroom are never reinforced).
rather than over days, years, or millennia. If re- In some cases this may be a problem, but in oth-
inforcers can do so much to behavior when con- ers it may instead be advantageous, such as when
tingencies are deliberately arranged over relatively new behavior emerges as a novel instance of the
short periods of time, it is reasonable to assume that higher-order class (e.g., the generalized imitation
they also affect behavior when natural contingen- of a movement the child has never seen before).
cies operate over substantial periods throughout Now consider a boy whose self-injurious behav-
an organism’s lifetime, as in the acquisition of ver- ior is reinforced by attention. Suppose we try to
bal behavior by children. Many contingencies may extinguish his self-injurious behavior by ignoring
take hold of behavior over the course of months it. We may have trouble from the start, because
or years in the life of a young child. Some may we cannot tolerate the damage he may do to him-
be subtle, especially given the very broad range self. We nevertheless persevere and discover that
of events that can serve as reinforcers. Some may his self-injurious behavior does not decrease. One
produce desirable behavior, and others may do the possibility is that we have not adequately identi-
opposite. The self-injurious behavior of an institu- fied the relevant response class. If the function
tionalized 9-year-old child may seem resistant to of this behavior is to produce attention, it may
52 B a s i c P r in c ip l e s a nd C o n c e p t s
be part of a much larger class of behavior that Verbal Behavior and the Hidden Costs of Reward
includes shouting obscenities, acting up, hitting
Reinforcement may be obscured when human
or otherwise abusing the caregivers in the treat-
verbal and nonverbal behavior interact. For ex-
ment center, and any number of other responses
ample, instruction-following behavior is more
that might get attention (Lalli, Mace, Wohn, &
than the following of particular instructions; it is
Livezey, 1995). This tells us how important atten-
a higher-order class of behavior held together by
tion is to this child. We must consider a treatment
common contingencies (e.g., Shimoff & Catania,
program that uses attention to reinforce more ef-
1998). Following orders in the military is a prod-
fective and appropriate behavior, but the example
also reminds us that we cannot define response uct of extensive and powerful social contingen-
classes by what they look like. cies, often based on aversive consequences, but in
Related topographies within higher-order class- actual combat, the long-term contingencies that
es may also enter into hierarchies in which some maintain instruction following in general as a
subclasses are more likely than others. For ex- higher-order class may be pitted against the imme-
ample, with severe self-injurious behavior or other diate consequences of following a particular order
dangerous forms of problem behavior, functional (Skinner, 1969).
analyses may involve risks of injury. But they Verbal behavior is involved in the distinction
may also reveal precursors of dangerous behavior between intrinsic and extrinsic reinforcers. An in-
within such a hierarchy. Reducing the likelihood trinsic reward or reinforcer is one that has a natural
of the precursors may also reduce the likelihood of relation to the responses that produce it, whereas
the more dangerous forms of behavior that follow an extrinsic one has an arbitrary relation to those
them (Fritz, Iwata, Hammond, & Bloom, 2013). responses (e.g., music is an intrinsic consequence
Assessments of hierarchical response classes there- of playing an instrument, but the music teacher’s
fore can lead to function-based interventions that praise is an extrinsic one). Events presumed to
reduce the risks of functional analyses by reducing function as reinforcers because their function has
the likelihood of the most dangerous topographies been instructed have been called extrinsic reinforc-
within the hierarchy. ers (e.g., as when a child is told that it is important
The criterion for defining response classes is to earn good grades), but labeling them so does
function, and common consequences are the glue not guarantee their effectiveness. It has been ar-
that holds classes of behavior together. The larger gued that extrinsic consequences undermine the
class is held together by the common consequenc- effectiveness of intrinsic ones, and despite much
es of its members, just as the various topographies evidence to the contrary, the argument has per-
of a rat’s lever presses (left or right paw, both paws, sisted and continues to influence the use of op-
rump) are held together by the common conse- erant contingencies in schools and other settings
quence of producing food. But the human case is (Cameron, Banko, & Pierce, 2001; Cameron &
distinguished by the embedding of one response Pierce, 1994; Eisenberger & Cameron, 1996).
class within another. The self-injurious behavior In one experiment (Lepper, Greene, & Nisbett,
in the example above is embedded in the larger 1973), one group of children received gold stars for
class of attention-getting behavior. When a re- artwork such as finger painting; after the gold stars
sponse class seems insensitive to its consequences, were discontinued, children in this group did less
such as when the self-injurious behavior seems not artwork than those in a second group that never
to extinguish, we must entertain the possibility received gold stars. The gold stars, extrinsic rein-
that we have improperly specified the class, and forcers, were said to have undermined the intrinsic
that it is part of a larger class whose other mem- reinforcers, the natural consequences of painting.
bers continue to have the consequences it once The children had been told to earn the gold stars,
shared with them. The hierarchical structure of however, and the experiment did not test the stars’
some classes of behavior may sometimes make it effectiveness as reinforcers. There were no data to
appear that reinforcement is not working, but it show that children painted more when they got
may be working on a response class larger than the gold stars.
one in which we have been interested. When re- The claimed deleterious effects are only in-
inforcement seems not to be working, we should consistently demonstrable, and they are small
consider whether the response class in which we and transient when they do occur (Cameron et
are interested is part of another larger class (Cata- al., 2001; Cameron & Pierce, 1994). Problems are
nia, 1995). more likely to arise with extrinsic reward that is
Basic Operant Contingencies 53
not contingent on performance than with contin- bribes will no doubt eventually find that the child
gent reward (Eisenberger & Cameron, 1996). In complies only when a bribe is offered.
any case, if there is an effect, its transience and The child will not learn to initiate appropri-
small size are hardly consistent with the argument ate behavior if the initiation rests with the one
that extrinsic reinforcement may ruin the lives who offers the bribe. Over the long run, therefore,
of children. Nonverbal effects of reinforcers must compliance with bribes will probably interfere
be distinguished from the social contingencies with the effects of more constructive contingen-
that maintain the verbal governance of behavior. cies. If reinforcement works at all in such cases, it
When situations involve verbal behavior, there is a is in strengthening compliance with bribes, which
good chance that verbal governance will override is hardly the best way to make use of reinforcers.
more direct effects of reinforcement. When such unintended stimulus control devel-
ops, it is important to teach the parent to rein-
force compliance without explicitly stating the
Reinforcers versus Bribes
contingency, or at least to reinforce compliance
In the literature of the “hidden costs of reward,” both when the contingency is explicitly stated and
reinforcers have sometimes been equated with when it is not.
bribes (Kohn, 1993), but it is unlikely that the ar- As for the parent who has heard the language
rangements described as bribes by such critics of of bribes applied to the practice of reinforcement
the practice of reinforcement involve the direct and is therefore reluctant to deliver reinforcers, it
effects of reinforcers. The language of bribery has is crucial to teach that parent not to accompany
an extensive history in law and ethics as an offer the arrangement of contingencies for a child’s
of goods or favors in exchange for favorable treat- behavior with statements of those contingencies.
ment in business, politics, or other human endeav- And that is probably good advice for teachers and
ors. Critics of the practice of reinforcement have clinicians, too.
extended this language to the common parental
practice of specifying a consequence when asking
a child to do something (e.g., “If you put away your REINFORCER CLASSES
toys, you can watch television”). There are good AND REINFORCER‑SPECIFIC EFFECTS
reasons to advise parents against the practice of
bribery in this sense, but the reasons are different Operant contingencies involve consequences, and
from those offered by the critics. They have cor- like responses, they can profitably be studied in
rectly recognized the potentially different effects terms of classes (Cuvo, 2000). Successive reinforc-
of natural and artificial consequences, but they ers arranged in experimental settings are ordinar-
have also seriously conflated cases of verbal stimu- ily similar but not identical. For example, individ-
lus control with those involving other varieties of ual pieces of grain made available when a pigeon’s
contingencies. pecks operate its feeder will differ slightly in shape
Parents sometimes complain that their child and color; a parent’s hugs or smiles or positive
only cooperates with requests when there is an im- comments that reinforce a child’s behavior will
mediate and explicit payoff. This problem is one undoubtedly vary from one instance to the next.
of stimulus control. The parent may sometimes The discussion of higher-order classes has ex-
say, “It is time to put your toys away,” and at other amined the different responses that may produce
times may say, “If you put away your toys, you can attention and thereby maintain the self-injurious
watch television.” But unless the child who has behavior of children with severe developmental
complied with the request gets an opportunity to disabilities. Because it shares its consequences
watch television whether or not the contingency with other responses, such as shouting obscenities
has been explicitly stated, the child will learn to or throwing things, the self-injurious behavior may
comply only when the parent states it. be part of a larger class we might call attention-get-
Given that a bribe specifies behavior and its ting behavior. Within this class, some types of re-
consequences, offers of bribes instead function as sponses may be more probable than others or may
stimuli that set the occasion for particular con- be differently available in different settings (Lalli
tingencies. The child who is frequently bribed in et al., 1995). For example, a child may be more
this sense will learn to discriminate between con- likely to engage in self-injury if nothing to throw
ditions in which bribes are in effect and those in is close at hand, or more likely to shout obscenities
which they are not, so the parent who often uses given one audience than given another. Neverthe-
54 B a s i c P r in c ip l e s a nd C o n c e p t s
less, their membership in a common class makes it 2015). The moral is that whenever possible, we
likely that these responses will vary together as a should arrange different reinforcers rather than a
function of establishing events or other variables. single reinforcer for the maintenance or the shap-
But what if attention from staff members on this ing of different response classes.
child’s hospital unit does not function like atten-
tion from the child’s mother when she visits the
unit? If we find that one kind of attention cannot REINFORCEMENT AND CULTURAL SELECTION
substitute for the other, we might best treat atten-
tion from these two different sources as two sepa- The relations among behavior and its consequenc-
rate reinforcer classes (cf. Kelly, Roscoe, Hanley, & es in operant contingencies seem simple, but they
Schlichenmeyer, 2014). This is important to know, have subtle properties, some of which become
because assessments of problem behavior on the evident only in special contexts. For example,
unit may yield different results from those taken at when side effects are not taken into account, con-
the child’s home; therefore, therapeutic interven- tingencies can appear to be ineffective. Side ef-
tions shaped by staff attention on the unit may be fects of operant contingencies may have affected
incompatible with the kinds of behavior shaped their acceptance because they allow the effects of
by the mother’s attention at home. An effective contingencies to be masked in various ways. It is
treatment program must deal with the mother’s therefore prudent to consider the circumstances
behavior as well as the child’s, or the treatment in which the properties of operant contingencies
gains realized on the unit will be lost soon after may mislead us as we deploy them and evaluate
the child’s discharge. their effects. In the interests of preventing mis-
The significance of reinforcer classes has been conceptions and misunderstandings, it is prob-
especially demonstrated in research on the acqui- ably even more important to remind ourselves of
sition of arbitrary matching by children and by them whenever we present what we know about
nonhuman organisms (Dube & McIlvane, 1995; operant contingencies to those outside behavior
Dube, McIlvane, Mackay, & Stoddard, 1987; analysis. To those who argue that these contin-
Kastak, Schusterman, & Kastak, 2001; Urcuioli, gencies should not be studied because they can be
1991, 2005). Experiments on arbitrary matching misused, the appropriate rejoinder is that detailed
typically incorporate correction procedures and familiarity with their properties may be the best
other features that reduce the proximity of errors defense against their misuse. Alone or in combi-
to later reinforcers, and that reduce the likelihood nation, the factors considered here may sometimes
that the organism will attend to dimensions of the give the appearance that operant contingencies do
environment that are irrelevant to the task. Nev- not work. On examination, we might instead con-
ertheless, some children, as well as some pigeons, clude that they work more ubiquitously and more
learn slowly. profoundly than we had originally imagined.
In a typical matching study, all correct re- Phylogenic selection is Darwinian selection
sponses, whether to one comparison stimulus or as it operates in the evolution of species. Onto-
the other, produce the same reinforcer. But if the genic selection is operant selection as it operates
reinforcers as well as the stimuli and responses in the shaping of behavior within an individual
of the arbitrary matching tasks enter into func- lifetime. A third level of selection is sociogenic or
tional classes, this may be a mistake. While the cultural selection, which involves the selection of
contingencies may work to separate the different behavior as it is passed on from one individual to
matching classes, such as green peck given square another (Skinner, 1981). Selection at any one of
sample and red peck given circle sample, the com- these levels need not be consistent with selection
mon reinforcers may work to keep them together. at the other two. For example, it may not matter
With the matching task modified for children so how valuable one way of doing things is relative
that correct responses from the different problem to some other way, if one is easy to pass on from
classes each produce a different visual reinforcer one individual to another, whereas the other can
(e.g., different cartoon pictures displayed on a be passed on only with difficulty. The one that is
video monitor), the acquisition of accurate arbi- easier to pass on may spread quickly and come to
trary matching usually proceeds far more rapidly dominate in a culture relative to the other, even
than when all responses produce the same rein- if the other would be more beneficial in the long
forcer (e.g., Urcuioli, 2005; Urcuioli & Swisher, term.
Basic Operant Contingencies 55
plied analyses of human behavior (pp. 23–37). Reno, (2013). Experimental analysis of precursors to severe
NV: Context Press. problem behavior. Journal of Applied Behavior Analy-
Catania, A. C. (2001). Positive psychology and positive sis, 46, 101–129.
reinforcement. American Psychologist, 56, 86–87. Green, C. W., Reid, D. H., White, L. K., Halford, R. C.,
Catania, A. C. (2005). The nonmaintenance of behav- Brittain, D. P., & Gardner, S. M. (1988). Identify-
ior by noncontingent reinforcement. European Jour- ing reinforcers for persons with profound handicaps:
nal of Behavior Analysis, 6, 89–94. Staff opinion versus systematic assessment of prefer-
Catania, A. C. (2013). Learning (5th ed.). Cornwall-on- ences. Journal of Applied Behavior Analysis, 21, 31–43.
Hudson, NY: Sloan. Greer, B. D., Fisher, W. W., Saini, V., Owen, T. M., &
Charlop, M. H., Kurtz, P. F., & Casey, F. G. (1990). Using Jones, J. K. (2016). Improving functional communi-
aberrant behaviors as reinforcers for autistic children. cation training during reinforcement schedule thin-
Journal of Applied Behavior Analysis, 23, 163–181. ning: An analysis of 25 applications. Journal of Ap-
Cuvo, A. J. (2000). Development and function of conse- plied Behavior Analysis, 49, 105–121.
quence classes in operant behavior. Behavior Analyst, Hagopian, L. P., Crockett, J. L., van Stone, M., DeLeon,
23, 57–68. I. G., & Bowman, L. G. (2000). Effects of noncontin-
Donahoe, J. W., Burgos, J. E., & Palmer, D. C. (1993). A gent reinforcement on problem behavior and stimu-
selectionist approach to reinforcement. Journal of the lus engagement: The role of satiation, extinction, and
Experimental Analysis of Behavior, 60, 17–40. alternative reinforcement. Journal of Applied Behavior
Donahoe, J. W., & Vegas, R. (2004). Pavlovian condi- Analysis, 33, 433–449.
tioning: The CS–UR relation. Journal of Experimen- Herrnstein, R. J., & Hineline, P. N. (1966). Negative
tal Psychology: Animal Behavior Processes, 30, 17–33. reinforcement as shock-frequency reduction. Journal
Dube, W. V., & McIlvane, W. J. (1995). Stimulus–re- of the Experimental Analysis of Behavior, 9, 421–430.
inforcer relations and emergent matching to sample. Hineline, P. N. (1970). Negative reinforcement without
Psychological Record, 45, 591–612. shock reduction. Journal of the Experimental Analysis
Dube, W. V., McIlvane, W. J., Mackay, H. A., & Stod- of Behavior, 14, 259–268.
dard, L. T. (1987). Stimulus class membership estab- Hineline, P. N. (1981). The several roles of stimuli
lished via stimulus–reinforcer relations. Journal of the in negative reinforcement. In P. Harzem & M. D.
Experimental Analysis of Behavior, 47, 159–175. Zeiler (Eds.), Predictability, correlation, and contiguity
Eisenberger, R., & Cameron, J. (1996). Detrimental ef- (pp. 203–246). New York: Wiley.
fects of reward: Reality or myth? American Psycholo- Holz, W. C., & Azrin, N. H. (1961). Discriminative
gist, 51, 1153–1166. properties of punishment. Journal of the Experimental
Favell, J. E., McGimsey, J. F., & Jones, M. L. (1978). The Analysis of Behavior, 4, 225–232.
use of physical restraint in the treatment of self-inju- Iversen, I. H., & Lattal, K. A. (1991a). Experimental
ry and as positive reinforcement. Journal of Applied analysis of behavior: Part 1. Amsterdam: Elsevier.
Behavior Analysis, 11, 225–241. Iversen, I. H., & Lattal, K. A. (1991b). Experimental
Ferrari, M., & Harris, S. L. (1981). The limits and mo- analysis of behavior: Part 2. Amsterdam: Elsevier.
tivating potential of sensory stimuli as reinforcers for Iwata, B. A., Pace, G. M., Kalsher, M. J., Cowdery, G. E.,
autistic children. Journal of Applied Behavior Analysis, & Cataldo, M. F. (1990). Experimental analysis and
14, 339–343. extinction of self-injurious escape behavior. Journal
Ferster, C. B. (1958). Control of behavior in chimpan- of Applied Behavior Analysis, 23, 11–27.
zees and pigeons by time out from positive reinforce- Kastak, C. R., Schusterman, R. J., & Kastak, D. (2001).
ment. Psychological Monographs, 72(8, Whole No. Equivalence classification by California sea lions
461). using class-specific reinforcers. Journal of the Experi-
Fisher, W. W., Pawich, T. L., Dickes, N., Paden, A. R., mental Analysis of Behavior, 76, 131–158.
& Toussaint, K. (2014). Increasing the saliency of be- Keller, F. S. (1941). Light aversion in the white rat. Psy-
havior–consequence relations for children with au- chological Record, 4, 235–250.
tism who exhibit persistent errors. Journal of Applied Kelly, M. A., Roscoe, E. M., Hanley, G. P., & Schlichen-
Behavior Analysis, 47, 738–748. meyer, K. (2014). Evaluation of assessment methods
Fisher, W. W., Piazza, C. C., Bowman, L. G., Hagopian, for identifying social reinforcers. Journal of Applied
L. P., Owens, J. C., & Slevin, I. (1992). A comparison Behavior Analysis, 47, 113–135.
of two approaches for identifying reinforcers for per- King, A. P., & West, M. J. (1985). Social metrics of song
sons with severe and profound disabilities. Journal of learning. Learning and Motivation, 15, 441–458.
Applied Behavior Analysis, 25, 491–498. Kohn, A. (1993). Punished by rewards. Boston: Hough-
Fisher, W. W., Rodriguez, N., & Owen, T. (2013). Func- ton Mifflin.
tional assessment and treatment of perseverative Lalli, J. S., Mace, F. C., Wohn, T., & Livezey, K. (1995).
speech about restricted topics in an adolescent with Identification and modification of a response-class
Asperger syndrome. Journal of Applied Behavior Anal- hierarchy. Journal of Applied Behavior Analysis, 28,
ysis. 46, 307–311. 551–559.
Fritz, J. N., Iwata, B. A., Hammond, J. L., & Bloom, S. E. Lattal, K. A. (1974). Combinations of response rein-
Basic Operant Contingencies 57
forcer dependence and independence. Journal of the maintenance of avoidance behavior by the white rat.
Experimental Analysis of Behavior, 22, 357–362. Journal of Comparative and Physiological Psychology,
Lepper, M. R., Greene, D., & Nisbett, R. E. (1973). Un- 46, 253–261.
dermining children’s intrinsic interest with extrinsic Sidman, M. (1958). By-products of aversive control.
reward: A test of the “overjustification” hypothesis. Journal of the Experimental Analysis of Behavior, 1,
Journal of Personality and Social Psychology, 28, 129– 265–280.
137. Sizemore, O. J., & Lattal, K. A. (1977). Dependency,
Lerman, D. C., Iwata, B. A., & Wallace, M. D. (1999). temporal contiguity, and response-independent re-
Side effects of extinction: Prevalence of bursting and inforcement. Journal of the Experimental Analysis of
aggression during the treatment of self-injurious be- Behavior, 27, 119–125.
havior. Journal of Applied Behavior Analysis, 32, 1–8. Skinner, B. F. (1935). The generic nature of the con-
Mahoney, M. J., & Bandura, A. (1972). Self-reinforce- cepts of stimulus and response. Journal of General
ment in pigeons. Learning and Motivation, 3, 293–303. Psychology, 12, 40–65.
Maier, S. F., Albin, R. W., & Testa, T. J. (1973). Failure Skinner, B. F. (1938). The behavior of organisms: An
to learn to escape in rats previously exposed to ines- experimental analysis. New York: Appleton-Century-
capable shock depends on nature of escape response. Crofts.
Journal of Comparative and Physiological Psychology, Skinner, B. F. (1953). Science and human behavior. New
85, 581–592. York: Macmillan.
Michael, J. (1982). Distinguishing between discrimina- Skinner, B. F. (1957). Verbal behavior. New York: Apple-
tive and motivational functions of stimuli. Journal of ton-Century-Crofts.
the Experimental Analysis of Behavior, 37, 149–155. Skinner, B. F. (1969). Contingencies of reinforcement.
Michael, J. (1989). Concepts and principles of behavior New York: Appleton-Century-Crofts.
analysis. Kalamazoo, MI: Association for Behavior Skinner, B. F. (1981). Selection by consequences. Sci-
Analysis. ence, 213, 501–504.
Neuringer, A. (2004). Reinforced variability in animals Skinner, B. F. (1999). Cumulative record (V. G. Laties &
and people. American Psychologist, 59, 891–906. A. C. Catania, Eds.). Cambridge, MA: B. F. Skinner
Nevin, J. A. (1992). An integrative model for the study Foundation.
of behavioral momentum. Journal of the Experimental Solnick, J. V., Rincover, A., & Peterson, C. R. (1977).
Analysis of Behavior, 57, 301–316. Some determinants of the reinforcing and punishing
Pavlov, I. P. (1927). Conditioned reflexes (G. V. Anrep, effects of timeout. Journal of Applied Behavior Analy-
Trans.). London: Oxford University Press. sis, 10, 415–424.
Perone, M. (2003). Negative effects of positive rein- Urcuioli, P. J. (1991). Retardation and facilitation of
forcement. Behavior Analyst, 26, 1–14. matching acquisition by differential outcomes. Ani-
Peterson, N. (1960). Control of behavior by presentation mal Learning and Behavior, 19, 29–36.
of an imprinted stimulus. Science, 132, 1395–1396. Urcuioli, P. J. (2005). Behavioral and associative effects
Premack, D. (1962). Reversibility of the reinforcement of differential outcomes in discrimination learning.
relation. Science, 136, 255–257. Learning and Behavior, 33, 1–21.
Seligman, M. E. P., & Csikszentmihalyi, M. (2000). Urcuioli, P. J., & Swisher, M. J. (2015). Transitive and
Positive psychology: An introduction. American Psy- anti-transitive emergent relations in pigeons: Sup-
chologist, 55, 5–14. port for a theory of stimulus-class formation. Behav-
Seligman, M. E. P., & Csikszentmihalyi, M. (2001). ioural Processes, 112, 49–60.
Reply to comments. American Psychologist, 56, 89–90. Weiss, B., & Laties, V. G. (1961). Behavioral thermo-
Shimoff, E., & Catania, A. C. (1998). The verbal gov- regulation. Science, 133, 1338–1344.
ernance of behavior. In K. A. Lattal & M. Perone Wolf, M., Risley, T., & Mees, H. (1986). Application
(Eds.), Handbook of research methods in human operant of operant conditioning procedures to the behavior
behavior (pp. 371–404). New York: Plenum Press. problems of an autistic child. Behaviour Research and
Sidman, M. (1953). Two temporal parameters in the Therapy, 1, 305–312.
CHAP TER 4
Schedules of Reinforcement
Rules that describe the relation between responses motivate the occurrence of discriminated oper-
and reinforcer deliveries are known as schedules of ants are known as motivating operations (Laraway,
reinforcement. Researchers can arrange these rules Snycerski, Michael, & Poling, 2003). Motivating
deliberately in the context of an experiment or be- operations are events preceding occurrences of
havioral treatment, or they can surmise them from discriminated operants that can have evocative
the pattern of responses to reinforcer deliveries or abative effects on behavior; that is, they in-
that occur naturally. In either case, schedules of crease or decrease their occurrence. They alter
reinforcement are important for applied behavior the effectiveness of consequences of behavior
analysts to consider, because each schedule has by establishing or abolishing their reinforcing or
predictable effects on one or more dimensions of punishing effects. Discriminative stimuli are ante-
behavior. With this knowledge, applied behavior cedent occurrences of stimuli or events that cor-
analysts can describe the conditions of reinforce- relate with or signal the increased or decreased
ment that maintain undesirable behavior, and de- availability of reinforcement or punishment and
sign interventions that have a higher likelihood of affect discriminated operants. Behavior analysts
increasing desirable behavior. say that discriminative stimuli set the occasion for
This chapter discusses the dimensions of behav- operants to occur, because they predict the likely
ior that schedules of reinforcement can affect, and consequences of responses. However, the effects of
presents descriptions and examples of basic and discriminative stimuli on behavior depend on the
combined schedules of reinforcement. presence or absence of related motivating opera-
tions (Laraway et al., 2003).
In natural human environments, individuals
SCHEDULES OF REINFORCEMENT IN CONTEXT are free to engage in any of several concurrently
available discriminated operants. Each one pro-
Understanding the effects schedules of reinforce- duces one or more consequences, and the individ-
ment have on behavior in the broader context in ual has a choice to engage in any alternative (Fisher
which they operate is important. The basic unit & Mazur, 1997; Mace & Roberts, 1993). Investi-
of analysis in applied behavior analysis (ABA) is gators have studied the variables that influence
the discriminated operant, which is a class of re- choice extensively and have quantified them in
sponses defined by both the effect the responses the generalized matching law (Baum, 1974), which
have on the environment and the stimuli present we briefly discuss later. However, one of the vari-
when responses occur (Catania, 2013). Events that ables that affects choice is the relative schedule
58
Schedules of Reinforcement 59
of reinforcement operating for each concurrently general types of contingencies. First, a ratio contin-
available alternative. We also provide a concep- gency is one between the number of responses that
tual framework for understanding how changes occur and the delivery of a reinforcer. Second, an
in relative motivation and relative history of re- interval contingency is one between the occurrence
inforcement or punishment influence concurrent of responses and the passage of time. Third, a time
discriminated operants in a dynamic manner (i.e., contingency is the passage of time with no relation
behavioral mass; Mace, Gritter, Johnson, Malley, to the occurrence of responses. The arrangement
& Steege, 2007; see Figure 4.1). Learning histories of ratio, interval, and time contingencies can af-
affect the relative resistance to change or momen- fect the response’s rate, periodicity, and resistance
tum each discriminated operant has, and can af- to change.
fect the relative value of concurrently available
alternatives and the choices individuals make.
Response Rate
Nevin and Grace (2000) refer to these histories as
behavioral mass in the context of their formulation Response rate is the ratio of the number of respons-
of behavioral-momentum theory. es per time, or response count divided by time. Its
The purpose of this preface to our discussion synonym is frequency. An alternative expression of
of schedules of reinforcement is to emphasize that response rate is the average time between respons-
the effects of reinforcement schedules on behav- es, or interresponse time (IRT), which is the mean
ior are relative, not absolute (Herrnstein, 1961, for the elapsed time between the offset of one re-
1970). That is, the influence of any given schedule sponse and the onset of the next response in the
of reinforcement on a discriminated operant will response class. Response rate and IRT have a per-
depend on the relative factors that affect choice, fect inverse correlation, such that a unit increase
including reinforcer rate, quality, and delay, as in response rate will have a corresponding unit
well as response effort, motivation, and behavioral decrease in interresponse time. In ABA practice,
mass. The practical significance of this conceptual response rate is often an important dimension that
model is that applied behavior analysts may need the behavior analyst aims to alter—generally by
to consider many factors that influence both de- attempting to increase the frequency of desirable
sirable and undesirable behavior to maximize the behaviors, to decrease the frequency of undesir-
effectiveness of their interventions. able ones, or both. Thus knowing the effects of
different schedules of reinforcement on response
rate is important in the design of interventions.
BEHAVIOR DIMENSIONS
AFFECTED BY SCHEDULES OF REINFORCEMENT Response Periodicity
In ABA practice, clinicians and researchers de- Response periodicity is the pattern of responses in
liver reinforcers contingently. There are three relation to the passage of time. Whereas interre-
MO MO MO MO
}
}
}
FIGURE 4.1. Framework for conceptualizing discriminated operants in context: The dynamic interplay between moti-
vating operations, behavioral mass, and the classic three-term contingency.
60 B a s i c P r in c ip l e s a nd c o n c e p t s
sponse time expresses the average time between onstrated the relevance of these schedules to the
responses, response periodicity is the pattern of assessment and treatment of human behavior.
times between individual responses or individual Basic schedules are single schedules of rein-
IRTs during a period. In general, schedules of re- forcement applied to one class of responses, which
inforcement promote four patterns of response form the building blocks for the more complex,
periodicity. The first pattern is a constant time combined schedules of reinforcement. Table 4.1
between responses with little variability in indi- summarizes the basic schedules discussed in this
vidual IRTs, which is characteristic of variable- chapter, schedule definitions, the response pat-
ratio (VR) and variable-interval (VI) schedules. terns each schedule promotes, and applications of
A second pattern is a pause in responding after a the schedule in ABA work.
comparatively higher response rate. Both fixed-ra-
tio (FR) and fixed-interval (FI) schedules can pro-
Ratio Schedules
mote temporary pauses in responding after rein-
forcement delivery before responding resumes. A Ratio schedules of reinforcement specify the num-
third pattern is the suspension of responding after a ber of responses the organism must emit to effect
given response pattern. Time, extinction, and dif- delivery of reinforcement, independent of the time
ferential reinforcement of other behavior (DRO) the organism takes to complete the schedule re-
can produce reductions in response rates to zero quirement. However, because slow response rates
or near-zero. Finally, response periodicity can show delay the time to reinforcement, ratio schedules
a pattern of celeration (Johnston & Pennypacker, generally promote relatively high response rates
2009), which refers to a progressive change in in- with relatively constant individual IRTs, with
dividual IRTs. Progressively shorter IRTs reflect ac- some exceptions noted below. Two schedule fea-
celeration in response rate, and progressively longer tures influence the response patterns promoted by
IRTs reflect deceleration. Many schedules of rein- ratio schedules: (1) the ratio of responses to rein-
forcement can promote these response patterns, as forcers and (2) the predictability of this ratio.
we discuss in the following sections.
FR Schedules
Resistance to Change
In an FR schedule, the number of responses re-
An important dimension of behavior that applied quired to produce a reinforcer is constant (e.g.,
behavior analysts increasingly consider is its resis- inputting a three-digit area code and seven-digit
tance to change, which is the rate of deceleration in phone number to make a call is an example of an
responding after the introduction of some response FR 10 schedule). When the ratio of responses to
disruptor (Nevin, 1974; Nevin, Mandell, & Atak, reinforcers is very low, as in the case of the FR
1983). Common events that disrupt the response– 1 schedule or continuous-reinforcement (CRF)
reinforcer relation include extinction, satiation, schedule, responses rates are typically low. How-
alternative reinforcement, punishment, dark-key ever, as reinforcer deliveries become less frequent
or between-session reinforcement, and distraction. (e.g., FR 5), the response rates the schedule pro-
Each operation can decelerate responding. This motes increase rapidly and eventually support
dimension of behavior is particularly relevant to comparatively high rates of responding. As the
ABA work aimed at strengthening the resistance ratio of responses to reinforcers increases, pauses
to change of desirable behavior and weakening the in responding after the reinforcer delivery, or
resistance to change of undesirable behavior. postreinforcement pauses, also increase (Felton &
Lyon, 1966). Finally, as the ratio of responses to
reinforcers becomes comparatively high, pauses in
BASIC SCHEDULES OF REINFORCEMENT responding can appear before the reinforcement
delivery. This is ratio strain, which can produce
Ferster and Skinner (1957) provided the founda- either the temporary interruption of responding
tional work for schedules of reinforcement in their or its cessation. The predictability of the ratio of
compilation of over 100 experimental demonstra- responses to reinforcers in FR schedules generally
tions of the patterns of responding promoted by promotes the highest response rates with uniform
various schedules. Numerous applications of vari- individual IRTs. However, humans may show FR
ous schedules of reinforcement in ABA have dem- response patterns even when the ratio of responses
Schedules of Reinforcement 61
TABLE 4.1. Basic Schedules of Reinforcement, ABA Example, and the Response Patterns Promoted
Schedule Definition ABA example Response pattern promoted
FR—fixed ratio Reinforcers are contingent on Cohen et al. (2001) High response rate with
every ith response (e.g., FR demonstrated FR patterns comparatively short and
4—every fourth response) by measuring muscle uniform IRTs. Pause in
contractions emitted by responding follows reinforcer
undergraduate students. deliveries.
VR—variable ratio Reinforcers are contingent DeLuca and Holburn (1992) High response rate with
on a variable number of showed VR patterns by comparatively short and
responses; the average number measuring obese children’s uniform IRTs.
of responses defines the rate of stationary bike
schedule. revolutions.
FI—fixed interval Reinforcers are contingent on Critchfield et al. (2003) Possible cumulative record
the first response following a analyzed the bill-passing scalloping when measurement
fixed time interval. behavior of the U.S. Congress. of the passage of time is
unavailable.
VI—variable Reinforcers are contingent on Martens, Lockner, and Moderate response rates with
interval the first response following Kelly (1992) demonstrated uniform but longer IRTs than
a variable interval of time; VI response patterns in the ratio schedules.
the average of these intervals academic engagement of
defines the schedule. typically developing 8-year-
olds.
FT–VT—fixed Reinforcers are delivered Vollmer et al. (1998) used FT When combined with EXT,
or variable time independently of any behavior schedules to reduce problem sharp drop in response rate to
schedules at FT or VT intervals. behaviors (i.e., aggression, self- near-zero or zero levels. When
injurious behavior, disruption, combined with ratio or interval
and tantrums) displayed by schedules, a reduction in the
both children and adults with reinforced class of behaviors.
intellectual disability disorder.
Mace and Lalli (1991) used
VT schedules to reduce
bizarre vocalizations emitted
by an adult with moderate
intellectual disability disorder.
DRA—differential Reinforcers are contingent Harding et al. (2004) used Comparatively higher response
reinforcement of on specific topographies of DRA schedules to increase rates for behaviors that produce
alternative behavior behavior and not others. adults’ correct execution reinforcers than for those that
Combines ratio or interval of various martial arts do not.
schedules with extinction. techniques.
(continued)
62 B a s i c P r in c ip l e s a nd c o n c e p t s
DRO—differential Reinforcers are contingent Heard and Watson (1999) Low or zero rates of the
reinforcement of on the absence of specified used an interval DRO target behavior omitted from
other behavior behavior(s) during a specified schedule to reduce wandering reinforcement. Behaviors
time interval. Also called behavior exhibited by geriatric other than the target behavior
omission training. patients. Kahng, Abt, and increase in rate.
Schonbachler (2001) used a
momentary DRO schedule to
reduce the rate of aggression
displayed by a woman with
developmental disabilities.
to reinforcers is not constant. For example, a par- the number of responses the organism emits. How-
ent may tell a child that he or she may engage in ever, in VR schedules, the interreinforcement re-
a leisure activity after the child completes a math sponse criterion varies for each reinforcer delivery.
homework assignment. Because the child knows The schedule value is the average ratio of responses
how many math problems he or she must com- to reinforcers over the course of the VR condition.
plete, the ratio of responses to reinforcers is pre- For example, reinforcers delivered after the second
dictable, and the characteristic FR response pat- response, then after the sixth response, and then
tern may occur. after four more responses would be a VR 4 sched-
Cohen, Richardson, Klebez, Febbo, and Tucker ule. Examples of human behaviors maintained
(2001) provided undergraduate psychology majors by VR schedules include sampling restaurants in
with auditory and visual biofeedback for electro- search of ones that suit one’s taste, purchasing lot-
myography readings from their forearms. Investi- tery tickets, looking for misplaced items, and an-
gators instructed participants to alternately tense swering a teacher’s questions.
and relax their forearms. Different groups of stu- VR schedules generally promote high rates of re-
dents received feedback on whether their electro- sponding, with short and uniform individual IRTs.
myography values moved from below a threshold However, when the ratio of responses to reinforc-
to above the threshold, which were the responses. ers exceeds a threshold, response rates decline as
Investigators provided feedback for these responses the ratio of responses to reinforcers increases. This
according to five different schedules of reinforce- value can exceed 200 key pecks per minute in pi-
ment. Two of the schedules were FR schedules: FR geons (Brandauer, 1958). In addition, ratio strain
1 and FR 4. The FR 4 schedule generated the high- can occur at lower ratios of response to reinforcer
est rates of responding, whereas the FR 1 schedule values when single interreinforcement intervals
produced response rates that were approximately become quite large. Finally, the unpredictability
half the higher ratio of responses to reinforcers. of the individual ratio of responses to reinforcers
tends to promote short postreinforcement pauses.
DeLuca and Holburn (1992) reinforced revolu-
VR Schedules
tions on an exercise bicycle with preferred objects
Like FR schedules, VR schedules or random-ratio on a VR schedule for three obese children. After
(RR) schedules deliver reinforcers contingent on calculation of baseline pedal revolutions per min-
Schedules of Reinforcement 63
ute, the investigators set an initial VR schedule Researchers have speculated whether FI scal-
value at 15% above baseline. Investigators imple- loping occurs in humans in natural environments.
mented two additional increases in VR values For example, Critchfield, Haley, Sabo, Colbert,
after participants demonstrated stability in each and Macropoulis (2003) analyzed the bill-passing
VR condition. The progressive arrangement of behavior of the U.S. Congress over a 52-year pe-
the VR schedule value, also known as a progressive riod. Each 2-year Congress comprises two sessions
VR schedule, produced an approximate doubling of of approximately equal duration. Critchfield et al.
pedal revolution rates. graphed bill passing in cumulative records, which
showed that bill passing escalated toward the end
of each session, in accord with the FI scalloping in
Interval Schedules
laboratory experiments with nonhuman species.
Interval schedules of reinforcement define the However, FI scalloping is a productive response
time that responses become eligible for reinforce- pattern only when there is no external means of
ment. The interval value indicates the minimum discriminating time. When the preparation clearly
time that must elapse before a response produces signals the end of an FI schedule, the most parsi-
reinforcement. Unlike ratio schedules, the rate of monious response pattern would be to emit a sin-
responding does not affect the rate of reinforce- gle response at the end of the interval. In the case
ment; response rate and reinforcement rate are of Congressional behavior, the contingency would
independent, in that higher response rates do not appear to be between the number of bills passed
produce higher rates of reinforcement in interval and the production of the reinforcing consequence
schedules. The contingency is between the re- of reelection or campaign donations, rather than
sponse periodicity and the delivery of reinforce- the timing of the completion of any one legislative
ment. Thus interval schedules generally support act. The consensus of other investigators is that FI
lower response rates than ratio schedules (Cohen scalloping in humans is a rare occurrence (Hyten
et al., 2001). & Madden, 1993; Ninness, Ozenne, & McCuller,
An adjunctive procedure, limited hold, is needed 2000).
sometimes for interval schedules to support a con-
sistent response pattern. A limited hold specifies
VI Schedules
the time that reinforcement is available once the
schedule sets it up. Thus a 5-second limited hold VI schedules or random-interval (RI) schedules
means that if a response does not occur within 5 make responses eligible for reinforcement based
seconds of becoming eligible for reinforcement, on a mean time interval that elapses. For example,
the organism forfeits the opportunity for rein- a VI 15-second schedule would program reinforce-
forcement. This procedure promotes a consistent ment delivery after the 10th second, then after 20
pattern of behavior that maximizes obtaining the more seconds, and then after 15 more seconds.
reinforcement that is available in a period. Hantula and Crowell (1994) provided a BASIC
program for deriving interval values based on the
formula by Fleshler and Hoffman (1962), such
FI Schedules
that the time between reinforcer deliveries is truly
In an FI schedule of reinforcement, the first re- random. Because the interreinforcement interval
sponse that occurs after the expiration of a con- is unpredictable, VI schedules promote consistent
stant time produces reinforcement. For example, response rates with uniform individual IRTs.
reinforcement is available every 5 minutes on an Teachers often deliver attention on a VI sched-
FI 5-minute schedule. The predictability of rein- ule based on when they can deliver attention,
forcement availability can promote a pattern of rather than on the number of responses a stu-
behavior that is sensitive to this periodicity. In dent makes. Martens, Lochner, and Kelly (1992)
the laboratory, especially with nonhuman partici- provided praise contingent on academic engage-
pants, FI schedules tend to promote an escalation ment for two 8-year-old students with low baseline
in response rates toward the end of the interval. rates of academic engagement. The investigator
When expressed in a cumulative record of re- delivered praise alternately on VI 5-minute and
sponses, the pattern has a scalloped appearance. A VI 2-minute schedules. Both schedules improved
postreinforcement pause that can last more than academic engagement; however, the VI 2-minute
half the interreinforcement interval occurs after schedule consistently produced higher levels of en-
delivery of each reinforcer. gagement for both students—a finding consistent
64 B a s i c P r in c ip l e s a nd c o n c e p t s
with basic research with nonhuman species (Cata- or emotional behavior, resumption of previously re-
nia & Reynolds, 1968). inforced behaviors, behavioral contrast (increased
occurrences of undesirable behavior in settings
in which individuals do not use extinction), and
Extinction Schedules
spontaneous recovery (recurrence of the extin-
Extinction schedules withhold reinforcement for guished target behavior). Lerman, Iwata, and Wal-
specified response topographies during certain lace (1999) examined the prevalence of extinction
times. Extinction schedules come in two general bursts and extinction-induced aggression for 41
forms. First, we can arrange an extinction sched- individuals with self-injurious behavior who re-
ule in which we discontinue reinforcement for re- ceived treatment using extinction alone or extinc-
sponses that we reinforced in the past. Scientists tion plus additional treatment components. They
use EXT+ to denote extinction that follows posi- found that response bursting was evident in 39%
tive reinforcement, and EXT– to denote extinc- of cases, 22% showed increased aggression, and
tion following negative reinforcement. Second, we 58% showed neither side effect. However, of the
can arrange an extinction schedule for responses cases treated with extinction alone, 69% showed
we have not reinforced explicitly in the past, and response bursting and 29% showed increased ag-
we should not reinforce during the process of gression, compared to only 15% for either side
teaching a new skill. For example, when teaching effect when treated with extinction plus another
a child with autism spectrum disorder (ASD) to treatment component. These findings suggest the
say the word apple, we would not reinforce vocal- importance of combining extinction with other
ized sounds unrelated to a, p, and l. schedules to avoid unwanted side effects.
Magee and Ellis (2001) used extinction alone
to reduce multiple undesirable behaviors for two
Time Schedules
children. A functional analysis showed that one
child’s out-of-seat behavior occurred at high lev- Time schedules arrange occurrences of reinforcer
els when a therapist discontinued instructions deliveries contingent on the passage of time and
contingent on the behavior. Escape extinction independently of behavior. These reinforcers are
(EXT–) consisted of the continuous presentation response-independent and delivered noncontingent-
of instructions every 10 seconds, regardless of oc- ly.1 In ABA, investigators and clinicians use time
currences of undesirable behavior. The second schedules (1) to enrich an environment and alter
child’s functional analysis showed that adult at- the motivation of individuals to engage in unde-
tention maintained his object mouthing. Positive sirable behavior to obtain reinforcement (Horner,
reinforcement extinction (EXT+) for this child 1980); (2) to serve as an experimental control pro-
involved withholding attention following any cedure to demonstrate the effects of a contingency
undesirable responses. Both extinction schedules between a response and a reinforcer (Thompson
were effective in reducing out-of-seat behavior and & Iwata, 2005); and (3) to reduce undesirable be-
hand mouthing. However, Magee and Ellis found havior. Our discussion here focuses on this last ap-
that when they placed these two behaviors on ex- plication of time schedules.
tinction, other topographies of undesirable behav-
iors emerged—first yelling for one child, and then
Fixed‑ or Variable‑Time Schedules
property destruction for the other. When Magee
and Ellis placed these behaviors on extinction, We can deliver reinforcers in time schedules at
additional topographies of undesirable behaviors regular or fixed intervals (FT schedules) or at ran-
began occurring. The sequential emergence of dom or variable intervals (VT schedules). Numer-
multiple undesirable behaviors demonstrated that
the behaviors were members of a response class 1 Researchers introduced the term noncontingent reinforce-
hierarchy (Lalli, Mace, Wohn, & Livezey, 1995). ment (NCR) to describe fixed-time (FT) and variable-
Magee and Ellis (2001) illustrated that extinc- time (VT) schedules (Vollmer, Iwata, Zarcone, Smith, &
tion schedules can have collateral effects, in ad- Mazaleski, 1993). However, Poling and Normand (1999)
dition to the reduction of responses subject to questioned the technical accuracy of the term to describe
the noncontingent delivery of reinforcement. Because non-
extinction. These include the extinction burst contingent delivery of reinforcement does not constitute a
(initial increases in responding), extinction-in- reinforcement operation or process (i.e., nothing is rein-
duced aggression (violent acts both related and forced), the term NCR does appear to be a misnomer, and
unrelated to the source of reinforcement), agitated we do not use it in our discussion of the topic.
Schedules of Reinforcement 65
ous studies have demonstrated that time schedules response bursting when they have faded the rate of
are effective and efficient for reducing many un- reinforcement delivery via time schedules, and it
desirable behaviors maintained by both positive apparently becomes too lean. Vollmer, Ringdahl,
and negative reinforcement (Vollmer, Marcus, & Roane, and Marcus (1997) found that undesirable
Ringdahl, 1995; for reviews, see Saini, Miller, & behavior escalated to approximately five times the
Fisher, 2016; Vollmer & Hackenberg, 2001; Wal- baseline rate during FT schedule thinning; Mace
lace, Iwata, Hanley, Thompson, & Roscoe, 2012). et al. (2008) reported similar findings.
Investigators have used time schedules to reduce There are several procedural variations of time
the duration and the frequency of exposure to schedules to consider in designing interventions.
aversive behavioral intervention. For example, First, the investigator or clinician must select
Luiselli, Pace, and Dunn (2006) compared behav- a time-schedule value that is sufficiently dense
ior-contingent release from therapeutic physical to suppress undesirable behavior. For example,
restraint with FT release from restraint to reduce Ringdahl, Vollmer, Borrero, and Connell (2001)
aggressive behavior. The FT release substantially evaluated the effectiveness of initial time-schedule
reduced the duration of individual restraints and values when they were similar or dissimilar to the
in some instances the frequency with which the rates of baseline reinforcement. They found that
investigators applied restraint. FT schedule values that were like baseline rates
In an interesting variation of time schedules to of reinforcement for undesirable behavior were
reduce undesirable behavior, Marsteller and St. less effective than those that were dissimilar. This
Peter (2014) effectively reduced problem behav- finding held even when the FT schedule values
ior maintained by social contingencies and in- were four to nine times leaner than the baseline
creased an alternative prosocial behavior, using a reinforcement rates. This counterintuitive finding
differential reinforcement of alternative behavior may be attributed to dissimilar rates being easier
(DRA) schedule with four children with disabili- to discriminate from baseline. A second proce-
ties. However, after effective DRA treatment, they dural question is whether time schedules must be
compared the effects of extinction to FT reinforce- used in conjunction with extinction to be effec-
ment. Extinction after DRA produced a marked tive. Lalli et al. (1997) compared FT schedules
resurgence of problem behavior. By contrast, the with and without extinction and found that they
FT schedule after DRA treatment prevented re- were comparably effective at reducing undesirable
surgence despite the discontinuation of the con- behavior; however, this finding was based on only
tingency between reinforcement and prosocial one participant receiving FT intervention with-
alternative behavior. out extinction. Third, most clinical studies using
Time schedules are an attractive treatment time schedules have evaluated FT rather than
alternative for several reasons. First, they often VT schedules, but studies comparing the efficacy
produce rapid suppression of undesirable behav- of FT versus VT schedules have found similar ef-
ior when the reinforcement maintaining undesir- fectiveness (Carr, Kellum, & Chong, 2001; Van
able behavior is time-contingent (Lalli, Casey, & Camp, Lerman, Kelley, Contrucci, & Vorndran,
Kates, 1997; Mace & Lalli, 1991). Second, adding 2000). We suggest the initial use of FT schedules
time-contingent reinforcement to a context can and a shift to VT schedules after initial treatment
reduce the motivation to engage in extreme acts, effects are established. The predictability of rein-
such as undesirable behavior, to obtain reinforce- forcement delivery in FI and VI schedules and the
ment. This also may increase the probability of characteristic response patterns they promote may
prosocial alternative behaviors that require less logically extend to time schedules.
effort to effect reinforcement (Ecott & Critch- There is some theoretical interest in which
field, 2004). Third, relative to extinction sched- behavioral process or processes are invoked in
ules, time schedules often obviate an extinction time schedules to make them effective. The shift
burst. For example, Vollmer et al. (1998) com- from contingent baseline reinforcement to time-
pared FT deliveries of maintaining reinforcement contingent reinforcement delivery involves two
with withholding that reinforcement (i.e., extinc- simultaneous operations. First, the response–re-
tion). A burst was evident during extinction for inforcement contingency in baseline is discontin-
the three participants in their study. By contrast, ued, constituting a procedural variation of extinc-
the FT schedule produced rapid or immediate sup- tion. Second, the motivating operations change by
pression of undesirable behavior without response supplying reinforcement on a time schedule. This
bursting. We note that investigators have reported presumably abolishes the consequence as effective
66 B a s i c P r in c ip l e s a nd c o n c e p t s
reinforcement for undesirable behavior and abates include specific forms or topographies of responses,
those same responses. Kahng, Iwata, Thompson, response rates and the periodicity of responding,
and Hanley (2000) examined response patterns in and the time spent engaging in specific behaviors.
the time immediately after FT intervention. They In ABA practice, the behavior analyst may
reasoned that if FT effects were due to extinction, change the criteria for reinforcement systemati-
responding would not resume after FT treatment, cally to promote gradual and progressive changes
because they did not reinstate the response–rein- in responding toward a target criterion. When this
forcement contingency. Alternatively, if FT effects involves the discriminative stimuli correlated with
were the result of altered motivating operations, reinforcement, the operation is known as fading.
they could expect response rates to resume when For example, Flood and Wilder (2004) used dif-
reinforcement shifted from being available to un- ferential reinforcement and fading to increase the
available. The findings of Kahng et al. were mixed, time an 11-year-old boy with separation anxiety
with one participant each supporting the extinc- disorder spent away from his mother without dis-
tion and motivating-operations accounts, and a tress behavior, such as crying, whining, or other
third showed a change in response patterns over emotional behavior. The investigators provided
time, from supporting the motivating-operations access to preferred items contingent on the boy’s
account to the extinction account. Finally, Ecott meeting his goals for time spent away from his
and Critchfield (2004) suggested that time sched- mother without distress behavior. The investiga-
ules may be effective because reinforcement deliv- tors faded the time goals from 3 minutes to 90
ery may coincide temporally with other behaviors minutes over 27 treatment sessions. The inves-
and result in adventitious reinforcement of those tigators used a similar fading and differential-
responses. In a laboratory demonstration with reinforcement procedure to increase the distance
undergraduate students, the investigators concur- the boy’s mother was from the therapy room.
rently reinforced two behaviors with points. The When changes in specific response topogra-
investigators reinforced target behavior on a VI phies or in response rates are subject to changing
10-second schedule and the alternative behavior criteria for reinforcement, the operation is known
on a VI 30-second schedule (see discussion of con- as shaping. Gillis, Hammond Natof, Lockshin, and
current schedules of reinforcement below). After Romanczyk (2009) used differential reinforcement
responding was stable in this baseline phase, they and gradual exposure with 18 children with ASD
systematically varied the proportion of reinforce- who showed extreme fear of medical procedures to
ment delivery for the target behavior that was increase cooperation with physical examinations.
response-contingent from 100 to 66 to 33 to 0%. The school nurse used a 17-step procedure to ex-
Results showed that as the proportion of time-con- pose the children gradually to the examination
tingent reinforcement delivery increased, the re- room and to the medical instruments, such as a
sponse rates for the alternative behavior increased. stethoscope, a sphygmomanometer, a thermom-
Ecott and Critchfield suggested that adventitious eter, and an otoscope. The nurse delivered various
reinforcement of alternative behavior is one pos- positive reinforcers for completion of each step in
sible account of the behavioral process involved in which a child showed minimal or no fear respons-
treatment effects of time schedules. es. Eighty-three percent of the children completed
the 17 steps showing minimal or no fear responses.
Tiger and Hanley discuss differential-reinforce-
Differential‑Reinforcement Schedules
ment procedures and their application more fully
Differential schedules of reinforcement specify in Chapter 14 of this book.
the dimensions of behavior that are and are not
eligible for reinforcement. They also may define
the stimuli that must be present when responses COMBINED SCHEDULES OF REINFORCEMENT
produce reinforcement (i.e., an SD), and the stim-
uli that must be present when responses do not Combined schedules of reinforcement comprise
produce reinforcement (i.e., an SD), thus defining two or more basic schedules. Each basic schedule
the discriminated operant. As such, differential- is a schedule component. Investigators may arrange
reinforcement schedules implicitly involve two these components to alternate, to be available at
types of operations: (1) positive or negative rein- the same time, to occur in a sequence, or in some
forcement and (2) extinction. The dimensions of combination of these arrangements. Combined
behavior subject to reinforcement and extinction schedules are particularly relevant to ABA, be-
Schedules of Reinforcement 67
cause they better represent the circumstances ternative methods of saying no to a child’s request
humans encounter in everyday life. Our coverage for tangible reinforcement. The child approached
of combined schedules includes definitions, ex- an adult using a computer and requested to play a
amples, and a discussion of the relation between computer game. During a functional analysis, de-
certain combined schedules and contemporary de- nied access to the computer produced high rates
velopments in ABA, such as behavioral contrast, of aggressive and disruptive behavior. During
matching theory, and behavioral momentum. treatment, a single therapist in a constant setting
Table 4.2 summarizes the combined schedules denied computer access, but alternately offered
discussed here, schedule definitions, response pat- another reinforcing activity or delayed access to
terns promoted by each schedule, and applications reinforcement contingent on completing a non-
in ABA work. preferred math worksheet. Use of a single thera-
pist in a constant setting did not correlate distinct
stimuli with the two alternative methods of deny-
Multiple and Mixed Schedules
ing computer access, making the arrangement a
In multiple and mixed schedules, two or more mixed schedule.
schedule components alternate in a random, coun- We (Pritchard, Hoerger, Mace, Penney, & Har-
terbalanced, or natural-temporal sequence. The ris, 2014) used a multiple schedule to evaluate the
difference between multiple and mixed schedules effects of different rates of reinforcement on treat-
is that schedule components are correlated with ment relapse. A functional analysis showed that
distinct stimuli in multiple schedules but not in attention maintained an adolescent’s problem be-
mixed schedules. As an individual experiences the havior. Treatment consisted of withholding atten-
multiple schedules, the correlated stimuli acquire tion for problem behavior and providing attention
stimulus control over responding and become dis- for communication responses or providing atten-
criminative stimuli. To the extent that the sched- tion on a VI VT schedule. One therapist provided
ule components differ, differential responding in VI VT attention at a rate four times greater than
the schedule components usually occurs more rap- that of a second therapist. Both rates of reinforce-
idly and is more pronounced in multiple schedules ment reduced problem behavior, but when the two
than in mixed schedules. therapists discontinued treatment and provided
Humans encounter multiple schedules regular- reinforcement for problem behavior at equal rates,
ly. Students in school who attend a sequence of the magnitude of treatment relapse was much
classes throughout the day experience a multiple higher for the therapist who used the higher rates
schedule. Each class is a schedule component and of VI VT reinforcement during treatment.
is correlated with distinct stimuli, such as different Tiger and Hanley (2005) used multiple and
teachers, textbooks, classrooms, and seating ar- mixed schedules to study variables that promoted
rangements. The teachers in each class undoubt- discriminative control of social-approach respons-
edly reinforce students’ participation in classroom es in preschool children. Two children sat facing
activities on different schedules; some use ratio the experimenter at tables containing academic
schedules, others use interval schedules, and still materials. The experimenter looked down except
others use differential reinforcement of high-rate when he or she delivered 5 seconds of attention
behavior (DRH) schedules. Humans also encoun- contingent on social-approach responses. The ex-
ter mixed schedules frequently. The first time we perimenter alternated presentation of three sched-
read a novel, watch a film, or drive through unfa- ule components in a randomized order to each
miliar countryside, our points of interest (i.e., the child. In the FR 1 component, the experimenter
availability of reinforcement) for attention to the provided reinforcement in the form of attention for
activity vary from one time to another. However, each social-approach response. In the extinction1
there is no indication that the reinforcing proper- component, the experimenter provided reinforce-
ties of the novel, film, or drive are about to shift. ment for one child’s social approaches but not for
Because we usually do not repeat these activities, the other child’s. In the extinction2 component,
any stimuli correlated with changes in reinforce- the experimenter did not provide reinforcement
ment do not develop stimulus control (i.e., the for either child’s social approaches. In the mul-
mixed schedule does not become a multiple sched- tiple schedule, denoted MULT FR 1 extinction1
ule). extinction2, the experimenter wore a different-
Mace, Pratt, Prager, and Pritchard (2011) used a colored floral lei during each component. The ex-
mixed schedule to evaluate the effectiveness of al- perimenter did not wear the leis during the mixed
68 B a s i c P r in c ip l e s a nd c o n c e p t s
TABLE 4.2. Combined Schedules of Reinforcement, ABA Example, and the Response Patterns Promoted
Schedule Definition ABA example Response pattern promoted
Multiple/mixed Alternation of two or more Tiger and Hanley (2005) Schedule-specific patterns of
schedules schedules of reinforcement. In a used MULT and MIX FR behavior are observed in each
multiple schedule, each schedule is 1 EXT1 EXT2 schedules to component. Schedule-specific
correlated with a distinct stimulus produce differential rates of response patterns are more
(e.g., a VR schedule in context A social approach responses pronounced in multiple than in
alternates with a DRL schedule in emitted by preschoolers. mixed schedules.
context B). In a mixed schedule,
no distinct stimuli are correlated
with each schedule (e.g., a VR
schedule and DRL schedule
alternate in the same context).
Concurrent/ Two or more schedules of Conger and Killeen (1974) Concurrent interval schedules
conjoint reinforcement are available at employed CONC VI VI promote allocation of
schedules the same time. In a concurrent schedules to demonstrate responding to each schedule
schedule, each schedule is college students’ allocation of in proportion to relative rates
correlated with a distinct stimulus attending responses. of reinforcement obtained on
(e.g., a choice between VR each schedule. Concurrent
reinforcement from source A and ratio schedules promote
VI reinforcement from source B in exclusive responding on the
the same context). In a conjoint relatively denser schedule of
schedule, no distinct stimuli are reinforcement.
correlated with each schedule
(e.g., behavior A produces VR
reinforcement and behavior B
produces DRH reinforcement in
the same context).
Chained/tandem Two or more schedules of Hoerger and Mace (2006) Schedule-specific patterns of
schedules reinforcement are available. used concurrent-chain behavior are observed in each
Completion of schedule A schedules to measure component. Schedule-specific
produces schedule B, and impulsive versus self- response patterns are more
completion of schedule B controlled choices made pronounced in chained than in
produces reinforcement. In a by male children with and tandem schedules.
chained schedule, each schedule without symptoms of ADHD.
component is correlated with a
distinct stimulus. In a tandem
schedule, no distinct stimuli are
correlated with each schedule
component.
Conjunctive Two or more schedules of Vollmer et al. (1997) used Schedule-specific patterns of
schedule reinforcement are arranged. a conjunctive FT DRO behavior are observed in each
All schedule requirements schedule to reduce aggression component.
must be completed to receive in an adolescent with
reinforcement. intellectual developmental
disorder.
Alternative Two or more schedules of Bowman et al. (1997) utilized Responding reflects a
schedule reinforcement are available an ALT FR 1 FR 1 EXT preference for one schedule
concurrently. The first schedule schedule to evaluate the component.
completed produces reinforcement. reinforcer preferences of
children with intellectual
developmental disorder.
Schedules of Reinforcement 69
Behavioral Contrast
compensate for the loss of reinforcement in the
Behavioral contrast is a phenomenon that re- extinction component. There are numerous ac-
sults from an interaction among schedule com- counts for behavioral contrast (see Catania, 2013,
ponents in a multiple schedule; that is, events in p. 221).
one schedule component affect responding in the There is growing evidence that humans show
other components. Reynolds (1961) first described behavioral contrast. For example, Hantula and
this interaction in an experiment with pigeons ex- Crowell (1994) presented undergraduate students
posed to two different multiple schedules. In the with a computerized investment task in which they
first schedule, key pecking was reinforced on a VI managed money for an investment group with the
3-minute schedule when the response keys were goal of maximizing financial returns. The experi-
alternately illuminated red and green, resulting menters gave the students numerous opportunities
in a multiple VI 3-minute VI 3-minute schedule. to invest between $100 and $10,000 in $100 incre-
After stable responding in this schedule, the ex- ments in two different markets. Investments either
perimenter introduced the second multiple sched- gained or lost money. Phases 1 and 3 were equal
ule. In this schedule, the experimenter changed multiple VI VI schedules. Phase 2 was a multiple
the green component from VI 3 minutes to ex- VI extinction schedule, and Phase 4 was a mul-
tinction, resulting in a multiple VI 3-minute ex- tiple extinction VI schedule. During the equal VI
tinction schedule. Figure 4.2 shows the results of VI schedules, investments were comparable in the
Reynolds’s experiment. Behavioral contrast is evi- two markets. However, when investments in one
dent in the second multiple schedule. Whereas re- market no longer made money (i.e., extinction),
sponding declined as expected during the extinc- behavioral contrast occurred in the market with
tion component, responding in the unchanged VI an unchanged rate of reinforcement.
3-minute component increased substantially. Note Tarbox and Parrot Hayes (2005) exposed un-
that the increased response rate evident in the un- dergraduate college students to multiple equal VI
changed component did not produce an increased VI, multiple VI extinction, multiple VI VI, and
rate of reinforcement, because response rate and multiple VI extinction schedules. They observed
reinforcement rate are largely independent in in- behavioral contrast consistently only when they
terval schedules. Behavioral contrast in interval provided participants with rules describing the
schedules represents an irrational expenditure of contingencies in the extinction component only.
responses and calories. This would not be the case They did not observe contrast when they did not
in multiple ratio schedules, in which increased give participants rules or when they gave partici-
responding in the unchanged component would pants rules for all components. The findings are
70 B a s i c P r in c ip l e s a nd c o n c e p t s
counterintuitive, but suggest that verbal behavior schedules ordinarily produce exclusive responding
interacts with direct contingencies in ways that on the richer of the two schedules (Herrnstein &
are not well understood yet. Loveland, 1975). The arrangement of asymmetri-
Behavioral contrast can be relevant in clinical cal schedules, such as concurrent VI FR, can pro-
ABA work (see Koegel, Egel & Williams, 1980, for duce a preference for the qualitative features of
an illustration). Interventions typically involve one schedule that is independent of the amount
the discontinuation of reinforcement for undesir- of reinforcement from the schedule (Baum, 1974).
able behavior. When a caregiver provides rein- Experiments arranging concurrent VI VI sched-
forcement for the undesirable behavior at a high ules generally include an adjunctive procedure
rate and then places the behavior on extinction in known as a change-over delay, which imposes a
one context (e.g., at school), contrast effects may brief interval during which responses cannot be
emerge in other contexts in which the caregiver reinforced immediately after the organism switch-
has not implemented extinction (e.g., home). This es from one schedule to another. The change-over
may be more likely if the reinforcement of pro- delay reduces the likelihood that schedule switch-
social alternative behavior does not compensate ing will be reinforced accidentally, relative to the
fully for the reduction in reinforcement from the probability of accidental reinforcement when the
extinction schedule. This possibility should guide first response after the schedule switch is eligible
the selection of reinforcement schedules for the for reinforcement.
prosocial behavior and the provision of advice to
parents, for example, about the possible side effects
Matching Theory
of intervention.
Concurrent schedules promote a pattern of re-
sponse allocation that is very orderly. Herrnstein
Concurrent and Conjoint Schedules
(1961, 1970) formulated the matching law, which
Concurrent and conjoint schedules arrange for the quantitatively describes the functional relation
simultaneous availability of two or more schedule between relative response rates for concurrent
components, such that the individual is free to al- alternatives and relative obtained rates of rein-
ternate among the components. This arrangement forcement. The matching law states that relative
permits the assessment of the relative preference for response rate will match or be equal to relative re-
the schedule components and the study of choice. inforcement rate. In its simplest form, the match-
As with multiple and mixed schedules, the differ- ing law is expressed as B1/B1 + B2 = r1/r1 + r2,
ence between concurrent and conjoint schedules where B1 and B2 are response rates for two behav-
is that components in concurrent schedules cor- iors, and r1 and r2 are the obtained reinforcement
relate with distinct stimuli, but components in rates for the two behaviors. We can reduce this
conjoint schedules do not. Concurrent schedules equation to B1/B2 = r1/r2, and we can fit a line to
are characteristic of human environments where logarithmic transformations of the obtained data
numerous alternatives are available and generally in the form of log (B1/B2) = a log (r1/r2) + log k,
are correlated with distinct stimuli. For example, where a is the slope of the line and log k is its inter-
a woman commuting to work with a friend on a cept at the y-axis (see Baum, 1974, and McDowell,
subway will have many different concurrently 1989, for full descriptions of mathematical trans-
available alternatives. She can converse with her formations of the simplified form of the matching
friend, read a newspaper, do a crossword puzzle, lis- law). When there is perfect matching, a = 1.0 and
ten to music, people watch, plan her work day, and log k = 0. Values of a > 1.0 show overmatching, and
so on. Each of these activities is correlated with values of a < 1.0 show undermatching, reflecting
distinct stimuli, and each provides reinforcement the individual’s sensitivity to relative reinforce-
according to some schedule. Because concurrent ment rate. Values of log k > 0 reflect a bias for B1,
schedules of reinforcement characterize human and values of log k < 0 show a bias for B2 due to
environments, this is our emphasis in this chapter. variables other than relative reinforcement rate
Experimenters can arrange concurrent sched- (see below).
ules for any combination of interval, ratio, or Conger and Killeen (1974) provided one of the
differential schedules of reinforcement (Davison first demonstrations of the matching law involv-
& McCarthy, 1988). However, most studies with ing human social behavior. Participants engaged
concurrent schedules have used concurrent VI in conversation with two experimenters, who
VI schedules. This is because concurrent-ratio provided comments of approval contingent on
Schedules of Reinforcement 71
participants’ statements. The experimenters pro- that this tool affords a more refined approach to
vided comments on different sets of concurrent develop interventions based on variables that af-
VI VI schedules. The results of the study showed fect choice.
that the relative amount of time participants di-
rected verbal statements to the two experimenters
Chained and Tandem Schedules
closely matched the relative rates of experimenter
attention. Numerous studies have established the Chained and tandem schedules organize sequenc-
generality of the matching law to human behav- es of behavior that produce reinforcement. Both
ior in the laboratory (Pierce & Epling, 1984) and schedules comprise two or more components ar-
to many socially relevant human behaviors, from ranged in a sequence. In a two-component ex-
academic engagement (Martens et al., 1992) to ample, completion of the schedule requirements
the performance of basketball players (Vollmer & for the initial link produces the onset of the sec-
Bourret, 2000). Although these findings are gen- ond component or terminal link. Completion of
erally robust, results of some studies show human the terminal-link schedule requirements produces
performance departing from matching (Pierce & delivery of a reinforcer. Schedule components in
Epling, 1984). For example, Mace, Neef, Shade, chained schedules correlate with distinct stim-
and Mauro (1994) needed to use several adjunc- uli, whereas components in tandem schedules
tive procedures, a 15-second limited hold, and a do not. Human behavior contacts chained and
change-over delay for adolescents to allocate their tandem schedules regularly. For example, numer-
time to arithmetic problems in accordance with ous sequences of behavior required to experience
matching. a vacation constitute a chained schedule, such
Choices are symmetrical when they differ only as planning the vacation (initial link), booking
by relative rate of reinforcement. However, human transportation (interim link), and transportation
choices in natural environments are most often to the desired location (terminal link). Comple-
asymmetrical. Response alternatives can differ tion of these schedule components produces ac-
along several parameters of reinforcement, includ- cess to the reinforcing events available during the
ing reinforcement quality, reinforcement delay, re- vacation.
inforcement amount, reinforcement schedule fea- As is true of the basic schedules of reinforce-
tures, and control of reinforcement. The response ment discussed earlier, chained and tandem
requirements or effort to obtain reinforcement schedules rarely operate in isolation. The more
can also differ. Baum (1974) proposed a match- common characteristic of natural human environ-
ing equation that accommodated independent ments is for initial links to consist of a concurrent
variables other than relative reinforcement rate. schedule; that is, humans typically have a choice
The generalized matching law expresses that B1/B2 of sequential activities and terminal reinforcers.
= V1/V2, where V refers to the value of the given This arrangement is known as a concurrent-chain
alternative as defined by the sum of the relative schedule. In laboratory experiments, the initial-
reinforcement parameters and response effort. link schedule requirements are usually identical
Mace and Roberts (1993) illustrated the ap- (e.g., concurrent VI 20-seconds VI 20-seconds).
plied relevance of the generalized matching law. However, terminal-link reinforcers and sometimes
They provided a conceptual framework to guide schedule requirements differ. Completion of the
the functional assessment of undesirable behaviors initial-link alternative produces the discrimina-
and the selection of behavioral treatments. They tive stimulus for the terminal link associated with
recommended using a descriptive analysis of unde- that alternative. For example, completion of initial
sirable behavior under natural conditions to iden- link A produces presentation of the discriminative
tify the quality of the reinforcing consequence, the stimulus for terminal link A, and completion of
magnitude of delay to reinforcement, the amount this schedule requirement produces reinforcer A.
of reinforcement provided, and the response re- A parallel sequence meets the schedule require-
quirement to produce reinforcement, and to esti- ments for initial link B.
mate the operative schedule of reinforcement. The
behavior analyst can use information from the
Self‑Control
descriptive analysis to design an intervention that
should effectively compete with the parameters of One contemporary development in ABA that
reinforcement and response effort that maintain uses concurrent-chain schedules is the behavioral
undesirable behavior. Mace and Roberts suggested model of self-control. Rachlin and Green (1972)
72 B a s i c P r in c ip l e s a nd c o n c e p t s
formally developed the model in an experiment delayed reinforcement. When the experimenters
with pigeons. Figure 4.3 diagrams the concurrent- varied the value of T, pigeons showed a shift in
chain procedure they used. The initial link was a their preference on the initial link. In general, as
concurrent FR 25 FR 25 schedule with both re- the value of T increased, so did the probability of
sponse keys illuminated white. Completion of the choosing the left initial-link FR 25 schedule that
right FR 25 schedule (top sequence) darkened the later produced no choice and the large-delayed-re-
response keys and house light for T seconds. After inforcement contingency only. Rachlin and Green
the blackout, they illuminated the right response called choosing the left initial-link key a commit-
key green and the left response key red. The ter- ment response—one that avoided the temptation
minal link was a concurrent CRF CRF schedule of small immediate reinforcement and exposed the
in which a single key peck on green produced a pigeon to the large-delayed-reinforcement contin-
4-second blackout followed by 4-second access to gency only. Self-control is said to occur when an
food, and a single key peck on red produced im- individual (1) chooses the large delayed reinforce-
mediate access to 2-second access to food followed ment over the small immediate reinforcement
by a 6-second blackout. Thus completion of the when exposed to both, or (2) makes the commit-
right FR 25 initial link produced later exposure to ment response in the initial link.
a choice between small immediate reinforcement Numerous applied studies have used concur-
and large delayed reinforcement. By contrast, rent-chain schedules to study impulsivity and
completion of the left initial-link FR 25 schedule self-control. The behavior of children with atten-
produced a similar blackout for T seconds, fol- tion-deficit/hyperactivity disorder (ADHD) is par-
lowed by the illumination of the green-key alter- ticularly relevant to this procedure, because a de-
native only with the large-delayed-reinforcement fining characteristic of this disorder is impulsivity,
contingency. When the time between completion and a primary clinical goal is the development of
of the initial link and onset of the terminal link self-control. For example, numerous studies have
was short (e.g., 0.5 second), the pigeons reliably found that children with ADHD are more likely
chose the red-key alternative. Rachlin and Green than their typically developing peers to choose
described this choice as impulsive, because pigeons small immediate reinforcement over large de-
forfeited an additional 2-second access to food layed reinforcement (e.g., Hoerger & Mace, 2006;
available for pecking the green key. Thus the delay Neef, Marckel, et al., 2005; Schweitzer & Sulzer-
to reinforcement discounted the value of the large Azaroff, 1988). Other studies have shown that
impulsive behavior is sensitive to variables other
than delayed reinforcement, such as response ef-
fort (Hoerger & Mace, 2006). Finally, investigators
have used this paradigm to guide the development
of interventions to promote self-control, such as
delay fading and commitment training (Binder,
Dixon, & Ghezzi, 2000; DuPaul & Ervin, 1996),
and to evaluate response to stimulant medication
(Neef, Bicard, Endo, Coury, & Aman, 2005). Thus
the concurrent-chain schedule has provided a con-
ceptual model for understanding impulsivity and
self-control; a procedure for objectively assessing
an individual’s sensitivity to delayed reinforcement
and other variables; and a model for identifying
specific interventions to promote self-control and
FIGURE 4.3. Rachlin and Green’s (1972) concurrent- evaluating pharmacological interventions.
chain schedule illustrates a behavioral model of self-
control. The initial link (CONC FR 25 FR 25) pre-
sented a choice between access to the terminal link Behavioral‑Momentum Theory
schedules. Completion of the top (right) link produced a
Behavioral momentum is a metaphor used by Nevin
T-second delay followed by CONC CRF CRF schedules
that presented a choice between small immediate rein- et al. (1983) to describe the tendency for baseline
forcement and large delayed reinforcement. Completion response rates to persist after some response dis-
of the bottom (left) link produced a T-second delay fol- ruptor (see the earlier discussion of resistance to
lowed by CONC CRF EXT schedules and the option change). As in Newton’s second law of motion,
only for large delayed reinforcement. behavioral momentum is the product of behavior-
Schedules of Reinforcement 73
al mass and behavioral velocity, where behavioral contingency), rather than the number of reinforc-
velocity is baseline response rate and behavioral ers delivered on the right key (peck-reinforcer
mass is the resistance of baseline response rate to contingency). Several researchers have replicated
change after application of varying amounts of Nevin et al.’s general findings with humans (e.g.,
some response disruptor. When we graph response Cohen et al., 2001; Dube & McIlvane, 2001; Mace
rates across varying amounts of the response dis- et al., 1990; Pritchard et al., 2014).
ruptor, such as sessions of extinction or amounts Ahearn, Clark, Gardenier, Chung, and Dube
of presession food, the height of the curve or func- (2003) illustrated the relevance of the Nevin et
tion on the y-axis is behavioral velocity, and the al. (1990) findings to clinically important human
slope of the function across the x-axis is resistance behavior. A functional analysis of the stereotypi-
to change; the total area under the curve repre- cal behavior of three children with ASD suggested
sents a response’s behavioral momentum. that automatic reinforcement maintained stereo-
Investigators have used various schedules of typy. Next, investigators identified preferred ob-
reinforcement to study behavioral momentum, jects via a preference assessment. They compared
including multiple schedules (Nevin et al., 1983), levels of stereotypical behavior in a test sequence
multiple concurrent schedules (Nevin, Tota, Tor- of conditions and a control sequence. The test
quato, & Shull, 1990), and concurrent-chain sequence consisted of baseline (no play materi-
schedules (Grace & Nevin, 1997). Numerous als available) → VT delivery of a preferred item
studies have shown that resistance to change is a → test (continuous access to a second preferred
function of the reinforcement conditions for these item) → baseline. The control sequence consisted
schedules. For example, Nevin (1974) and Nevin et of baseline → baseline → test → baseline. Ahern
al. (1983) used a two-component multiple sched- et al. found that although both the VT and test
ule to demonstrate that resistance to change is a conditions reduced stereotypical behavior relative
positive function of baseline rate of reinforcement. to baseline, due to the availability of alternative
Different pairs of MULT VI VI schedules arranged reinforcement, levels of stereotypy were higher
a higher rate of reinforcement in one component in the test condition that followed VT reinforcer
than in the other. During conditions of extinc- deliveries than in the test condition that followed
tion, satiation, and dark-key food, key pecking in baseline with no toys available. Mace (2000) and
pigeons was more persistent in the component Ahern et al. pointed out that although interven-
with the higher baseline reinforcement rate. In tions based on DRA and FT or VT schedules reli-
a subsequent series of experiments, Nevin et al. ably reduce occurrences of undesirable behavior,
(1990) tested the competing hypotheses that resis- these same interventions can have persistence-
tance to change is a function of baseline response– strengthening effects on undesirable behavior.
reinforcer relations versus baseline stimulus–rein- Grace and Nevin (2000) proposed a unifying
forcer relations. In their Experiment 2, baseline theory of choice and behavioral momentum in
consisted of a three-component multiple concur- which the variables functionally related to prefer-
rent schedule (MULT). In each component of the ence or choice were the same as those related to
MULT, two concurrent schedules operated, where resistance to change. Grace and Nevin (1997) con-
the left key was the first concurrent schedule and ducted one study forming the basis of this theory.
the right key the second: concurrent VI 45/hour Grace and Nevin (1997) randomly alternated a
VI 15/hour (green), concurrent extinction VI 15/ concurrent-chain procedure and a multiple-sched-
hour (red), and concurrent extinction VI 60/hour ule procedure in a single experimental session.
(white). In this arrangement, the response–rein- Three response keys were mounted on a wall above
forcer contingencies were equal in the green- and a food magazine. In the concurrent-chain proce-
red-key components (15/hour each) and less than dure, the side keys illuminated white in the initial
in the white-key component (60/hour). By con- link consisting of equal concurrent VI 20-second
trast, the stimulus–reinforcer contingencies (i.e., VI 20-second schedules. Initial-link reinforcement
the total reinforcers delivered in the presence of consisted of terminal-link entry and darkening of
each color) were equal in the green- and white- the side keys, and illumination of the center key
key components (60/hour each) and less than in (either green or red, depending on whether termi-
the red-key component (15/hour). Tests of the nal-link entry was contingent on a left or a right
resistance of right-key pecking to extinction and initial-link key peck). The terminal-key colors cor-
satiation showed that resistance to change was a related with a higher- or lower-rate VI schedule.
positive function of the total number of reinforc- Investigators presented 36 cycles of the concur-
ers delivered in each component (color-reinforcer rent-chain arrangement in each session. Thus the
74 B a s i c P r in c ip l e s a nd c o n c e p t s
concurrent-chain procedure assessed preference that the FT schedule could have adventitiously
for the terminal link as a function of choice in the reinforced aggressive behavior. To avoid this pos-
initial link. The multiple-schedule procedure in sibility, Vollmer et al. introduced a conjunctive FT
the experimental session involved the usual alter- DRO 10-second schedule. The FT schedule set up
nation of green and red keys correlated with the access to the preferred magazine; however, inves-
same VI schedules used in the concurrent-chain tigators delivered the magazine only if the partici-
procedure. After this baseline arrangement, in- pant had not engaged in aggression during the last
vestigators tested resistance to change by dark-key 10 seconds of the FT interval. That is, the partici-
food deliveries between components in the mul- pant had to satisfy both the FT and DRO schedule
tiple schedule. Grace and Nevin found that rela- requirements to gain access to the magazine. After
tive rate of reinforcement comparably predicted an initial response burst, the conjunctive schedule
preference in the concurrent-chain schedule and reduced aggression to low levels, and the investiga-
resistance to change in the multiple schedule. tors then thinned the FT schedule to a conjunctive
Mace, Mauro, Boyajian, and Eckert (1997) dem- FT 5-minute DRO 10-second schedule.
onstrated the applied significance of Grace and Tiger, Hanley, and Hernandez (2006) used an
Nevin’s work. They modified the high-p proce- alternative schedule of reinforcement in the con-
dure, a behavioral-momentum-inspired procedure, text of a concurrent-chain schedule. In a concur-
to increase its effectiveness. Knowing that rein- rent-chain schedule, two response alternatives (A
forcer quality affects choice, Mace et al. reasoned and B) are available concurrently in the initial
that supplying a higher-quality reinforcer (food) link. The participant can choose one alternative.
contingent on compliance to high-p instructions Completion of the schedule requirements for ini-
would increase the resistance of compliance to tial link A leads to terminal link A, and comple-
change when they presented a low-p instruction. tion of the schedule requirements for terminal link
The high-p procedure with food increased compli- A produces a reinforcer. The same is true for ini-
ance to low-p instructions that were unresponsive tial and terminal link B. In general, the choice to
to the high-p procedure without food. complete initial link A versus B reflects a prefer-
ence for the reinforcer associated with the respec-
tive terminal link. Tiger et al. evaluated preschool
Conjunctive and Alternative Schedules
children’s preference for choice as a reinforcer
Both conjunctive and alternative schedules com- independent of food reinforcers. The initial link
prise two or more schedule components. In con- presented three different-colored worksheets, each
junctive schedules, the subject must satisfy sched- with an identical academic task. Worksheet choice
ule requirements for all components to produce in the initial link was followed by a terminal-link
a reinforcer delivery. Unlike chained schedules, worksheet. Correct responses produced the conse-
the order of schedule completion is irrelevant in quences associated with each colored worksheet:
conjunctive schedules. By contrast, alternative- (1) a choice of one of five identical food reinforcers
schedule components are available concurrently. (choice), (2) a single food reinforcer identical to
Reinforcement is contingent on completion of ei- the choice option (no choice), and (3) no mate-
ther component, whichever occurs first. rial reinforcer. Five of the six children showed a
Vollmer et al. (1997) evaluated the effectiveness preference for the choice alternative in the initial
of FT schedules to reduce the severe aggressive link, although this preference did not persist for
behavior of a 13-year-old girl with severe intel- two of the five children. Although the study used
lectual developmental disorder for whom tangible a concurrent-chain procedure, it also represented
reinforcement maintained aggression. After a an alternative schedule, because completion of all
functional-analytic baseline, investigators deliv- schedule requirements for each alternative led to
ered access to a preferred magazine continuously, the reinforcers associated with each alternative.
resulting in zero occurrences of aggression. Aggres-
sion reemerged when the investigators attempted
to implement an FT schedule to thin the reinforce- SUMMARY AND CONCLUSION
ment schedule. A within-session analysis of the
temporal relation between FT reinforcer deliveries We have reviewed basic and combined schedules
and occurrences of aggression showed that sched- of reinforcement, and have provided definitions
uled reinforcer deliveries often coincided (within for each schedule and illustrations of the applica-
10 seconds) with aggressive acts. This suggested tions of the schedules in the ABA research litera-
Schedules of Reinforcement 75
ture. Schedules of reinforcement promote specific Critchfield, T. S., Haley, R., Sabo, B., Colbert, J., &
patterns of responding, but do so only in a broader Macropoulis, G. (2003). A half century of scallop-
context of available concurrent discriminated ing in the work habits of the United States Congress.
operants. This broader context includes the tem- Journal of Applied Behavior Analysis, 36, 465–486.
Davison, M., & McCarthy, D. (1988). The matching law:
porary motivational conditions for each discrimi-
A review. Hillsdale, NJ: Erlbaum.
nated operant and its particular history of rein- DeLuca, R. B., & Holburn, S. W. (1992). Effects of a
forcement or behavioral mass. We have provided variable-ratio reinforcement schedule with changing
an overview of some contemporary developments criteria on exercise in obese and nonobese boys. Jour-
in ABA, such as behavioral contrast, matching nal of Applied Behavior Analysis, 25, 671–679.
theory, self-control, and behavioral-momentum Dube, W. V., & McIlvane, W. J. (2001). Behavioral mo-
theory, and have illustrated that these topics are mentum in computer presented discriminations in
related directly to specific schedules of reinforce- individuals with severe mental retardation. Journal of
ment. Deliberate use of schedules of reinforcement the Experimental Analysis of Behavior, 75, 15–23.
offers applied behavior analysts a powerful tool to DuPaul, G. J., & Ervin, R. A. (1996). Functional assess-
understand the conditions that maintain behavior ment of behaviors related to attention-deficit/hyper-
activity disorder: Linking assessment to intervention
and to design highly effective interventions.
design. Behavior Therapy, 27, 601–622.
Ecott, C. L., & Critchfield, T. S. (2004). Noncontingent
REFERENCES reinforcement, alternative reinforcement, and the
matching law: A laboratory demonstration. Journal
Ahearn, W. H., Clark, K. M., Gardenier, N. C., Chung, of Applied Behavior Analysis, 37, 249–265.
B. I., & Dube, W. V. (2003). Persistence of stereotypic Felton, M., & Lyon, D. O. (1966). The post-reinforce-
behavior: Examining the effects of external reinforc- ment pause. Journal of the Experimental Analysis of
ers. Journal of Applied Behavior Analysis, 36, 439–448. Behavior, 9, 131–134.
Baum, W. M. (1974). On two types of deviation from the Ferster, C. B., & Skinner, B. F. (1957). Schedules of rein-
matching law: Bias and undermatching. Journal of the forcement. New York: Appleton-Century-Crofts.
Experimental Analysis of Behavior, 22, 231–242. Fisher, W. W., & Mazur, J. E. (1997). Basic and applied
Binder, L. M., Dixon, M. R., & Ghezzi, P. M. (2000). A research on choice responding. Journal of Applied Be-
procedure to teach self-control to children with at- havior Analysis, 30, 387–410.
tention deficit hyperactivity disorder. Journal of Ap- Fleshler, M., & Hoffman, H. S. (1962). A progression for
plied Behavior Analysis, 33, 233–237. generating variable-interval schedules. Journal of the
Bowman, L. G., Piazza, C. C., Fisher, W. W., Hagopian, Experimental Analysis of Behavior, 5, 529–530.
L. P., & Kogan, J. S. (1997). Assessment of preference Flood, W. A., & Wilder, D. A. (2004). The use of dif-
for varied versus constant reinforcement. Journal of ferential reinforcement and fading to increase time
Applied Behavior Analysis, 30, 451–458. away from a caregiver in a child with separation anxi-
Brandauer, C. (1958). The effects of uniform probabilities ety disorder. Education and Treatment of Children, 27,
of reinforcement on the response rate of the pigeon. Un- 1–8.
published doctoral dissertation, Columbia Univer- Gillis, J. M., Hammond Natof, T., Lockshin, S. B., &
sity. Romanczyk, R. G. (2009). Fear of routine physical
Carr, J. E., Kellum, K. K., & Chong, I. M. (2001). The exams in children with autism spectrum disorders.
reductive effects of noncontingent reinforcement: Focus on Autism and Other Developmental Disabilities,
Fixed-time versus variable-time schedules. Journal of 24, 156–168.
Applied Behavior Analysis, 34, 505–509. Grace, R. C., & Nevin, J. A. (1997). On the relation
Catania, A. C. (2013). Learning (5th ed.). Cornwall-on- between preference and resistance to change. Journal
Hudson, NY: Sloan. of the Experimental Analysis of Behavior, 67, 43–65.
Catania, A. C., & Reynolds, G. S. (1968). A quantita- Grace, R. C., & Nevin, J. A. (2000). Comparing pref-
tive analysis of the responding maintained by inter- erence and resistance to change in constant- and
val schedules of reinforcement. Journal of the Experi- variable-duration schedule components. Journal of
mental Analysis of Behavior, 11, 327–383. the Experimental Analysis of Behavior, 74, 165–188.
Cohen, S. L., Richardson, J., Klebez, J., Febbo, S., & Hantula, D. A., & Crowell, C. R. (1994). Matching and
Tucker, D. (2001). EMG biofeedback: The effects of behavioral contrast in a two-option repeated invest-
CRF, FR, VR, FI, and VI schedules of reinforcement ment simulation. Journal of Applied Behavior Analysis,
on the acquisition and extinction of increases in 27, 607–617.
forearm muscle tension. Applied Psychophysiology and Harding, J. W., Wacker, D. P., Berg, W. K., Rick, G., &
Biofeedback, 26, 179–194. Lee, J. F. (2004). Promoting response variability and
Conger, R., & Killeen, P. (1974). Use of concurrent stimulus generalization in martial arts training. Jour-
operants in small group research: A demonstration. nal of Applied Behavior Analysis, 37, 185–196.
Pacific Sociological Review, 17, 399–416. Heard, K., & Watson, T. S. (1999). Reducing wander-
76 B a s i c P r in c ip l e s a nd c o n c e p t s
ing by persons with dementia using differential rein- aggression during the treatment of self-injurious be-
forcement. Journal of Applied Behavior Analysis, 32, havior. Journal of Applied Behavior Analysis, 32, 1–8.
381–384. Luiselli, J. K., Pace, G. M., & Dunn, E. K. (2006). Effects
Herrnstein, R. J. (1961). Relative and absolute strength of behavior-contingent and fixed-time release con-
of a response as a function of frequency of reinforce- tingencies on frequency and duration of therapeutic
ment. Journal of the Experimental Analysis of Behavior, restraint. Behavior Modification, 30, 442–455.
4, 267–272. Mace, F. C. (2000). The clinical importance of choice
Herrnstein, R. J. (1970). On the law of effect. Journal and resistance to change. Behavioral and Brain Sci-
of the Experimental Analysis of Behavior, 13, 243–266. ences, 23, 105–106.
Herrnstein, R. J., & Loveland, D. H. (1975). Maximizing Mace, F. C., Gritter, A. K., Johnson, P. E., Malley, J. L.,
and matching on concurrent ratio schedules. Journal & Steege, M. W. (2007). Contingent reinforcement
of the Experimental Analysis of Behavior, 24, 107–116. in context. European Journal of Behavior Analysis, 7,
Hoerger, M. L., & Mace, F. C. (2006). A computerized 115–120.
test of self-control predicts classroom behavior. Jour- Mace, F. C., & Lalli, J. S. (1991). Linking descriptive
nal of Applied Behavior Analysis, 39, 147–159. and experimental analyses in the treatment of bi-
Horner, R. D. (1980). The effects of an environmental zarre speech. Journal of Applied Behavior Analysis, 24,
enrichment program on the behavior of institution- 553–562.
alized profoundly retarded children. Journal of Ap- Mace, F. C., Lalli, J. S., Shea, M. C., Lalli, E. P., West,
plied Behavior Analysis, 13, 473–491. B. J., Roberts, M., et al. (1990). The momentum of
Hyten, C., & Madden, G. J. (1993). The scallop in fixed- human behavior in a natural setting. Journal of the
interval research: A review with data description. Experimental Analysis of Behavior, 54, 163–172.
Psychological Record, 43, 471–500. Mace, F. C., Mauro, B. C., Boyajian, A. E., & Eckert, T.
Johnston, J. M., & Pennypacker, H. S., Jr. (2009). Strate- L. (1997). Effects of reinforcer quality on behavioral
gies and tactics of behavioral research (3rd ed.). New momentum: Coordinated applied and basic research.
York: Routledge. Journal of Applied Behavior Analysis, 30, 1–20.
Kahng, S., Abt, K. A., & Schonbachler, H. E. (2001). Mace, F. C., Neef, N. A., Shade, D., & Mauro, B. C.
Assessment and treatment of low-rate high-intensity (1994). Limited matching on concurrent schedule re-
problem behavior. Journal of Applied Behavior Analy- inforcement of academic behavior. Journal of Applied
sis, 34, 225–228. Behavior Analysis, 27, 585–596.
Kahng, S., Iwata, B. A., Thompson, R. H., & Hanley, G. Mace, F. C., Pratt, J. L., Prager, K. L., & Pritchard, D.
P. (2000). A method for identifying satiation versus (2011). An evaluation of three methods of saying ‘no’
extinction effects under noncontingent reinforce- to avoid an escalating response class hierarchy. Jour-
ment schedules. Journal of Applied Behavior Analysis, nal of Applied Behavior Analysis, 44, 83–94.
33, 419–432. Mace, F. C., & Roberts, M. L. (1993). Factors affecting
Koegel, R. L., Egel, A. L., & Williams, J. A. (1980). Be- selection of behavioral interventions. In J. Reichle
havioral contrast and generalization across settings & D. Wacker (Eds.), Communicative alternatives to
in the treatment of autistic children. Journal of Ex- challenging behavior: Integrating functional assessment
perimental Child Psychology, 30, 422–437. and intervention strategies (pp. 113–133). Baltimore:
Lalli, J. S., Casey, S. D., & Kates, K. (1997). Noncontin- Brookes.
gent reinforcement as treatment for severe problem Mace, F. C., Zangrillo, A. N., Prager, K., Carolan, E.,
behavior: Some procedural variations. Journal of Ap- Hoerger, M., Thomas, K., et al. (2008). A method-
plied Behavior Analysis, 30, 127–137. ology for maintaining low levels of attention-main-
Lalli, J. S., Mace, F. C., Wohn, T., & Livezey, K. (1995). tained problem behaviors following variable-time
Identification and modification of a response-class schedule thinning. European Journal of Behavior
hierarchy. Journal of Applied Behavior Analysis, 28, Analysis, 9, 149–156.
551–559. Magee, S. K., & Ellis, J. (2001). The detrimental effects
Lane, L. L., Rogers, L. A., Parks, R. J., Weisenbach, J. L., of physical restraint as a consequence for inappropri-
Mau, A. C., Merwin, M. T., et al. (2007). Function- ate classroom behavior. Journal of Applied Behavior
based interventions for students who are nonrespon- Analysis, 34, 501–504.
sive to primary and secondary efforts: Illustrations at Marsteller, T. M., & St. Peter, C. C. (2014). Effects of
the elementary and middle school levels. Journal of fixed-time reinforcement schedules on resurgence of
Emotional and Behavioral Disorders, 15, 169–183. problem behavior. Journal of Applied Behavior Analy-
Laraway, S., Snycerski, S., Michael, J., & Poling, A. sis, 47, 455–469.
(2003). Motivating operations and terms to describe Martens, B. K., Lochner, D. G., & Kelly, S. Q. (1992).
them: Some further refinements. Journal of Applied The effects of variable-interval reinforcement on
Behavior Analysis, 36, 407–414. academic engagement: A demonstration of match-
Lerman, D. C., Iwata, B. A., & Wallace, M. D. (1999). ing theory. Journal of Applied Behavior Analysis, 25,
Side effects of extinction: Prevalence of bursting and 143–151.
Schedules of Reinforcement 77
McDowell, J. J. (1989). Two modern developments in trol: Teaching tolerance for delay in impulsive chil-
matching theory. Behavior Analyst, 12, 153–166. dren. Journal of the Experimental Analysis of Behavior,
Neef, N. A., Bicard, D. F., Endo, S., Coury, D. L., & 50, 173–186.
Aman, M. G. (2005). Evaluation of pharmacological Tarbox, J., & Parrot Hayes, L. (2005). Verbal behavior
treatment of impulsivity in children with attention and behavioral contrast in human subjects. Psycho-
deficit hyperactivity disorder. Journal of Applied Be- logical Record, 55, 419–437.
havior Analysis, 38, 135–146. Thompson, R. H., & Iwata, B. A. (2005). A review of
Neef, N. A., Marckel, J., Ferreri, S. J., Bicard, D. F., Endo, reinforcement control procedures. Journal of Applied
S., Aman, M. G., et al. (2005). Behavioral assessment Behavior Analysis, 38, 257–278.
of impulsivity: A comparison of children with and Tiger, J. H., & Hanley, G. P. (2005). An example of
without attention deficit hyperactivity disorder. Jour- discovery research involving the transfer of stimu-
nal of Applied Behavior Analysis, 38, 23–37. lus control. Journal of Applied Behavior Analysis, 38,
Nevin, J. A. (1974). Response strength in multiple 499–509.
schedules. Journal of the Experimental Analysis of Be- Tiger, J. H., Hanley, G. P., & Hernandez, E. (2006). An
havior, 21, 389–408. evaluation of the value of choice with preschool chil-
Nevin, J. A., & Grace, R. C. (2000). Preference and re- dren. Journal of Applied Behavior Analysis, 39, 1–16.
sistance to change with constant-duration schedule Van Camp, C. M., Lerman, D. C., Kelley, M. E., Con-
components. Journal of the Experimental Analysis of trucci, S. A., & Vorndran, C. M. (2000). Variable-
Behavior, 74, 79–100. time reinforcement schedules in the treatment of
Nevin, J. A., Mandell, C., & Atak, J. R. (1983). The socially maintained problem behavior. Journal of Ap-
analysis of behavioral momentum. Journal of the Ex- plied Behavior Analysis, 33, 545–557.
perimental Analysis of Behavior, 39, 49–59. Vollmer, T. R., & Bourret, J. (2000). An application of
Nevin, J. A., Tota, M. E., Torquato, R. D., & Shull, R. the matching law to evaluate the allocation of two-
L. (1990). Alternative reinforcement increases resis- and three-point shots by college basketball players.
tance to change: Pavlovian or operant contingen- Journal of Applied Behavior Analysis, 33, 137–150.
cies? Journal of the Experimental Analysis of Behavior, Vollmer, T. R., & Hackenberg, T. D. (2001). Reinforce-
53, 359–379. ment contingencies and social reinforcement: Some re-
Ninness, H. A. C., Ozenne, L., & McCuller, G. (2000). ciprocal relations between basic and applied research.
Fixed-interval responding during computer interac- Journal of Applied Behavior Analysis, 34, 241–253.
tive problem solving. Psychological Record, 50, 387– Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R.
401. G., & Mazaleski, J. L. (1993). The role of attention in
Pierce, W. D., & Epling, W. F. (1984). On the persis- the treatment of attention-maintained self-injurious
tence of cognitive explanation: Implications for be- behavior: Noncontingent reinforcement and differ-
havior analysis. Behaviorism, 12, 15–27. ential reinforcement of other behavior. Journal of Ap-
Poling, A., & Normand, M. (1999). Noncontingent re- plied Behavior Analysis, 26, 9–21.
inforcement: An inappropriate description of time- Vollmer, T. R., Marcus, B. A., & Ringdahl, J. E. (1995).
based schedules that reduce behavior. Journal of Ap- Noncontingent escape as a treatment for self-injuri-
plied Behavior Analysis, 32, 237–238. ous behavior maintained by negative reinforcement.
Pritchard, D., Hoerger, M., Mace, F. C., Penney, H., Journal of Applied Behavior Analysis, 28, 15–26.
& Harris, B. (2014). Clinical translation of animal Vollmer, T. R., Progar, P. R., Lalli, J. S., Van Camp, C.
models of treatment relapse. Journal of the Experimen- M., Sierp, B. J., Wright, C. S., et al. (1998). Fixed-
tal Analysis of Behavior, 101, 442–449. time schedules attenuate extinction-induced phe-
Rachlin, H., & Green, L. (1972). Commitment, choice nomena in the treatment of severe aberrant behavior.
and self-control. Journal of the Experimental Analysis Journal of Applied Behavior Analysis, 31, 529–542.
of Behavior, 17, 15–22. Vollmer, T. R., Ringdahl, J. E., Roane, H. S., & Marcus,
Reynolds, G. S. (1961). Behavioral contrast. Journal of B. A. (1997). Negative side effects of noncontingent
the Experimental Analysis of Behavior, 4, 57–71. reinforcement. Journal of Applied Behavior Analysis,
Ringdahl, J. E., Vollmer, T. R., Borrero, J. C., & Con- 30, 161–164.
nell, J. E. (2001). Fixed-time schedule effects as a Wallace, M. D., Iwata, B. A., Hanley, G. P., Thompson,
function of baseline reinforcement rate. Journal of R. H., & Roscoe, E. M. (2012). Noncontingent rein-
Applied Behavior Analysis, 34, 1–15. forcement: A further examination of schedule effects
Saini, V., Miller, S. A., & Fisher, W. W. (2016). Multiple during treatment. Journal of Applied Behavior Analy-
schedules in practical application: Research trends sis, 45, 709–719.
and implications for future investigation. Journal of Wright, C. S., & Vollmer, T. R. (2002). Evaluation of a
Applied Behavior Analysis, 49, 1–24. treatment package to reduce rapid eating. Journal of
Schweitzer, J. B., & Sulzer-Azaroff, B. (1988). Self-con- Applied Behavior Analysis, 35, 89–93.
CHAP TER 5
A teenage boy may readily acquiesce to his mother’s forcing functions of stimuli (Skinner, 1938). We
request for a hug and a kiss in the privacy of their have restricted the discussion for this chapter
home but vehemently refuse the same request in primarily to discriminative control—the stimulus
front of his peers. We often yell and cheer at sport- control that develops when we present positive re-
ing events, sit quietly but periodically sing during a inforcement or withdraw negative reinforcement
church service, and whisper when we wish to con- contingent on a response in the presence of a
vey information to one individual without being stimulus.
overheard by others. These examples illustrate It is difficult to discuss the topic of stimulus dis-
how the behavior of humans and other species crimination without also discussing its counteref-
often changes, depending on the circumstances. fect—generalization. Researchers have used the
When our behavior changes in response to such term generalization in many ways. Basic research-
environmental circumstances, we say that it is ers working with nonhumans in the laboratory
under stimulus control. Stimulus control is highly typically refer to primary stimulus generalization.
relevant to applied behavior analysts because most For example, Jenkins and Harrison (1960) condi-
behavior is under some degree of stimulus control. tioned a response to occur when they presented
As such, in this chapter we discuss a broad range a 1,000-cycles-per-second (1,000-cps) tone and
of phenomena that fall within the topic of stimu- showed that the response also occurred when they
lus control. These include discriminations that presented a 1,500- or a 670-cps tone. Applied be-
are learned through direct training or experience, havior analysts have used the term generalization
such as simple discriminations and conditional in a much broader sense. Stokes and Baer (1977)
discriminations; they also include ones learned defined generalization as “any occurrence of rel-
through generalization processes, such as stimulus evant behavior under nontraining conditions (i.e.,
generalization, stimulus equivalence, and recombi- across subjects, settings, people, behaviors, and/or
native generalization. time) without the scheduling of the same events
Researchers have used the term stimulus control in those conditions as had been scheduled in the
in many ways, which have broad connotations. training conditions” (p. 350). We use the term
We may define stimulus control in terms of the generalization in the current chapter in a slightly
changes in the probability of a form or rate of a different way. Generalization for us includes more
behavior that occur due to the presentation of a rapid learning of new discriminations, based on
stimulus. Defined in this way, stimulus control the learning of similar discriminations in the past
would include discriminative, eliciting, and rein- (i.e., learning to learn; Harlow, 1949).
78
Stimulus Control and Generalization 79
properties of the relevant stimulus after target ing, 2016). The odors of narcotics and explosives
responding occurs primarily when the stimulus is are salient stimuli for dogs and rats, and research-
present and not when it is absent. One possibil- ers have trained these animals to respond to the
ity is that reinforcement may have developed dis- odors by reinforcing a specific response, such as sit
criminative properties, because reinforcement of and stay, when the odor is present and not when
target responding is likely to be followed by more it is absent.
reinforcement, and nonreinforcement is likely to
be followed by more nonreinforcement. In this
case, target responding may occur more after rein- DIFFERENTIAL STIMULUS CONTROL
forcement delivery than following a response that BY STIMULI PRESENTED SUCCESSIVELY
was not reinforced if the delivery of the reinforcer
functions as a discriminative stimulus (Blough, Although the mere detection of a stimulus is im-
1966). We can be more confident that the relevant portant under some conditions, stimulus control
stimulus controls target responding if responding more often involves a discrimination between two
begins immediately after stimulus presentation stimuli: a positive stimulus correlated with rein-
and stops immediately after stimulus termination. forcement, and a negative stimulus correlated with
Even when we use a salient stimulus and have nonreinforcement. The researcher alternates the
established control over a response, other, similar positive stimulus (such as a red light) with a nega-
stimuli also may control the response. Stimulus tive stimulus (such as a green light) of the same
generalization allows for the reinforcement of re- intensity and presentation duration, to establish a
sponses in the presence of stimuli that initially may successive discrimination. We use the term succes-
have been difficult to condition. For example, Ful- sive discrimination because the positive and nega-
ton and Spradlin (1971) initially established con- tive stimuli are present at different times. Howev-
trol over a button-pressing response to a 70-deci- er, discriminative responding under this condition
bel, 500-Hertz tone, which is a salient stimulus for can be slow and may not occur for all experimental
people with normal hearing, to assess the hearing participants.
of children with intellectual developmental disor- Fading is likely to establish a more rapid dis-
der. Control occurred for a target 70-decibel, 500- crimination among the stimuli (Terrace, 1966),
Hertz tone for most participants, and less intense because we can maximize the difference between
tones also controlled responding. Responding to the positive and negative stimuli initially and then
less intense tones than the target 70-decibel tone decrease the differences between stimuli gradually.
allowed the researchers to provide reinforcement Methods for increasing the difference between
of responses to lower-intensity tones. Generaliza- the stimuli include maximizing the saliency of the
tion occurred to tones of still lower intensity after positive stimulus (such as a bright red light) and
researchers provided reinforcement for responses minimizing the saliency of the negative stimulus
to lower-intensity tones. The tone maintained (such as a faint green light), and presenting the
stimulus control over button pressing due to stimu- positive stimulus for a long duration and the nega-
lus generalization and reinforcement of responses tive stimulus for a brief duration. Maximizing the
to tones with progressively lower intensity until differences between the stimuli increases the prob-
the tone reached a threshold level, which is the ability of responding in the presence of the posi-
magnitude level at which responding is no longer tive stimulus and not responding in the presence
discriminative. of the negative stimulus. Fading initially requires
Researchers have conducted most studies of a salient positive stimulus (such as a bright red
simple stimulus control like those described above light) and a nonsalient negative stimulus (such as
with auditory and visual stimuli. We also can use a faint green light). We establish responding in
these procedures to establish control with tactile, the presence of the positive stimulus and not in
gustatory, and olfactory stimuli. Although control the presence of the negative stimulus. We then in-
with a single stimulus has few parallels in indi- crease the intensity of the negative stimulus (such
viduals’ daily lives, single-stimulus control is use- as increasing the brightness of the green light)
ful for conducting hearing evaluations for people until the intensity of the positive and negative
with intellectual developmental disorder (Fulton stimuli is equal, and responding occurs only in the
& Spradlin, 1971) and for teaching dogs and rats presence of the positive stimulus. We then gradu-
to find people, narcotics, and explosives by smell ally increase the duration of the negative-stimulus
(e.g., Edwards, La Londe, Cox, Weetjens, & Pol- presentation until the durations of positive- and
Stimulus Control and Generalization 81
negative-stimulus presentations are equal, and re- (1967) established fine circle–ellipse discrimina-
sponding occurs only in the presence of the posi- tions by using a fading procedure with an appa-
tive stimulus. ratus with eight translucent keys on which they
Macht (1971) used a similar procedure to teach projected light and figures. Initially, there was
a discrimination between forward and reversed E’s great disparity between the positive stimulus (a
to children with intellectual developmental dis- circle on a fully illuminated white key) and the
order. He taught each child to press a lever when negative stimulus (no figures and nonilluminated
he presented the letter E (which was the positive keys). Training began with the presentation of one
stimulus) and not to press the lever when he pre- positive and seven negative stimuli. Touching the
sented a solid black square (which was the nega- positive stimulus produced reinforcement, whereas
tive stimulus). Initially, presentations of the E’s touching any negative stimulus did not. The re-
were long in duration, and he presented the black searchers reinstated the previous fading step when
square for a brief period when responding paused. participants made errors. Gradually, the research-
He increased the duration of subsequent black- ers illuminated the seven negative stimuli until
square presentations and punished responses in its they were the same intensity as the positive stimu-
presence. He established discrimination between lus. Next, the researchers introduced a horizon-
the forward E and the reversed E by using varia- tal line on the negative-stimulus keys. After par-
tions of this procedure after responding between ticipants were responding to the positive stimulus,
the E and the square was discriminative. He then the researchers gradually changed the negative-
evaluated each child’s vision by moving the child stimulus figure (i.e., the horizontal line morphed
farther and farther from the apparatus that pre- into a very narrow ellipse). The ellipse morphed
sented the forward and reversed E’s. into an ellipse-like circle like that on the positive-
When we establish the first discrimination (e.g., stimulus key after the researchers illuminated the
forward and reversed E’s in Macht’s experiment) by negative-stimulus keys with an ellipse. Eventually,
using a fading technique, the next discrimination they established the threshold for the circle–el-
we establish (e.g., A vs. F) may not require exten- lipse discrimination.
sive programming, or we may be able to establish Sometimes extending stimulus control from
it in fewer stimulus presentations. In addition, we stimuli that already control differential respond-
may establish future discriminations (e.g., N vs. ing to stimuli that currently do not is desirable.
M) more quickly even without fading. Dinsmoor One method researchers have used to transfer
(1995a) attributed this type of improvement in dis- control from existing stimulus dimensions to novel
criminative learning to the control of attending re- ones is a delayed-prompting procedure (Touchette,
sponses. We also may consider it as a special case of 1971). Touchette (1971) demonstrated the delayed-
learning to learn or a learning set (Harlow, 1949). prompting procedure with three adult students
Even though establishing discrimination be- with intellectual developmental disorder. He es-
tween a forward and reversed E may be sufficient tablished a discrimination between the positive
for the evaluation of subjective vision, it is insuf- stimulus (a red key) and the negative stimulus (a
ficient for most educational purposes. For example, white key) by reinforcing responses to the red key
teaching the discrimination between a forward and not reinforcing responses to the white key.
and reversed E is unlikely to establish a discrimi- Next, he established a new discrimination between
nation between the letters E and F. Therefore, we two stimuli with minimal disparity (a horizontal E
should use several different negative stimuli, such with the legs pointing up vs. a horizontal E with
as F, P, B, and I, and variations of the positive stim- the legs pointing down). Initially he superimposed
ulus when teaching such a discrimination, such as the positive and negative stimuli on the red and
E and E (see Engelmann & Carnine, 1982, for a white stimuli, which already controlled behavior.
more complete discussion). He delayed the onset of the red background light
behind the positive stimulus for 0.5 second after
the first correct response. He delayed the onset of
DIFFERENTIAL STIMULUS CONTROL BY TWO the red background stimulus by an additional 0.5
OR MORE SIMULTANEOUSLY PRESENTED STIMULI second and decreased its presentation duration by
0.5 second for each incorrect response on each ad-
A simple simultaneous discrimination involves ditional trial. Correct responding occurred before
the discrimination between two or more stimuli the delivery of the prompt (i.e., the red light) in 10
presented at the same time. Sidman and Stoddard trials with all three students. However, such rapid
82 B a s i c P r in c ip l e s a nd C o n c e p t s
acquisition of correct responding does not always auditory stimulus “one” to teach a child with imi-
occur. Subsequent research suggests that the stu- tative speech to name printed numerals. The be-
dents may have been atypical (e.g., Oppenheimer, havior analyst could present reinforcement for the
Saunders, & Spradlin, 1993). vocalization when the child imitates the auditory
Oppenheimer et al. (1993), using a similar stimulus “one.” The behavior analyst should pres-
procedure with 30 adults with intellectual de- ent the auditory stimulus “one” after a few-second
velopmental disorder, obtained three differ- delay after presenting the printed numeral 1 on
ent outcomes. First, some participants were like subsequent trials. The behavior analyst should
Touchette’s (1971) participants who responded present reinforcement if the child names the nu-
before the onset of the red light and were cor- meral before the behavior analyst presents the
rect. Second, some participants responded before auditory stimulus; ideally, the probability of rein-
the onset of the red light, but were correct on forcement would be greater for such preemptive re-
only 50% of trials. Third, some participants never sponses (Touchette & Howard, 1984). Training on
responded before the onset of the red light. The the numeral 2 could begin after the child consis-
researchers conducted further testing with this tently names the numeral 1. Initially, the behavior
last group to determine (1) whether responding analyst should present the auditory stimulus “two”
did not occur because they had not established simultaneously with the printed numeral 2. The
discrimination between the horizontal E’s; or (2) behavior analyst should provide reinforcement
whether they had established the discrimination when the child imitates “two.” Presentations of
between the two horizontal E’s, but the partici- the printed numeral 2 should continue with a brief
pants were waiting until the red light came on be- delay before the auditory stimulus. This procedure
fore responding. Oppenheimer et al. presented the should continue until the child consistently re-
red light simultaneously on the keys behind both sponds before the auditory stimulus. Then the be-
the correct and the incorrect E. Some participants havior analyst should present the printed numerals
did not respond before the red light came on be- 1 and 2 in a random order. Initially, the behavior
cause they did not make the discrimination; that analyst should present the auditory stimuli simul-
is, their performance was at chance level when the taneously with the printed numerals to prevent
researcher projected the red light on both keys. initial errors. Then the behavior analyst should
However, other participants responded correctly present the printed numerals (1 or 2) 4–5 sec-
when the researcher projected the red light on onds before the auditory stimulus. This procedure
both E’s. Therefore, the red light functioned as should produce the discrimination quite rapidly for
a “go” stimulus for discriminative responding be- most children. The behavior analyst can introduce
tween the E’s for some participants. other printed numerals with this same procedure.
The procedure just described is useful in the In many cases, the behavior analyst may begin by
laboratory study of discrimination and sensory intermixing the printed numerals when he or she
processes. However, many discrimination situa- uses the delayed-prompt technique. In fact, tact
tions involve making different responses to dif- training, in which researchers intersperse three
ferent stimuli. Researchers have used variants of targets simultaneously, produced rapid acquisition
the delayed-prompt technique, such as progres- of multiple target responses and promoted general-
sive and constant time or prompt delay, to teach ization in children with autism spectrum disorder
a two-choice visual discrimination (e.g., Handen and intellectual developmental disorder (e.g., Leaf
& Zane, 1987). Researchers have used delayed- et al., 2016; Marchese, Carr, LeBlanc, Rosati, &
prompt procedures to teach skills such as selection Conroy, 2012; Wunderlich, Vollmer, Donaldson,
of letters and numbers (e.g., Bradley-Johnson, Sun- & Phillips, 2014).
derman, & Johnson, 1983; Touchette & Howard,
1984), sight word reading (e.g., Gast, Ault, Wolery,
Doyle, & Belanger, 1988; Knight, Ross, Taylor, CONDITIONAL STIMULUS CONTROL
& Ramasamy, 2003), and naming and requesting
(Charlop, Schreibman, & Thibodeau, 1985; Char- The discriminations we have discussed previously
lop & Walsh, 1986; Halle, Baer, & Spradlin, 1981; have been either simple successive or simple si-
Halle, Marshall, & Spradlin, 1979; Knight et al., multaneous discriminations, which are essential
2003). in daily life. Many discriminations that individu-
A behavior analyst could present the printed als make during their daily activities, however, are
numeral 1 while simultaneously presenting the conditional discriminations, in which reinforcement
Stimulus Control and Generalization 83
of a response in the presence of a stimulus depends tity matching (Alexander, Ayres, Smith, Shepley,
on the presence or absence of other stimuli. An & Mataras, 2013; Farber, Dube, & Dixon, 2016).
example of a conditional discrimination is passing We consider generalized identity matching as
the salt when someone asks you to pass the salt, another example of generalization. Preacademic
and passing the bread when someone asks you to workbooks use identity matching extensively to
pass the bread. That is, the discriminated behavior teach letter and number discrimination to stu-
is conditional because the positive stimulus (e.g., dents. Usually, the workbook presents the sample
salt, bread) changes, depending on the request letter or number at the left margin, and the choice
(Serna, Dube, & McIlvane, 1997). letters or numbers in a row to the right of the sam-
Simultaneous identity matching is a very simple ple. The student’s response is to mark the correct
conditional-discrimination procedure. Research- choice. This workbook task is an example of iden-
ers have studied this procedure widely in labora- tity matching; however, it is not an ideal teaching
tories with humans and nonhumans. In the labo- technique, because it involves delayed reinforce-
ratory, the typical procedure involves presenting ment for correct responses.
a visual sample stimulus (e.g., the numeral 2) to Simultaneous identity matching requires dis-
which the experimental participant must respond. crimination of the sample stimuli from the re-
The researcher then presents two or more compar- maining comparison stimuli. It neither requires
ison stimuli (such as 2 and 3, or 2, 3, 4, and 5) after nor ensures successive discrimination among
the participant’s response to the sample. One of sample stimuli, because the sample stimulus re-
the comparison stimuli (e.g., 2) is identical to the mains available throughout the trial. However, it
sample stimulus; the remaining stimulus (e.g., 3) is a delayed-matching-to-sample procedure if the
or stimuli (e.g., 3, 4, and 5) differ from the sample teacher removes the sample after the onset of the
stimulus. The disparity between the correct com- comparisons and probably requires successive dis-
parison stimulus and the other stimuli may be large crimination of the sample stimuli (e.g., Constan-
or small. The researcher presents different stimuli tine & Sidman, 1975).
as samples from trial to trial; thus the correct com- Closely related to the identity-matching task
parison stimulus is conditional on which sample is the oddity procedure. In this discrimination,
is present. Identity-matching experiments have the teacher presents the student with an array of
involved simple trial-and-error procedures (Saun- three or more stimuli, one of which is different.
ders, Johnson, Tompkins, Dutcher, & Williams, Sidman and Stoddard (1967) presented such a
1997; Saunders & Sherman, 1986) and fading pro- display to their students; however, they used only
cedures (Dube, Iennaco, & McIlvane, 1993). The one stimulus (a circle) as the correct odd stimulus.
researcher may present only the single comparison Other researchers have presented a series of trials
stimulus that matches the sample stimulus, or the on which the odd stimulus (i.e., the positive stimu-
other comparisons may be blank on the first few lus) was different on various trials (e.g., Dickerson
trials of an identity-matching-to-sample task using & Girardeau, 1971; Ellis & Sloan, 1959; Smeets
a fading procedure. The researcher may begin to & Striefel, 1974; Soraci et al., 1987; Stromer &
fade in the nonmatching stimulus or stimuli after Stromer, 1989). For example, the researcher might
a few trials. The nonmatching stimuli become present the numeral 1 as the single positive stimu-
more visible until the intensity of the nonmatch- lus, with two or more numeral 2’s as the negative
ing stimuli matches that of the sample stimulus stimuli on one trial. The researcher might present
with each successive trial of correct responding. a single 2 with two or more 1’s on other trials. In
Participants typically match new stimuli on their that case, the oddity task is a conditional-discrim-
first presentation after they have matched a few ination task, because the stimulus designated as
comparison stimuli to samples; that is, they exhib- correct depends on the other stimuli. Preacademic
it generalized identity matching. Nonhuman par- workbooks have used the oddity procedure exten-
ticipants and some participants with intellectual sively.
developmental disorder do not exhibit generalized A more complex type of conditional discrimi-
matching readily. Researchers have used video nation is one in which the comparison stimuli are
modeling, error correction, and fading from a not similar physically to the sample stimulus. Au-
simple tabletop sorting task to compound identity ditory–visual matching (e.g., receptive labeling)
matching on a computer screen with participants consists of the presentation of an auditory sample,
with autism spectrum disorder and intellectual de- such as dictated object names, with visual com-
velopmental disorder to facilitate generalized iden- parisons, such as an array of objects. For example,
84 B a s i c P r in c ip l e s a nd C o n c e p t s
the correct response is conditional on the pre- 3) with previously trained comparisons (e.g., the
sented sample if a teacher is training a student to numeral 1 or 2). Researchers have called this type
touch specific printed numerals (e.g., 1 and 2) in of performance exclusion (Dixon, 1977; McIlvane,
response to corresponding auditory samples (e.g., Kledaras, Lowry, & Stoddard, 1992). Researchers
“one” or “two”). That is, the answer is the numeral have demonstrated the emergence of exclusion-
1 if the auditory sample is “one,” and the correct ary relations between three-dimensional visual
response is the numeral 2 if the auditory sample is and auditory stimuli (e.g., McIlvane & Stoddard,
“two.” Such a conditional discrimination depends 1981, 1985), between two-dimensional visual and
on (1) a successive discrimination of the audi- auditory stimuli (e.g., Wilkinson & Green, 1998;
tory stimuli “one” and “two,” (2) a simultaneous Wilkinson & McIlvane, 1994), and between text
discrimination between the printed numerals 1 and auditory stimuli (e.g., de Rose, de Souza, &
and 2, and (3) the correspondence of the auditory Hanna, 1996). However, testing for control by the
stimulus “one” with the numeral 1. Not all typi- emerged relation under different conditions (e.g.,
cally developing children learn such conditional multiple trials with more than two previously
discriminations without first learning the compo- unknown comparisons) is necessary to ensure
nent discriminations. In those cases, the behavior control by the sample (Carr, 2003). The teacher
analyst can establish conditional responding by may train new relations by using a variation of the
teaching each component discrimination in isola- procedure used to teach the initial conditional
tion (see Saunders & Spradlin, 1993). discrimination if exclusion does not establish the
The first component of a component-teaching discrimination.
procedure is to establish successive discrimina-
tions among the auditory number samples. A
teacher may have the student echo the auditory EQUIVALENCE CLASSES
stimuli if the student emits vocal behavior. The
second component is to establish discriminated re- Configurations of shared physical properties deter-
sponding to the comparison stimuli (e.g., printed mine many stimulus classes (e.g., balls, cars, cats,
numerals). The teacher can conduct a matching- humans, men, women, red objects). The actual
to-sample probe with the numerals. The teacher configurations of shared physical characteristics
can implement the third component if the student determining class membership have been the
exhibits matching (e.g., 1 to 1, and 2 to 2). If not, focus of research by psycholinguists, cognitive psy-
the teacher should establish the discrimination be- chologists (Medin & Smith, 1984; Rosch, 1973),
tween the printed numerals. The third component and behavior analysts (e.g., Fields et al., 2002;
is to establish the relation between the auditory Fields & Reeve, 2001; Galizio, Stewart, & Pilgrim,
stimulus (e.g., “one”) and the corresponding nu- 2004). However, shared physical properties do
meral comparison (e.g., 1). The teacher may train not define many important stimulus classes (e.g.,
this conditional discrimination by using a delayed- lawyers, letters, medical doctors, numbers, tools,
prompt procedure (Grow, Carr, Kodak, Jostad, toys). Whether each member is substitutable, and
& Kisamore, 2011; Grow, Kodak, & Carr, 2014; whether they evoke new, untrained responses in
Touchette & Howard, 1984) or a blocking proce- certain contexts, define the members of these
dure (Saunders & Spradlin, 1993; Saunders, Wil- classes (Saunders & Green, 1992; Urcuioli, 2013).
liams, & Spradlin, 1995; Smeets & Striefel, 1994). For example, we may define toys as a stimulus class
The teacher typically teaches the second con- because they are items that children manipulate,
ditional discrimination with less careful program- and we may store them in a toy box. In addition, a
ming and fewer trials after he or she has taught child is likely to engage in exploratory and novel
the first conditional discrimination (e.g., between play behavior without any direct training when
“one” and 1 and “two” and 2). The teacher can he or she finds a new item in the toy box. Medi-
then teach the next discrimination between the cal doctors are a stimulus class because we call
printed numeral (e.g., 3) and the corresponding them doctor, and any member with the appropri-
auditory sample (e.g., “three”), intermixed with ate credentials may practice medicine. In addition,
trials of the previously discriminated numer- people are much more likely to follow the health-
als. The selection of the novel comparison may related advice of someone called doctor than some-
emerge without direct training when the teacher one called waiter.
presents the novel sample and comparison stimuli Sidman (1971) established an equivalence class
(e.g., the auditory sample “three” and the numeral using a symbol-matching procedure with a 17-year-
Stimulus Control and Generalization 85
old student with microcephaly and intellectual or nonsense stimuli (e.g., vek, zog; Fields & Ver-
developmental disorder. The student selected 20 have, 1990; Sidman et al., 1974; Sidman & Tailby,
pictures when the researcher presented their cor- 1982; Spradlin et al., 1973). A general procedure
responding dictated words (AB), and could name consists of teaching an AB conditional discrimi-
20 pictures when the researcher presented the pic- nation (i.e., present sample stimulus A1, provide
tures (BD) before the study. However, he did not reinforcement for selection of comparison stimulus
name the 20 printed words related to the pictures B1; present sample stimulus A2, provide reinforce-
(CD), select printed words in response to their ment for selection of comparison stimulus B2).
dictated names (AC), select printed words that The researcher teaches a new discrimination (BC)
named the pictures (BC), or select the pictures after he or she establishes the AB discrimination.
when presented with the printed words (CB). Sid- In this case, the researcher presents the stimuli, B1
man trained selection of printed words when he and B2, as samples and introduces two new com-
presented corresponding dictated words (AC) to parison stimuli, C1 and C2. The researcher inter-
the student. He conducted probes after training mixes the AB and BC conditional-discrimination
to test whether the student would select printed trials after the student demonstrates the BC dis-
words when given the respective pictures (BC), crimination. The researcher introduces probe tri-
and whether the student would select the pictures als when discrimination is nearly perfect on the
when given the respective printed words (CB). AB and BC discriminations. The researcher usu-
Not only did the student select the printed words ally introduces probe trials without reinforcement
when Sidman presented dictated words, but the or differential feedback and in a series of AB and
BC and CB relations between pictures and printed BC trials. The first probes are often for symmetry
words emerged, regardless of whether Sidman pre- (i.e., A1 is the correct comparison when B1 is the
sented the printed words as sample or comparison sample stimulus, and A2 is the correct comparison
stimuli. In addition, the student named many of when B2 is the sample). In addition, the researcher
the printed words (CD) after the initial AC train- conducts the CB probe to determine whether the
ing. This training established 20 stimulus classes; student will select the appropriate comparison, B1
each class consisted of the spoken word, the print- and B2, when presented with the samples C1 and
ed word, and the pictures. C2. Typically, the student demonstrates symmetry
Although the results of Sidman’s (1971) study (Fields & Verhave, 1990; Pilgrim & Galizio, 1990).
were remarkable, the design was less than ideal for The researcher conducts another probe to test
demonstrating the development of new stimulus for transitivity (e.g., C1 is the correct comparison
classes, because the stimuli were common (e.g., when A1 is the sample, and C2 is the correct com-
car, cat, dog) and only one student participated. parison when A2 is the sample). In addition, the
Sidman’s experiment led to a flurry of research di- researcher conducts probes to determine whether
rected toward the development of stimulus classes the student will select A1 as the comparison when
comprising stimuli that shared no defining physical C1 is the sample, and whether the student will se-
properties (Saunders, Saunders, Kirby, & Spradlin, lect A2 as the comparison when C2 is the sample.
1988; Saunders, Wachter, & Spradlin, 1988; Sid- This final probe is a combined test for symmetry
man, Cresson, & Willson-Morris, 1974; Sidman, and transitivity, because it cannot be positive un-
Kirk, & Willson-Morris, 1985; Sidman & Tailby, less both symmetry and transitivity are present.
1982; Spradlin, Cotter, & Baxley, 1973; Spradlin The student demonstrates equivalence when these
& Saunders, 1986; Wetherby, Karlan, & Spradlin, tests are positive (Sidman & Tailby, 1982). Most
1983). These studies and others led to numerous students given this training pass these tests, even
theoretical discussions concerning the necessary though some percentage of students who learn the
and sufficient conditions for the development of AB and BC conditional discriminations fail the
such classes (Baer, 1982; Fields & Verhave, 1987; equivalence tests.
Hayes, Barnes-Holmes, & Roche, 2001; Horne & AB and BC training can establish an equiva-
Lowe, 1996; Saunders & Green, 1992; Sidman, lence class; however, these are not the only combi-
1994, 2000; Sidman & Tailby, 1982). nations that produce an equivalence class. Train-
Most matching-to-sample studies after Sid- ing AB and AC conditional discriminations or BA
man (1971) used procedures with better experi- and CA conditional discriminations also produces
mental control (Sidman et al., 1974; Sidman & equivalence classes. In fact, some research suggests
Tailby, 1982; Spradlin et al., 1973). Experimental that teaching students to select a single compari-
stimuli are typically abstract forms (e.g., #, @, ?) son in response to multiple samples is a more effec-
86 B a s i c P r in c ip l e s a nd C o n c e p t s
tive procedure for demonstrating stimulus equiva- addition, four new conditional discriminations
lence than the other two procedures (Spradlin & (BC, CB, BA, and CA) may emerge when there
Saunders, 1986). The teacher can use all of the are three visual stimuli in each class, and we train
tests for equivalence when the stimuli are visual. AB and BC conditional discriminations. Teaching
However, the teacher typically does not present three conditional discriminations (AB, AC, AD)
A1 and A2 as comparisons if they are auditory may produce nine new emergent discriminations
stimuli, because of problems in presenting and dis- (BC, CB, BD, DB, CD, DC, BA, CA, DA) when
criminating two auditory stimuli simultaneously. the number of visual stimuli in each class increas-
When we are studying the development of es to four stimuli; that is, the number of potential
equivalence classes in a laboratory, the procedures untaught discriminations increases dramatically
and outcomes may appear quite remote from the as the number of stimuli in each class increases.
problems encountered in our daily lives. However, Moreover, Saunders, Wachter, et al. (1988) have
they may not appear quite as remote with an ex- established equivalence classes with as many as
ample of equivalence we encounter daily: number nine members in each class.
equivalence. Although many students learn to Researchers have established equivalence
select the numerals 1 through 4 in response to classes with stimuli that have potential social
dictated number names and to name the numer- use, including equivalence among printed names,
als 1 through 4, some do not. We can teach these dictated names, and faces of therapists (Cowley,
students three conditional discriminations (AB, Green, & Braunling-McMorrow, 1992); among
AC, and AD), and perhaps nine additional con- dictated words, objects, and manual signs (Van-
ditional discriminations will emerge through sym- Biervliet, 1977); among reading-relevant stimuli
metry and transitivity (BA, BC, CA, CB, CD, DA, (de Rose et al., 1996; de Rose, de Souza, Rossito, &
DC, BD, and DB). Additionally, naming numer- de Rose, 1992; Wultz & Hollis, 1980); in prearith-
als, sets (i.e., quantity), and words (BE, CE, DE) metic skills (Gast et al., 1979); in money relations
may emerge. Naming responses are likely if a stu- (Stoddard, Brown, Hurlbert, Manoli, & McIlvane,
dent names either the numerals or the sets before 1989); and in spelling and reading (Mackay, 1985;
training (Gast, VanBiervliet, & Spradlin, 1979). Mueller, Olmi, & Saunders, 2000; Stromer &
There are many ways that we can establish equiva- Mackay, 1992).
lence classes with numbers. For example, a teacher Research has demonstrated several interesting
might teach the student to select (1) the numerals findings about subsequent performances after the
in response to their dictated names (AB), (2) sets establishment of equivalence classes. First, we can
in response to their printed numerals (BC), and add new members readily to a class (e.g., Gast et
(3) printed words in response to their respective al., 1979; Saunders, Wachter, et al., 1988). For ex-
sets (CD). An effective procedure for establishing ample, Saunders, Wachter, et al. (1988) established
the performances if the student is verbal would be two eight-member classes of visual stimuli by using
to teach the (1) selection of printed numerals in procedures analogous to those described earlier.
response to presentation of dictated words (AB), Researchers taught the students to select two vi-
(2) names of the printed numerals (BE), (3) selec- sual stimuli, one member from each equivalence
tion of sets in response to presentation of printed class, in response to two different auditory stimuli
numerals (BC), (4) selection of the printed words (i.e., nonsense syllables) after they had established
in response to presentation of sets (CD), and (5) the two classes of eight stimuli each. Three of the
appropriate responding to an intermix of trials four students selected the seven remaining stimuli
from the trained discriminations (1, 2, and 3). in each class in response to their respective spoken
This procedure may be redundant if students name nonsense syllables when the researchers presented
printed numerals after learning to select printed probe trials. They taught the fourth student to se-
numerals in response to dictated number names lect a second stimulus in response to each audi-
(AB). However, this procedure ensures a student’s tory stimulus. Students selected the remaining six
familiarity with the generalization-testing formats, stimuli of each class in response to their respective
and the naming response is in the student’s rep- auditory stimuli after this training.
ertoire. In addition, intermixing trials maintains Second, equivalence classes are durable. After
prerequisite conditional discriminations (Saun- training and probing, Saunders, Wachtel, et al.
ders, Wachter, et al., 1988; Spradlin et al., 1973). (1988) dismissed their students for 2–5 months
There are numerous ways of establishing equiva- and then retested to see whether the auditory
lence classes via conditional discriminations. In stimuli still controlled the selection of the same
Stimulus Control and Generalization 87
comparison stimuli. They provided no differential with eight students. The researchers used two
reinforcement during the pretest for baseline train- stimuli (B1 and B2) in a separate classical-condi-
ing or for probe trials. Correct selections occurred tioning setting after they had established the two
immediately for three of four students. Respond- four-member equivalence classes. They presented
ing for the fourth student gradually recovered to B1 and B2 successively during this phase. Shock
prebreak levels, even though the researchers did always followed B1, but never followed B2. B1 came
not deliver differential reinforcement for baseline to elicit a galvanic skin response; B2 did not. Six of
training or for probe trials. the eight students exhibited galvanic skin respons-
Third, performances that indicate equivalence es to the stimulus class of which B1 was a member,
are resistant to change as a function of changes and did not exhibit galvanic skin responses to the
in baseline reinforcement conditions (Pilgrim & stimulus class including B2, when the researchers
Galizio, 1990; Saunders, Saunders, et al., 1988). presented the six remaining stimuli from the two
Pilgrim and Galizio (1990) taught five college classes. This study demonstrated that emotional
students AB and AC conditional discriminations, responses can develop without direct conditioning
and students demonstrated equivalence after train- if stimuli that evoke those responses are members
ing. The researchers then reversed the contingen- of an equivalence class.
cies for the AC discrimination (i.e., selecting C2 The preceding text provides a sample of equiva-
was reinforced when a student was presented with lence studies and some of their implications for
sample A1, and selecting C1 was reinforced when understanding the development of stimulus con-
a student was presented with sample A2). Sym- trol. Procedures other than conditional discrimi-
metry responding to CA relations was reversed for nations have also resulted in the development of
some students, but no BC or CB responses were equivalence classes. For example, if we teach a stu-
reversed. Saunders, Saunders, et al. (1988) found dent to say a specific word in response to a set of
that equivalence classes were difficult to disrupt previously unrelated stimuli, those stimuli are like-
once they had established them. ly to function as a class (Reese, 1972). Researchers
Fourth, conditioning an operant response in have demonstrated that stimuli presented in the
the presence of a member of the class produces same position in a sequence of stimuli function as
generalization to other members of the class after ordinal-equivalence classes (Mackay, Stoddard, &
the establishment of an equivalence class. Barnes Spencer, 1989; Sigurdardottir, Green, & Saunders,
and Keenan (1993) demonstrated the transfer of 1990). For example, if we teach a student to place
operant responding from one member of an equiv- three stimuli (N, B, and X) in a one–two–three se-
alence class to other members. The researchers es- quence while we teach him or her to place stimuli
tablished two equivalence classes (A1, B1, and C1; E, L, and Z in a one–two–three sequence, stimuli
A2, B2, and C2) by training AB and AC relations N and E are likely to become members of one
with college students. The researchers trained the equivalence class. Stimuli B and L will probably
students to respond slowly when they presented become members of another, and stimuli X and Z
one stimulus (B1) and rapidly when they presented will become members of a third class. In addition,
the second stimulus (B2) after the students dem- stimuli that occur contiguously also may come
onstrated equivalence classes. Then the research- to function as members of an equivalence class
ers presented the four remaining stimuli (A1, C1, (Stromer & Stromer, 1990a, 1990b).
A2, and C2). Students emitted slow responses in Finally, researchers have shown that adding
the presence of A1 and C1, and rapid responses in stimuli to a previously established equivalence
the presence of A2 and C2. Gast et al. (1979) dem- class is possible merely on the basis of shared con-
onstrated a similar transfer of control from some sequences (Dube & McIlvane, 1995). In summary,
members of an equivalence class to a remaining there are many ways that we can establish equiva-
member of that class. lence classes. We would venture that two or more
Fifth, if we condition one or more members stimuli will become members of the same equiva-
of an equivalence class to elicit an emotional re- lence class in a context without a change in con-
sponse, other members of that class will elicit that tingencies if those stimuli are substitutable.
response (Dougher, Auguston, Markham, Wulfert, Note that context determines the formation of
& Greenway, 1994). Dougher et al. (1994) estab- a stimulus class. For example, if a teacher asks a
lished two four-member equivalence classes (A1, student to put toys in one box and tools in an-
B1, C1, and D1; A2, B2, C2, and D2) by training other, the student will respond differently than if
AB, AC, and AD conditional discriminations the teacher asks the student to put soft items in
88 B a s i c P r in c ip l e s a nd C o n c e p t s
one box and hard items in another. Bush, Sidman, Striefel et al. (1976, 1978) extended previous work
and de Rose (1989) demonstrated such contextual by demonstrating recombinative generalization of
control over equivalence class membership with stimulus–response units with nonvocal children
traditional matching-to-sample procedures. with intellectual developmental disorder. They
Research on the applied implications of pro- conducted a series of studies to establish instruc-
cedures that promote the formation of stimulus tion following with verb–noun combinations (e.g.,
classes has proceeded somewhat slowly in com- “Push glass,” “Drop car”). They first taught imi-
parison to the basic research on this topic (McLay, tative responding by using instructions, models,
Sutherland, Church, & Tyler-Merrick, 2013; Re- reinforcement, and time out. They presented si-
hfeldt, 2011). However, a randomized clinical trial multaneous vocal instructions (e.g., “Push glass”)
showed that college students learned the generic and modeling after the participants displayed imi-
and brand names of 32 pharmacological agents in tation independently. They used a delayed-prompt
both written and spoken forms more quickly when procedure to train independent responding after
equivalence-based instruction than when a more consistent performance. They taught new nouns
traditional approach based on flash card instruc- using the same procedure (e.g., “Push glass,” then
tion was used (Zinn, Newland, & Ritchie, 2015). “Push car”) after independent responding oc-
In addition, research on increasing complex ver- curred for the first instruction. They intermixed
bal behavior has increasingly examined the extent the two instructions after observing accurate re-
to which various training procedures affect con- sponding with each instruction in isolation. They
sistent and rapid improvement in the specifically introduced a new verb instruction with the first
targeted responses and in the emergence of novel trained noun and later intermixed it with the pre-
responses (DeSouza, Akers, & Fisher, 2017; Tin- viously taught verb instruction (e.g., “Drop glass”
cani, Miguel, Bondy, & Crozier, Chapter 16, this vs. “Push glass”). Training continued with other
volume). nouns until responding was accurate with all 12
As noted earlier, this is far from a complete nouns for both verb instructions. They trained
discussion of stimulus equivalence and related novel verbs by using the same procedure.
phenomena. We refer the reader to Hayes (1991), Other researchers have extended recombinative
Hayes et al. (2001), Horne and Lowe (1996), and generalization to more complex forms of instruc-
Sidman (1994, 2000) for extensive discussions and tion following and accompanying grammatical
theoretical interpretations of the necessary and verbal phrases describing actions (e.g., Frampton,
sufficient conditions for the development of such Wymer, Hansen, & Shillingsburg, 2016; Goldstein
stimulus control. & Mousetis, 1989) and reading skills and spell-
ing (Mueller et al., 2000; Saunders, O’Donnell,
Vaidya, & Williams, 2003). See Goldstein (1984),
STIMULUS CONTROL BASED ON RECOMBINATION Saunders (2011), and Wetherby and Striefel (1978)
OF STIMULUS–RESPONSE COMPONENTS for a more complete discussion of research on re-
combinative generalization and its potential value
One of the remarkable characteristics of human in understanding the development of complex be-
behavior is the degree to which responding to havior and guiding teaching programs.
complex stimuli occurs without previous direct Studies on the recombination of stimulus–re-
experience. For example, young children develop sponse units (instruction following) with non-
generalized imitation so that their behavior can vocal children with intellectual developmental
approximate that of a model, even though they disorder may suggest a model for understanding
have never had direct training on imitating the the development of generalized imitation. Baer
specific model (e.g., Baer, Peterson, & Sherman, et al. (1967) taught imitation of several different
1967; Peterson, 1968). In addition, individuals may behaviors to three children with intellectual de-
respond appropriately to specific verbal instruc- velopmental disorder. The children imitated novel
tions they have never encountered previously (e.g., behaviors after training, and they demonstrated
Stewart, McElwee, & Ming, 2013; Striefel, Weth- generalized imitation. The researchers discussed
erby, & Karlan, 1976). Recombinative generalization these phenomena in terms of behavior similarity
occurs when a student recombines responses tar- and response classes. However, the similarity in-
geted during training in novel ways. For example, terpretation acknowledges that the similarity only
a student may “push car” and “drop glass” if we occurs for a third person who observes both the
teach him or her to “push glass” and “drop car.” response modeled by the training and the imita-
Stimulus Control and Generalization 89
tive response of the child. From the perspective of tones with other frequencies and volumes. Hence
the child learning to imitate, little similarity exists extending the stimulus control across a total range
between the modeled stimuli and the child’s re- of frequencies and volumes accessible to human
sponse, because the child only observes the mod- hearing is easy. Therefore, more learning occurs
el’s behavior. In fact, a more recent study found than what we teach directly even in simple suc-
that allowing the child to observe him- or herself cessive discriminations. However, the equivalence
using a mirror during imitation training might fa- paradigm provides even more examples of how an
cilitate acquisition of imitation skills (Miller, Ro- extensive repertoire of discriminations can emerge
driguez, & Rourke, 2015). from very little teaching. In the hypothetical num-
In a sense, imitation training is another instruc- ber example, researchers only teach three condi-
tion-following task. However, the instructional tional discriminations before the nine additional
stimuli are visual rather than auditory. Siegel and conditional discriminations, and potentially three
Spradlin (1978) speculated that generalized imita- stimuli names emerge. Saunders, Wachter, et al.
tion might involve a similar process to that dem- (1988) taught seven conditional discriminations,
onstrated by Striefel et al. (1976); that is, children and a total repertoire of 56 conditional discrimi-
learn individual components and then recombine nations emerged. In other words, they taught seven
them when they imitate a novel motor response conditional discriminations, and 49 emerged.
after receiving imitation training with specific Recombinative generalization provides an ad-
motor responses. Siegel and Spradlin noted that ditional example of how a little training produces
21 of the 131 motor imitations taught to a student, an extensive repertoire. Striefel et al. (1976) taught
who required the most exemplars before show- 31 noun–verb instructions to a student with intel-
ing generalized imitations, involved the action of lectual developmental disorder, and 113 emerged
tapping. They suggested that the student might without training. The recombination of letter–
tap new objects or body parts without additional sound units makes it possible for students to re-
training after they taught him to tap a few objects spond appropriately to almost any new printed
or body parts. In addition, the researchers sug- English word after being taught only a limited
gested that generalized imitation occurs only for number of letter–sound units. Kohler and Malott
models in the general training domain (e.g., vocal, (2014) recently replicated these finding with chil-
motor). Therefore, children may not imitate fine dren with autism spectrum disorder.
motor movements if training involves following In short, research on primary stimulus general-
gross motor movements, and children may not ization, stimulus equivalence, and recombinative
imitate vocal models if training involves nonvocal generalization provides examples of how behav-
motor movements. ioral repertoires are acquired rapidly and suggests
methods for the effective teaching of such vast
repertoires.
SOME CONCLUDING REMARKS
Baer, D. M., Peterson, R. F., & Sherman, J. A. (1967). Dinsmoor, J. A. (1995b). Stimulus control: Part 2. Be-
The development of imitation by reinforcing behav- havior Analyst, 18, 253–269.
ioral similarity to a model. Journal of the Experimental Dixon, L. S. (1977). The nature of control by spoken
Analysis of Behavior, 10, 405–416. words over visual stimulus selection. Journal of the
Barnes, D., & Keenan, M. (1993). A transfer of func- Experimental Analysis of Behavior, 29, 433–442.
tions through derived arbitrary and nonarbitrary Dougher, M. J., Auguston, E. M., Markham, M. R.,
stimulus relations. Journal of the Experimental Analy- Wulfert, E., & Greenway, D. E. (1994). The trans-
sis of Behavior, 59, 61–81. fer of respondent eliciting and extinction functions
Blough, D. S. (1966). The study of animal sensory pro- through stimulus equivalence classes. Journal of the
cesses by operant methods. In W. K. Honig (Ed.), Experimental Analysis of Behavior, 62, 331–351.
Operant behavior: Areas of research and application Dube, W. V., Iennaco, F. M., & McIlvane, W. J. (1993).
(pp. 345–379). New York: Appleton-Century-Crofts. Generalized identity matching to sample of two-
Bradley-Johnson, S., Sunderman, P., & Johnson, C. M. dimensional forms in individuals with intellectual
(1983). Comparison of delayed prompting and fading disabilities. Research in Developmental Disabilities, 14,
for teaching preschoolers easily confused letters and 457–477.
numbers. Journal of School Psychology, 21, 327–335. Dube, W. V., & McIlvane, W. J. (1995). Stimulus–re-
Bush, K. M., Sidman, M., & de Rose, T. (1989). Con- inforcer relations and emergent matching to sample.
textual control of emergent equivalence relations. Psychological Record, 45, 591–612.
Journal of the Experimental Analysis of Behavior, 51, Edwards, T. L., La Londe, K. B., Cox, C., Weetjens, B.,
29–45. & Poling, A. (2016). Effects of schedules of reinforce-
Carr, D. (2003). Effects of exemplar training in exclu- ment on pouched rats’ performance in urban search-
sion responding on auditory–visual discrimination and-rescue training. Journal of Applied Behavior Anal-
tasks with children with autism. Journal of Applied ysis, 49, 199–204.
Behavior Analysis, 36, 507–524. Ellis, N. R., & Sloan, W. (1959). Oddity learning as a
Charlop, M. H., Schreibman, L., & Thibodeau, M. G. function of mental age. Journal of Comparative and
(1985). Increasing spontaneous verbal responding in Physiological Psychology, 52, 228–230.
autistic children using a time delay procedure. Jour- Engelmann, S., & Carnine, D. (1982). Theory of instruc-
nal of Applied Behavior Analysis, 18, 155–166. tion: Principles and applications. New York: Irvington.
Charlop, M. H., & Walsh, M. E. (1986). Increasing au- Farber, R. S., Dube, W. V., & Dickson, C. A. (2016).
tistic children’s spontaneous verbalizations of affec- A sorting-to-matching method to teach compound
tion: An assessment of time delay and peer modeling matching to sample. Journal of Applied Behavior Anal-
procedures. Journal of Applied Behavior Analysis, 19, ysis, 49, 294–307.
307–314. Ferster, C. B., & Perrott, M. C. (1968). Behavior prin-
Constantine, B., & Sidman, M. (1975). The role of nam- ciples. New York: New Century.
ing in delayed matching to sample. American Journal Fields, L., Matneja, P., Varelas, A., Belanich, J., Fitzer,
of Mental Deficiency, 79, 680–689. A., & Shamoun, K. (2002). The formation of linked
Cowley, B. J., Green, G., & Braunling-McMorrow, D. perceptual classes. Journal of the Experimental Analy-
(1992). Using stimulus equivalence procedures to sis of Behavior, 78, 271–290.
teach name–face matching to adults with brain inju- Fields, L., & Reeve, K. F. (2001). A methodological in-
ries. Journal of Applied Behavior Analysis, 25, 461–475. tegration of generalized equivalence classes, natural
de Rose, J. C., de Souza, D. G., & Hanna, E. S. (1996). categories, and crossmodal perception. Psychological
Teaching reading and spelling: Exclusion and stimu- Record, 51, 67–87.
lus equivalence. Journal of Applied Behavior Analysis, Fields, L., & Verhave, T. (1987). The structure of equiv-
29, 451–469. alence classes. Journal of the Experimental Analysis of
de Rose, J. C., de Souza, D. G., Rossito, A. L., & de Rose, Behavior, 48, 317–332.
T. M. S. (1992). Stimulus equivalence and generaliza- Fields, L., & Verhave, T. (1990). The effects of nodality
tion in reading after matching to sample by exclu- on the formation of equivalence classes. Journal of the
sion. In S. C. Hayes & L. J. Hayes (Eds.), Understand- Experimental Analysis of Behavior, 53, 345–358.
ing verbal relations (pp. 69–82). Reno, NV: Context Fisher, W. W., Greer, B. D., Fuhrman, A. M., & Querim,
Press. A. C. (2015). Using multiple schedules during func-
DeSouza, A. A., Akers, J. S., & Fisher, W. W. (2017). tional communication training to promote rapid
Empirical application of Skinner’s verbal behavior transfer of treatment effects. Journal of Applied Be-
to interventions for children with autism: A review. havior Analysis, 48, 713–733.
Analysis of Verbal Behavior, 33, 229–259. Frampton, S. E., Wymer, S. C., Hansen, B., & Shil-
Dickerson, D. S., & Girardeau, F. L. (1971). Oddity pref- lingsburg, M. A. (2016). The use of matrix training
erence by mental retardates. Journal of Experimental to promote generative language with children with
Child Psychology, 10, 28–32. autism. Journal of Applied Behavior Analysis, 49,
Dinsmoor, J. A. (1995a). Stimulus control: Part 1. Be- 869–883.
havior Analyst, 18, 51–68. Fulton, R. T., & Spradlin, J. E. (1971). Operant audiom-
Stimulus Control and Generalization 91
etry with severely retarded children. Audiology, 10, Jenkins, H. M., & Harrison, R. H. (1960). Effects of
203–211. discrimination training on auditory generalization.
Galizio, M., Stewart, K. L., & Pilgrim, C. (2004). Typi- Journal of Experimental Psychology, 59, 246–253.
cality effects in contingency-shaped generalization Knight, M. G., Ross, D. E., Taylor, R. L., & Ramasamy,
classes. Journal of the Experimental Analysis of Behav- R. (2003). Constant time delay and interspersal of
ior, 82, 253–273. known items to teach sight words to students with
Gast, D. L., Ault, M. J., Wolery, M., Doyle, P. M., & Be- mental retardation and learning disabilities. Educa-
langer, S. (1988). Comparison of constant delay and tion and Training in Developmental Disabilities, 38,
the system of least prompts in teaching sight word 179–191.
reading to students with moderate retardation. Edu- Kohler, K. T., & Malott, R. W. (2014). Matrix train-
cation and Training in Mental Retardation, 23, 117–128. ing and verbal generativity in children with autism.
Gast, D. L., VanBiervliet, A., & Spradlin, J. E. (1979). Analysis of Verbal Behavior, 30(2), 170–177.
Teaching number–word equivalences: A study of Landa, R., & Hanley, G. P. (2016). An evaluation of
transfer. American Journal of Mental Deficiency, 83, multiple-schedule variations to reduce high-rate re-
524–527. quests in the picture exchange communication sys-
Goldstein, H. (1984). Enhancing language generaliza- tem. Journal of Applied Behavior Analysis, 49, 388–
tion using matrix and stimulus equivalence training. 393.
In S. F. Warren & A. K. Rogers-Warren (Eds.), Teach- Leaf, J. B., Townley-Cochran, D., Mitchell, E., Milne,
ing functional language (Vol. 9, pp. 225–249). Balti- C., Alcalay, A., Leaf, J., et al. (2016). Evaluation of
more: University Park Press. multiple-alternative prompts during tact training.
Goldstein, H., & Mousetis, L. (1989). Generalized lan- Journal of Applied Behavior Analysis, 49, 399–404.
guage learning by children with severe mental retar- Macht, J. (1971). Operant measurement of subjective vi-
dation: Effects of peers’ expressive modeling. Journal sual acuity in nonverbal children. Journal of Applied
of Applied Behavior Analysis, 22, 245–259. Behavior Analysis, 4, 23–36.
Grow, L. L., Carr, J. E., Kodak, T. M., Jostad, C. M., & Mackay, H. A. (1985). Stimulus equivalence in rudi-
Kisamore, A. N. (2011). A comparison of methods mentary reading and spelling. Analysis and Interven-
for teaching receptive labeling to children with au- tion in Developmental Disabilities, 5, 373–387.
tism spectrum disorders. Journal of Applied Behavior Mackay, H. A., Stoddard, L. T., & Spencer, T. J. (1989).
Analysis, 44, 475–498. Symbols and meaning classes: Multiple sequence
Grow, L. L., Kodak, T. M., & Carr, J. E. (2014). A com- production and the emergence of ordinal stimulus
parison of methods for teaching receptive labeling to classes. Experimental Analysis of Human Behavior
children with autism spectrum disorders: A system- Bulletin, 7, 16–17.
atic replication. Journal of Applied Behavior Analysis, Marchese, N. V., Carr, J. E., LeBlanc, L. A., Rosati, T.
47, 600–605. C., & Conroy, S. A. (2012). The effects of the ques-
Halle, J. W., Baer, D. M., & Spradlin, J. E. (1981). Teach- tion “What is this?” on tact-training outcomes of
ers’ generalized use of delay as a stimulus control children with autism. Journal of Applied Behavior
procedure to increase language use in handicapped Analysis, 45, 539–547.
children. Journal of Applied Behavior Analysis, 14, McIlvane, W. J., Kledaras, J. B., Lowry, M. W., & Stod-
398–409. dard, L. T. (1992). Studies of exclusion in individuals
Halle, J. W., Marshall, A. M., & Spradlin, J. E. (1979). with severe mental retardation. Research in Develop-
Time delay: A technique to increase language use mental Disabilities, 13, 509–532.
and facilitate generalization in retarded children. McIlvane, W. J., & Stoddard, L. T. (1981). Acquisition
Journal of Applied Behavior Analysis, 12, 431–439. of matching-to-sample performances in severe men-
Handen, B. J., & Zane, T. (1987). Delayed prompting: A tal retardation: Learning by exclusion. Journal of
review of procedural variations and results. Research Mental Deficiency Research, 25, 33–48.
in Developmental Disabilities, 8, 307–330. McIlvane, W. J., & Stoddard, L. T. (1985). Complex
Harlow, H. F. (1949). The formation of learning sets. stimulus relations and exclusion in mental retarda-
Psychological Review, 56, 51–65. tion. Analysis and Intervention in Developmental Dis-
Hayes, S. C. (1991). A relational control theory of stim- abilities, 5, 307–321.
ulus equivalence. In L. J. Hayes & P. N. Chase (Eds.), McLay, L. K., Sutherland, D., Church, J., & Tyler-Mer-
Dialogues on verbal behavior (pp. 19–40). Reno, NV: rick, G. (2013). The formation of equivalence classes
Context Press. in individuals with autism spectrum disorder: A re-
Hayes, S. C., Barnes-Holmes, D., & Roche, B. (Eds.). view of the literature. Research in Autism Spectrum
(2001). Relational frame theory: A post-Skinnerian ac- Disorders, 7(2), 418–431.
count of language and cognition. New York: Kluwer Medin, D. L., & Smith, E. E. (1984). Concepts and
Academic/Plenum. concept formation. Annual Reviews of Psychology, 35,
Horne, P. J., & Lowe, C. F. (1996). On the origins of 113–138.
naming and other symbolic behavior. Journal of the Miller, S. A., Rodriguez, N. M., & Rourke, A. J. (2015).
Experimental Analysis of Behavior, 65, 185–241. Do mirrors facilitate acquisition of motor imitation
92 B a s i c P r in c ip l e s a nd C o n c e p t s
in children diagnosed with autism? Journal of Applied lection of comparison stimuli. Journal of the Experi-
Behavior Analysis, 48, 194–198. mental Analysis of Behavior, 50, 145–162.
Mueller, M., Olmi, J., & Saunders, K. J. (2000). Re- Saunders, R. R., & Sherman, J. A. (1986). Analysis of
combinative generalization of within-syllable units the “discrimination failure hypothesis” in generalized
in prereading children. Journal of Applied Behavior matching and mismatching behavior. Analysis and
Analysis, 33, 515–531. Intervention in Developmental Disabilities, 6, 89–107.
Oppenheimer, M., Saunders, R. R., & Spradlin, J. E. Saunders, R. R., Wachter, J., & Spradlin, J. E. (1988).
(1993). Investigating the generality of the delayed- Establishing auditory stimulus control over an eight-
prompt effect. Research in Developmental Disabilities, member equivalence class via conditional discrimi-
14, 425–444. nation procedures. Journal of the Experimental Analy-
Peterson, R. F. (1968). Some experiments on the orga- sis of Behavior, 49, 95–115.
nization of a class of imitative behaviors. Journal of Serna, R. W., Dube, W. V., & McIlvane, W. J. (1997).
Applied Behavior Analysis, 1, 225–235. Assessing same/different judgments in individuals
Pilgrim, C., & Galizio, M. (1990). Relations between with severe intellectual disabilities: A status report.
baseline contingencies and equivalence probe per- Research in Developmental Disabilities, 18, 342–368.
formances. Journal of the Experimental Analysis of Sidman, M. (1971). Reading and auditory–visual equiv-
Behavior, 54, 213–224. alences. Journal of Speech and Hearing Research, 14,
Reese, H. W. (1972). Acquired distinctiveness and 5–13.
equivalence of cues in young children. Journal of Ex- Sidman, M. (1994). Equivalence relations and behavior: A
perimental Child Psychology, 13, 171–182. research story. Boston: Authors Cooperative.
Rehfeldt, R. A. (2011). Toward a technology of derived Sidman, M. (2000). Equivalence relations and the re-
stimulus relations: An analysis of articles published in inforcement contingency. Journal of the Experimental
the Journal of Applied Behavior Analysis, 1992–2009. Analysis of Behavior, 74, 127–146.
Journal of Applied Behavior Analysis, 44, 109–119. Sidman, M., Cresson, O., Jr., & Willson-Morris, M.
Rosch, E. H. (1973). Natural categories. Cognitive Psy- (1974). Acquisition of matching to sample via medi-
chology, 4, 328–350. ated transfer. Journal of the Experimental Analysis of
Saunders, K. J. (2011). Designing instructional program- Behavior, 22, 261–273.
ming for early reading skills. In W. W. Fisher, C. C. Sidman, M., Kirk, B., & Willson-Morris, M. (1985). Six-
Piazza, & H. S. Roane (Eds.), Handbook of applied member stimulus classes generated by conditional-
behavior analysis (pp. 92–109). New York: Guilford discrimination procedures. Journal of the Experimen-
Press. tal Analysis of Behavior, 43, 21–42.
Saunders, K. J., Johnson, M. D., Tompkins, B. F., Sidman, M., & Stoddard, L. T. (1967). The effective-
Dutcher, D. L., & Williams, D. C. (1997). General- ness of fading in programming a simultaneous form
ized identity matching of two-dimensional forms by discrimination for retarded children. Journal of the
individuals with moderate to profound mental retar- Experimental Analysis of Behavior, 10, 3–15.
dation. American Journal of Mental Retardation, 102, Sidman, M., & Tailby, W. (1982). Conditional discrimi-
285–291. nation vs. matching-to-sample: An expansion of the
Saunders, K. J., O’Donnell, J., Vaidya, M., & Williams, testing paradigm. Journal of the Experimental Analysis
D. C. (2003). Recombinative generalization of with- of Behavior, 37, 5–22.
in-syllable units in non-reading adults with mental Siegel, G. M., & Spradlin, J. E. (1978). Programming
retardation. Journal of Applied Behavior Analysis, 36, for language and communication therapy. In R. L.
95–99. Schiefelbusch (Ed.), Language intervention strategies
Saunders, K. J., & Spradlin, J. E. (1993). Conditional (pp. 357–398). Baltimore: University Park Press.
discrimination in mentally retarded subjects: Pro- Sigurdardottir, Z. G., Green, G., & Saunders, R. R.
gramming acquisition and learning set. Journal of the (1990). Equivalence classes generated by sequence
Experimental Analysis of Behavior, 60, 571–585. training. Journal of the Experimental Analysis of Be-
Saunders, K. J., Williams, D. C., & Spradlin, J. E. (1995). havior, 53, 47–63.
Conditional discrimination by adults with mental re- Skinner, B. F. (1938). The behavior of organisms. New
tardation: Establishing relations between physically York: Appleton-Century-Crofts.
identical stimuli. American Journal on Mental Retar- Smeets, P. M., & Striefel, S. (1974). Oddity and match-
dation, 99, 558–563. to-sample tasks as the components of a chained
Saunders, R. R., & Green, G. (1992). The nonequiva- schedule with retarded children. American Journal of
lence of behavioral and mathematical equivalence. Mental Deficiency, 4, 462–470.
Journal of the Experimental Analysis of Behavior, 57, Smeets, P. M., & Striefel, S. (1994). A revised blocked-
227–241. trial procedure for establishing arbitrary matching in
Saunders, R. R., Saunders, K. J., Kirby, K. C., & Spra- children. Quarterly Journal of Experimental Psychol-
dlin, J. E. (1988). The merger and development of ogy, 47B, 241–261.
equivalence classes by unreinforced conditional se- Soraci, S. A., Jr., Deckner, C. W., Haenlein, M., Bau-
Stimulus Control and Generalization 93
meister, A. A., Murata-Soraci, K., & Blanton, R. L. Terrace, H. S. (1966). Stimulus control. In W. K. Honig
(1987). Oddity performance in preschool children (Ed.), Operant behavior: Areas of research and applica-
at risk for mental retardation: Transfer and main- tion (pp. 271–344). New York: Appleton–Century–
tenance. Research in Developmental Disabilities, 8, Crofts.
137–151. Touchette, P. E. (1971). Transfer of stimulus control:
Spradlin, J. E., Cotter, V. W., & Baxley, N. (1973). Es- Measuring the moment of transfer. Journal of the Ex-
tablishing a conditional discrimination without di- perimental Analysis of Behavior, 15, 347–354.
rect training: A study of transfer with retarded ado- Touchette, P. E., & Howard, J. S. (1984). Errorless learn-
lescents. American Journal of Mental Deficiency, 77, ing: Reinforcement contingencies and stimulus con-
556–566. trol transfer in delayed prompting. Journal of Applied
Spradlin, J. E., & Saunders, R. R. (1986). The devel- Behavior Analysis, 17, 175–188.
opment of stimulus classes using match-to-sample Urcuioli, P. J. (2013). Stimulus control and stimulus
procedures: Sample classification versus comparison class formation. In G. J. Madden, W. V. Dube, T. D.
classification. Analysis and Intervention in Develop- Hackenberg, G. P. Hanley, & K. A. Lattal (Eds.),
mental Disabilities, 6, 41–58. APA handbook of behavior analysis (Vol. 1, pp. 361–
Stewart, I., McElwee, J., & Ming, S. (2013). Language 386). Washington, DC: American Psychological As-
generativity, response generalization, and derived re- sociation.
lational responding. Analysis of Verbal Behavior, 29, VanBiervliet, A. (1977). Establishing words and ob-
137–155. jects as functionally equivalent through manual sign
Stoddard, L. T., Brown, J., Hurlbert, B., Manoli, C., training. American Journal of Mental Deficiency, 82,
& McIlvane, W. J. (1989). Teaching money skills 178–186.
through stimulus class formation, exclusion, and Wetherby, B., Karlan, G. R., & Spradlin, J. E. (1983).
component matching methods: Three case studies. The development of derived stimulus relations
Research in Developmental Disabilities, 10, 413–439. through training in arbitrary-matching sequences.
Stokes, T. F., & Baer, D. M. (1977). An implicit tech- Journal of the Experimental Analysis of Behavior, 40,
nology of generalization. Journal of Applied Behavior 69–78.
Analysis, 10, 349–367. Wetherby, B., & Striefel, S. (1978). Application of
Striefel, S., Wetherby, B., & Karlan, G. R. (1976). Estab- miniature linguistic system or matrix-training pro-
lishing generative verb–noun instruction-following cedures. In R. L. Schiefelbusch (Ed.), Language inter-
skills in retarded children. Journal of Experimental vention strategies (pp. 317–356). Baltimore: University
Child Psychology, 22, 247–260. Park Press.
Striefel, S., Wetherby, B., & Karlan, G. R. (1978). Devel- Wilkinson, K. M., & Green, G. (1998). Implications of
oping generalized instruction-following behavior in fast mapping for vocabulary expansion in individuals
severely retarded people. In C. E. Meyers (Ed.), Qual- with mental retardation. Augmentative and Alterna-
ity of life in severely and profoundly mentally retarded tive Communication, 14, 162–170.
people: Research foundations for improvement (Vol. 3, Wilkinson, K. M., & McIlvane, W. J. (1994). Stimulus
pp. 267–326). Washington, DC: American Associa- organization and learning by exclusion: A prelimi-
tion on Mental Deficiency. nary experimental analysis. Experimental Analysis of
Stromer, R., & Mackay, H. A. (1992). Delayed con- Human Behavior Bulletin, 12, 21–25.
structed–response identity matching improves the Wultz, S. V., & Hollis, J. H. (1980). Word identifica-
spelling performance of students with mental retar- tion and comprehension training for exceptional
dation. Journal of Behavioral Education, 2, 139–156. children. In R. L. Schiefelbusch (Ed.), Nonspeech
Stromer, R., & Stromer, J. B. (1989). Children’s identity language and communication: Analysis and interven-
matching and oddity: Assessing control by specific tion (pp. 359–387). Baltimore: University Park Press.
and general sample-comparison relations. Journal of Wunderlich, K. L., Vollmer, T. R., Donaldson, J. M., &
the Experimental Analysis of Behavior, 51, 47–64. Phillips, C. L. (2014). Effects of serial and concurrent
Stromer, R., & Stromer, J. B. (1990a). The formation training on acquisition and generalization. Journal of
of arbitrary stimulus classes in matching to complex Applied Behavior Analysis, 47, 1–15.
samples. Psychological Record, 41, 51–66. Zinn, T. E., Newland, M. C., & Ritchie, K. E. (2015).
Stromer, R., & Stromer, J. B. (1990b). Matching to The efficiency and efficacy of equivalence-based
complex samples: Further study of arbitrary stimulus learning: A randomized controlled trial. Journal of
classes. Psychological Record, 40, 505–516. Applied Behavior Analysis, 48, 865–882.
CHAP TER 6
Christopher A. Podlesnik, Corina Jimenez-Gomez,
and Michael E. Kelley
The matching law and behavioral momentum With all the ways practitioners might manipu-
theory are quantitative theoretical frameworks late reinforcement contingencies, those contin-
developed to understand how reinforcement af- gencies can influence behavior in complex ways.
fects behavior. The matching law describes how In some cases, the immediate effects of contingen-
changes in relative reinforcement variables influ- cies differ from long-term effects. To address these
ence the allocation of operant behavior or choice. different effects of reinforcement, the purpose of
Behavioral momentum theory asserts that the the present chapter is to describe research showing
matching law governs the allocation of operant how reinforcement contingencies affect (1) the al-
behavior, but proposes that Pavlovian processes location of behavior, (2) the persistence of behavior,
govern the persistence of behavior. This chapter and (3) the allocation and persistence of behav-
describes matching and behavioral momentum as ior in different and perhaps counterintuitive ways.
quantitative frameworks for understanding vari- Thus we will argue that behavior analysts should
ables influencing behavior. We also discuss their attend to both the allocation and persistence of
implications for behavioral treatments. behavior, which they can understand through two
The primary goal of behavioral science is to distinct but related areas of research. Research
identify systematic relations between environ- on choice and the matching law provides insight
mental events and behavior (e.g., Nevin, 1984). into the variables influencing the allocation of
behavior. Research on persistence and behavioral
Essentially, one asks how changing something in
momentum theory provides insight into variables
the environment (i.e., the independent variable)
influencing both the allocation and persistence of
affects behavior (i.e., the dependent variable). Re-
behavior. The literature on the matching law and
inforcement is one of the primary tools of environ-
behavioral momentum are extensive, and the pres-
mental change that applied behavior analysts use
ent review is quite selective and directed toward
to effect behavior change. Contingencies between introducing these areas.1
behavior and reinforcement can have powerful ef-
fects, as demonstrated by the many ways behavior 1 Those interested in a contemporary theoretical integra-
analysts use it in treatments to increase desirable tion of choice and persistence should find the framework
behavior and decrease problem behavior. described by Shahan and Craig (2017) informative.
94
Matching and Behavioral Momentum 95
inforcing every instance of appropriate behavior. relative to R2, B1 increased relative to B2, as de-
In routine clinical settings (e.g., homes), caregiv- scribed by Equation 1:
ers might inadvertently reinforce instances of
problem behavior and not reinforce instances of B1 R1
= (1)
appropriate behavior (Athens & Vollmer, 2010). B1 + B2 R1 + R2
Therefore, these methods of arranging intermit-
tent reinforcement of problem and appropriate be- Herrnstein (1961) introduced Equation 1 to
havior simulate situations in which treatment in- describe pigeons’ choices in situations resembling
tegrity is compromised (St. Peter Pipkin, Vollmer, those arranged by Borrero et al. (2010). Specifical-
& Sloman, 2010). ly, pigeons pecked between two lighted keys, with
These methods also promote understanding of food reinforcement concurrently available on VI
how reinforcement processes influence the alloca- VI schedules (i.e., R1 vs. R2) that changed across
tion of problem and appropriate behavior. After all, conditions. Parametric manipulations of relative
the variables that influence behavioral phenome- reinforcement rates (i.e., across a range of differ-
na do so at the level of fundamental processes. Any ent levels of the independent variable) produced
treatment or corresponding decrease in treatment proportional rates of behavior that approximately
integrity affects clients’ behavior at the level of matched the obtained relative reinforcement rates;
behavioral processes, which is the level at which these findings were generally consistent with Fig-
studies of choice and quantitative analyses, begin- ure 6.1. Thus Borrero et al.’s findings in a clinical
ning with the matching law, become relevant. setting resembled those with pigeons in a con-
trolled laboratory setting.
Equation 1 suggests that both the rate of rein-
THE STRICT MATCHING LAW forcement for the target response and the rate of
reinforcement for other responses controls the rate
The findings of Borrero et al. (2010) fit not only of the target behavior. The effects of a given rein-
with the general notion that reinforcement influ- forcement rate (R1) depend on the context of other
ences clinically relevant behavior, but also with concurrently available sources of reinforcement.
quantitative assessments of choice behavior. Spe- Response rates might be high or low, depending on
cifically, the proportion of responding allocated whether an alternative source of reinforcement is
between problem behavior (B1) and appropriate low or high, respectively (e.g., Findley, 1958; Her-
behavior (B2) approximately matched the propor- rnstein, 1961, 1970).
tion of reinforcement allocated between problem
behavior (R1) and appropriate behavior (R2). Fig-
ure 6.1 shows an idealized version of such findings, THE QUANTITATIVE LAW OF EFFECT
to which we refer as we go along. As R1 increased
R2 is the only alternative source of reinforcement
for which Equation 1 accounts. Nevertheless, most
environments will include multiple sources of rein-
forcement; even in experimental situations, there
are likely to be sources of reinforcement other than
those arranged explicitly. For a pigeon in a largely
barren operant chamber, these might include re-
inforcers for grooming or tending to other bodily
functions. Humans might whistle or daydream.
Under natural conditions, the possible alternative
sources of reinforcement are endless. Herrnstein’s
important insight was that even simple schedules
of reinforcement—situations with only one explic-
itly arranged reinforcement schedule—necessarily
involve choice between the scheduled source of
FIGURE 6.1. The basic matching relation predicted from reinforcement and all other sources of reinforce-
Equation 1. The equation predicts that the proportion ment. That is, organisms choose whether to en-
of responses (B) should equal the proportion of reinforc- gage in the target response or in an undefined
ers (R). number of other responses.
Matching and Behavioral Momentum 97
log b parameter in Equation 3 is the y-intercept variance in a dataset a model accounts for with
of the line and provides an index of bias of the measures like r2. An imperfect model fit could be
behavior ratio toward one source of reinforcement due to random variation in behavior, but a model
over another. also could make systematic errors in prediction,
If the log–behavior ratios on the y-axis approxi- thereby indicating that the model makes incorrect
mately equal the log–reinforcement ratios on the assumptions. Another important consideration
x-axis, this is strict matching, as the dotted line in is whether a model’s parameters make sense, and
Figure 6.3 shows. Log axes represent each 10-fold thus whether we can use them to say something
increase with equal spacing: A ratio of 1:10 equals meaningful about the variables controlling be-
–1, 1:1 equals 0, 10:1 equals 1, 100:1 equals 2, and havior. That is, do the free parameters reflect a
so forth. With strict matching, (1) the ar parame- relevant and realistic aspect of behavior, or the en-
ter, indicated by the slope, equals 1; and (2) the log vironment that guides the endeavors of behavior
b parameter, indicated by the y-intercept, equals 0. analysts? These issues are beyond the scope of this
The other functions in Figure 6.3 reveal deviations chapter, but others have discussed them in detail
from strict matching. If the range of log–behav- elsewhere (e.g., Dallery & Soto, 2013; Davison &
ior ratios is less than the range of log–reinforce- McCarthy, 1988; Shull, 1991).
ment ratios, the a parameter is less than 1, and log Much of our discussion of choice and matching
b equals 0 (white data points). We often call this thus far has focused on clarifying how the equa-
function undermatching, because the log–behavior tion works. Now we discuss which variables can
ratios are less extreme than the log–reinforcer ra- influence choice, how the equation accounts for
tios. For example, Borrero et al. (2010) arranged 3 variables influencing choice, and why an under-
times more reinforcement for appropriate behavior standing of matching can be useful to behavioral
than problem behavior in some conditions and clinicians.
vice versa; undermatching would have occurred
if the behavior ratio was less than 3 times greater
Variables Affecting Sensitivity
(e.g., only 2:1). A bias for B1 occurs if the range
of log–behavior ratios spans the same range as The slope of a function when fitting the general-
log–reinforcement ratios, and the function shifts ized matching equation to data expresses sensitiv-
toward B1 and log b is greater than 0 (gray data ity of choice to reinforcement conditions (Equa-
points). tion 3). In Borrero et al. (2010), one participant,
Figure 6.3 reveals only two of the most basic Greg, engaged in problem behavior and appropri-
deviations from strict matching. Specifically, the ate behavior within 1 second of each other when
slope of the equation can be greater than 1, which both behaviors produced access to tangibles. The
is called overmatching; the y-intercept can be nega- researchers observed that the responses formed a
tive (bias for B2); and Equation 3 can describe mul- chain, even though they had arranged separate
tiple deviations from strict matching. That is, both VI schedules for the two responses. That is, the
slope and y-intercept might deviate simultaneously VI schedules did not require the successive emis-
from the dotted line in Figure 6.3 (e.g., ar < 1 and sion of both responses to produce reinforcement.
log b > 1). When we use Equation 3 to plot devia- In Equation 1, rapid alternation between two re-
tions from strict matching, changes in slope and sponses produces a slope approaching 0, despite
y-intercept characterize these deviations. Plotting parametric changes in relative reinforcement
the same data using Equation 1 produces curvilin- rates. Borrero et al. introduced a 5-second change-
ear functions that are very difficult to characterize over delay to separate the two VI reinforcement
intuitively. Thus the added complexity of intro- schedules for problem and appropriate behavior in
ducing logarithms with Equation 3 is more than time. Introducing the change-over delay success-
justified by simplifying the data paths and the fully increased sensitivity of choice to the different
interpretations and conclusions stemming from VI schedules (Herrnstein, 1961; Shull & Pliskoff,
them. 1967). In general, researchers believe that change-
Finally, fitting any equation to data requires over delays increase the discriminability between
some assessment of how well the model accounts the two concurrently available reinforcement
for the data. This aspect of model fitting can be- schedules (Davison & Nevin, 1999) and function
come rather complex because a model fit to data analogously to increasing the spatial separation
can be less than perfect for multiple reasons. between responses (Baum, 1982). Longer duration
Assessing models involves assessing how much change-over delays (Shull & Pliskoff, 1967) and
100 B a s i c P r in c ip l e s a nd C o n c e p t s
more obstructive barriers between responses (e.g., sitive to changes in the log–reinforcer ratio, but
Aparicio & Baum, 1997; Baum, 1982) increase there is a shift in choice toward one alternative.
sensitivity to changes in relative reinforcement In Borrero et al. (2010), the relative rate of prob-
rates. lem and appropriate behavior with the participant
Antecedent conditions also can influence sen- named Greg showed sensitivity to changes in rela-
sitivity. An increase in the discriminability be- tive reinforcement rates. However, Greg reliably
tween antecedent stimuli that signals the response engaged in a higher rate of appropriate behavior
alternatives also increases sensitivity to changes than problem behavior than Equation 1 predicted.
in relative reinforcement rate (Alsop & Davison, Thus bias expressed by the y-intercept shifted to-
1991; Miller, Saunders, & Bourland, 1980). When ward appropriate behavior. Interestingly, escape
a change-over response controlled pigeons’ choic- from demands maintained a different topography
es between VI schedules in these studies, greater of Greg’s problem behavior, and relative response
differences between discriminative stimuli sig- rates showed bias toward problem behavior over
naling the two VI schedules increased sensitivity appropriate behavior. Thus bias can differ between
to changes in relative reinforcement rate. These problem and appropriate behaviors, depending
findings with change-over requirements and dif- on the circumstances—even for the same indi-
ferences in antecedent stimuli suggest that manip- vidual. How does the matching framework inform
ulations increasing the discriminability between us about variables producing bias toward one re-
reinforcement contingencies will increase sensi- sponse over another?
tivity to changes in relative reinforcement rates The cause of bias for Greg in Borrero et al.
(Davison & Nevin, 1999; Fisher, Pawich, Dickes, (2010) for one response over another might not
Paden, & Toussaint, 2014). be clear. It could have been due to differences
Lastly, organismic variables influence sensitiv- between problem and appropriate behavior in
ity to changes in relative reinforcement rates. For response variables (e.g., effort) or reinforcer vari-
example, Oscar-Berman, Heyman, Bonner, and ables, such as magnitude, immediacy, and qual-
Ryder (1980) parametrically assessed a range of ity. In these cases, log b in Equation 3 quantifies
concurrently available VI schedules of monetary any inherent bias for one response over another.
reinforcement between participants with Korsa- When we know the differences in response and
koff syndrome versus control participants. Rela- reinforcement variables, we can expand Equation
tive response rates for participants with Korsakoff 3 to account for the effects of variables other than
syndrome showed less sensitivity to changes in changes in relative reinforcer rate (Baum & Rach-
relative reinforcement rates than those of the con- lin, 1969; Davison & McCarthy, 1988; Grace &
trol participants did. Similarly, Buckley and Ras- Hucks, 2013; Killeen, 1972):
mussen (2012) assessed matching in two strains
of rats: obese and lean Zucker rats. Obese Zucker B R X
rats express an obese phenotype and eat more log 1 = a r log 1 + a x log 1 + log b (4)
B2 R2 X2
when freely fed than lean Zucker rats. Buckley and
Rasmussen found that obese Zucker rats’ relative Note that the only difference between Equation
response rates were more sensitive to parametric 3 and Equation 4 is the addition of the bracketed
changes to concurrently available VI schedules of portion of Equation 4. With R representing rein-
food reinforcement. These findings reveal the use- forcer rate, X is a generic variable that may rep-
fulness of the sensitivity parameter from Equation resent differences in the response and reinforcer
3 for characterizing differences in responsiveness variables mentioned above (e.g., response effort,
to changes in reinforcement conditions due to or- reinforcer magnitude). The X variable may also
ganismic variables, such as differences in genes or represent differences in reinforcement history
chronic versus acute drug effects (e.g., Newland, (Davison & McCarthy, 1988) in the sense that
Reile, & Langston, 2004). problem behavior could have a longer history of
reinforcement than appropriate behavior, thereby
causing the bias for problem behavior over appro-
Variables Affecting Bias
priate behavior. As with relative reinforcer rate,
The y-intercept of a function when the generalized the equation includes a sensitivity parameter, ax,
matching equation is fitted to data expresses bias to account for sensitivity to the difference in X1
for one reinforced alternative over others (Equa- and X2. Assuming no inherent bias (log b = 0), if
tion 3). Thus the log–response ratio remains sen- the relative reinforcer magnitude is 4 times greater
Matching and Behavioral Momentum 101
for X1 than X2, but the relative response ratio is choice, but are missing entirely from Equations
only 2 times greater for B1 than B2, the estimated 1–4. The requirement for nearly perfectly dis-
sensitivity to reinforcer magnitude (i.e., am) from criminated response–reinforcer contingencies to
log b would be 0.5. observe log–response ratios closely matching log–
The important point is that Equation 4 can ac- reinforcer ratios is directly relevant to the findings
count for how differences in independent variables described above from Alsop and Davison (1991).
can bias responding toward or away from one re- Reducing differences in antecedent stimuli signal-
sponse alternative or another when we manipulate ing two response alternatives reduced sensitivity
some other independent variable parametrically. to changes in relative reinforcement rates. Thus,
We often manipulate R1 and R2 parametrically despite the fact that Equations 3 and 4 provide ad-
across conditions, and hold other variables con- equate descriptions of choice across an impressive
stant and equal. We should hold only one variable range of conditions (Davison & McCarthy, 1988;
constant and different (e.g., relative reinforcer Grace & Hucks, 2013), these variables affecting
magnitude) while parametrically manipulating bias and sensitivity show that the matching rela-
another (e.g., relative reinforcer rate) to determine tion between behavior and reinforcement is not
whether variables influence bias Note that we can the only process governing choice. Researchers
hold R1 and R2 constant while we vary X1 and X2 continue to debate whether other processes are
parametrically to obtain an estimate of sensitiv- necessary and, if so, what those processes might be
ity to the changes in a variable characterized by (Baum, 2010; Cording, McLean, & Grace, 2011;
X in Equation 4 (e.g., response effort, reinforcer Elliffe, Davison, & Landon, 2008; Sutton, Grace,
magnitude). McLean, & Baum, 2008); this debate goes beyond
the scope of this chapter.
CHALLENGES TO GENERALIZED MATCHING
CHOICE AND PERSISTENCE
Equations 3 and 4 have been useful for describ-
ing how changes to response and reinforcement Matching, characterized by Equations 1–4, de-
variables influence choice. However, these equa- scribes how reinforcement variables influence the
tions predict only data falling along straight lines allocation of behavior in an environment. Thus
because they use linear regression. Several find- matching provides a general framework for under-
ings produce functions that violate this required standing how behavior becomes distributed among
relation between response and reinforcer ratios. existing sources of reinforcement. We observe that
Researchers have used such findings to evaluate a given response rate depends on reinforcement
processes potentially underlying choice. contingent on that response (e.g., reinforcement
In one example, Davison and Jones (1995) as- rate or magnitude) and on all other available sourc-
sessed a range of reinforcer ratios beyond what re- es of reinforcement (e.g., Herrnstein, 1961, 1970).
searchers typically examine. A straight line with Thus a given response, such as problem behavior,
sensitivity estimates approaching 1.0 character- will decrease in rate if an alternative response,
ized the data well when they assessed a typical such as appropriate behavior, produces a greater
range of reinforcer–rate ratios from 1:10 to 10:1 reinforcement rate (e.g., Borrero et al., 2010). The
(i.e., log values of –1 to +1). However, arranging a allocation of behavior relative to available sources
very rich versus a very lean reinforcement sched- of reinforcement is only one consideration, es-
ule extended the reinforcer ratios well beyond the pecially when we are planning for the long-term
normal range—out to 1:100 and 100:1 (i.e., log effectiveness of behavioral treatments. What are
values of –2 to 2). This produced a flattening of the factors determining whether problem behavior
the function between log values of ±1 to ±2. Davi- will persist or relapse, or whether treatments will
son and Jones suggested that discrimination of the continue to be effective? The theoretical frame-
response–reinforcer contingencies decreased at work of behavioral momentum primarily concerns
these more extreme ratios. Specifically, reinforc- the role of learning or training conditions in how
ers obtained from the currently richer alternative likely reinforced behavior is to continue, despite
became misallocated to the leaner alternative (see disruptive challenges to treatment.
Davison & Jenkins, 1985). This contingency–dis- Mace et al. (2010) arranged reinforcement rates
criminability interpretation suggests that stimu- for problem behavior and appropriate behavior
lus control processes play a fundamental role in like those Borrero et al. (2010) arranged. Rate of
102 B a s i c P r in c ip l e s a nd C o n c e p t s
appropriate behavior was greater when reinforce- intermittent access to a different preferred toy in-
ment rate for appropriate behavior was greater dependently of behavior during an NCR schedule.
than for problem behavior. These findings gen- Even though levels of stereotypy decreased during
erally are consistent with the predictions of the the initial NCR schedule, levels of stereotypy were
matching law: Relative response rates track rela- higher when the persistence assessment followed
tive reinforcement rates. In addition, these find- a phase of NCR versus a phase of baseline. These
ings are consistent with the goals of DRA treat- findings resemble those of Mace et al. (2010) with
ments to reduce problem behavior. Thus DRA DRA treatment: Additional reinforcement pre-
treatments are successful in altering the alloca- sented either contingent on a different response
tion of appropriate and problem behavior, as the (DRA) or noncontingent on response (NCR) de-
matching law describes. creased the rate or level of problem behavior dur-
Mace et al. (2010) assessed another important ing treatment implementation, but enhanced the
component of behavior: persistence. Mace et al. persistence of problem behavior when exposed to
arranged baseline conditions in which problem a disruptor. Therefore, two very commonly imple-
behavior produced reinforcement and appropriate mented behavioral treatments appeared effective
behavior produced no differential consequence. initially, but compromised long-term treatment
After baseline, Mace et al. then arranged a se- effectiveness by making problem behavior more
quence of phases in which a phase of extinction resistant to other environmental challenges to
for problem and appropriate behavior followed ei- treatment.
ther a baseline phase of reinforcement for problem The findings of Mace et al. (2010) and Ahearn
behavior and extinction for appropriate behavior, et al. (2003) are similar because they appear to re-
or a phase of DRA plus reinforcement for prob- flect common behavioral processes. Nevertheless,
lem behavior. The researchers counterbalanced these findings on their own might not convince
the order of phases so that each participant expe- readers of the reliability or importance of these ef-
rienced an extinction phase preceded by a phase fects. However, researchers have obtained similar
of either baseline or DRA plus reinforcement for effects repeatedly across many experimental ar-
problem behavior. Problem behavior decreased rangements and species—from fish, rats, and pi-
to near-zero rates in approximately 10 sessions in geons, to humans with intellectual developmental
extinction phases that followed baseline phases. disorder and to neurotypical humans (e.g., Cohen,
By contrast, problem behavior persisted nearly 3 1996; Igaki & Sakagami, 2004; Mace et al., 1990;
times longer and occurred at higher rates in ex- Mauro & Mace, 1996; Nevin, Tota, Torquato,
tinction phases that followed DRA plus reinforce- & Shull, 1990; Nevin & Wacker, 2013; Shahan
ment for problem behavior phases. Thus rates of & Burke, 2004). Moreover, several studies with
problem behavior during DRA plus reinforcement human and nonhumans also have shown that
for problem behavior were lower than those in even when alternative reinforcement decreases
baseline, but the DRA inadvertently increased the behavior, that behavior is likely to relapse later
persistence of problem behavior during an extinc- in the absence of the contingencies for the alter-
tion challenge. native reinforcement (e.g., Kuroda, Cançado, &
Ahearn, Clark, Gardenier, Chung, and Dube Podlesnik, 2016; Miranda-Dukoski, Bensemann,
(2003) observed similar response persistence when & Podlesnik, 2016; Podlesnik & Shahan, 2009,
they used NCR schedules to treat automatically 2010). Therefore, arranging alternative sources
maintained stereotypy in three children with in- of reinforcement that decrease the rate of a target
tellectual developmental disorder. They provided behavior but enhance its persistence is a common
participants with intermittent access to a pre- and robust finding. Behavioral momentum theory
ferred toy independently of behavior during the provides a framework within which to understand
NCR schedule, which decreased levels of stereo- these effects and potential ways to ameliorate
typy relative to the absence of NCR. Equation 2 them.
predicts the reduction in stereotypy with NCR.
Specifically, NCR increased Re, which was in
competition with reinforcement for engaging in BEHAVIORAL MOMENTUM THEORY
stereotypy, R. Ahearn et al. then assessed the per-
sistence of stereotypy in the presence and absence Behavioral momentum theory assumes that two
of the NCR schedule. During the assessment of separate environmental relations govern behavior
this persistence, they provided participants with in the same way that two separate variables govern
Matching and Behavioral Momentum 103
the movement of physical objects: mass and veloc- bursts of responses. The patterns with DRL and
ity. Just as objects of high or low mass can move at DRH schedules reflect the effects of the different
high or low velocity, behaviors also have proper- contingencies in shaping response rates. Essential-
ties like mass and velocity. The rate of a behavior is ly, the contingencies alter the functional unit of
analogous to the velocity of an object, and the per- responding. DRL schedules result in a pause-then-
sistence of a behavior is analogous to the mass of an respond response unit, and DRH produces bursts
object. Thus different behaviors can occur at high of responses as a unit. These units could change
or low rates that are easy or difficult to disrupt. in frequency with changes in reinforcement rate,
Research on behavioral momentum evaluates the as described by the matching law (Staddon, 1968).
variables influencing response rates (behavioral Similarly, variable-ratio (VR) and VI schedules
velocity) and persistence (behavioral mass). produce different response rates even when rein-
forcement rate is controlled. The important con-
sideration for the present purpose is the idea that
Response Rate and Operant
response rate itself is a conditionable dimension of
Response–Reinforcer Relations
behavior (Nevin, 1974). Again, behavioral veloc-
The discriminated operant is the fundamental ity is an operant process.
unit of operant behavior, which includes a dis- As Equation 2 describes, the increases in re-
criminative stimulus, a response, and a conse- sponse rates with increases in reinforcement rates
quence, according to Skinner (1969). Behavioral led some to conclude that response rate provides
momentum theory proposes that the operant an inappropriate measure of the fundamental pro-
relation between responding and consequences cesses underlying reinforcement effects, response
govern response rate in a manner consistent strength (Herrnstein, 1961, 1970; Nevin & Grace,
with the matching law (Nevin, 2015; Nevin & 2000). For these reasons, Nevin (1974) offered a
Grace, 2000). As described in Equations 1–4, different approach to assessing the processes un-
the rate of an operant behavior depends on the derlying reinforcement.
reinforcement rate for that response and the rein-
forcement rate for all other responses. Consistent
Persistence and Pavlovian
with Equation 1, Mace et al. (2010) observed that
Stimulus–Reinforcer Relations
decreases in rates of problem behavior (B1) were
reinforced at low rates (R1) when appropriate al- Nevin (1974) suggested that a more appropriate
ternative behavior (B2) was reinforced at a higher method for evaluating response strength is to as-
rate (R2). Consistent with Equation 2, Ahearn et sess the way responding changes when some dis-
al. (2003) observed decreases in rates of stereotypy ruptive event challenges it—a measure he called
(B) maintained by automatic reinforcement (R) resistance to change (see also Nevin & Wacker,
when providing response-independent access to a 2013). Resistance to change assesses response
toy (Re). Therefore, available alternative sources rates during conditions of disruption compared to
of reinforcement degrade the effectiveness of the previous baseline response rates; it is a measure of
operant response–reinforcer relation. Specifically, behavioral persistence. For example, introducing
they reduce the proportion of reinforcers delivered extinction, distracting stimuli, or providing addi-
contingently on the target response, and thereby tional food before or during sessions can disrupt
decrease the correlation between the response and food-maintained responding under steady-state
the reinforcer. The velocity of behavior is an oper- conditions (e.g., Mace et al., 1990, 2010; Nevin
ant process. et al., 1990). These disruptive events decrease re-
Furthermore, response–reinforcer contingen- sponse rates, which we can compare among mul-
cies can shape different patterns of responding tiple responses. Persistent responses are ones that
(see Morse, 1966). For example, differential rein- decrease less rapidly with more disruptive events
forcement of low behavior rates (DRL) produces (e.g., greater amounts of additional food or suc-
low rates of responding by arranging reinforce- cessive sessions of extinction), and we call those
ment with sufficiently spaced interresponse times. responses resistant to change.
Thus DRL schedules increase the likelihood of As mass and velocity are separable aspects of
pausing between responses. On the other hand, physical momentum, response rates and persis-
differential reinforcement of high behavior rates tence are products of separable processes in behav-
(DRH) produces high rates of responding by ar- ioral momentum (Bell, 1999; Grace, Schwendi-
ranging reinforcement with little or no pausing or man, & Nevin, 1998; Podlesnik, Jimenez-Gomez,
104 B a s i c P r in c ip l e s a nd C o n c e p t s
Ward, & Shahan, 2006; Podlesnik & Shahan, that are lower in baseline when alternative rein-
2008). Behavioral momentum theory asserts that forcement is available than when it is not available
the persistence of behavior results from a Pavlov- can be greater under conditions of disruption (e.g.,
ian process (i.e., respondent or classical condition- Ahearn et al., 2003; Grimes & Shull, 2001; Mace
ing). In Pavlovian conditioning, a conditioned et al., 2010; Nevin et al., 1990).
response (CR) occurs to a conditioned stimulus Mace et al. (2010) compared resistance to ex-
(CS) when the CS forms a predictive relation with tinction following a DRA schedule that arranged
an evolutionarily relevant unconditioned stimulus low reinforcement rates for problem behavior
(US; Domjan, 2005, 2016; Lattal, 2013; Rescorla, and greater reinforcement rates for appropriate
1988). A dog salivates (CR) in the presence of a behavior. Problem behavior persisted more dur-
tone (CS) when the tone reliably predicts access ing extinction following the DRA schedule than
to food (US). Similarly, behavioral momentum following the baseline condition when only prob-
theory proposes that behavioral persistence is lem behavior produced reinforcement. Thus add-
functionally a CR expressed in the presence of ing alternative reinforcement decreased the rate
the environmental discriminative-stimulus con- of problem behavior as described by matching in
text governing the target behavior. As CSs that Equation 2, but only by degrading the operant
more reliably predict USs produce more robust response–reinforcer relation. The additional al-
CRs, environmental stimuli more predictive of re- ternative reinforcement with the DRA schedule
inforcement will produce greater persistence. From increased the total overall reinforcement rate in
the discriminated operant, behavioral momentum the current environment, compared with the ab-
theory asserts that Pavlovian stimulus–reinforcer sence of the DRA schedule. Because the DRA
relations govern persistence (Nevin & Shahan, schedule increased the extent to which the cur-
2011): rent environmental context predicted reinforce-
−x ment overall, DRA enhanced the Pavlovian stimu-
Bx b lus–reinforcer relation, compared to the absence
= 10 r (5)
B0 of the DRA schedule (see Nevin, 1997; Nevin
& Wacker, 2013). According to Equation 5, the
Bx is response rate during disruption, and B0 is DRA increased r, thereby countering the disrup-
training response rate; these terms represent the tive effects of extinction more than when r was
change in behavioral velocity from baseline to lower during the baseline without DRA. Similarly,
disruption. On the right side of the equation, any Ahearn et al. (2003) showed that NCR decreased
terms in the numerator of the exponent contrib- the ongoing rate of stereotypy but increased its
ute to the disruption of target responding relative persistence. The addition of a preferred toy as
to training response rates; this is disruptive force, NCR increased the total overall reinforcement
represented by the generic term x. Terms in the rate in the current environment, compared to the
denominator contribute to countering those dis- environment without NCR (i.e., r in the denomi-
ruptive effects; this is behavioral mass and is a nator in Equation 5). As with DRA in Mace et al.,
function of the baseline reinforcement rate, r. As Ahearn et al. increased persistence by enhancing
described above, r can be contingent on the target the Pavlovian stimulus–reinforcer relation with
response or from other sources. The free param- NCR. Greater reinforcement rates in the presence
eter b is the sensitivity parameter, which scales the of discriminative stimuli reliably produce greater
persistence-enhancing effects of r on resistance persistence when we arrange (1) all reinforcers
to change on the left side of the equation. There- presented dependently on the target response,
fore, greater values of x (e.g., time in extinction (2) a proportion of reinforcers independently of
or greater satiation) increase the disruptive impact responding, and (3) a proportion of reinforcers
of terms in the numerator, and these effects are dependently on engaging in a different response
countered by all sources of reinforcement in the (Nevin & Wacker, 2013; Podlesnik & DeLeon,
denominator. Importantly, Equation 5 accounts 2015). Thus behavioral mass is a Pavlovian pro-
for data that Equation 2 cannot. Specifically, one cess and can be accounted for quantitatively with
could conceptualize the effects of disruptors as models like Equation 5. The clinical implications
increases in Re in Equation 2, especially common of the findings from Mace et al. and Ahearn et al.
disruptors like free access to the reinforcer main- are considerable: Common behavioral treatments
taining behavior. However, Equation 2 fails to ac- like DRA and NCR can decrease the rate of prob-
count for the common finding that response rates lem behavior, but inadvertently increase its persis-
Matching and Behavioral Momentum 105
tence. We can understand how behavioral treat- might use a concurrent-chain schedule to assess
ments like DRA and NCR can reduce the rate of a client preference among discriminative stimuli
problem behavior while simultaneously enhancing (Tiger, Hanley, & Heal, 2006) as a relatively rapid
its persistence by separating response–reinforcer way for determining which environments or treat-
and stimulus–reinforcer relations. ment approaches might produce more persistent
behavior.
Behavioral Momentum and Matching
Behavioral Momentum and Relapse
Despite the response–reinforcer relation govern-
ing response rate and stimulus–reinforcer relations Persistence correlates with another measure of
governing persistence, response rate and persis- importance concerning outcomes of behavioral
tence are functions of situational reinforcement treatments for problem behavior: relapse. When
parameters. Equation 4 describes how relative re- behavioral treatments eliminate problem behav-
sponse rates match the distribution of reinforcer ior, environmental circumstances can produce
parameters between two options. For example, a recurrence of problem behavior, or treatment
parametrically manipulating relative reinforce- relapse (Mace et al., 2010; Pritchard, Hoerger, &
ment rate produces a positive relation between the Mace, 2014). Many events contribute to treatment
log–response ratio and log–reinforcer ratio—the relapse under natural conditions, which explains
slope of the function expressed as ar. Simultane- why many preclinical models exist for assessing
ously arranging a larger reinforcer for Alternative treatment relapse (Bouton, Winterbauer, & Todd,
1 (x1) than Alternative 2 (x2) will produce a shift 2012; Marchant, Li, & Shaham, 2014; Podlesnik
in the y-intercept through a shift in responding & Kelley, 2015; Wathen & Podlesnik, 2018). We
toward B1 (see Cording et al., 2011). We observe discuss a model with direct relevance to behav-
similar effects with persistence. ioral treatments arranging alternative sources of
In a study by Grace, Bedell, and Nevin (2002), reinforcement—a model called resurgence.
food reinforcement maintained pigeons’ respond- Resurgence is the return of a previously rein-
ing. The researchers manipulated relative rein- forced and extinguished target response when
forcement rates parametrically between two alter- we extinguish a more recently reinforced alterna-
nating discriminative stimuli. Thus sometimes one tive response (Epstein, 1983; Podlesnik & Kelley,
discriminative stimulus would signal a higher rein- 2014). For example, Volkert, Lerman, Call, and
forcement rate than the other, and vice versa. At Trosclair-Lasserre (2009) reinforced instances of
each relative reinforcer rate, Grace et al. disrupted problem behavior according to a fixed-ratio (FR)
responding by presenting an alternative source 1 schedule in children diagnosed with autism
of reinforcement response independently during spectrum disorder or intellectual developmental
time outs from the procedure. As with sensitivity disorder. Next, extinction of problem behavior
in choice, persistence between the discriminative and reinforcement of a functional communica-
stimuli was related positively to reinforcement tion response essentially eliminated problem be-
rate. Moreover, when the different discriminative havior. Volkert et al. then introduced extinction
stimuli arranged constant but different reinforcer of the communication response and continued
magnitudes across relative reinforcement rates, extinction of problem behavior, which produced
persistence was biased toward the discriminative an increase in problem behavior. That is, problem
stimulus presenting the larger-magnitude rein- behavior increased or resurged when the research-
forcer. Moreover, arranging choices between the ers discontinued reinforcement for the appropriate
two discriminative stimuli with concurrent-chain alternative response. Therefore, resurgence models
schedules produced changes in the log–response the relapse of problem behavior due to treatment-
ratio consistent with Equation 4 and correlated integrity-produced errors of omission—specifical-
with relative persistence, as just described. There- ly, failures to reinforce appropriate communication
fore, preference for discriminative stimuli and per- responses or other alternative behavior (St. Peter
sistence within discriminative stimuli appear to Pipkin et al., 2010).
correlate as a function of the current stimulus–re- Obviously, most omission errors likely will not
inforcer relations. As such, Grace et al. suggested be as extreme as complete extinction, as in re-
that preference and persistence provide converg- surgence procedures. However, any decrease in
ing expressions of the same underlying construct treatment integrity increases the likelihood prob-
of response strength or behavioral mass. Thus we lem behavior will return and potentially contact
106 B a s i c P r in c ip l e s a nd C o n c e p t s
reinforcement. Additionally, similar effects occur of the target response; (2) alternative reinforce-
when response-independent reinforcement is re- ment (e.g., DRA, NCR) arranged during extinc-
moved, suggesting that removal of NCR will also tion of target responding in Phase 2 disrupts target
produce resurgence (Marsteller & St. Peter, 2014; responding (i.e., weakens the response–reinforcer
Saini, Fisher, & Pisman, 2017; Winterbauer & relation); (3) alternative reinforcement arranged
Bouton, 2010). Therefore, the findings with re- during Phase 2 (e.g., DRA, NCR) also enhances
surgence and those described earlier show that the Pavlovian stimulus–reinforcer relation of the
behavioral treatments enhance the persistence of target response; and (4) removing the alterna-
problem behavior (Ahearn et al., 2003; Mace et tive reinforcement removes a disruptor of target
al., 2010). Behavioral momentum theory recon- behavior, thereby producing resurgence of target
ciles these different effects through the same be- responding. Thus alternative reinforcement (e.g.,
havioral processes (Nevin & Shahan, 2011; Swee- DRA, NCR) can both strengthen and disrupt
ney & Shahan, 2011). target responding. Shahan and Sweeney’s quanti-
Behavioral momentum theory asserts that rein- tative model of resurgence formalizes these asser-
forcement obtained in the presence of a discrimi- tions in Equation 6:
native stimulus contributes to the persistence of − t (kRa + c + dr )
the responses occasioned by that discriminative Bt b
= 10 ( a )
r+R
(6)
stimulus, even when some reinforcement sources B0
might decrease response rates. During DRA and
NCR treatments, the participant obtains differen- Bt is response rate at time t in extinction (Phas-
tial or noncontingent reinforcement in the same es 2 and 3), and B0 is the training response rate
context as he or she obtains reinforcement for (Phase 1). As in Equation 5, terms in the numera-
problem behavior, thereby enhancing the stimu- tor of the exponent contribute to the disruption
lus–reinforcer relation. Additionally, DRA and of target responding, and terms in the denomi-
NCR treatments serve as disruptors of problem nator counter those disruptive effects. In effect,
behavior, along with the extinction contingency Equation 6 expands on the factors contributing
(when we implement one).2 Thus, greater rates to persistence and relapse, compared to Equation
of alternative reinforcement tend to be more ef- 5. During extinction of target responding, c is
fective in decreasing target problem behavior the effect of removing the contingency between
(e.g., Carr & Durand, 1985; Kelley, Lerman, & responding and reinforcement; d scales the gener-
Van Camp, 2002; Leitenberg, Rawson, & Mulick, alization decrement from eliminating the training
1975; Sweeney & Shahan, 2013), which explains reinforcement rate r as stimuli (i.e., the salience of
why DRA and NCR decrease problem behavior. removing reinforcement); and k scales the disrup-
During the extinction challenge of resurgence tive effect of alternative reinforcement, Ra. In the
procedures, removing DRA or NCR functionally denominator, b is the sensitivity parameter scaling
eliminates a disruptor for problem behavior, ac- the persistence-enhancing effects of r and Ra on
cording to behavioral momentum theory. Thus resistance to extinction and resurgence. There-
problem behavior resurges in the absence of the fore, time in extinction increases the disruptive
disruptive force of DRA or NCR, which previously impact of terms in the numerator, but all sources of
suppressed responding. reinforcement in the denominator counter those
Shahan and Sweeney (2011) developed a quan- disruptive effects. Equation 6 accounts for resur-
titative model of resurgence based on behavioral gence of target responding by (1) setting Ra to the
momentum theory, building on Equation 5. This alternative reinforcement rate in the numerator
model asserts the following: (1) Baseline reinforce- and denominator during Phase 2, and (2) setting
ment rate in Phase 1 of a resurgence procedure en- Ra only in the numerator to zero when alternative
hances the Pavlovian stimulus–reinforcer relation reinforcement is removed in Phase 3. Researchers
have used Equation 6 to describe resurgence across
a range of experiments involving rats, pigeons, and
2 Note that conceptualizing DRA and NCR as disruptors children (Nevin et al., 2017; Shahan & Sweeney,
of problem behavior differs from the assumptions of the 2011). By contrast, Shahan and Craig (2017) is a
matching law. With the matching law, alternative rein-
forcement competes with problem behavior by increasing
useful resource describing limitations in applying
the allocation of behavior toward the alternative response. behavioral momentum theory to resurgence.
With behavioral momentum, the alternative reinforcement Overall, Equation 6 provides a set of assump-
itself disrupts problem behavior directly. tions based on behavioral momentum theory from
Matching and Behavioral Momentum 107
which to assess the environmental factors con- the alternative response (e.g., Hagopian, Toole,
tributing to resurgence. Specifically, we can make Long, Bowman, & Lieving, 2004). Specifically, re-
and test predictions about how common and novel searchers have gradually increased the response re-
aspects of behavioral treatments influence the quirement for reinforcement of the alternative re-
persistence and relapse of problem behavior. For sponse to produce more manageable reinforcement
example, Equation 6 predicts that higher rates of schedules for alternative behavior. Unfortunately,
DRA or NCR will produce greater resurgence be- these methods have not eliminated resurgence in
cause they enhance stimulus–reinforcer relations clinical and translational studies (Hagopian et al.,
more than lower rates of DRA or NCR, which is 2004; Lieving & Lattal, 2003; Sweeney & Shahan,
what happens (Leitenberg et al., 1975; Sweeney & 2013; Volkert et al., 2009; Winterbauer & Bouton,
Shahan, 2013). For instance, Pritchard, Hoerger, 2012). Resurgence tends to occur as the reinforce-
Mace, Penney, and Harris (2014) reinforced ag- ment schedule for the alternative response be-
gressive behavior maintained by attention accord- comes leaner; this has prompted translational re-
ing to a VI 60-second schedule in children with searchers to identify ways to decrease persistence
intellectual developmental disorder in the pres- and the likelihood of relapse of problem behavior
ence of two different therapists. Later, they extin- (DeLeon, Miller, & Podlesnik, 2015; Nevin &
guished reinforcement for problem behavior, and Wacker, 2013; Podlesnik & DeLeon, 2015).
two different therapists reinforced a communica- Wacker et al. (2011) demonstrated another ap-
tion response either every 30 seconds or every 120 proach to reduce relapse of problem behavior by
seconds on average (Ra in Equation 6) in a mul- manipulating reinforcement contingencies (Leit-
tielement design. The researchers discontinued enberg et al., 1975). They provided differential
reinforcement for the communication response reinforcement for communication responses and
after problem behavior decreased by setting Ra = implemented extinction for the problem behavior
0 in the numerator, which produced resurgence of eight children with intellectual developmen-
in problem behavior with both therapists. Resur- tal disorder. They assessed resurgence repeatedly
gence was greater with the therapist whose rein- across sessions by arranging extinction for commu-
forcement rate for the communication response nication responses. Importantly, resurgence gener-
was higher in the previous phase, due to greater Ra. ally decreased across successive resurgence sessions
Thus the common practice of delivering alterna- (Shahan & Sweeney, 2011). These findings sug-
tive sources of reinforcement with DRA and NCR gest that more extensive exposure to differential
treatments at higher rates effectively decreases reinforcement for alternative or communication
problem behavior (Carr & Durand, 1985), but it behavior reduces the likelihood of resurgence.
may make resurgence more likely. Given these These findings suggest that maintaining DRA
problems with common behavioral treatments, we treatments for extended periods may improve
next discuss some approaches for decreasing the the maintenance of treatment gains. Behavioral
persistence and relapse of problem behavior with momentum theory suggests that the disruptive
DRA and NCR treatments. effects of extinction and alternative reinforce-
ment increase with additional exposure to those
contingencies. In other words, the disruptive force
Procedures for Mitigating the Persistence
of DRA treatment increases with additional ex-
and Relapse of Problem Behavior
posure, thereby increasing the long-term effective-
Translational researchers have developed several ness of such treatment. We do not know whether
approaches to mitigate the persistence and re- the same effects would occur for NCR treatments
lapse of problem behavior that DRA and NCR combined with extinction.
treatments cause. These approaches manipulate One limitation of this approach is that doing
reinforcement contingencies and environmental the treatment must involve few or no errors of
stimuli. commission or instances of reinforcing problem
behavior. Otherwise, prior stimulus–reinforcer re-
lations might reinstate responding, thereby elimi-
Methods Focusing on Reinforcement Contingencies
nating progress with DRA treatment. Clearly,
Manipulating reinforcement contingencies is a treatment-integrity-produced errors are not always
general method of reducing the likelihood that predictable in clinical situations, so maintain-
problem behavior will resurge. One specific ap- ing high treatment fidelity for long periods could
proach is to fade the reinforcement schedule for prove difficult or impossible in certain circum-
108 B a s i c P r in c ip l e s a nd C o n c e p t s
stances and with certain caregivers. Neverthe- traditional-training condition. In the modified-
less, these data suggest that extending exposure training procedure, therapists wore lanyards with
to DRA treatments could increase their long-term different-colored cards to signal a multiple sched-
effectiveness, but the conditions under which this ule. Signaled components alternated between 60
is a useful approach remain to be elucidated. seconds of reinforcement for functional-communi-
cation responses (S+) and 30 seconds of extinction
of functional-communication responses (S–). The
Methods Focusing on Environmental Stimuli
duration of S– intervals increased across sessions
Mace et al. (2010) introduced an approach that until the extinction component was 240 seconds
relied primarily on contextual stimulus control and the reinforcement component was 60 seconds.
to mitigate the persistence-enhancing effects of During the resurgence test, the therapists extin-
DRA treatments. Because reinforcing an alterna- guished functional-communication responses and
tive response in the presence of the same discrimi- presented NCR according to a VT 200-second
native stimulus as problem behavior enhances the schedule to decrease discrimination of the extinc-
Pavlovian stimulus–reinforcer relation, Mace et tion contingency. Therapists only presented the
al. reasoned that training alternative responding S– stimulus during the resurgence test in the mod-
separately from problem behavior could be an ef- ified-training condition. In most cases, the modi-
fective approach. Specifically, providing reinforce- fied procedure reduced resurgence to a greater ex-
ment for the alternative response in a context tent than the traditional procedure did. Therefore,
separate from the context in which problem be- thinning reinforcement schedules for functional-
havior produces reinforcement allows us to teach communication responses while incorporating sig-
the alternative response without enhancing the naled periods of S– could promote generalization
Pavlovian stimulus–reinforcer relation. After we and maintenance of behavioral treatments.
train the alternative response in the alternative In a related laboratory study with pigeons, Bland
stimulus context, we can combine the alternative et al. (2016) reinforced an alternative response in
stimulus context with a stimulus context associ- the presence of the same discriminative stimulus as
ated with problem behavior. Combining stimulus a target response in two different ways. In the first,
contexts should disrupt problem behavior and pro- a distinct stimulus change occurred when alterna-
vide the individual with the opportunity to engage tive reinforcement was available according to a VI
in appropriate alternative responses and receive schedule. In the other, no stimulus change accom-
reinforcement. Mace et al.’s results provided sup- panied the availability of alternative reinforce-
port for this approach, and researchers have repli- ment. Although target response rates were higher
cated and extended these findings in studies with with the signaled alternative reinforcement, target
pigeons (Podlesnik & Bai, 2015; Podlesnik, Bai, responding was less persistent in the component
& Elliffe, 2012; Podlesnik, Bai, & Skinner, 2016; with the signaled alternative. Thus signaling al-
Podlesnik, Miranda-Dukoski, Chan, Bland, & Bai, ternative reinforcement appears to have separated
2017) and through telehealth with children diag- the stimulus–reinforcer relation of the alternative
nosed with ASD (Suess, Schieltz, Wacker, Detrick, reinforcement from that of the target response.
& Podlesnik, 2020). Therefore, these findings and those of Fuhrman et
A multiple schedule is another stimulus con- al. (2016) show that discriminable periods of non-
trol approach researchers have used to reduce the reinforcement during functional-communication
persistence of problem behavior. This approach training may reduce the persistence and relapse
has preliminary support from clinical studies and of problem behavior during behavioral treatments
laboratory models with pigeons (Bland, Bai, Fuller- (Nevin et al., 2016). Nevertheless, future research
ton, & Podlesnik, 2016; Fuhrman, Fisher, & Greer, should assess methods for also reducing the rate of
2016; Nevin et al., 2016). Fuhrman et al. (2016) the target response in the presence of the signaled
found that signaling periods in which functional- alternative reinforcement, which is a key compo-
communication responses would and would not nent of behavioral treatments aimed at reducing
produce reinforcement could mitigate resurgence problem behavior. The methods of Fuhrman et al.
of problem behavior, compared to traditional func- (2016) provide a promising avenue to explore (see
tional-communication training. Therapists rein- also Fisher, Fuhrman, Greer, Mitteer, and Piazza,
forced two children’s functional-communication 2020).
responses according to VI 20-second schedules, The findings from these studies highlight an
and extinguished problem behavior during the important point. Translational and applied studies
Matching and Behavioral Momentum 109
produce evidence for individual procedures that treatments in conditions separate from nontreat-
show promise as effective treatment. Most effec- ment, this issue has implications for the long-term
tive treatments, however, are likely to incorporate efficacy of behavioral treatments. Therefore, the
multiple approaches developed in parallel, like conditions in which persistence and relapse are re-
those described here and elsewhere (e.g., DeLeon liably functions of reinforcement rate remains an
et al., 2015). Determining the best methods to important question for additional research.
incorporate into behavioral treatments with the Another area of concern is the partial-rein-
goal of mitigating the persistence and likelihood forcement extinction effect (Mackintosh, 1974;
of treatment relapse is an important aim for trans- Nevin, 1988). This effect occurs when responses
lational researchers. reinforced intermittently produce more respond-
ing during extinction than responses reinforced
continuously on an FR 1 schedule do. These find-
Additional Problems to Address in Future Research
ings appear counter to behavioral momentum
The matching law and behavioral momentum theory, because FR 1 schedules should enhance
theory provide translational researchers and clini- Pavlovian stimulus–reinforcer relations more
cians with powerful tools to understand and assess than any intermittent-reinforcement schedule.
factors contributing to the effectiveness of behav- Behavioral momentum theory (e.g., Equation 6)
ioral treatments. Nevertheless, the science behind accounts for the partial-reinforcement extinction
the factors contributing to treatment effectiveness effect as follows. Extinction terminates the contin-
remains incomplete. We discuss two areas in de- gency between responding and reinforcement with
tail in which additional basic and translational parameter c. In addition, the removal of reinforce-
research could improve our capacity to understand ment is more salient following FR 1 than following
and effectively implement behavioral treatments. intermittent reinforcement, parameter d. The gen-
Typically, researchers assessing persistence and eralization decrement is the discriminable change
relapse from the framework of behavioral momen- in stimulus conditions and is a disruptive force in
tum theory use multiple schedules of reinforcement Equation 6 because it is in the numerator (Nevin,
(see Nevin, 1974; Nevin & Grace, 2000). Analo- McLean, & Grace, 2001). Despite the many in-
gous to multielement designs, multiple schedules stances of the partial-reinforcement extinction
rapidly alternate conditions, such as different re- effect in within-participant designs assessing oper-
inforcement rates in the presence of dramatically ant behavior (e.g., Nevin & Grace, 2005; Shull &
different antecedent stimuli. These methods are Grimes, 2006), there are many examples in which
effective for establishing strong antecedent dis- FR 1 reinforcement produces greater resistance
criminative stimulus control, but are dissimilar to to extinction than intermittent-reinforcement
the delivery of many behavioral treatments. Sev- schedules (e.g., Lerman, Iwata, Shore, & Kahng,
eral researchers arranged different reinforcement 1996; Schmid, 1988; Theios, 1962). Exactly which
rates between multiple-schedule components and features contribute to changes in resistance to
assessed the persistence of responding in stud- extinction following continuous or intermittent
ies with pigeons (Cohen, 1998; Cohen, Riley, & reinforcement is not entirely evident. Given the
Weigle, 1993) and in translational research with prevalent use of extinction and continuous-rein-
children with intellectual developmental disorder forcement schedules in behavioral treatments, un-
(Lionello-DeNolf & Dube, 2011). When resistance derstanding the factors contributing to the persis-
to disruption between components of a multiple tence of behavior during extinction is important.
schedule was assessed, persistence was greater in Nevertheless, extinction is only one source of
the component arranging greater reinforcement disruption, and behavioral momentum theory pre-
rates. However, arranging the same reinforcement dicts that continuously reinforced behavior should
rates between extended conditions did not produce be more resistant to all sources of disruption other
reliable differences in persistence as a function of than extinction (e.g., satiation, distraction).
reinforcement rate (Shull & Grimes, 2006). Never-
theless, researchers assessed the greater persistence
in the presence of NCR (Ahearn et al., 2003) and CONCLUSIONS
DRA (Mace et al., 2010) across successive con-
ditions arranging reinforcement for problem be- The matching law and behavioral momentum
havior in the absence of those treatments. Given theory provide two quantitative frameworks to
that most behavior analysts implement behavioral describe and explain operant behavior. Readers
110 B a s i c P r in c ip l e s a nd C o n c e p t s
may want to choose between the two frameworks Baum, W. M. (1982). Choice, changeover, and travel.
or conclude whether one is superior. However, Journal of the Experimental Analysis of Behavior, 38,
theories are like different tools; we use theories 35–49.
Baum, W. M. (2010). Dynamics of choice: A tutorial.
for different things. We use a lawn mower to cut
Journal of the Experimental Analysis of Behavior, 94,
a larger section of grass, but a weed whacker to get 161–174.
the edges. Neither one is better in general; they do Baum, W. M., & Rachlin, H. C. (1969). Choice as time
related things (i.e., shorten grass), but have differ- allocation. Journal of the Experimental Analysis of Be-
ent purposes. As suggested by Killeen (1999), havior, 12, 861–874.
Bell, M. C. (1999). Pavlovian contingencies and resis-
If you think models are about the truth, or that there tance to change in a multiple schedule. Journal of the
is a best timing model, then you are in trouble. There Experimental Analysis of Behavior, 72, 81–96.
is no best model, any more than there is a best car Bland, V. J., Bai, J. Y. H., Fullerton, J. A., & Podlesnik,
model or a best swimsuit model, even though each of C. A. (2016). Signaled alternative reinforcement and
us may have our favorites. It all depends on what you the persistence of operant behavior. Journal of the Ex-
want to do with the model. (p. 275) perimental Analysis of Behavior, 106, 22–33.
Borrero, C. S., Vollmer, T. R., Borrero, J. C., Bourret, J.
Thus matching and behavioral momentum theory C., Sloman, K. N., Samaha, A. L., et al. (2010). Con-
both have limitations, but we should not evaluate current reinforcement schedules for problem behav-
them for how poorly they account for behavior ior and appropriate behavior: Experimental applica-
tions of the matching law. Journal of the Experimental
outside the relevant domain—which is the allo-
Analysis of Behavior, 93, 455–469.
cation of behavior for the matching law, and the Bouton, M. E., Winterbauer, N. E., & Todd, T. P. (2012).
persistence and relapse of behavior for behavioral Relapse processes after the extinction of instrumen-
momentum theory. Instead, we should evaluate tal learning: Renewal, resurgence, and reacquisition.
them for how well they account for behavior in Behavioural Processes, 90, 130–141.
the relevant domains (Mazur, 2006). They also Buckley, J. L., & Rasmussen, E. B. (2012). Obese and
serve as frameworks for understanding behavior lean Zucker rats demonstrate differential sensitivity
and posing questions regarding variables that may to rates of food reinforcement in a choice procedure.
be relevant to the effectiveness of behavioral treat- Physiology and Behavior, 108, 19–27.
Carr, E. G., & Durand, V. M. (1985). Reducing behavior
ments (Critchfield & Reed, 2009).
problems through functional communication train-
ing. Journal of Applied Behavior Analysis, 18, 111–126.
Cohen, S. L. (1996). Behavioral momentum of typ-
REFERENCES ing behavior in college students. Journal of Behavior
Analysis and Therapy, 1, 36–51.
Ahearn, W. H., Clark, K. M., Gardenier, N. C., Chung, Cohen, S. L. (1998). Behavioral momentum: The ef-
B. I., & Dube, W. V. (2003). Persistence of stereo- fects of the temporal separation of rates of reinforce-
typic behavior: Examining the effects of external ment. Journal of the Experimental Analysis of Behavior,
reinforcers. Journal of Applied Behavior Analysis, 36, 69, 29–47.
439–448. Cohen, S. L., Riley, D. S., & Weigle, P. A. (1993). Tests
Alsop, B., & Davison, M. (1991). Effects of varying stim- of behavioral momentum in simple and multiple
ulus disparity and the reinforcer ratio in concurrent- schedules with rats and pigeons. Journal of the Experi-
schedule and signal-detection procedures. Journal of mental Analysis of Behavior, 60, 255–291.
the Experimental Analysis of Behavior, 56, 67–80. Cording, J. R., McLean, A. P., & Grace, R. C. (2011).
Aparicio, C. F., & Baum, W. M. (1997). Comparing lo- Testing the linearity and independence assumptions
comotion with lever-press travel in an operant simu- of the generalized matching law for reinforcer magni-
lation of foraging. Journal of the Experimental Analysis tude: A residual meta-analysis. Behavioural Processes,
of Behavior, 68, 177–192. 87, 64–70.
Athens, E. S., & Vollmer, T. R. (2010). An investigation Critchfield, T. S., & Reed, D. D. (2009). What are we
of differential reinforcement of alternative behavior doing when we translate from quantitative models?
without extinction. Journal of Applied Behavior Anal- Behavior Analyst, 32(2), 339–362.
ysis, 43, 569–589. Dallery, J., & Soto, P. L. (2013). Quantitative descrip-
Baum, W. M. (1974). On two types of deviation from the tion of environment–behavior relations. In G. J.
matching law: Bias and undermatching. Journal of the Madden (Ed.) & W. V. Dube, T. Hackenberg, G. P.
Experimental Analysis of Behavior, 22, 231–242. Hanley, & K. A. Lattal (Assoc. Eds.), APA handbook
Baum, W. M. (1979). Matching, undermatching, and of behavior analysis: Vol. 1. Methods and principles
overmatching in studies of choice. Journal of the Ex- (pp. 219–249). Washington, DC: American Psycho-
perimental Analysis of Behavior, 32, 269–281. logical Association.
Matching and Behavioral Momentum 111
Davison, M., & Jenkins, P. E. (1985). Stimulus discrim- of operant behavior. In G. J Madden (Ed.) & W. V.
inability, contingency discriminability, and sched- Dube, T. Hackenberg, G. P. Hanley, & K. A. Lattal
ule performance. Animal Learning and Behavior, 13, (Assoc. Eds.), APA handbook of behavior analysis: Vol.
77–84. 1. Methods and principles (pp. 219–249). Washington,
Davison, M., & Jones, B. M. (1995). A quantitative DC: American Psychological Association.
analysis of extreme choice. Journal of the Experimen- Grace, R. C., Schwendiman, J. W., & Nevin, J. A.
tal Analysis of Behavior, 64, 147–162. (1998). Effects of unsignaled delay of reinforcement
Davison, M., & McCarthy, D. (1988). The matching law: on preference and resistance to change. Journal of the
A research review. Hillsdale, NJ: Erlbaum. Experimental Analysis of Behavior, 69, 247–261.
Davison, M., & Nevin, J. A. (1999). Stimuli, reinforcers, Grimes, J. A., & Shull, R. L. (2001). Response-indepen-
and behavior: An integration. Journal of the Experi- dent milk delivery enhances persistence of pellet-
mental Analysis of Behavior, 71, 439–482. reinforced lever pressing by rats. Journal of the Experi-
DeLeon, I. G., Podlesnik, C. A., & Miller, J. R. (2015). mental Analysis of Behavior, 76, 179–194.
Applied research on resistance to change: Implica- Hagopian, L. P., Toole, L. M., Long, E. S., Bowman, L.
tions for treatment in autism spectrum disorders. In G., & Lieving, G. A. (2004). A comparison of dense-
F. D. DiGennaro Reed & D. D. Reed (Eds.), Bridging to-lean and fixed-lean schedules of alternative rein-
the gap between science and practice in autism service forcement and extinction. Journal of Applied Behavior
delivery (pp. 353–374). New York: Springer. Analysis, 37, 323–338.
Domjan, M. (2005). Pavlovian conditioning: A func- Herrnstein, R. J. (1961). Relative and absolute strength
tional perspective. Annual Review of Psychology, 56, of response as a function of frequency of reinforce-
179–206. ment. Journal of the Experimental Analysis of Behavior,
Domjan, M. (2016). Elicited versus emitted behavior: 4, 267–272.
Time to abandon the distinction. Journal of the Ex- Herrnstein, R. J. (1970). On the law of effect. Journal
perimental Analysis of Behavior, 105, 231–245. of the Experimental Analysis of Behavior, 13, 243–266.
Elliffe, D., Davison, M., & Landon, J. (2008). Relative Igaki, T., & Sakagami, T. (2004). Resistance to change
reinforcer rates and magnitudes do not control con- in goldfish. Behavioural Processes, 66, 139–152.
current choice independently. Journal of the Experi- Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., &
mental Analysis of Behavior, 90, 169–185. Richman, G. S. (1994). Toward a functional analysis
Epstein, R. (1983). Resurgence of previously reinforced of self-injury. Journal of Applied Behavior Analysis, 27,
behavior during extinction. Behaviour Analysis Let- 197–209. (Reprinted from Analysis and Intervention in
ters, 3, 391–397. Developmental Disabilities, 2, 3–20, 1982)
Fantino, E., Squires, N., Delbrück, N., & Peterson, C. Jacobs, E. A., Borrero, J. C., & Vollmer, T. R. (2013).
(1972). Choice behavior and the accessibility of the Translational applications of quantitative choice
reinforcer. Journal of the Experimental Analysis of Be- models. In G. J. Madden (Ed.) & W. V. Dube, T. D.
havior, 18, 35–43. Hackenberg, G. P. Hanley, & K. A. Lattal (Assoc.
Findley, J. D. (1958). Preference and switching under Eds.), APA handbook of behavior analysis: Vol. 2.
concurrent scheduling. Journal of the Experimental Translating principles into practice (pp. 165–190).
Analysis of Behavior, 1, 123–144. Washington, DC: American Psychological Associa-
Fisher, W. W., Fuhrman, A. M., Greer, B. D., Mitteer, D. tion.
R., & Piazza, C. C. (2020). Mitigating resurgence of Kelley, M. E., Lerman, D. C., & Van Camp, C. M. (2002).
destructive behavior using the discriminative stimuli The effects of competing reinforcement schedules on
of a multiple schedule. Journal of the Experimental the acquisition of functional communication. Journal
Analysis of Behavior, 113, 263–277. of Applied Behavior Analysis, 35, 59–63.
Fisher, W. W., Pawich, T. L., Dickes, N., Paden, A. R., Killeen, P. (1972). The matching law. Journal of the Ex-
& Toussaint, K. (2014). Increasing the saliency of perimental Analysis of Behavior, 17, 489–495.
behavior-consequence relations for children with au- Killeen, P. R. (1999). Modeling modeling. Journal of the
tism who exhibit persistent errors. Journal of Applied Experimental Analysis of Behavior, 71, 275–280.
Behavior Analysis, 47, 1–11. Kollins, S. H., Newland, M. C., & Critchfield, T. S.
Fuhrman, A. M., Fisher, W. W., & Greer, B. D. (2016). (1997). Human sensitivity to reinforcement in oper-
A preliminary investigation on improving functional ant choice: How much do consequences matter? Psy-
communication training by mitigating resurgence chonomic Bulletin and Review, 4, 208–220.
of destructive behavior. Journal of Applied Behavior Kuroda, T., Cançado, C. R. X., & Podlesnik, C. A.
Analysis, 49, 884–899. (2016). Resistance to change and resurgence in hu-
Grace, R. C., Bedell, M. A., & Nevin, J. A. (2002). Pref- mans engaging in a computer task. Behavioural Pro-
erence and resistance to change with constant- and cesses, 125, 1–5.
variable-duration terminal links: Independence of Lattal, K. M. (2013). Pavlovian conditioning. In G. J.
reinforcement rate and magnitude. Journal of the Ex- Madden (Ed.) & W. V. Dube, T. D. Hackenberg, G. P.
perimental Analysis of Behavior, 77, 233–255. Hanley, & K. A. Lattal (Assoc. Eds.), APA handbook
Grace, R. C., & Hucks, A. D. (2013). The allocation of behavior analysis: Vol. 1. Methods and principles
112 B a s i c P r in c ip l e s a nd C o n c e p t s
(pp. 283–306). Washington, DC: American Psycho- and application (pp. 52–108). New York: Appleton-
logical Association. Century-Crofts.
Leitenberg, H., Rawson, R. A., & Mulick, J. A. (1975). Nevin, J. A. (1974). Response strength in multiple
Extinction and reinforcement of alternative behav- schedules. Journal of the Experimental Analysis of Be-
ior. Journal of Comparative and Physiological Psychol- havior, 21, 389–408.
ogy, 88, 640–652. Nevin, J. A. (1984). Quantitative analysis. Journal of the
Lerman, D. C., Iwata, B. A., Shore, B. A., & Kahng, S. Experimental Analysis of Behavior, 42, 421–434.
W. (1996). Responding maintained by intermittent Nevin, J. A. (1988). Behavioral momentum and the par-
reinforcement: Implications for the use of extinction tial reinforcement effect. Psychological Bulletin, 103,
with problem behavior. Journal of Applied Behavior 44–56.
Analysis, 29, 153–171. Nevin, J. A. (1997). Choice and momentum. In W.
Lieving, G. A., & Lattal, K. A. (2003). Recency, repeat- O’Donohue (Ed.), Learning and behavior therapy
ability, and reinforcer retrenchment: An experimen- (pp. 230–251). Boston: Allyn & Bacon.
tal analysis of resurgence. Journal of the Experimental Nevin, J. A. (2015). Behavioral momentum: A scientific
Analysis of Behavior, 80, 217–233. metaphor. Vineyard Haven, MA: Author.
Lionello-DeNolf, K. M., & Dube, W. V. (2011). Contex- Nevin, J. A., Craig, A. R., Cunningham, P. J., Podle-
tual influences on resistance to disruption in chil- snik, C. A., Shahan, T. A., & Sweeney, M. M. (2017).
dren with intellectual disabilities. Journal of the Ex- Quantitative models of persistence and relapse from
perimental Analysis of Behavior, 96, 317–327. the perspective of behavioral momentum theory: Fits
Mace, F. C., Lalli, J. S., Shea, M. C., Lalli, E. P., West, and misfits. Behavioural Processes, 141, 92–99.
B. J., Roberts, M., et al. (1990). The momentum of Nevin, J. A., & Grace, R. C. (2000). Behavioral mo-
human behavior in a natural setting. Journal of the mentum and the law of effect. Behavioral and Brain
Experimental Analysis of Behavior, 54, 163–172. Sciences, 23, 73–130.
Mace, F. C., McComas, J. J., Mauro, B. C., Progar, P. Nevin, J. A., & Grace, R. C. (2005). Resistance to ex-
R., Taylor, B., Ervin, R., et al. (2010). Differential tinction in the steady state and in transition. Journal
reinforcement of alternative behavior increases of Experimental Psychology: Animal Behavior Process-
resistance to extinction: Clinical demonstration, es, 31, 199–212.
animal modeling, and clinical test of one solution. Nevin, J. A., Mace, F. C., DeLeon, I. G., Shahan, T. A.,
Journal of the Experimental Analysis of Behavior, 93, Shamlian, K. D., Lit, K., et al. (2016). Effects of signaled
349–367. and unsignaled alternative reinforcement on persis-
Mackintosh, N. J. (1974). The psychology of animal learn- tence and relapse in children and pigeons. Journal of
ing. New York: Academic Press. the Experimental Analysis of Behavior, 106, 34–57.
Marchant, N. J., Li, X., & Shaham, Y. (2013). Recent Nevin, J. A., McLean, A. P., & Grace, R. C. (2001).
developments in animal models of drug relapse. Cur- Resistance to extinction: Contingency termination
rent Opinion in Neurobiology, 23, 675–683. and generalization decrement. Animal Learning and
Marsteller, T. M., & St. Peter, C. C. (2014). Effects of Behavior, 29, 176–191.
fixed-time reinforcement schedules on resurgence of Nevin, J. A., & Shahan, T. A. (2011). Behavioral mo-
problem behavior. Journal of Applied Behavior Analy- mentum theory: Equations and applications. Journal
sis, 47, 455–469. of Applied Behavior Analysis, 44, 877–895.
Mauro, B. C., & Mace, F. C. (1996). Differences in Nevin, J. A., Tota, M. E., Torquato, R. D., & Shull, R.
the effect of Pavlovian contingencies upon behav- L. (1990). Alternative reinforcement increases resis-
ioral momentum using auditory versus visual stimuli. tance to change: Pavlovian or operant contingen-
Journal of the Experimental Analysis of Behavior, 65, cies? Journal of the Experimental Analysis of Behavior,
389–399. 53, 359–379.
Mazur, J. E. (2006). Mathematical models and the ex- Nevin, J. A., & Wacker, D. P. (2013). Response strength
perimental analysis of behavior. Journal of the Experi- and persistence. In G. J. Madden (Ed.) & W. V. Dube,
mental Analysis of Behavior, 85, 275–291. T. D. Hackenberg, G. P. Hanley, & K. A. Lattal
McDowell, J. J. (1989). Two modern developments in (Assoc. Eds.), APA handbook of behavior analysis: Vol.
matching theory. Behavior Analyst, 12, 153–166. 2. Translating principles into practice (pp. 109–128).
Miller, J. T., Saunders, S. S., & Bourland, G. (1980). Washington, DC: American Psychological Associa-
The role of stimulus disparity in concurrently avail- tion.
able reinforcement schedules. Animal Learning and Newland, M. C., Reile, P. A., & Langston, J. L. (2004).
Behavior, 8, 635–641. Gestational exposure to methylmercury retards
Miranda-Dukoski, L., Bensemann, J., & Podlesnik, C. choice in transition in aging rats. Neurotoxicology
A. (2016). Training reinforcement rates, resistance to and Teratology, 26, 179–194.
extinction, and the role of context in reinstatement. Oscar-Berman, M., Heyman, G. M., Bonner, R. T., &
Learning and Behavior, 44, 29–48. Ryder, J. (1980). Human neuropsychology: Some dif-
Morse, W. H. (1966). Intermittent reinforcement. In W. ferences between Korsakoff and normal operant per-
K. Honig (Ed.), Operant behavior: Areas of research formance. Psychological Research, 41, 235–247.
Matching and Behavioral Momentum 113
Petscher, E. S., Rey, C., & Bailey, J. A. (2009). A review Rescorla, R. A. (1988). Pavlovian conditioning: It’s
of empirical support for differential reinforcement of not what you think it is. American Psychologist, 43,
alternative behavior. Research in Developmental Dis- 151–160.
abilities, 30, 409–425. Saini, V., Fisher, W. W., & Pisman, M. D. (2017). Per-
Podlesnik, C. A., & Bai, J. Y. H. (2015). Method of sistence during and resurgence following noncontin-
stimulus combination impacts resistance to extinc- gent reinforcement implemented with and without
tion. Journal of the Experimental Analysis of Behavior, extinction. Journal of Applied Behavior Analysis, 50,
104, 30–47. 377–392.
Podlesnik, C. A., Bai, J. Y. H., & Elliffe, D. (2012). Resis- Schmid, T. L. (1988). A comparison of two behavior re-
tance to extinction and relapse in combined stimulus duction procedures: Traditional extinction alone and
contexts. Journal of the Experimental Analysis of Be- interpolated reinforcement followed by extinction.
havior, 98, 169–189. Journal of Mental Deficiency Research, 32, 289–300.
Podlesnik, C. A., Bai, J. Y. H., & Skinner, K. A. (2016). Shahan, T. A., & Burke, K. A. (2004). Ethanol-main-
Greater alternative response and reinforcement rates tained responding of rats is more resistant to change
increase disruption of target responding when com- in a context with added non-drug reinforcement. Be-
bining stimuli. Journal of the Experimental Analysis of havioral Pharmacology, 15, 279–285.
Behavior, 105, 427–444. Shahan, T. A., & Craig, A. R. (2017). Resurgence as
Podlesnik, C. A., & DeLeon, I. G. (2015). Behavioral choice. Behavioural Processes, 141, 100–127.
momentum theory: Understanding persistence and Shahan, T. A., & Sweeney, M. M. (2011). A model of
improving treatment. In F. D DiGennaro Reed & D. resurgence based on behavioral momentum theory.
D. Reed (Eds.), Bridging the gap between science and Journal of the Experimental Analysis of Behavior, 95,
practice in autism service delivery (pp. 327–351). New 91–108.
York: Springer. Shull, R. L. (1991). Mathematical description of operant
Podlesnik, C. A., Jimenez-Gomez, C., Ward, R. D., & behavior: An introduction. In I. H. Iversen & K. A.
Shahan, T. A. (2006). Resistance to change of re- Lattal (Eds.), Experimental analysis of behavior (Part
sponding maintained by unsignaled delays to rein- 2) (pp. 243–282). Amsterdam: Elsevier.
forcement: A response-bout analysis. Journal of the Shull, R. L., & Grimes, J. A. (2006). Resistance to ex-
Experimental Analysis of Behavior, 85, 329–347. tinction following variable-interval reinforcement:
Podlesnik, C. A., & Kelley, M. E. (2014). Resurgence: Reinforcer rate and amount. Journal of the Experi-
Response competition, stimulus control, and rein- mental Analysis of Behavior, 85, 23–39.
forcer control. Journal of the Experimental Analysis of Shull, R. L., & Pliskoff, S. S. (1967). Changeover delay
Behavior, 102, 231–240. and concurrent schedules: Some effects on relative
Podlesnik, C. A., & Kelley, M. E. (2015). Translational performance measures. Journal of the Experimental
research on the relapse of operant behavior. Mexican Analysis of Behavior, 10, 517–527.
Journal of Behavior Analysis, 41, 226–251. Skinner, B. F. (1938). The behavior of organisms: An ex-
Podlesnik, C. A., Miranda-Dukoski, L., Chan, J. C. K., perimental analysis. New York: Appleton-Century.
Bland, V. J., & Bai, J. Y. H. (2017). Generalization Skinner, B. F. (1969). Contingencies of reinforcement.
of the disruptive effects of alternative stimuli when Englewood Cliffs, NJ: Prentice-Hall.
combined with target stimuli in extinction. Journal of St. Peter Pipkin, C., Vollmer, T. R., & Sloman, K. N.
the Experimental Analysis of Behavior, 108, 255–268. (2010). Effects of treatment integrity failures during
Podlesnik, C. A., & Shahan, T. A. (2008). Response– differential reinforcement of alternative behavior:
reinforcer relations and resistance to change. Behav- A translational model. Journal of Applied Behavior
ioural Processes, 77, 109–125. Analysis, 43, 47–70.
Podlesnik, C. A., & Shahan, T. A. (2009). Behavioral Staddon, J. E. R. (1968). Spaced responding and choice:
momentum and relapse of extinguished operant re- A preliminary analysis. Journal of the Experimental
sponding. Learning and Behavior, 37, 357–364. Analysis of Behavior, 11, 669–682.
Podlesnik, C. A., & Shahan, T. A. (2010). Extinction, Suess, A. N., Schieltz, K. M., Wacker, D. P., Detrick, J.,
relapse, and behavioral momentum. Behavioural Pro- & Podlesnik, C. A. (2020). An evaluation of resur-
cesses, 84, 400–411. gence following functional communication training
Pritchard, D., Hoerger, M., & Mace, F. C. (2014). Treat- conducted in alternative antecedent contexts via
ment relapse and behavioral momentum theory. telehealth. Journal of the Experimental Analysis of Be-
Journal of Applied Behavior Analysis, 47, 814–833. havior, 113, 278–301.
Pritchard, D., Hoerger, M., Mace, F. C., Penney, H., Sutton, N. P., Grace, R. C., McLean, A. P., & Baum, W.
& Harris, B. (2014). Clinical translation of animal M. (2008). Comparing the generalized matching law
models of treatment relapse. Journal of the Experimen- and contingency discriminability model as accounts
tal Analysis of Behavior, 101, 442–449. of concurrent schedule performance using residual
Reed, D. D., & Kaplan, B. A. (2011). The matching law: meta-analysis. Behavioural Processes, 78, 224–230.
A tutorial for practitioners. Behavior Analysis in Prac- Sweeney, M. M., & Shahan, T. A. (2013). Behavioral
tice, 4, 15–24. momentum and resurgence: Effects of time in extinc-
114 B a s i c P r in c ip l e s a nd C o n c e p t s
tion and repeated resurgence tests. Learning and Be- models of treatment relapse and mitigation tech-
havior, 41, 414–424. niques. Behavior Analysis: Research and Practice,
Theios, J. (1962). The partial reinforcement effect sus- 18(4), 362–387.
tained through the blocks of ontinuous reinforce- Wearden, J. H., & Burgess, I. S. (1982). Matching since
ment. Journal of Experimental Psychology, 64, 1–6. Baum (1979). Journal of the Experimental Analysis of
Tiger, J. H., Hanley, G. P., & Heal, N. A. (2006). The ef- Behavior, 38, 339–348.
fectiveness of and preschoolers’ preferences for varia- Williams, B. A. (1988). Reinforcement, choice, and re-
tions of multiple-schedule arrangements. Journal of sponse strength. In R. C. Atkinson, R. J. Herrnstein,
Applied Behavior Analysis, 39, 475–488. G. Lindzey, & R. D. Luce (Eds.), Stevens’ handbook of
Volkert, V. M., Lerman, D. C., Call, N. A., & Trosclair- experimental psychology: Vol. 2. Learning and cognition
Lasserre, N. (2009). An evaluation of resurgence (pp. 167–244). New York: Wiley.
during treatment with functional communication Winterbauer, N. E., & Bouton, M. E. (2010). Mecha-
training. Journal of Applied Behavior Analysis, 42, nisms of resurgence of an extinguished instrumental
145–160. behavior. Journal of Experimental Psychology: Animal
Wacker, D. P., Harding, J. W., Berg, W. K., Lee, J. F., Behavior Processes, 36, 343–353.
Schieltz, K. M., Padilla, Y. C., et al. (2011). An evalu- Winterbauer, N. E., & Bouton, M. E. (2012). Effects of
ation of persistence of treatment effects during long- thinning the rate at which the alternative behavior
term treatment of destructive behavior. Journal of the is reinforced on resurgence of an extinguished instru-
Experimental Analysis of Behavior, 96, 261–282. mental response. Journal of Experimental Psychology:
Wathen, S. N., & Podlesnik, C. A. (2018). Laboratory Animal Behavior Processes, 3, 279–291.
CHAP TER 7
Behavioral Economics
Principles and Applications
Behavioral economics is the application of psycho- tives (i.e., comparing “apples to oranges”). This
logical and behavioral principles to understand may be more relevant in applied contexts, because
choice or decision making, but the specific prin- human decisions about response allocation more
ciples differ according to the school of thought often involve responses leading to qualitatively
and the direction of the flow of influence. The different outcomes (Fuqua, 1984); thus microeco-
behavioral economics most familiar to behavior nomic and operant theories converge at multiple
analysts draws principally from the integration points. For example, both are concerned with the
of microeconomic theory and operant psychol- factors that influence choice under conditions of
ogy (Kagel & Winkler, 1972). This approach ex- constraint and the relative value of goods (Hursh
tends both concepts from microeconomic theory et al., 2013; Hursh & Roma, 2016). The parallels
to study consumption in various species in the between microeconomic theory and behavior
laboratory, and concepts of operant conditioning analysis suggest a wealth of relations only hereto-
to understand demand for economic commodities fore considered by economists—new phenomena
(Hursh, Madden, Spiga, DeLeon, & Francisco, previously ignored and functional relations previ-
2013). In this chapter, we concentrate on operant ously unnamed by behavior analysts.
behavioral economics and use the term behavioral In this chapter, we begin by describing those
economics in the operant and not the traditional concepts and tools, translating from the language
vein. of microeconomics to the language of behavior
One appeal of importing microeconomic theory analysis when necessary. We then review how be-
into behavior analysis is that it gives behavior ana- havior analysts have applied those concepts and
lysts a framework for understanding choices among tools to understand and promote beneficial choic-
concurrently available but asymmetrical alterna- es. Much of our discussion describes applications
115
116 B a s i c P r in c ip l e s a nd C o n c e p t s
to persons with neurodevelopmental disorders with humans may not be feasible. Thus investiga-
because of the predominance of research on this tors often evaluate consumption during smaller
population in applied behavior analysis. We also time units or until a participant earns a maximum
reference basic or other applied research in which number of reinforcers available.
behavioral economics has gained a foothold, such
as studies of addictions and health-related behav-
Assessing Stimulus Value with Demand Functions
ior.
A demand function relates consumption of a rein-
forcer to its cost. Specifically, demand functions
EFFORT AS COST: TERMS, CONCEPTS, quantify reinforcer consumption amidst increas-
AND METHODS ing costs. We derive this function by varying the
reinforcer price (i.e., the schedule requirement)
Behavioral economists assume that how individu- and measuring how much of that reinforcer an in-
als behave to procure and consume reinforcers dividual consumes at each price. Thus a behavior
follows certain microeconomic principles that analyst can examine how changes in the amount
govern population dynamics. Behavior analysts of work the individual must complete to produce a
have adopted principles largely concerned with reinforcer alters how much of that reinforcer the
stimulus value, or whether a reinforcer will main- individual earns. For example, if a single response
tain responding under varying constraint condi- produces the reinforcer on a fixed-ratio (FR) 1
tions. In this context, value involves the contex- schedule, the demand analysis would quantify how
tual variables that affect the reinforcing efficacy consumption decreases as reinforcer availability
of the stimulus, not an intrinsic stimulus feature. decreases (i.e., as the reinforcement schedule shifts
Constraint involves contingencies that dictate from FR 1 to increasingly intermittent schedules).
how, how many, when, and if reinforcers will be We could vary the response requirement from FR
delivered. This approach assumes that the effec- 5, to FR 10, FR 15, FR 20, and so on. The resulting
tiveness of a reinforcer is not a static, inherent demand function would show the mean number
property of the stimulus; many factors can influ- of reinforcers earned at each FR value. Behavioral
ence its effectiveness. Below, we define some of the economists often interpret the resulting demand
more critical microeconomic terms and principles function as the degree to which an individual will
in this behavioral context. defend consumption during increased prices versus
when the same reinforcer was inexpensive (i.e., at
the FR 1 schedule).
Commodities and Unit Price
The left panel of Figure 7.1, reproduced from
In behavioral economics, a commodity is a rein- Hursh and Roma (2013), depicts demand func-
forcer purchased with behavior. That is, an indi- tions for food pellets and units of saccharin solu-
vidual pays the cost of the reinforcer by meeting tion as a function of the number of responses re-
a behavioral contingency. Unit price describes the quired to earn a unit of each reinforcer, aggregated
cost–benefit ratio of a commodity. In behavioral for two monkeys. Consumption (reinforcers per
terms, unit price describes the number of respons- day) decreased for each reinforcer as price (the FR
es to the amount of a reinforcer. For example, a requirement) increased for both functions. Behav-
person might earn one token (benefit) for five cor- ioral economists characterize such a relation as the
rect responses (cost) on a worksheet; thus the unit law of demand, which describes the characteristic
price of one token is five responses. inverse relation between consumption and unit
price. This inverse relation implies that higher
unit prices produce less consumption. Stated more
Consumption
simply, consumption of a reinforcer decreases as
Consumption in a behavioral-economic context the number of responses required to earn that re-
is a measurement of the quantity of a reinforcer inforcer increases, and vice versa.
earned during a given time, like an observation Figure 7.1 shows that the slopes of the gener-
session. Consumption often reflects the number of ated functions are nonlinear, as is typical of de-
reinforcers earned per day in basic behavioral-eco- mand functions. Initial price increases may have
nomic research. Assessing consumption through- little effect, producing only small decreases in
out a day or during long periods in applied settings consumption. The effect of additional price in-
Behavioral Economics 117
FIGURE 7.1. Demand functions (left panel) and work functions (right panel) for food and saccharin across a range of
prices for two rhesus monkeys. From Hursh and Roma (2013). Copyright © 2013 Society for the Experimental Analy-
sis of Behavior. Reprinted by permission.
creases is greater as the price increases. In the left we do not set reinforcement schedules that main-
panel of Figure 7.1, changes from FR 10 to FR 100 tain responding too high or too low.1
produced small decreases in the number of food The effect of price changes on consumption can
pellets earned per day. Consumption began to de- differ across reinforcers. In Figure 7.1, saccharin
crease more steeply as the cost exceeded FR 100. and food pellets have different demand profiles;
Graphing demand functions on double-logarith- consumption of saccharin decreased more rapidly
mic axes facilitates visual and quantitative analy- than consumption of food pellets across similar
sis of the proportional relation between changes price increases. Thus a subject may consume two
in price and changes in consumption (Hursh et reinforcers similarly when both are inexpensive,
al., 2013). but consumption of one may decrease more rap-
Price elasticity of demand is a measure of the idly than consumption of the other as prices in-
change in consumption that accompanies a price crease. The rate at which consumption decreases
change. When the decrease in consumption is depends upon the necessity and initial value of the
proportionally less than the increase in price, the reinforcer and external factors, like the availabil-
slope of the double-logarithmic demand function ity of alternative reinforcers and constraints in the
is shallower than –1. Across this range of price economies.
increases, we describe demand as inelastic. In con- The right panel of Figure 7.1 depicts work func-
trast, the portion of the food function after FR 100 tions, or total responding for the two reinforcers
in Figure 7.1 shows elastic demand. At higher costs, across unit prices. The work function character-
further increases in price produce suprapropor- istically shows a positive slope at low prices as the
tional changes in consumption. We should expect monkeys emitted more responses to defend con-
more rapid decreases in consumption with further sumption (i.e., to earn the same number of rein-
increases in price after demand elasticity shifts
from inelastic to elastic. In Figure 7.1, the lines
1 Readersinterested in the important topics of quantifying
labeled Pmax indicate the price at which demand
shifted from inelastic to elastic for each reinforcer. and extrapolating the shape of demand functions, deriving
Pmax, and so on should see Hursh et al. (2013) and Hursh
This shift in demand corresponded to the price and Silberberg (2008). Those interested in relevant treat-
at which consumption became highly sensitive ments related to neurodevelopmental disorders should see
to further price increases. Knowing the price at Reed, Kaplan, and Becirevic (2015) and Gilroy, Kaplan,
which such a shift in demand occurs ensures that and Leader (2018).
118 B a s i c P r in c ip l e s a nd C o n c e p t s
forcers), even though it became increasingly ex- al. (2017) used progressive-ratio schedules to com-
pensive. The demand function turns from inelastic pare the reinforcing efficacy of a solitary or a social
to elastic (Pmax) at the peak of the work function, activity. Children with autism spectrum disorder
which begins to slope downward with additional earned 30 seconds to do an activity alone on a pro-
price increases. Thus work functions generally as- gressive-ratio schedule that began at one response
sume an inverted-U shape. Work functions may for the first unit of reinforcement, but increased
estimate how high an applied behavior analyst arithmetically by a step size of 10 responses for
could set a reinforcement schedule requirement each additional unit. The progressive-ratio sched-
before negatively affecting levels of responding. ule continued until a child stopped responding for
30 seconds or verbally indicated that he or she was
finished responding. Investigators repeated the
Assessing Stimulus Value
procedure, but the reinforcer was doing the activ-
with Progressive‑Ratio Schedules
ity with a parent. The investigators presented the
A related method for assessing stimulus value in proportion of children who “purchased” each addi-
relation to price is the progressive-ratio schedule tional 30-second reinforcement unit at each price
(Hodos, 1961). Progressive-ratio schedules, like in each condition. This method of presenting the
demand curves, determine reinforcer strength data is similar to a survival analysis. Figure 7.2
by measuring the amount of work an individual shows the data, which resembles a demand func-
will complete to earn a reinforcer. However, cost tion relating consumption (proportion of children
increases in progressive-ratio schedules usually paying each price) to increasing prices. A higher
occur within an experimental session after each proportion of children paid most prices to do the
reinforcer delivery (but see Jarmolowicz & Lattal, activity with a parent, perhaps contradicting con-
2010). The response requirement might begin at ventional notions of the value of social interaction
FR 1 in each session and then increase after each in children with autism spectrum disorder.
successive reinforcer delivery, often in arithme- Progressive-ratio schedules are a quick and ef-
tic or geometric progressions. The step size is the fective way to assess stimulus value. Behavior ana-
number of or formula for additionally required lysts can identify breakpoints across progressive-
responses. The breakpoint is the principal depen- ratio schedules, directly compare the reinforcing
dent measure, which is the last schedule value the efficacy of two or more stimuli, and use this in-
individual completes before responding stops for a formation to develop more robust reinforcement-
specified time (e.g., 2 or 5 minutes). We might rea- based procedures. Behavior analysts can construct
sonably conclude that Stimulus A is a more effec- progressive-ratio schedules in a variety of ways,
tive reinforcer than Stimulus B if responding stops and depicting the results graphically can show use-
after completion of an FR 20 schedule for Stimulus ful economic relations between behavior and the
A and after an FR 10 schedule for Stimulus B. environment.
Investigators have shown the results of progres-
sive-ratio schedules in applied research by depicting
Factors That Influence Demand
(1) the breakpoint of each progressive-ratio session
as a line graph across sessions and conditions (e.g., When prices increase, demand for some goods is
Francisco, Borrero, & Sy, 2008, Figure 3; Jerome & generally more elastic than demand for others. The
Sturmey, 2008; Russell, Ingvarsson, Haggar, & Jes- two curves in Figure 7.1 reflect this relation. In this
sel, 2018, Figure 2); (2) the total response output figure, consumption of food pellets remained rela-
per session (e.g., Fiske et al., 2015; Tiger, Toussaint, tively unaffected through price increases as high
& Roath, 2013); (3) the mean breakpoint across as FR 100. By contrast, saccharin consumption de-
repeated progressive-ratio sessions (Call, Trosclair- creased more rapidly across equal price increases.
Lasserre, Findley, Reavis, & Shillingsburg, 2012, Behavioral economists have evaluated several in-
Figure 7.3; DeLeon, Frank, Gregory, & Allman, fluences on the shape of these curves, including
2009); and (4) the proportion of opportunities the nature of the economy (open vs. closed) and
that a participant or group of participants earned the substitutability of reinforcers.
reinforcement at each schedule value across multi-
ple progressive-ratio sessions (Hoffmann, Samaha,
Open versus Closed Economies
Bloom, & Boyle, 2017; Roane, Lerman, & Vorn-
dran, 2001; Trosclair-Lasserre, Lerman, Call, Ad- In a closed economy, the reinforcer or a close substi-
dison, & Kodak, 2008). For example, Goldberg et tute is only available in the earning environment
Behavioral Economics 119
1.0
Proportion of Participants
0.6
0.4
0.2
Solitary Play
Activity Embedded in Social Context
0.0
0 10 20 30 40 50 60 70 80 90 100
FIGURE 7.2. Survivor analysis of solitary play and social play as a proportion of participants who completed each sched-
ule requirement. Data replotted from Goldberg et al. (2017).
(Reed, Niileksela, & Kaplan, 2013). For example, were about twice as high as those in a condition
a child can earn access to a highly preferred movie with postsession reinforcement. These results sug-
for finishing schoolwork, but that movie is never gest that reinforcers will sustain more responding
available otherwise in a closed economy. In an if they are only available in the earning environ-
open economy, the reinforcer or a close substitute ment versus when reinforcement is available else-
is available in and out of the earning environment. where.
For example, a child in an open economy can earn
Skittles candies by touching a communication
The Continuum of Substitutability
card in the clinic, but the caregiver allows free ac-
cess to Skittles at home. Human environments are Substitutability is a variable that influences demand
typically open economies with multiple sources of elasticity on a continuum (Green & Freed, 1993).
reinforcement. Stimuli are highly substitutable when decreased
Several applied behavioral-economic studies consumption of one stimulus results in increased
have examined the effects of economy type on re- consumption of the other stimulus when income
sponding across price increases. In the open econ- constrains choice or the number of reinforcers
omy in Roane, Call, and Falcomata (2005), correct an individual can earn. For example, decreased
responding produced reinforcement; however, the consumption of coffee may be associated with
investigators gave the participants supplemental increased consumption of tea. Demand for com-
reinforcement if they did not earn the maximum modities with many substitutes tends to be more
number of possible reinforcers during the session. elastic than demand for commodities with few or
Correct responding was the only way participants no substitutes. Thus many people may switch to
accessed reinforcement in the closed economy. tea if the price of coffee increases dramatically. By
Both arrangements increased adaptive respond- contrast, consumption of heating oil tends to re-
ing relative to baseline; however, demand was main relatively stable regardless of price, because
more elastic in the open economy. Participants it has no close substitute.
were less likely to work for reinforcement at the We can only determine substitutability by ex-
higher-response requirements when reinforcement amining changes in consumption experimentally.
was available outside the work environment in the Specifically, we increase the price of one reinforcer,
open economy. Kodak, Lerman, and Call (2007) hold the price of the other constant, and measure
found that mean progressive-ratio breakpoints in relative consumption of each reinforcer at each
a condition without postsession reinforcement concurrent-schedule value. Lea and Roper (1977)
120 B a s i c P r in c ip l e s a nd C o n c e p t s
examined rats’ demand for mixed-diet pellets dur- than using a doll and a dollhouse. However, using
ing increases in ratio-schedule requirements (FR the doll and dollhouse may be necessary if they
1, FR 6, FR 11, and FR 16), and separate responses are complementary, and the doll is reinforcing
produced no food, sucrose, or mixed-diet pellets on only with the dollhouse or another independent
a constant FR 8 schedule. As the price of pellets stimulus. Similar problems exist when substitut-
increased, demand for pellets decreased most, and able reinforcers are readily available. For example,
demand for the alternative reinforcer increased acquisition may be slower with popcorn as the pro-
most, when pellets were concurrently available grammed reinforcer if other salty snacks (i.e., close
and least when the most dissimilar consequence substitutes) are available freely.
(no food) was concurrently available. Thus the
identical edible stimulus was most substitutable,
the other edible stimulus was intermediately sub- THE APPLIED RELEVANCE
stitutable, and the absence of an edible stimulus OF ASSESSING STIMULUS VALUE
provided no substitute.
Individuals tend to consume complementary Analyses of reinforcer consumption under differ-
reinforcers together; consumption of one tends ing response requirements are particularly relevant
to vary positively with consumption of the other. in applied settings. Reinforcement schedule thin-
Increased consumption of tea may produce corre- ning from rich contingencies to more stringent re-
sponding increases in consumption of sugar. If the sponse requirements improves treatment practical-
price of one complement decreases its consump- ity and is a common activity for behavior analysts.
tion, consumption of the other will also decrease. Essentially, the behavior analyst has created a
Increases in the price of tea may produce both demand function when the response requirements
decreased tea and sugar consumption. Comple- to produce the same reinforcer increase across ses-
mentary reinforcers can influence stimuli hypoth- sions.
esized to function as reinforcers in several ways.
The removal of one stimulus that was reinforcing
Implications for Selecting Reinforcers
behavior may decrease the reinforcing effective-
ness of other stimuli also present. For example, a Investigators have used demand curves and pro-
child may play with a basketball when a hoop is gressive-ratio schedules to validate predictions of
available, but not when it is unavailable. The bas- the relative value of stimuli identified with other
ketball alone may not maintain responding if the preference assessment procedures (e.g., Glover,
basketball and hoop are complementary reinforc- Roane, Kadey, & Grow, 2008; Martin, Franklin,
ers. Similarly, when a reinforcer has many different Perlman, & Bloomsmith, 2018; Penrod, Wallace,
stimulus properties, removal of one property may & Dyer, 2008). For example, DeLeon et al. (2009)
alter the stimulus’s reinforcing value. For example, showed that stimuli identified as highly preferred in
attention from one therapist may function as a re- a paired-stimulus preference assessment produced
inforcer, but attention from another therapist may higher mean break points in a progressive-ratio
not. Alternatively, praise with tickles may func- schedule than stimuli identified as less preferred.
tion as a reinforcer, but praise with a pat on the Fiske et al. (2015) showed that total response out-
back may not. In these examples, the combination put and breakpoints in progressive-ratio schedules
of stimuli or stimulus features interact in such a matched for token reinforcers and the primary re-
way that an increase in effort required to gain ac- inforcers with which they had paired the tokens.
cess to either stimulus will decrease responding By contrast, response output and breakpoints for
towards both stimuli. Conversely, a decrease in ef- unpaired tokens approximated those of the no-re-
fort required to gain access to either stimulus will inforcement control conditions, which were much
increase responding toward both stimuli. lower. Researchers have also used demand curves
Independent commodities are those in which and progressive-ratio schedules to evaluate the ac-
the price of one commodity has no effect on the curacy of different preference assessment formats
consumption of another. Increasing the response (e.g., Call et al., 2012; Reed et al., 2009).
requirements to earn access to a tablet should not Choice responses in stimulus-preference assess-
change the effectiveness of food as a reinforcer, so ments typically require minimal effort from sub-
consumption of food will remain constant. Inde- jects, like a reaching response (DeLeon & Iwata,
pendence is often desired in application. For ex- 1996; Fisher et al., 1992; Pace, Ivancic, Edwards,
ample, using a doll as a reinforcer would be easier Iwata, & Page, 1985; Roane, Vollmer, Ringdahl, &
Behavioral Economics 121
FIGURE 7.3. Own-price demand functions showing total number of reinforcer deliveries across various prices (ratio
values) of each stimulus. From Reed et al. (2009). Reprinted by permission of the publisher (Taylor & Francis Ltd,
www.tandfonline.com).
Marcus, 1998). Unfortunately, relative reinforcer Reed et al. (2009) and other studies of participants
effectiveness identified under such low-cost condi- with neurodevelopmental disabilities (e.g., DeLe-
tions does not always remain constant when re- on, Iwata, Goh, & Worsdell, 1997; Tustin, 1994,
sponse requirements increase. Investigators have 2000) suggest that behavior analysts should evalu-
used demand curves and progressive-ratio sched- ate reinforcer effectiveness under conditions that
ules to determine relative value among reinforcers approximate their use in practice.
when the effort required to produce them increas- Tustin (1994) was the first to explicitly use a
es. Some studies have shown that distinct reinforc- behavioral-economic framework with participants
ers can maintain equivalent response levels when with a neurodevelopmental disorder. The demand
unit price is low, but increases in unit price reveal function for a visual reinforcer had a steeper slope
important differences across stimuli in maintain- when an auditory reinforcer was concurrently
ing responding, as in Figure 7.1. For example, Reed available relative to the demand function for the
et al. (2009) evaluated six edible reinforcers inde- same visual reinforcer when 5 seconds of social
pendently during progressive-ratio schedules that interaction were concurrently available. This find-
began with FR 1 and increased after every second ing suggested that the two sensory reinforcers were
reinforcer. The participant had six opportunities more substitutable than the sensory reinforcer and
to earn each reinforcer.2 Figure 7.3 shows that the attention.
participant earned Cheerios and Goldfish five Individuals with neurodevelopmental disorders
times, and the other foods six times at FR 1. No- in DeLeon et al. (1997) could press two differ-
table differences in responding between Cheerios, ent microswitch panels to produce two different
Goldfish, and Sour Patch Kids and the other ed- reinforcers according to progressively increasing,
ibles began at FR 2, with no responding at FR 2 concurrent-ratio schedules. Participants did not
for Cheerios, at FR 5 for Goldfish, and at FR 20 show a clear preference for one reinforcer over the
for Sour Patch Kids. Pringles, Doritos, and Gush- other when the two were functionally dissimilar,
ers maintained a decreasing number of responses like an edible item and a toy. By contrast, partici-
until FR 30, when responding stopped. Results of pants showed a clear preference for one reinforcer
as the schedule requirements increased from FR 1
2 Evaluations of this sort (i.e., assessing demand for each to FR 5 or FR 10 when the concurrently available
reinforcer independently of other reinforcers) determine reinforcers were functionally similar, like two food
own-price demand. items. DeLeon et al. (1997) suggested that the sub-
122 B a s i c P r in c ip l e s a nd C o n c e p t s
stitutability between functionally similar but not unit price had increased). In a second analysis,
functionally dissimilar reinforcers produced this they again increased the DRO interval, but also
effect. They also proposed that increased schedule proportionally increased the reinforcement com-
requirements may magnify small differences in rel- ponent. Problem behavior was maintained at low
ative preference for reinforcers that share physical rates up to a 180-second DRO interval and a cor-
characteristics (like the two food items) because responding 360-second reinforcement interval.
they are also likely to share functional properties, Thus a strategy for increasing the effectiveness of
like hunger reduction. Participants could exclu- behavioral treatments across price increases in-
sively consume the more preferred or less costly volves corresponding changes in the magnitude of
reinforcer without experiencing deprivation when the reinforcer to preserve unit price.
the two concurrently available reinforcers served Others have similarly manipulated unit price
the same function (e.g., either food reduced hun- parametrically by varying either the schedule re-
ger). By contrast, when the two stimuli served quirements or reinforcer magnitude to produce
different functions, exclusive consumption of one more favorable response–reinforcer ratios (Bor-
reinforcer resulted in deprivation of the other rero, Francisco, Haberlin, Ross, & Sran, 2007;
(e.g., sustenance, but no fun or leisure). If each Trosclair-Lasserre et al., 2008). Trosclair-Lasserre
reinforcer is important or valued, individuals may et al. (2008) showed less elastic demand for larger-
continue to allocate responding to both options as magnitude reinforcers (120 seconds of attention)
schedule requirements increase. than smaller-magnitude reinforcers (10 seconds of
attention) for two of four children whose problem
behavior was maintained by attention. Other in-
Implications for Weakening Undesirable Behavior
vestigators have used unit-price manipulations to
Investigators have adopted economic concepts to inform treatment development (DeLeon et al.,
treat problem behavior through parametric evalua- 2000; Roane et al., 2001; Wilson & Gratz, 2016).
tions of differential consequences for problem and DeLeon et al. (2000) used an abbreviated progres-
appropriate behaviors (DeLeon, Fisher, Herman, sive-ratio schedule to accurately predict the differ-
& Cros land, 2000; DeLeon, Neidert, Anders, ence between reinforcement schedules for aggres-
& Rodriguez-Catter, 2001; Delmendo, Borrero, sive and alternative behaviors needed to maintain
Beauchamp, & Francisco, 2009; Kerwin, Ahearn, more alternative behavior than aggression. Roane
Eicher, & Burd, 1995; Perry & Fisher, 2001). Some et al. (2001) showed that reinforcers with less
researchers have manipulated the costs and ben- elastic demand profiles were generally more effec-
efits of responding to treat problem behavior or tive as reinforcers in DRO schedules used to treat
to maintain treatment effects across unit-price problem behavior than were reinforcers with more
changes when the reinforcers for problem and al- elastic profiles.
ternative behaviors are symmetrical. Others have
used substitutability concepts. We consider these
Stimulus Substitutability
approaches in separate sections.
Numerous human and nonhuman behavioral
pharmacology and addiction studies have used
Unit‑Price Manipulations
economic analyses to examine whether certain
Investigators have manipulated the cost and ben- compounds can be substituted for drugs of abuse
efit components of unit price and assessed its ef- (e.g., Johnson & Bickel, 2003; Petry & Bickel,
fects on problem behavior. For example, Roane, 1998; Pope et al., 2019; Shahan, Odum, & Bickel,
Falcomata, and Fisher (2007) showed that a 2000; Smethells, Harris, Burroughs, Hursh, &
schedule for differential reinforcement of other Lesage, 2018). For example, Pope et al. (2019) re-
behavior (DRO) in which a 10-second absence cently used demand curves to examine the substi-
of inappropriate vocalizations produced 20 sec- tutability of nicotine gum, chewing tobacco, and
onds of reinforcement was effective. Next, they e-liquid to cigarettes. Only e-liquid, the substance
increased the DRO interval and held the amount in electronic cigarettes, substituted for actual ciga-
of reinforcement constant. The DRO became in- rettes. Substitutability increased as the nicotine
effective when the interval reached 23 seconds. A concentration in the e-liquid increased, but the
behavioral-economic interpretation suggests that overall substitution effects were marginal.
the participant was “paying” with a longer delay Investigators have also examined whether
for the same amount of the reinforcer (i.e., the foods that promote obesity and either healthier
Behavioral Economics 123
foods or activity reinforcers are substitutable. For after every fifth trial. Some participants chose
example, Salvy, Nitecki, and Epstein (2009) ex- work regardless of the reinforcement schedule or
amined whether social activities with a friend or the specific reinforcer. Others chose to work when
with an unfamiliar peer would substitute for food the response requirements were low, but switched
in overweight and nonoverweight children. Social to escape as the response requirements increased.
interaction with an unfamiliar but not a familiar Furthermore, different edible reinforcers produced
peer substituted for food when the cost of food different demand profiles, as in Figure 7.1. Results
increased for overweight and nonoverweight chil- suggested that behavior analysts could use this
dren. Temple, Legierski, Giacomelli, Salvy, and procedure to predict the relative effectiveness of
Epstein (2008) evaluated demand for food on an reinforcers to treat escape-maintained problem
escalating reinforcement schedule versus nonfood behavior during reinforcement schedule thinning.
activities on a constant, low schedule. Generally,
overweight children completed higher schedule
values for food as the price increased than non- DELAY AND PROBABILITY AS COST
overweight children did. In addition, demand
was less elastic for food in overweight children as Thus far, we have mostly considered cost as the
prices increased when alternative activities were effort required to produce a reinforcer. Delay to
concurrently available. reinforcement is another “cost” in choice arrange-
Shore, Iwata, DeLeon, Kahng, and Smith ments. In fact, the few studies that have directly
(1997) were perhaps the first to explicitly examine compared effort and delay as costs have found
substitutability in treating problem behavior for parallel effects: Increases in either can result in
individuals with neurodevelopmental disorders. decreased reinforcer consumption (e.g., Bauman,
Participants switched from self-injurious behavior 1991). Thus, just as increases in response require-
(SIB) to engagement with preferred leisure items ments can shift responding away from more ex-
that were available continuously and noncontin- pensive items, increases in delay can have the
gently, suggesting that leisure items were substitut- same effect.
able for the automatic reinforcement from SIB. In
a second experiment, participants had leisure-item
Assessing Delay Discounting
access on a DRO schedule, and SIB increased. In
a third experiment, the investigators tied the lei- Much of what is known about the impact of delay
sure item to a string and measured item engage- has been determined in experiments on delay dis-
ment as a proportion of string length. Thus string counting. This term usually refers to the decrease
length was the measure of effort (i.e., how far a in a reinforcer’s present subjective value as delay
participant had to reach to procure the item). SIB increases (Mazur, 1987; Critchfield & Kollins,
increased and leisure-item engagement decreased 2001). In other words, the efficacy of a reinforcer
when the string was half of its original length. may decrease as a function of delays to its deliv-
Similarly, Zhou, Goff, and Iwata (2000) assessed ery. A delay-discounting task assesses preferences
preference for automatically reinforced problem between a smaller–sooner reward and a larger–
behavior versus preferred leisure-item engagement. later reward at variable delays. Often participants
Participants engaged in problem behavior when choose the larger–later reward when the delivery
both responses were freely available. Leisure-item of both the smaller–sooner reward and larger–later
engagement substituted for problem behavior reward are set in the future. Investigators often ob-
when the participants wore a restrictive sleeve that serve a change in preference to the smaller–sooner
increased response effort for problem behavior. reward as its delivery time draws closer. For ex-
Results of these studies demonstrate that increases ample, given a choice between $50 in 9 months
in effort for one response will likely increase re- or $100 in 1 year, most participants would prefer
sponding that produces a substitutable reinforcer. the larger-magnitude reward. By contrast, prefer-
Frank-Crawford, Castillo, and DeLeon (2018) ence reversal occurs when, given a choice between
examined whether edible reinforcers would sub- $50 now or $100 in 3 months, preference for the
stitute for escape from instructional activities. smaller, more immediate reward emerges (Tversky
Participants could choose to complete a task and & Thaler, 1990).
earn an edible reinforcer, or to take a 30-second Investigators have replicated the results de-
break with no edible reinforcer. The response scribed above across species (e.g., Kirby & Herrn-
requirements for the edible reinforcer increased stein, 1995; Mazur, 1987; Mazur & Biondi, 2009;
124 B a s i c P r i n c ip l e s a nd C o n c e p t s
Rachlin, Raineri, & Cross, 1991); among indi- $1,000 delivered immediately. These investigators
viduals with substance abuse (e.g., Bickel, Odum, additionally examined probability discounting, or
& Madden, 1999; Coffey, Gudleski, Saladin, & the extent to which the odds of actually receiv-
Brady, 2003); and across monetary gains and loss- ing the larger reinforcer would influence choices
es (Odum, Madden, & Bickel, 2002), which are against certain, but smaller rewards. Participants
predicted by a hyperbolic discounting function. in the probability-discounting group could choose
Mazur (1987) was one of the first investigators to between receiving $1,000 in probabilities that
assess delay-discounting rates in pigeons by using varied from 5 to 100%, or an amount that varied
an adjusting-delay procedure. Mazur held the from $1 to $1,000 delivered for certain. For ex-
amounts of the smaller and larger rewards con- ample, investigators asked participants to choose
stant and adjusted the delay to the larger reward between having a 50% chance of receiving $1,000
delivery across trials. Pigeons selected between or receiving $1 for certain. Results showed that
2-second access to grain with a 2-second delay and the hyperbolic discounting function described
6-second access with an adjusted delay. The delay discounting for both delayed choices and choices
to the larger–later reward in the next trial in- with probabilistic outcomes.
creased if a pigeon selected the larger–later reward
and decreased if the pigeon selected the smaller–
Determinants of Delay Discounting
sooner reward. This procedure continued until the
pigeon reached an indifference point, in which it The magnitude effect is the effect of reward mag-
selected either alternative 50% of the time. Thus nitude on delay and probability discounting and
the subjective value for the smaller–sooner reward is a robust finding in the discounting literature.
and larger–later reward were about equal at the in- In delay discounting, individuals discount smaller
difference point. Mazur observed that indifference reward magnitudes more steeply than larger ones,
points follow a hyperbolic function, and suggested with responding for smaller rewards decreas-
the following hyperbolic discounting equation: ing more rapidly across similar increases in delay
(Green, Fry, & Myerson, 1994; Green, Myerson,
A & McFadden, 1997; Odum, Baumann, & Rim-
V=
1 + kD ington, 2006). In other words, the decrease in
the subjective value of smaller delayed rewards is
where V is the subjective value of a reward; A is greater than the decrease in the subjective value of
the amount of the reward; d is the delay to the re- larger delayed rewards. Investigators have observed
ward; and k is a free parameter that describes how the opposite effect when rewards are probabilistic:
much the delay affects the value of a reward, also Individuals discount smaller probabilistic rewards
called the discounting rate. The discounting rate less steeply than larger ones (Estle, Green, Myer-
describes how steeply delay decreases the value of son, & Holt, 2007). Investigators often use hypo-
a reward, as measured by response rate or choice thetical rewards in these procedures, to facilitate
for one reward over another. In other words, the assessment of large-magnitude rewards that inves-
subjective value of a reward decreases as a function tigators typically cannot pay due to financial con-
of the delay to the delivery of that reward as k in- straints (e.g., $1,000). Hypothetical rewards also
creases. As a fitted parameter, k often serves as the allow investigators to assess long delays (e.g., 50
dependent variable in delay-discounting studies years) that may be impractical with real rewards.
and may be considered an individual-differences Importantly, studies have not found differences
parameter, sensitive to both state and trait influ- in discounting between real and hypothetical re-
ences (see Odum, 2011). wards across preparations (Kirby & Maraković,
Alternatively, investigators sometimes hold the 1995; Johnson & Bickel, 2002; Lagorio & Mad-
delay and amount of the larger–later reward con- den, 2005).
stant and adjust the amount of the smaller–sooner The domain effect, or the type of rewards evalu-
reward until an individual demonstrates indiffer- ated, can also influence discounting. The hyper-
ence in the adjusting-amount procedure. For exam- bolic function also describes discounting of non-
ple, Rachlin et al. (1991) assessed choices between monetary rewards like alcohol (Petry, 2001; Estle
large delayed and small immediate hypothetical et al., 2007), cigarettes (Bickel et al., 1999; Odum
rewards. Participants could choose between re- et al., 2002; Reynolds, Richard, Horn, & Karraker,
ceiving $1,000 delivered at delays from 1 month 2004), food (Charlton & Fantino, 2008; Odum &
to 50 years, or an amount that varied from $1 to Rainaud, 2003), and sexual outcomes (Johnson &
Behavioral Economics 125
Bruner, 2012; Lawyer, Williams, Prihodova, Rol- nation hyperactive/impulsive) display steeper dis-
lins, & Lester, 2010), but differences in the rate counting rates for monetary rewards than children
of discounting can be observed across reinforcer diagnosed with ADHD, inattentive type (ADHD-
classes. For example, Odum and Rainaud (2003) I) and their typically developing peers (Scheres,
found that participants discounted money less Tontsch, Thoeny, & Kaczkurkin, 2010; Scheres,
steeply than preferred food and alcohol, which they Tontsch, & Thoeny, 2013; Sonuga-Barke, Taylor,
discounted at similar rates. Charlton and Fantino Sembi, & Smith, 1992). Rosch and Mostofsky
(2008) theorized that the domain effect observed (2016) assessed discounting for leisure activities
across consumable commodities was due to meta- in children with ADHD-C and typically develop-
bolic function. They observed that participants ing children. They found that girls with ADHD-C
discounted money less steeply than entertainment displayed a greater discounting rate for leisure ac-
media, and discounted entertainment media less tivities than boys with ADHD-C and the typically
steeply than preferred food. Thus participants tend developing group.
to discount more steeply those rewards that have a Delays in the delivery of reinforcers are ubiq-
direct metabolic function (e.g., food, water, drugs) uitous in the natural environment. For example,
relative to those that do not (e.g., money, enter- teachers often use tokens to bridge the delay be-
tainment media). However, Charlton and Fantino tween the target response and reinforcer delivery
also offered that domain effects may be related to (Hackenberg, 2009). Reed and Martens (2011)
commodity fungibility, or the degree to which a conducted a discounting assessment for hypotheti-
commodity is exchangeable for other commodi- cal monetary rewards with 46 typically developing
ties. Money, for example, is exchangeable for vari- sixth graders. They then implemented a classwide
ous primary and secondary reinforcers, which may intervention where students could earn tokens ex-
account for the shallower rate of discounting for changeable for preferred items for on-task behav-
money versus less fungible commodities, like food ior. Token exchange occurred immediately after
and commodity-specific gift cards (Holt, Glodows- the session or at the beginning of the next day’s
ki, Smits-Seeman, & Tiry, 2016). session. Reed and Martens found that discount-
ing rates correlated with on-task behavior during
the intervention. Students who displayed higher
Delay Discounting in Clinical Populations
scores on the discounting assessment engaged in
Studies have shown that populations with sub- less on-task behavior when it produced delayed
stance use disorders discount their preferred drugs token exchange. Therefore, the delay-discounting
more steeply than monetary rewards (Coffey et assessment for hypothetical rewards was predic-
al., 2003; Madden, Petry, Badger, & Bickel, 1997). tive of the effects of delayed rewards on classroom
Delay-discounting studies with obese adults have behavior. Investigators have used similar delay-dis-
found that they discount monetary rewards more counting preparations with young children; results
steeply than average-weight adults do (Weller, suggest that these assessments may be useful for
Cook, Avsar, & Cox, 2008; Epstein et al., 2014; examining developmental progressions, academic
Bickel, Wilson, et al., 2014). These results suggest disorders, and/or neurodevelopmental disorders in
that individuals with substance use disorders and children (see Staubitz, Lloyd, & Reed, 2018)
obesity prefer immediate over delayed rewards Leon, Borrero, and DeLeon (2016) assessed
more than their non-substance-misusing and non- the effects of delays for primary and conditioned
overweight peers. reinforcers on the responding of individuals with
Characteristics like age (Green et al., 1994; neurodevelopmental disorders. Responding was
Scheres et al., 2006), IQ (de Wit, Flory, Acheson, maintained at high levels when the investigators
McCloskey, & Manuck, 2007), and diagnosis (De- delivered food or tokens immediately after the re-
murie, Roeyers, Baeyens, & Sonuga-Barke, 2012) sponse or exchanged tokens for food immediately
can affect discounting rates. Individuals with after a participant earned the token. Respond-
impulse-control disorders may be particularly af- ing decreased most as delays to token delivery
fected by delays to reinforcement; thus studying increased, followed by delays to token exchange.
discounting with this population is important. Delays to food reinforcers maintained responding
Research using hypothetical and real monetary at longer delays than delays to token delivery or
rewards delivered in real-time have found that token exchange.
children diagnosed with attention-deficit/hyperac- Delayed reinforcement may also affect skill ac-
tivity disorder, combined type (ADHD-C; combi- quisition in individuals with neurodevelopmental
126 B a s i c P r in c ip l e s a nd C o n c e p t s
disorders. Sy and Vollmer (2012) found that al- price densities (e.g., Reed, Kaplan, Roma, & Hursh,
though some participants learned conditional dis- 2014) and sequences (e.g., Amlung & MacKillop,
criminations during delayed-reinforcement condi- 2012). The sound psychometric properties and
tions, others did not. Similarly, Carroll, Kodak, correspondence of hypothetical purchase tasks
and Adolf (2015) showed that delays to reinforce- with known behavioral-economic variables is im-
ment decreased the efficiency and effectiveness of portant, because these procedures are helpful in
discrete-trial instruction for a receptive-identifica- studies that would be unethical or infeasible to do
tion task. otherwise.
In summary, delays to reinforcement may have Investigators have used hypothetical purchase
a negative effect on some individuals with neuro- tasks to predict how consumption for various com-
developmental disorders. Investigators have used modities change as a function of price. For ex-
progressive schedules to teach such individuals ample, Kaplan and Reed (2018) used an alcohol
to tolerate delay to reinforcement. For example, purchase task to examine the effects of happy-hour
Dixon et al. (1998) gave three adults with neuro- drink pricing on excessive alcohol consumption.
developmental disorders a concurrent choice be- Higgins et al. (2017) demonstrated the utility of
tween a smaller- and larger-magnitude reward, and cigarette purchase tasks in simulating effects of
asked these adults to engage in a concurrent activ- cigarette nicotine content regulations and pric-
ity (e.g., exercise) during the delay to the larger ing policies on cigarette smokers. These studies
reward. Dixon et al. gradually increased the delay and others show that hypothetical purchase tasks
to the larger reward, and found that selection of provide a safe simulation of various clinically im-
the larger delayed reward and the time engaging portant consumer behaviors and open the door for
in the concurrent activity increased for the three large-scale applications of behavioral economics at
participants. Further research demonstrated that the policy level.
individuals with autism spectrum disorder pre-
ferred the activity over waiting without an activity
Reinforcer Pathology
(Dixon & Cummings, 2001; Dixon, Rehfeldt, &
Randich, 2003). Investigators have used a reinforcer pathology model
to characterize some forms of maladaptive behav-
ior. This model classifies maladaptive behavior
EMERGING THEMES IN BEHAVIORAL ECONOMICS based on responses to reinforcers and manipula-
tion of reinforcement parameters, such as exces-
Hypothetical Purchase Tasks
sive demand (i.e. willingness to pay much higher
Recent research has studied hypothetical choices prices for a commodity) or excessive sensitivity to
among commodities with different prices rather delay (i.e., discounting functions for the commod-
than delays (see discussion by Roma, Reed, Di- ity that indicate excessive value of immediacy)
Gennaro Reed, & Hursh, 2017). Investigators have relative to the norm (Bickel, Jarmolowicz, Mueller,
used a hypothetical purchase task, which asks par- & Gatchalian, 2011).
ticipants to indicate the quantity of a commodity Reinforcer pathology has only recently been de-
they would purchase at varying price points. Hy- fined as the combined effects of (1) a consistent
pothetical purchase tasks demonstrate adequate high valuation of a particular reinforcer or com-
relations with actual consumption (e.g., Amlung, modity and (2) excessive preference for immediate
Acker, Stojek, Murphy, & MacKillop, 2012), pre- reinforcement despite possible negative outcomes
dictive validity (e.g., MacKillop & Murphy, 2007), (Carr, Daniel, Lin, & Epstein, 2011). For example,
construct validity (e.g., MacKillop et al., 2010), cigarette smokers’ demand for nicotine is inelastic
and temporal stability (e.g., Few, Acker, Murphy, under conditions of deprivation (Madden & Bick-
& MacKillop, 2012). Behavioral-economic find- el, 1999); current cigarette smokers also discount
ings with hypothetical purchase tasks reproduce the value of delayed money more steeply than
those observed in operant laboratories, suggesting never- and ex-smokers, and more so for delayed
that they reflect real-world behavior. For example, cigarettes than for delayed money (Bickel et al.,
these tasks appear sensitive to economy type (e.g., 1999). Therefore, the relation between excessive
Kaplan et al., 2017) and availability of substitutes discounting and demand for the misused com-
(e.g., Grace, Kivell, & Laugesen, 2014); they also modity may account for substance use disorders,
demonstrate consistent elasticity across differing but research on this relation has been inconclusive
Behavioral Economics 127
(Bickel, Johnson, Koffarnus, MacKillop, & Mur- depending upon whether one is a cognitively or
phy, 2014). Rats with steep delay discounting for behaviorally oriented theorist, behavior analysts
food demonstrated inelastic demand for nicotine should always be interested in procedures that reli-
self-administration (Diergaarde, van Mourik, Pat- ably influence human choice.
tij, Schoffelmeer, & De Vries, 2012). Conversely,
Field, Santarcangelo, Sumnall, Goudie, and Cole
(2006) found that human participants demon- CONCLUSION
strated steeper delay discounting for monetary
rewards, but that demand for cigarettes was unaf- Behavioral economics represents the intersection
fected during nicotine deprivation. of microeconomic concepts and behavioral psy-
According to the reinforcer pathology model, chology. Both traditional and operant behavioral
intervening in substance use should produce economics are gaining notoriety in mainstream
change in measures of demand and discounting. media and policy making (Hursh & Roma, 2013),
Madden and Kalman (2010) monitored smokers’ in part due to the face-valid approaches and social-
cigarette demand by using a purchase task dur- ly significant outcomes across a range of applica-
ing a course of either bupropion, a medication tions (see the bibliometric analysis by Costa, Car-
that dampens the reinforcing effects of nicotine, valho, & Moreira, 2019). The operant approach to
or counseling treatment. Shifts in simulated de- behavioral economics is entirely compatible with
mand after 1 week of either treatment predicted the dimensions of behavior analysis and offers a
eventual cessation. In a related study, McClure, unique lens to aid in basic (Hursh & Roma, 2016)
Vandrey, Johnson, and Stitzer (2013) found that and practical (Reed et al., 2013) issues. Although
participants receiving varenicline (a medication behavior analysis has increasingly used behavioral
that reduces nicotine craving and withdrawal) economics to address applied issues (see Gilroy
demonstrated greater demand elasticity for simu- et al., 2018), the field remains relatively small.
lated cigarette purchases in a programmed relapse Emerging concepts in behavioral economics, like
period than those receiving a placebo. Additional reinforcer pathology theory (e.g., Bickel et al.,
research on the effects of substance use treatment 2011) and hypothetical purchase tasks (e.g., Roma
on simulated demand is needed. et al., 2017), provide exciting new frontiers for
behavioral-economic approaches to societal con-
cerns. Given the unique insights afforded by be-
On Other Variants of Behavioral Economics
havioral economics, behavior analysts can affect
We would like to end by emphasizing that behav- meaningful change that is conceptually systematic
ior analysts should not ignore the very interest- with behavior-analytic principles.
ing findings of more cognitively oriented behav-
ioral economists. Popular books written from this
perspective, such as Predictably Irrational (Ariely, ACKNOWLEDGMENTS
2008), Thinking, Fast and Slow (Kahneman, 2011),
Grant Nos. 5R01HD079113, 5R01HD083214, and
and Nudge: Improving Decisions about Health,
1R01HD093734 from the National Institute of Child
Wealth, and Happiness (Thaler & Sunstein, 2008), Health and Human Development provided partial sup-
contain a wealth of information about the factors port to Brian D. Greer for this work.
that influence choice and tactics for changing
human behavior in meaningful ways. For example,
Thaler and Sunstein introduce the concept of REFERENCES
choice architecture, which involves how we can ar-
range environments to promote choices that are Amlung, M. T., Acker, J., Stojek, M. K., Murphy, J. G.,
beneficial to humans and our cultures. Behavior & MacKillop, J. (2012). Is talk “cheap”?: An initial
analysts have begun to deconstruct many of these investigation of the equivalence of alcohol pur-
chase task performance for hypothetical and actual
principles from more behavioral perspectives, like
rewards. Alcoholism: Clinical and Experimental Re-
sunk costs (e.g., Navarro & Fantino, 2005), fram- search, 36, 716–724.
ing effects (e.g., Naudé, Kaplan, Reed, Henley, Amlung, M., & MacKillop, J. (2012). Consistency of
& DiGennaro Reed, 2018), and prospect theory self-reported alcohol consumption on randomized
(e.g., Rachlin et al., 1991). Despite the fact that and sequential alcohol purchase tasks. Frontiers in
the interpretation of such phenomena may differ Psychiatry, 3, 65.
128 B a s i c P r in c ip l e s a nd C o n c e p t s
Ariely, D. (2008). Predictably irrational: The hidden cially important behavior. Journal of Applied Behavior
forces that shape our decisions. New York: Harper Pe- Analysis, 34, 101–122.
rennial. de Wit, H., Flory, J. D., Acheson, A., McCloskey, M., &
Bauman, R. (1991). An experimental analysis of the Manuck, S. B. (2007). IQ and nonplanning impulsiv-
cost of food in a closed economy. Journal of the Ex- ity are independently associated with delay discount-
perimental Analysis of Behavior, 56, 33–50. ing in middle-aged adults. Personality and Individual
Bickel, W. K., Jarmolowicz, D. P., Mueller, E. T., & Differences, 42, 111–121.
Gatchalian, K. M. (2011). The behavioral econom- DeLeon, I. G., Fisher, W. W., Herman, K. M., & Cro-
ics and neuroeconomics of reinforcer pathologies: sland, K. C. (2000). Assessment of a response bias
Implications for etiology and treatment of addiction. for aggression over functionally equivalent appropri-
Current Psychiatry Reports, 13, 406–415. ate behavior. Journal of Applied Behavior Analysis, 33,
Bickel, W. K., Johnson, M. W., Koffarnus, M. N., MacK- 73–77.
illop, J., & Murphy, J. G. (2014). The behavioral eco- DeLeon, I. G., Frank, M. A., Gregory, M. K., & Allman,
nomics of substance use disorders: Reinforcement pa- M. J. (2009). On the correspondence between pref-
thologies and their repair. Annual Review of Clinical erence assessment outcomes and progressive-ratio
Psychology, 10, 641–677. schedule assessments of stimulus value. Journal of
Bickel, W. K., Odum, A. L., & Madden, G. J. (1999). Im- Applied Behavior Analysis, 42, 729–733.
pulsivity and cigarette smoking: Delay discounting in DeLeon, I. G., & Iwata, B. A. (1996). Evaluation of a
current, never, and ex-smokers. Psychopharmacology, multiple-stimulus presentation format for assessing
146, 447–454. reinforcer preferences. Journal of Applied Behavior
Bickel, W. K., Wilson, A. G., Franck, C. T., Mueller, E. Analysis, 29, 519–533.
T., Jarmolowicz, D. P., Koffarnus, M. N., et al. (2014). DeLeon, I. G., Iwata, B. A., Goh, H., & Worsdell, A. S.
Using crowdsourcing to compare temporal, social (1997). Emergence of reinforcer preference as a func-
temporal, and probability discounting among obese tion of schedule requirements and stimulus similarity.
and non-obese individuals. Appetite, 75, 82–89. Journal of Applied Behavior Analysis, 30, 439–449.
Borrero, J. C., Francisco, M. T., Haberlin, A. T., Ross, N. DeLeon, I. G., Neidert, P. L., Anders, B. M., & Rodri-
A., & Sran, S. K. (2007). A unit price evaluation of guez-Catter, V. (2001). Choices between positive and
severe problem behavior. Journal of Applied Behavior negative reinforcement during treatment for escape-
Analysis, 40, 463–474. maintained behavior. Journal of Applied Behavior
Call, N. A., Trosclair-Lasserre, N. M., Findley, A. J., Analysis, 34, 521–525.
Reavis, A. R., & Shillingsburg, M. A. (2012). Cor- Delmendo, X., Borrero, J. C., Beauchamp, K. L., & Fran-
respondence between single versus daily preference cisco, M. T. (2009). Consumption and response out-
assessment outcomes and reinforcer efficacy under put as a function of unit price: Manipulation of cost
progressive-ratio schedules. Journal of Applied Behav- and benefit components. Journal of Applied Behavior
ior Analysis, 45, 763–777. Analysis, 42, 609–625.
Carr, K. A., Daniel, T. O., Lin, H., & Epstein, L. H. Demurie, E., Roeyers, H., Baeyens, D., & Sonuga-Barke,
(2011). Reinforcement pathology and obesity. Cur- E. (2012). Temporal discounting of monetary rewards
rent Drug Abuse Reviews, 4, 190–196. in children with ADHD and autism spectrum disor-
Carroll, R. A., Kodak, T., & Adolf, K. J. (2016). Effect ders. Developmental Science, 15, 791–800.
of delayed reinforcement on skill acquisition during Diergaarde, L., van Mourik, Y., Pattij, T., Schoffelmeer,
discrete-trial instruction: Implications for treatment- A. N., & De Vries, T. J. (2012). Poor impulse control
integrity errors in academic settings. Journal of Ap- predicts inelastic demand for nicotine but not alco-
plied Behavior Analysis, 49, 176–181. hol in rats. Addiction Biology, 17, 576–587.
Charlton, S. R., & Fantino, E. (2008). Commodity spe- Dixon, M. R., & Cummings, A. (2001). Self-control in
cific rates of temporal discounting: Does metabolic children with autism: Response allocation during
function underlie differences in rates of discounting? delays to reinforcement. Journal of Applied Behavior
Behavioural Processes, 77, 334–342. Analysis, 34, 491–495.
Coffey, S. F., Gudleski, G. D., Saladin, M. E., & Brady, Dixon, M. R., Hayes, L. J., Binder, L. M., Manthey, S.,
K. T. (2003). Impulsivity and rapid discounting of Sigman, C., & Zdanowski, D. M. (1998). Using a
delayed hypothetical rewards in cocaine-dependent self-control training procedure to increase appropri-
individuals. Experimental and Clinical Psychopharma- ate behavior. Journal of Applied Behavior Analysis, 31,
cology, 11, 18–25. 203–210.
Costa, D. F., Carvalho, F. D. M., & Moreira, B. C. D. Dixon, M. R., Rehfeldt, R. A., & Randich, L. (2003).
M. (2019). Behavioral economics and behavioral fi- Enhancing tolerance to delayed reinforcers: The role
nance: A bibliometric analysis of the scientific fields. of intervening activities. Journal of Applied Behavior
Journal of Economic Surveys, 33, 3–24. Analysis, 36, 263–266.
Critchfield, T. S., & Kollins, S. H. (2001). Temporal Epstein, L. H., Jankowiak, N., Fletcher, K. D., Carr, K.
discounting: Basic research and the analysis of so- A., Nederkoorn, C., Raynor, H. A., et al. (2014).
Behavioral Economics 129
Women who are motivated to eat and discount the Green, L., Fry, A. F., & Myerson, J. (1994). Discounting
future are more obese. Obesity, 22, 1394–1399. of delayed rewards: A life-span comparison. Psycho-
Estle, S. J., Green, L., Myerson, J., & Holt, D. D. (2007). logical Science, 5, 33–36.
Discounting of monetary and directly consumable Green, L., Myerson, J., & McFadden, E. (1997). Rate of
rewards. Psychological Science, 18, 58–63. temporal discounting decreases with amount of re-
Few, L. R., Acker, J., Murphy, C., & MacKillop, J. (2012). ward. Memory and Cognition, 25, 715–723.
Temporal stability of a cigarette purchase task. Nico- Hackenberg, T. D. (2009). Token reinforcement: A re-
tine Tobacco Research, 14, 761–765. view and analysis. Journal of the Experimental Analy-
Field, M., Santarcangelo, M., Sumnall, H., Goudie, A., sis of Behavior, 91, 257–286.
& Cole, J. (2006). Delay discounting and the behav- Higgins, S. T., Heil, S. H., Sigmon, S. C., Tidey, J. W.,
ioural economics of cigarette purchases in smokers: Gaalema, D. E., Hughes, J. R., et al. (2017). Addiction
The effects of nicotine deprivation. Psychopharma- potential of cigarettes with reduced nicotine content
cology, 186, 255–263. in populations with psychiatric disorders and other
Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. vulnerabilities to tobacco addiction. JAMA Psychia-
P., Owens, J. C., & Slevin, I. (1992). A comparison of try, 74, 1056–1064.
two approaches for identifying reinforcers for persons Hodos, W. (1961). Progressive ratio as a measure of re-
with severe and profound disabilities. Journal of Ap- ward strength. Science, 134, 943–944.
plied Behavior Analysis, 25, 491–498. Hoffmann, A. N., Samaha, A. L., Bloom, S. E., & Boyle,
Fiske, K. E., Isenhower, R. W., Bamond, M. J., Delmo- M. A. (2017). Preference and reinforcer efficacy of
lino, L., Sloman, K. N., & LaRue, R. H. (2015). As- high- and low-tech items: A comparison of item type
sessing the value of token reinforcement for individu- and duration of access. Journal of Applied Behavior
als with autism. Journal of Applied Behavior Analysis, Analysis, 50, 222–237.
48, 448–453. Holt, D. D., Glodowski, K., Smits-Seemann, R. R., &
Francisco, M. T., Borrero, J. C., & Sy, J. R. (2008). Eval- Tiry, A. M. (2016). The domain effect in delay dis-
uation of absolute and relative reinforcer value using counting: The roles of fungibility and perishability.
progressive-ratio schedules. Journal of Applied Behav- Behavioural Processes, 131, 47–52.
ior Analysis, 41, 189–202. Hursh, S. R., Madden, G. J., Spiga, R., DeLeon, I. G.,
Frank-Crawford, M. A., Castillo, M. I., & DeLeon, I. G. & Francisco, M. T. (2013). The translational utility
(2018). Does preference rank predict substitution for of behavioral economics: The experimental analysis
the reinforcer for problem behavior?: A behavioral of consumption and choice. In G. J. Madden (Ed.)
economic analysis. Journal of Applied Behavior Analy- & W. V. Dube, T. D. Hackenberg, G. P. Hanley, &
sis, 51, 276–282. K. A. Lattal (Assoc. Eds.), APA handbook of behav-
Fuqua, R. W. (1984). Comments on the applied rel- ior analysis: Vol. 2. Translating principles into practice
evance of the matching law. Journal of Applied Be- (pp. 109–128). Washington, DC: American Psycho-
havior Analysis, 17, 381–386. logical Association.
Gilroy, S. P., Kaplan, B. A., & Leader, G. (2018). A Hursh, S. R., & Roma, P. G. (2013). Behavioral econom-
systematic review of applied behavioral economics ics and empirical public policy. Journal of the Experi-
in assessments and treatments for individuals with mental Analysis of Behavior, 99, 98–124.
developmental disabilities. Review Journal of Autism Hursh, S. R., & Roma, P. G. (2016). Behavioral eco-
and Developmental Disorders, 5, 247–259. nomics and the analysis of consumption and choice.
Glover, A. C., Roane, H. S., Kadey, H. J., & Grow, L. L. Managerial and Decision Economics, 37, 224–238.
(2008). Preference for reinforcers under progressive- Hursh, S. R., & Silberberg, A. (2008). Economic de-
and fixed-ratio schedules: A comparison of single mand and essential value. Psychological Review, 115,
and concurrent arrangements. Journal of Applied Be- 186–198.
havior Analysis, 41, 163–176. Jarmolowicz, D. P., & Lattal, K. A. (2010). On distin-
Goldberg, M. C., Allman, M. J., Hagopian, L. P., Triggs, guishing progressively increasing response require-
M. M., Frank-Crawford, M. A., Mostofsky, S. H., et ments for reinforcement. Behavior Analyst, 33, 119–
al. (2017). Examining the reinforcing value of stim- 125.
uli within social and non-social contexts in children Jerome, J., & Sturmey, P. (2008). Reinforcing efficacy
with and without high-functioning autism. Autism, of interactions with preferred and nonpreferred staff
21, 881–895. under progressive-ratio schedules. Journal of Applied
Grace, R. C., Kivell, B. M., & Laugesen, M. (2014). Es- Behavior Analysis, 41, 221–225.
timating cross-price elasticity of e-cigarettes using a Johnson, M. W., & Bickel, W. K. (2002). Within-subject
simulated demand procedure. Nicotine and Tobacco comparison of real and hypothetical money rewards
Research, 17, 592–598. in delay discounting. Journal of the Experimental
Green, L., & Freed, D. E. (1993). The substitutability Analysis of Behavior, 77, 129–146.
of reinforcers. Journal of the Experimental Analysis of Johnson, M. W., & Bickel, W. K. (2003). The behavioral
Behavior, 60, 141–158. economics of cigarette smoking: The concurrent
130 B a s i c P r in c ip l e s a nd C o n c e p t s
presence of a substitute and an independent reinforc- G., Ray, L. A., Tidey, J. W., & Monti, P. M. (2010).
er. Behavioural Pharmacology, 14, 137–144. Behavioral economic analysis of cue-elicited craving
Johnson, M. W., & Bruner, N. R. (2012). The sexual dis- for alcohol. Addiction, 105, 1599–1607.
counting task: HIV risk behavior and the discount- Madden, G. J., & Bickel, W. K. (1999). Abstinence and
ing of delayed sexual rewards in cocaine dependence. price effects on demand for cigarettes: A behavioral-
Drug and Alcohol Dependence, 123, 15–21. economic analysis. Addiction, 94, 577–588.
Kagel, J. H., & Winkler, R. C. (1972). Behavioral eco- Madden, G. J., & Kalman, D. (2010). Effects of bupro-
nomics: Areas of cooperative research between eco- pion on simulated demand for cigarettes and the
nomics and applied behavioral research. Journal of subjective effects of smoking. Nicotine and Tobacco
Applied Behavior Analysis, 5, 335–342. Research, 12, 416–422.
Kahneman, D. (2011). Thinking, fast and slow. New York: Madden, G. J., Petry, N. M., Badger, G. J., & Bickel, W.
Farrar, Straus & Giroux. K. (1997). Impulsive and self-control choices in opi-
Kaplan, B. A., & Reed, D. D. (2018). Happy hour drink oid-dependent patients and non-drug-using control
specials in the alcohol purchase task. Experimental patients: Drug and monetary rewards. Experimental
and Clinical Psychopharmacology, 26, 156–167. and Clinical Psychopharmacology, 5, 256–262.
Kaplan, B. A., Reed, D. D., Murphy, J. G., Henley, A. Martin, A. L., Franklin, A. N., Perlman, J. E., & Bloom-
J., Reed, F. D. D., Roma, P. G., et al. (2017). Time smith, M. A. (2018). Systematic assessment of food
constraints in the alcohol purchase task. Experimen- item preference and reinforcer effectiveness: En-
tal and Clinical Psychopharmacology, 25(3), 186–197. hancements in training laboratory-housed rhesus
Kerwin, M. E., Ahearn, W. H., Eicher, P. S., & Burd, macaques. Behavioural Processes, 157, 445–452.
D. M. (1995). The costs of eating: A behavioral eco- Mazur, J. E. (1987). An adjusting procedure for study-
nomic analysis of food refusal. Journal of Applied Be- ing delayed reinforcement. In M. L. Commons, J. E.
havior Analysis, 28, 245–260. Mazur, J. A. Nevin, & H. Rachlin (Eds.), Quantitative
Kirby, K. N., & Herrnstein, R. J. (1995). Preference re- analysis of behavior: Vol. 5. The effect of delay and of
versals due to myopic discounting of delayed reward. intervening events on reinforcement value (pp. 55–73).
Psychological Science, 6, 83–89. Hillsdale, NJ: Erlbaum.
Kirby, K. N., & Maraković, N. N. (1995). Modeling Mazur, J. E., & Biondi, D. R. (2009). Delay–amount
myopic decisions: Evidence for hyperbolic delay-dis- tradeoffs in choices by pigeons and rats: Hyperbolic
counting within subjects and amounts. Organization- versus exponential discounting. Journal of the Experi-
al Behavior and Human Decision Processes, 64, 22–30. mental Analysis of Behavior, 91, 197–211.
Kodak, T., Lerman, D. C., & Call, N. (2007). Evaluating McClure, E. A., Vandrey, R. G., Johnson, M. W., &
the influence of postsession reinforcement on choice Stitzer, M. L. (2012). Effects of varenicline on absti-
of reinforcers. Journal of Applied Behavior Analysis, nence and smoking reward following a programmed
40, 515–527. lapse. Nicotine and Tobacco Research, 15, 139–148.
Lagorio, C. H., & Madden, G. J. (2005). Delay discount- Naudé, G. P., Kaplan, B. A., Reed, D. D., Henley, A.
ing of real and hypothetical rewards: III. Steady-state J., & DiGennaro Reed, F. D. (2018). Temporal fram-
assessments, forced-choice trials, and all real rewards. ing and the hidden-zero effect: Rate-dependent out-
Behavioural Processes, 69, 173–187. comes on delay discounting. Journal of the Experimen-
Lawyer, S. R., Williams, S. A., Prihodova, T., Rollins, tal Analysis of Behavior, 109, 506–519.
J. D., & Lester, A. C. (2010). Probability and delay Navarro, A. D., & Fantino, E. (2005). The sunk cost
discounting of hypothetical sexual outcomes. Behav- effect in pigeons and humans. Journal of the Experi-
ioural Processes, 84, 687–692. mental Analysis of Behavior, 83, 1–13.
Lea, S. E. G., & Roper, T. J. (1977). Demand for food on Odum, A. L. (2011). Delay discounting: I’m a k, you’re
fixed-ratio schedules as a function of the quality of a k. Journal of the Experimental Analysis of Behavior,
concurrently available reinforcement. Journal of the 96, 427–439.
Experimental Analysis of Behavior, 27, 371–380. Odum, A. L., Baumann, A. A., & Rimington, D. D.
Leon, Y., Borrero, J. C., & DeLeon, I. G. (2016). Para- (2006). Discounting of delayed hypothetical money
metric analysis of delayed primary and conditioned and food: Effects of amount. Behavioural Processes,
reinforcers. Journal of Applied Behavior Analysis, 49, 73, 278–284.
639–655. Odum, A. L., Madden, G. J., & Bickel, W. K. (2002).
Mach, E. (1914). The analysis of sensations and the re- Discounting of delayed health gains and losses by
lation of the physical to the psychical. Chicago: Open current, never- and ex-smokers of cigarettes. Nicotine
Court. and Tobacco Research, 4, 295–303.
MacKillop, J., & Murphy, J. G. (2007). A behavioral Odum, A. L., & Rainaud, C. P. (2003). Discounting of
economic measure of demand for alcohol predicts delayed hypothetical money, alcohol, and food. Be-
brief intervention outcomes. Drug and Alcohol De- havioural Processes, 64, 305–313.
pendence, 89, 227–233. Pace, G. M., Ivancic, M. T., Edwards, G. L., Iwata, B. A.,
MacKillop, J., O’Hagen, S., Lisman, S. A., Murphy, J. & Page, T. J. (1985). Assessment of stimulus prefer-
Behavioral Economics 131
ence and reinforcer value with profoundly retarded Applying the behavioral economics principle of unit
individuals. Journal of Applied Behavior Analysis, 18, price to DRO schedule thinning. Journal of Applied
249–255. Behavior Analysis, 40, 529–534.
Penrod, B., Wallace, M. D., & Dyer, E. J. (2008). Assess- Roane, H. S., Lerman, D. C., & Vorndran, C. M. (2001).
ing potency of high- and low-preference reinforcers Assessing reinforcers under progressive schedule re-
with respect to response rate and response patterns. quirements. Journal of Applied Behavior Analysis, 34,
Journal of Applied Behavior Analysis, 41, 177–188. 145–167.
Perry, A. C., & Fisher, W. W. (2001). Behavioral eco- Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus,
nomic influences on treatments designed to decrease B. A. (1998). Evaluation of a brief stimulus preference
destructive behavior. Journal of Applied Behavior assessment. Journal of Applied Behavior Analysis, 31,
Analysis, 34, 211–215. 605–620.
Petry, N. M. (2001). Delay discounting of money and Roma, P. G., Reed, D. D., DiGennaro Reed, F. D., &
alcohol in actively using alcoholics, currently absti- Hursh, S. R. (2017). Progress of and prospects for hy-
nent alcoholics, and controls. Psychopharmacology, pothetical purchase task questionnaires in consumer
154, 243–250. behavior analysis and public policy. Behavior Analyst,
Petry, N. M., & Bickel, W. K. (1998). Polydrug abuse in 40, 329–342.
heroin addicts: A behavioral economic analysis. Ad- Rosch, K. S., & Mostofsky, S. H. (2016). Increased delay
diction, 93, 321–335. discounting on a novel real-time task among girls,
Pope, D. A., Poe, L., Stein, J. S., Kaplan, B. A., Heck- but not boys, with ADHD. Journal of the International
man, B. W., Epstein, L. H., et al. (2019). Experimen- Neuropsychological Society, 22, 12–23.
tal tobacco marketplace: Substitutability of e-cig- Russell, D., Ingvarsson, E. T., Haggar, J. L., & Jessel, J.
arette liquid for cigarettes as a function of nicotine (2018). Using progressive ratio schedules to evaluate
strength. Tobacco Control, 28, 206–211. tokens as generalized conditioned reinforcers. Journal
Rachlin, H., Raineri, A., & Cross, D. (1991). Subjec- of Applied Behavior Analysis, 51, 40–52.
tive probability and delay. Journal of the Experimental Salvy, S. J., Nitecki, L. A., & Epstein, L. H. (2009). Do
Analysis of Behavior, 55, 233–244. social activities substitute for food in youth? Annals
Reed, D. D., Kaplan, B. A., & Becirevic, A. (2015). Basic of Behavioral Medicine, 38, 205–212.
research on the behavioral economics of reinforcer Scheres, A., Dijkstra, M., Ainslie, E., Balkan, J., Reyn-
value. In F. D. DiGennaro Reed & D. D. Reed (Eds.), olds, B., Sonuga-Barke, E., et al. (2006). Temporal
Autism service delivery: Bridging the gap between sci- and probabilistic discounting of rewards in children
ence and practice (pp. 279–306). New York: Springer. and adolescents: Effects of age and ADHD symp-
Reed, D. D., Kaplan, B. A., Roma, P. G., & Hursh, S. toms. Neuropsychologia, 44, 2092–2103.
R. (2014). Inter-method reliability of progression Scheres, A., Tontsch, C., & Thoeny, A. L. (2013). Steep
sizes in a hypothetical purchase task: Implications temporal reward discounting in ADHD-combined
for empirical public policy. Psychological Record, 64, type: Acting upon feelings. Psychiatry Research, 209,
671–679. 207–213.
Reed, D. D., Luiselli, J. K., Magnuson, J. D., Fillers, S., Scheres, A., Tontsch, C., Thoeny, A. L., & Kaczkurkin,
Vieira, S., & Rue, H. C. (2009). A comparison be- A. (2010). Temporal reward discounting in atten-
tween traditional economical and demand curve tion-deficit/hyperactivity disorder: The contribution
analyses of relative reinforcer efficacy in the valida- of symptom domains, reward magnitude, and session
tion of preference assessment predictions. Develop- length. Biological Psychiatry, 67, 641–648.
mental Neurorehabilitation, 12, 164–169. Shahan, T. A., Odum, A. L., & Bickel, W. K. (2000).
Reed, D. D., & Martens, B. K. (2011). Temporal dis- Nicotine gum as a substitute for cigarettes: A behav-
counting predicts student responsiveness to ex- ioral economic analysis. Behavioural Pharmacology,
change delays in a classroom token system. Journal of 11, 71–79.
Applied Behavior Analysis, 44, 1–18. Shore, B. A., Iwata, B. A., DeLeon, I. G., Kahng, S., &
Reed, D. D., Niileksela, C. R., & Kaplan, B. A. (2013). Smith, R. G. (1997). An analysis of reinforcer substi-
Behavioral economics. Behavior Analysis in Practice, tutability using object manipulation and self-injury
6, 34–54. as competing responses. Journal of Applied Behavior
Reynolds, B., Richards, J. B., Horn, K., & Karraker, K. Analysis, 30, 21–40.
(2004). Delay discounting and probability discount- Smethells, J. R., Harris, A. C., Burroughs, D., Hursh, S.
ing as related to cigarette smoking status in adults. R., & LeSage, M. G. (2018). Substitutability of nico-
Behavioural Processes, 65, 35–42. tine alone and an electronic cigarette liquid using a
Roane, H. S., Call, N. A., & Falcomata, T. S. (2005). concurrent choice assay in rats: A behavioral eco-
A preliminary analysis of adaptive responding under nomic analysis. Drug and Alcohol Dependence, 185,
open and closed economies. Journal of Applied Behav- 58–66.
ior Analysis, 38, 335–348. Sonuga-Barke, E. J. S., Taylor, E., Sembi, S., & Smith,
Roane, H. S., Falcomata, T. S., & Fisher, W. W. (2007). J. (1992). Hyperactivity and delay aversion: 1. The
132 B a s i c P r in c ip l e s a nd C o n c e p t s
effect of delay on choice. Journal of Child Psychology magnitude: An evaluation of preference and rein-
and Psychiatry, 33, 387–398. forcer efficacy. Journal of Applied Behavior Analysis,
Staubitz, J. L., Lloyd, B. P., & Reed, D. D. (2018). A sum- 41, 203–220.
mary of methods for measuring delay discounting in Tustin, D. (2000). Revealed preference between rein-
young children. Psychological Record, 68, 239–253. forcers used to examine hypotheses about behavioral
Sy, J. R., & Vollmer, T. R. (2012). Discrimination ac- consistencies. Behavior Modification, 24, 411–424.
quisition in children with developmental disabilities Tustin, R. D. (1994). Preference for reinforcers under
under immediate and delayed reinforcement. Journal varying schedule arrangements: A behavioral eco-
of Applied Behavior Analysis, 45, 667–684. nomic analysis. Journal of Applied Behavior Analysis,
Temple, J. L., Legierski, C. M., Giacomelli, A. M., Salvy, 27, 597–606.
S. J., & Epstein, L. H. (2008). Overweight children Tversky, A., & Thaler, R. H. (1990). Anomalies: Pref-
find food more reinforcing and consume more energy erence reversals. Journal of Economic Perspectives, 4,
than do nonoverweight children. American Journal of 201–211.
Clinical Nutrition, 87, 1121–1127. Weller, R. E., Cook, E. W., III, Avsar, K. B., & Cox, J. E.
Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improv- (2008). Obese women show greater delay discounting
ing decisions about health, wealth, and happiness. New than healthy-weight women. Appetite, 51, 563–569.
Haven, CT: Yale University Press. Wilson, A. N., & Gratz, O. H. (2016). Using a progres-
Tiger, J. H., Toussaint, K. A., & Roath, C. T. (2010). An sive ratio schedule of reinforcement as an assessment
evaluation of the value of choice-making opportuni- tool to inform treatment. Behavior Analysis in Prac-
ties in single-operant arrangements: Simple fixed- tice, 9, 257–260.
and progressive-ratio schedules. Journal of Applied Zhou, L., Goff, G. A., & Iwata, B. A. (2000). Effects
Behavior Analysis, 43, 519–524. of increased response effort on self-injury and object
Trosclair-Lasserre, N. M., Lerman, D. C., Call, N. A., manipulation as competing responses. Journal of Ap-
Addison, L. R., & Kodak, T. (2008). Reinforcement plied Behavior Analysis, 33, 29–40.
PAR T III
MEASUREMENT, DESIGN,
AND METHODOLOGICAL ISSUES
133
CHAP TER 8
A hallmark of applied behavior analysis is the pre- to ineffective action. Validity of a measurement
cise measurement of observable behavior (Baer, system refers to the extent to which the system
Wolf, & Risley, 1968). Measurement precision measures what it purports to measure. In general,
involves the reliable and accurate quantification direct observation of the target behavior is likely
of some dimension of the target behavior. The to yield highly valid measures. However, validity
response dimensions, which provide the basis for can suffer if (1) the behavior analyst uses indirect
systems of measurement, encompass specific char- measures, (2) the behavior analyst infers the tar-
acteristics of behavior such as frequency, duration, get behavior from other events, or (3) operational
and latency. An observation system consists of definitions specify irrelevant aspects of behavior.
formalized rules for extracting information from For example, measuring headache frequency by
a behavior stream. These rules specify the target counting how often the participant complains of
behavior; the way the behavior analyst samples headaches may not be a valid measure, because
the events; the dimensions of the events that the the participant may not always complain when a
behavior analyst assesses; the way the behavior headache is present and may sometimes complain
analyst records the data; and other pragmatic is- when no headache is present. Finally, we define
sues, such as observational setting, observers, and reliability as the extent to which a measurement
cost (Hartmann & Wood, 1982). system yields consistent outcomes. A measure-
Three characteristics of a measurement sys- ment is reliable if it produces the same outcome
tem—accuracy, validity, and reliability—gauge when we apply it to the same behavior, or when
its predictive value (Poling, Methot, & LeSage, different observers produce the same measurement
1995). A measurement system is accurate if the outcomes while independently scoring the same
yielded values reflect the true values of the target behavior (Johnston & Pennypacker, 1993; Poling
behavioral dimension. Any discrepancy between et al., 1995).
the obtained values and the true values constitutes Given that multiple factors (e.g., expectation
measurement error. Implications of this relation are bias, effects of agreement checking, adequacy of
considerable, because observational systems that behavioral definitions, complexity of coding sys-
are high in measurement error produce data that tems) can affect the measurement of behavior, one
fail to reflect true behavioral events and may lead of the primary challenges for behavior analysts
135
136 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
is to obtain a record of an individual’s behavior hands. Topographical definitions describe the form
that is both complete and accurate (Johnston & of behavior and specify the physical properties of
Pennypacker, 1993). The purpose of this chapter responding. For instance, we might define door
is to provide descriptions of observation and mea- opening as using one’s hands to turn the doorknob
surement systems and rationales for their use. while standing upright and facing forward (John-
ston & Pennypacker, 1993).
Topographical definitions are often important
OPERATIONALLY DEFINING BEHAVIOR in applied behavior analysis, because socially ap-
propriate and inappropriate behavior may be func-
General Characteristics of Behavioral Definitions
tionally identical but topographically distinct. To
Behavioral definitions or observational codes are demonstrate, aggressive behavior and appropriate
descriptions of target behaviors that should control verbal requests may each produce increased con-
observers’ scoring behavior. Generally, we should trol over the social environment, including the
judge the adequacy of behavioral definitions by behavior of others and access to preferred items
how consistently and accurately they control ob- and activities (e.g., Bowman, Fisher, Thompson, &
serving behavior and whether the resulting data Piazza, 1997; Torres-Viso, Strohmeier, & Zarcone,
allow the behavior analyst to take effective action 2018). Topographical definitions that differenti-
(i.e., gain control of socially meaningful behavior). ate these two response categories are important,
Attention to the following three aspects can help because appropriate requesting is more socially ap-
to increase the adequacy of behavioral definitions: propriate than aggression. Nevertheless, function-
(1) objectiveness, or whether definitions refer to ob- al definitions are often useful in applied behavior
servable events with which two or more observers analysis. For example, researchers might place be-
can agree (i.e., behavioral definitions should refer havior that produces property destruction, regard-
to observable behavior such as hitting, rather than less of topography, in a single definitional category
abstract concepts such as anger); (2) clarity, or the (e.g., Fisher, Lindauer, Alterson, & Thompson,
ability of observers to read and paraphrase the 1998). Behaviors that are different topographical-
definition accurately; and (3) completeness, or the ly, such as crying and throwing toys, may produce
inclusion of relevant aspects and exclusion of ir- the same consequence, such as parent attention
relevant aspects in the definition (Cooper, Heron, (e.g., Fritz, Iwata, Hammond, & Bloom, 2013). We
& Heward, 2020). can group these behaviors into the same response
Additionally, definitions might include a de- class when we define them by function.
scription of the conditions under which the target
behavior occurs. For example, question answering
can occur only when an individual asks a question, HUMANS AND MACHINES AS OBSERVERS
and hitting another individual as an instance of ag-
Humans as Observers
gression can occur only when another individual is
in physical proximity. Finally, we can enhance the In applied behavior analysis, we commonly use
effectiveness of behavioral definitions by includ- human observers to collect data. Human observers
ing relevant examples and nonexamples of target can collect data in many ways, including marking
behavior, potentially improving the control over data sheets with a pencil (e.g., Taravella, Lerman,
observers’ scoring behavior (Cooper et al., 2020). Contrucci, & Roane, 2000), operating a stop-
watch (e.g., Hoch, McComas, Johnson, Faranda,
& Guenther, 2002), or pressing keys on a com-
Topographical versus Functional Definitions
puter keyboard (e.g., Fuhrman, Fisher, & Greer,
We may define behavior according to either func- 2016). Humans as opposed to machines as data
tion or topography. Functional definitions describe collectors have the advantage of greater flexibil-
the effects of responding on some aspect of the en- ity, because we can train humans to take data on
vironment. For example, a behavior analyst might many responses, and humans can adapt to novel
define opening a door solely by the result of door- environments relatively easily (Page & Iwata,
opening behavior (e.g., the distance between the 1986). However, human observers are prone to the
door and the door frame is at least 2 meters). The influence of numerous errors and biases that can
movements that bring this result about are not rel- produce unreliable and inaccurate data collection.
evant with functional definitions; an individual We describe the most common potential biases in
could open the door with his or her toes, teeth, or the following sections.
Defining and Measuring Behavior 137
behavior samples on videos (e.g., role-played ses- servers scored in a session (Poling et al., 1995).
sions) can be helpful, because the trainer can eval- The disadvantage of this procedure is that it does
uate observers’ accuracy by comparing their scores not determine whether the two observers scored
to predetermined values. The trainer then can the same occurrences of behavior (Bijou, Peterson,
calibrate observers’ scoring behavior by conduct- & Ault, 1968). We do not recommend this index if
ing training to improve accuracy, if needed (John- agreement about specific instances of behavior (as
ston & Pennypacker, 1993; Mudford et al., 2011). opposed to overall occurrence) is important. How-
Computer technology may also facilitate observer ever, total reliability can be useful and sufficient to
training. For example, Bass (1987) described a estimate observer accuracy if the primary focus is
computerized system in which observers watched overall occurrence rather than the distribution of
videotapes and received automatic and immediate responding in a session (Rolider, Iwata, & Bullock,
feedback on their scoring. 2012).
Interobserver Agreement Interval Agreement
Applied behavior analysts expect that researchers More stringent indices compare scoring in speci-
will report interobserver agreement whenever they fied, relatively short intervals (e.g., 5–10 seconds)
use human observers (Page & Iwata, 1986). To ob- to better evaluate whether the observers are scor-
tain interobserver agreement scores, two persons, ing the same behavior at the same time (Bijou
a primary and a secondary observer, indepen- et al., 1968; Page & Iwata, 1986). We calculate
dently but simultaneously score the same behavior interval agreement by assessing agreement or dis-
episode and calculate agreement between the two agreement for each interval or trial and dividing
records. A common recommendation is to obtain the number of agreements by the total number of
interobserver agreement for a minimum propor- intervals or trials (e.g., Cariveau & Kodak, 2017).
tion of sessions (e.g., 25–30%) across conditions or This approach is perhaps the most common in-
phases (Bailey & Burch, 2002; Poling et al., 1995). terobserver agreement calculation method in ap-
High interobserver agreement scores (e.g., above plied behavior analysis (e.g., Hanley, Iwata, Lind-
80–90%) do not in and of themselves indicate the berg, & Conners, 2003; Lannie & Martens, 2004;
accuracy of a measurement system, because both Thiemann & Goldstein, 2001). We can use it with
observers could be inaccurate but still agree (John- any collection method involving interval data
ston & Pennypacker, 1993). However, low interob- (e.g., partial- and whole-interval recording or mo-
server agreement scores indicate the need for ad- mentary-time sampling; see below) or trial-by-trial
justments (e.g., additional observer training, more data (i.e., correct vs. incorrect responses, compli-
specific definitions) before observers can produce ance with teacher instructions). Each interval or
useful data (Poling et al., 1995). Therefore, ap- trial typically generates one score—the behavior
plied behavior analysts should assess interobserver either occurred or did not—except in the case of
agreement scores early in the process. Waiting to block-by-block agreement (see below).
calculate interobserver agreement after many ses- Exact agreement is a variant of the interval ap-
sions or after the study, if observers score sessions proach. Intervals in which observers scored the
from videos, is risky. If researchers discover serious same number of occurrences are exact agreements
flaws in the measurement system after the conclu- (e.g., Hagopian, Contrucci Kuhn, Long, & Rush,
sion of the manipulation, the only option may be 2005). The formula for calculating the exact-
to discard the data and start over. Thus researchers agreement coefficient is number of intervals with
should use interobserver agreement as a proactive exact agreement divided by the total number of
way to improve the measurement system and as a intervals. Although precise, this method may be
retroactive method to evaluate its adequacy. The too stringent for many practical purposes (Page &
following are several common methods of calcu- Iwata, 1986; Rolider et al., 2012).
lating interobserver agreement. Block-by-block agreement is a slightly less con-
servative but frequently sufficient method. We cal-
culate total agreement for each interval, calculate
Total Agreement
the mean of the quotients across the whole session,
The simplest index of interobserver agreement is and convert the ratio to a percentage (Bailey &
total agreement, calculated by dividing the smaller Burch, 2002; Page & Iwata, 1986; North & Iwata,
by the larger number of occurrences the two ob- 2005). Exact and block-by-block agreements are
Defining and Measuring Behavior 139
only applicable when observers can score multiple tages over paper-and-pencil methods, including
occurrences of behavior in each interval, such as increased efficiency of data analysis and graph-
discrete target behaviors. ing, and reduced risk of compromised reliability
A less commonly applied variant that is appli- and accuracy due to ambiguity of written records
cable to free-operant data is time-window analy- (Dixon, 2003; Jackson & Dixon, 2007; Kahng &
sis (Mudford, Martin, Hui, & Taylor, 2009). This Iwata, 1998; Whiting & Dixon, 2013).
approach requires time-stamped data streams for In addition, computerized systems that facilitate
which we can compare the data files of two ob- the coding of video and audio media are available.
servers second by second. An example of an agree- Such systems allow for the measurement of many
ment for frequency data is when both files contain simultaneously occurring events, precise detection
a scored response in a 1-second window, and an of response duration, and automatic data analysis
example of an agreement for duration data is if and interobserver agreement calculations (e.g.,
the 1-second window in both data files contains Tapp, 2003; Tapp & Walden, 1993). This approach
an ongoing occurrence of the behavior. We can suffers from the limitation that the observer may
modify the stringency of the time window analysis need to view each behavioral episode multiple
by including a tolerance interval. For instance, if we times; therefore, the time expenditure may in-
use a tolerance interval of ±2, we would count an crease considerably. Behavior analysts should use
agreement if the secondary observer scored a re- these kinds of computer programs only if they pro-
sponse in a window of 2 seconds before and 2 sec- vide sufficiently important improvements in data
onds after a response the primary observer scored. collection considering the target behavior and the
Researchers have not studied time-window analy- ultimate goals of data collection.
sis extensively. The available data suggest that it
may inflate accuracy estimates for high-rate and
Videos and Other Visual Media
low-duration behavior, but provide relatively accu-
rate estimates at lower rates and durations (Mud- We may enhance human data collection by vid-
ford et al., 2009). eo-recording sessions for later viewing. Video-re-
corded sessions enable the observer to score many
simultaneously occurring responses. Such sessions
Occurrence and Nonoccurrence Agreement
also are useful for observer training (Miltenberger,
Low-rate target behavior produces data streams Rapp, & Long, 1999). Using videos allows the ob-
with many intervals in which the behavior does server to score duration measures more precisely
not occur and few intervals in which the be- than live recording (Miltenberger et al., 1999;
havior occurs. In this case, the behavior analyst Tapp & Walden, 1993). Research has suggested
should assess occurrence agreement by examin- that accuracy and reliability generally are not af-
ing the intervals in which both observers agreed fected adversely by scoring sessions from videos,
on the occurrence of the behavior (e.g., Lerman although low-quality audio recording may affect
et al., 2005). The behavior analyst should assess scoring of vocal responses (Kent, O’Leary, Dietz,
nonoccurrence agreement for high-rate behavior by & Diament, 1979). However, use of video record-
examining the intervals in which both observers ings for data collection is time-consuming; it can
agreed on the nonoccurrence of the behavior. Oc- create a backlog of media to be scored; and valu-
currence and nonoccurrence agreement indices able data may be lost if recorded sessions are lost
eliminate the risk of inflating interobserver agree- or damaged before they can be viewed. Overall,
ment due to disproportionate numbers of intervals behavior analysts should weigh these costs against
in which the target behavior either did or did not potential benefits when deciding whether to use
occur (Bijou et al., 1968; Page & Iwata, 1986). live recording or to score behavior with the assis-
tance of visual media.
Computerized Data Collection Systems
Mechanical Recording
The use of laptops and handheld computerized
devices (e.g., tablets, smartphones) for data collec- Much laboratory research in behavior analysis in-
tion is common in applied behavior analysis (e.g., volves measures of target behaviors via mechanical
Slocum, Grauerholz, Peters, & Vollmer, 2018). devices (e.g., automatic recording of key strokes or
Computerized data collection systems operated lever presses). Machines are free of many human
by human observers have several potential advan- shortcomings (e.g., expectancy bias, observer drift,
140 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
illness) and may be less costly (Panos & Freed, ting. Direct observation maximizes the validity of
2007). Therefore, behavior analysts may prefer measurement, because it decreases the discrepan-
machines over human observers whenever the tar- cy between actual behavioral events and sampled
get behavior lends itself to mechanical recording behavior (Hersen & Barlow, 1976). Direct obser-
(Johnston & Pennypacker, 1993). However, many vation is preferred whenever possible in behav-
behaviors of interest to applied behavior analysts ior analysis; nonetheless, situations may arise in
are not amenable to mechanical recording with which indirect measures are necessary. For exam-
current technology. Nevertheless, behavior ana- ple, the target behavior may occur covertly, so that
lysts are using mechanical recording via computer direct observation is difficult or impossible (e.g.,
software increasingly in certain areas, especially Grace, Thompson, & Fisher, 1996). Furthermore,
those that use computer-based instruction. For behavior analysts may want data on past events, in
example, Connell and Witt (2004) used computer- which case the only recourse may be to examine
based instruction to teach relations among printed archival information or collect informant reports
uppercase and lowercase letters, letter names, and via interviews, surveys, or questionnaires.
letter sounds to kindergartners. The computer
software automatically collected data during the
Indirect Measurement
study, obviating the need for human observers.
Mechanical recording can be useful particularly Indirect measurement involves drawing infer-
for dependent variables that may be difficult for ences about the behavior of interest from other
human observers to quantify objectively. For ex- behaviors or environmental events. The validity
ample, Wilson, Iwata, and Bloom (2012) used a and accuracy of measurement may suffer because
computerized measurement of wound-surface area we do not observe the target behavior directly, but
to measure the severity of injuries produced by self- this approach may be of value to behavior analysts
injurious behavior. Sigurdsson et al. (2011) used when direct observation is impossible or impracti-
an automated system to measure noise levels in a cal.
therapeutic workplace. In these examples, human
observers would have had to rely on subjective es-
Permanent‑Product Recording
timates of injury severity or ambient-sound levels
in the absence of mechanical measures. Measuring outcomes of behavior, such as adminis-
Mechanical data collection systems have the trative records, completed assignments, or manu-
advantage of completely removing human judg- factured goods, may supplement or replace direct
ment from the data collection process, thereby measurement (Kazdin, 1979). Permanent-product
eliminating human sources of error and potential- recording involves recording the tangible outcome
ly increasing the reliability and accuracy of data of a behavior rather than the behavior itself. In
collection. We should not, however, automatically some instances, the products may be of primary
trust data gathered through these means. Ma- interest. For example, we might score completed
chines can malfunction, and human designers and homework for accuracy (Kelly, 1976; Miller & Kel-
programmers can make mistakes that render the ley, 1994). In other cases, permanent products are
systems inaccurate or unreliable. Behavior analysts not of primary interest, but are important indica-
should conduct calibration tests when using a new tors of target behavior that is not amenable to di-
data collection system, in which they check output rect observation. A study by Grace et al. (1996)
against input of known values. Replication of find- illustrated the latter approach. They conducted
ings across different laboratories and with different physical examinations to identify signs of self-in-
data collection systems is also important. jury (e.g., bruises, scratches, swelling, tissue dam-
age). The researchers chose this approach because
the participant’s self-injury was covert, and observ-
INDIRECT VERSUS DIRECT MEASUREMENT ers were unable to observe the behavior directly.
Maglieri, DeLeon, Rodriguez-Catter, and Sevin
Direct observation requires measuring the target (2000) provided another example: The research-
behavior without inferring its occurrence from ers implemented permanent-product recording to
other events (e.g., products of behavior, verbal evaluate covert food stealing. Observers recorded
recollections), usually by observing the behavior the amount of food presented at the start of a ses-
as it occurs in either a natural or contrived set- sion, left the participant alone in the room, and
Defining and Measuring Behavior 141
returned later to record the amount of remaining convenient and require little effort, are likely to
food. help in this matter. One valuable training strategy
The disadvantage of permanent-product re- is to place contingencies initially on accurate re-
cording lies in its very nature: It measures out- porting rather than on clinically desirable changes
comes rather than the behavior itself. This may in behavior, so that the behavior analyst rewards
be problematic because the outcome may be due data collection that agrees with independent
to something other than the target behavior, re- sources, regardless of other outcomes (Hartmann,
sulting in inaccurate data. For example, measuring 1984; Nelson, 1977). Second, observer reactivity
performance through examination of worksheets (i.e., the effects of observation on the target be-
will produce inaccurate measures if someone other havior) is of special concern in self-monitoring.
than the participant has completed the work (e.g., The behavior analyst should attempt to evaluate
a parent has done the homework for the child). In the effects of intervention via self-monitoring and
addition, the target behavior may not produce the via less intrusive data collection methods, to en-
measured outcome (e.g., self-injury may not pro- sure that the effects of the independent variable
duce tissue damage; Grace et al., 1996). Therefore, were not due exclusively or primarily to reactiv-
the behavior analyst should exercise caution when ity (Cooper et al., 2020). Although the reactivity
implementing product recording. Noting alterna- inherent in self-monitoring is a threat to effective
tive events that may produce the outcome of inter- measurement, the behavior analyst can use it as
est may enhance the accuracy of data collection. a treatment protocol. The very act of systemati-
For example, if a child falls during outdoor play- cally keeping track of one’s own behavior can be
time, the resulting bruises and scratches could be an effective tool for self-management and behavior
difficult to distinguish from injuries produced by change (Cooper et al., 2020; Critchfield, 1999).
self-injurious behavior.
Interviews, Surveys, and Rating Scales
Self‑Monitoring
The behavior analyst can use structured inter-
Self-monitoring or self-recording data is like direct views, surveys, or rating scales to collect important
observation, in that the observer records events of information about various aspects of the target
interest as they occur. However, the difference is behavior, such as topography and environmental
that the client or participant serves as his or her correlates (Beaver & Busse, 2000; Merrell, 2000).
own observer. Self-monitoring is particularly use- Depending on the situation, the behavior analyst
ful in the measurement of private events that only might gather this information from the client or
the client can access directly, such as headaches from significant others in the client’s environment
and thinking (Hartmann, 1984; Nelson, 1977). (e.g., parents, teachers). Indirect measures of this
Behavior analysts can use this approach with be- sort can generate information that may be help-
havior that is overt and observable by more than ful in the initial stages of assessment (McComas
one individual, but is not amenable to direct & Mace, 2000; Nay, 1979). Researchers have used
measurement due to practical constraints. For ex- behavioral interviews, surveys, and rating scales
ample, adult participants in VanWormer (2004) to generate hypotheses about potential behav-
carried pedometers, which counted the number ioral function (e.g., Applegate, Matson, & Cherry,
of steps each participant took, and weighed them- 1999) and to identify potential reinforcers to use
selves regularly to produce measures of physical ac- in subsequent treatments (Fisher, Piazza, Bowman,
tivity and weight loss. Family members and friends & Amari, 1996; Northup, 2000). Researchers have
served as secondary observers, and the partici- then used this information to determine what
pants periodically e-mailed data to the researcher. stimuli to include in subsequent preference or re-
Self-monitoring produces data that otherwise inforcer assessments, and which environmental
would be unobtainable. Nevertheless, there are conditions to include in functional assessments.
at least two major shortcomings of this approach. Importantly, behavior analysts should confirm
First, self-recorded data may be less accurate than hypotheses these methods generate through more
directly observed data, and the accuracy and re- stringent observation methods, preferably involv-
liability of these data are more difficult to verify. ing direct observation, because informant reports
Directly training clients to observe their own be- often do not correspond well with the behavior of
havior, and using measurement systems that are interest. Direct observation is the gold standard
142 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
against which to evaluate the accuracy of indirect Frequency. Frequency recording, also called the
measures. tally method, trial scoring, and event recording, en-
tails continuous recording of separate instances of
behavior as they occur (Hartmann & Wood, 1982;
Direct Observation
Kazdin, 2001; Repp, Roberts, Slack, Repp, & Berk-
Direct observation is more consistent with be- ler, 1976). Frequency recording is most appropri-
haviorism’s epistemological emphasis on overt be- ate when the number of times a behavior occurs is
havior, public events, quantification, low levels of the response dimension of interest (Hartmann &
inference, and the importance of environmental Wood, 1982; Schinke & Wong, 1977). However,
contingencies (Hartmann & Wood, 1982). Ad- the behavior analyst cannot compare frequencies
ditional advantages of direct observation include obtained in different observation sessions mean-
an increase in predictive validity as the discrep- ingfully if observation times are unequal. In those
ancy between sampled behavior and predicted instances, the behavior analyst should convert fre-
behavior decreases, and a close relation between quencies to response rate, defined as the frequency
the target behavior and intervention (Hersen & of behavior per some unit of time (e.g., minute,
Barlow, 1976). Behavior analysts prefer direct hour, day; Kazdin, 2001). When behavior is op-
measures to study behavior because of these ad- portunity-bound or trial-based, so that the target
vantages, and use them widely in numerous con- responses cannot occur in the absence of speci-
texts (e.g., Alevizos, DeRisi, Liberman, Eckman, fied stimuli (e.g., instructions, stimulus cards), the
& Callahan, 1978; Bijou, Peterson, Harris, Allen, behavior analyst should convert obtained frequen-
& Johnston, 1969; Cordes, 1994; Doll & Elliott, cies to percentages or report the number of re-
1994; Gresham, Gansle, & Noell, 1993; Odom & sponses relative to the number of opportunities for
Ogawa, 1992; Prinz, 1982; Test, Rose, & Corum, the response to occur. For example, Gutowski and
1990; Wasik & Loven, 1980). Stromer (2003) measured selection of comparison
Direct measures can be continuous or discontinu- stimuli that occurred in the presence of specific
ous. Continuous measures record all instances of sample stimuli, and then converted the frequen-
behavior. Discontinuous measures, also termed cies into percentages of correct selections.
intermittent measures, sample from all possible oc- Frequency recording is best suited for measur-
currences of behavior (Johnston & Pennypacker, ing responses that are relatively discrete and have
1993). Continuous measures—frequency, duration, a constant duration. Examples include hitting,
latency, and intensity—provide the most complete throwing objects, pressing buttons or keys, and
record of behavior, but may prove expensive and placing items in containers. In addition, responses
otherwise impractical. Therefore, intermittent or that tend to have variable durations can be ame-
discontinuous methods, such as interval recording nable to frequency recording if the behavior ana-
and momentary time sampling, are often valuable. lyst precisely defines their onset and offset. How-
However, discontinuous recording produces an in- ever, the behavior analyst should consider using
complete behavior record and consequently raises duration instead of frequency recording given
several questions about the representativeness of responses with varying durations. Frequency re-
the data. As such, discontinuous observational cording can generate much information about be-
procedures that appropriately sample from all pos- havior (Repp et al., 1976), and observers can use
sible occurrences of behavior to yield representa- it easily if responding is not excessive. Frequency
tive and accurate measures are important. measures reflect the amount of responding, which
is important when the goal is to increase or de-
crease the number of times the target behavior oc-
Continuous Recording Procedures
curs (Kazdin, 2001).
Continuous behavior recording is the most rigor-
ous and powerful measurement procedure avail- Duration. Duration involves measuring the en-
able to behavior analysts. This approach involves tire time that the participant performs or engages
recording behavior based on its occurrence in an in the target response (Kazdin, 2001). Duration re-
uninterrupted, natural time flow (Sanson-Fisher, cording is appropriate when temporal characteris-
Poole, Small, & Fleming, 1979; Hartmann & tics of a behavior are of interest, and when behav-
Wood, 1982). We discuss the application of con- iors vary in the length of time the client engages in
tinuous recording to relevant behavioral dimen- them. This approach is best suited to continuous,
sions in the following paragraphs. ongoing behaviors rather than to discrete, short-
Defining and Measuring Behavior 143
duration behaviors (Hartmann & Wood, 1982). sponse topographies, or if observers must measure
Observers can measure duration by turning on a many individuals’ behavior simultaneously. Under
stopwatch or a timer at the onset of behavior and these circumstances, the behavior analyst may use
turning it off at the offset of behavior. Most com- appropriate time-sampling techniques (e.g., mo-
puterized data collection systems allow observers mentary time sampling; Kazdin, 2001; see below).
to press designated keys at the onset and offset of
behavior, and many accommodate multiple, simul- Latency. Latency refers to the amount of time
taneous duration measures. The behavior analyst that elapses between the onset of a specific cue or
can calculate the proportion of session time dur- stimulus and the target behavior (Kazdin, 2001).
ing which a participant performed the behavior by Like duration, latency can be recorded with the
dividing the total duration of behavior by the total assistance of timing devices or computerized sys-
observation time. tems that measure duration. Latency recording is
Altmann (1974) suggested that duration re- appropriate when we are interested in the relation
cording is appropriate for behavioral states versus between a certain event and the initiation of a spe-
behavioral events. Behavioral states are those be- cific response. This includes situations in which
haviors that have appreciable duration. Behavioral the goal is either to reduce or to increase time be-
events are instantaneous occurrences of behavior. tween antecedent events and target behavior. For
Duration recording is useful in measuring time example, behavior analysts are frequently inter-
allocation, defined as the amount of time spent ested in reducing the latency between instruction
in certain environmental areas (e.g., sitting in a delivery and compliance. Wehby and Hollahan
chair) or engaged in certain response alternatives (2000) measured the seconds that passed between
(e.g., doing homework, practicing a musical instru- the delivery of instructions to engage in academic
ment; Kazdin, 2001). activities and compliance with the instructions.
As with other continuous measurement proce- They used these measures to evaluate the effec-
dures, duration recording is appealing because it tiveness of providing a series of high-probability
produces a complete measurement of the response requests (i.e., requests that the participants were
dimension of interest. Like frequency, duration is likely to engage in immediately) before the deliv-
comprehensible, is socially acceptable, and can be ery of academic instructions. Latency measures
accomplished without complicated observational have recently been extended to the assessment of
technology. The reliability of duration measures, challenging behavior (e.g., elopement), which may
however, may be low if the precise onset and off- be beneficial when the target behavior is not con-
set of behavior are difficult to identify reliably. tinuous or is specific to a particular evocative event
For instance, observers measuring speech content (Hansen et al., 2019; Traub & Vollmer, 2019).
may have difficulty determining the precise point Like frequency and duration recording, latency
at which changes in content (e.g., topics) occur, measures consist of complete measurement of
and the reliability of duration measures can suffer. the target behavior, and behavior analysts prefer
However, observers might agree more easily about them over discontinuous measures unless practical
whether a conversation topic occurred in an in- constraints dictate otherwise. In addition, latency
terval. Thus partial-interval recording (see below) recording is simple, is relatively straightforward,
may be easier to implement (e.g., Lepper, Devine, and is likely to be acceptable to consumers and
& Petursdottir, 2016). Alternatively, behavior observers. On the other hand, latency recording
analysts could define changes in speech topics by is only appropriate for measuring a specifically
the occurrence of specific keywords, enabling the defined relation between antecedent conditions
use of duration measures (e.g., Fisher, Rodriguez, and target responding. As with other continuous
& Owen, 2013). With both approaches, however, measures, practical constraints (e.g., simultaneous
there is some risk of failing to capture the entirety measurement of multiple responses or observation
of the target behavior. For example, partial-inter- of multiple clients) can necessitate the use of dis-
val recording might overestimate the occurrence continuous measures.
of behavior, and keywords may not always corre-
spond with actual changes in content. However, Intensity. Intensity recording involves mea-
both approaches may produce data that are suffi- sures of magnitude, strength, amplitude, force, or
ciently accurate. effort of a response. Observers occasionally can
In addition, duration recording can be effort- record intensity through automated mechanical
ful, particularly if we are interested in multiple re- devices, but degrees of intensity may require judg-
144 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
ments that are difficult for observers to make reli- inant-activity sampling may provide accurate esti-
ably without these devices (Kazdin, 2001). Inten- mates of the time long-duration behavior occurs,
sity recording may be appropriate for measuring but may underestimate instantaneous behavior
shouting, throwing an object, hitting, swearing, (Bailey & Burch, 2002; Murphy & Goodall, 1980;
tantrums, and noise levels. Wilson and Hopkins Powell, Martindale, & Kulp, 1975). By contrast,
(1973) recorded intensity while examining the partial-interval recording tends to overestimate
effects of quiet-contingent music on the general overall occurrence of long-duration behavior, but
noise levels of middle school classrooms. Intensity tends to underestimate the occurrence of instan-
recording was particularly useful in this situation, taneous, high-rate behavior (Harrop & Daniels,
because the goal of intervention was to decrease 1993; Murphy & Goodall, 1980; Repp et al., 1976).
general noise levels in the classrooms. When overall duration is the response dimension
of interest, partial-interval recording almost al-
ways provides overestimates (Suen, Ary, & Covalt,
Discontinuous Recording Procedures
1991), suggesting that the behavior analyst should
Continuous measures yield the most accurate be- use whole-interval recording, predominant-activ-
havioral data, but these methods can be impracti- ity sampling, or momentary time sampling when
cal, particularly in applied settings where the be- duration is the response dimension of interest
havior analyst may have multiple important tasks (but see Harrop & Daniels, 1993, for an alterna-
and limited time to complete them. Therefore, tive interpretation). Behavior analysts can achieve
behavior analysts may choose discontinuous mea- more accurate estimates of frequency with shorter
surement procedures, which frequently increase interval lengths (e.g., 10 seconds; Devine, Rapp,
the efficiency of data collection. We discuss the Testa, Henrickson, & Schnerch, 2011; Rapp,
most common varieties of discontinuous recording Colby-Dirksen, Michalski, Carroll, & Lindenberg,
in the following paragraphs. 2008). However, shorter interval lengths do not
systematically increase the accuracy of estimates
Interval Recording. The behavior analyst di- of long-duration behavior (Sanson-Fisher, Poole,
vides the observation session into periods of equal & Dunn, 1980).
length, such as 10 seconds, and observers score Figures 8.1, 8.2, and 8.3 provide hypotheti-
each interval as positive (i.e., occurrence) or nega- cal demonstrations of the sensitivity of interval-
tive (i.e., nonoccurrence) according to specific recording conventions with behavioral streams
criteria. Varieties include whole-interval recording, consisting of occurrences of various rates and
in which the observer scores a positive interval if durations. Figure 8.1 shows that partial-interval
the target behavior occurs for the entire observa- recording overestimates and whole-interval re-
tional interval; partial-interval recording, in which cording underestimates frequency and duration of
the observer scores a positive interval if the target responses of moderate rate and varying duration.
behavior occurs at any point during the observa- Predominant-activity sampling, on the other hand,
tional interval; and predominant-activity sampling, provides a close estimate. Figure 8.2 shows that
in which the observer scores a positive interval if partial-interval recording provides a relatively ac-
the target behavior occurs for more than half the curate but slightly conservative estimate of the fre-
interval (Adams, 1991; Harrop & Daniels, 1986; quency of an instantaneous response of moderate
Poling et al., 1995; Saudargas & Zanolli, 1990; rate, but grossly overestimates its duration. Both
Tyler, 1979). Like other discontinuous measure- predominant-activity sampling and whole-interval
ment procedures, interval recording provides an recording are unable to detect any responses in a
estimate of behavior rather than a representation behavioral stream of this sort. Figure 8.3 shows
of the entire behavioral episode. Therefore, be- that partial-interval recording somewhat overes-
havior analysts should design interval-recording timates duration and frequency of long-duration
procedures to minimize error and maximize mea- behavior, whereas predominant-activity sampling
surement accuracy. and whole-interval recording slightly underesti-
Researchers have evaluated whether interval mate its duration and overestimate frequency.
recording influences the accuracy of measure- Interval measures frequently introduce substan-
ment systems when applied to behaviors of differ- tial error into the estimates of behavior, and the
ent rates and durations (Wirth, Slaven, & Taylor, amount and type of error depend on the relative
2014). Both whole-interval recording and predom- rate and duration of behavior and the parameters
Defining and Measuring Behavior 145
PIR + + - + + + + + + -
MTS + - - + + + - - - -
PAS + - - - + + + - + -
WIR - - - - + + - - - -
FIGURE 8.1. A demonstration of hypothetical results obtained using different time-sampling conventions by means of
an analysis of an event-recorded tape. The width of the figure represents 150 seconds, and the vertical lines represent
15-second intervals. The table presents the results of an analysis of this tape. PIR, partial interval recording; MTS,
momentary time sampling; PAS, predominant-activity sampling; WIR, whole-interval recording. Adapted in part
from Tyler (1979). Copyright by Elsevier. Adapted by permission.
PIR + + - + + - - + + +
MTS - - - + - - - - + -
PA - - - - - - - - - -
WIR - - - - - - - - - -
FIGURE 8.2. A demonstration of hypothetical results obtained using different time-sampling conventions by means of
an analysis of an event-recorded tape. The width of the figure represents 150 seconds, and the vertical lines represent
15-second intervals. The table presents the results of an analysis of this tape. Abbreviations as in Figure 8.1. Adapted
in part from Tyler (1979). Copyright by Elsevier. Adapted by permission.
146 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
PIR + + + + + + + - - -
MTS + + + + + + - - - -
PAS + + + - + + - - - -
WIR - + + - + + - - - -
FIGURE 8.3. A demonstration of hypothetical results obtained using different time-sampling conventions by means of
an analysis of an event-recorded tape. The width of the figure represents 150 seconds, and the vertical lines represent
15-second intervals. The table presents the results of an analysis of this tape. Abbreviations as in Figure 8.1. Adapted
in part from Tyler (1979). Copyright by Elsevier. Adapted by permission.
of the interval-recording system (Powell, Martin- Daniels, 1986; Suen et al., 1991). Unlike interval
dale, Kulp, Martindale, & Bauman, 1977; Wirth measures, it does not make assumptions about
et al., 2014). Nevertheless, interval recording, par- portions of observations that the observer did not
ticularly partial-interval recording, is among the sample. The observer scores a positive interval if a
measurement procedures behavior analysts use response occurs at any point in the interval with
most frequently (Bailey & Burch, 2002). The con- partial-interval recording. The observer scores
venience of recording and its sensitivity to detect a negative interval unless the behavior occurred
changes in the relevant dimensions of behavior as during the entire interval with whole-interval re-
a function of intervention are likely reasons for its cording. Momentary time sampling, on the other
popularity (Harrop & Daniels, 1986, 1993). For hand, only makes assumptions about the momen-
example, partial-interval recording is likely to un- tary periods when the observer samples behavior,
derestimate the magnitude of change in high-rate, and the behavior analyst calculates proportional
instantaneous behavior and to provide a conserva- duration estimates by dividing the number of
tive estimate of behavior change (Harrop & Dan- scored occurrences by the total number of inter-
iels, 1986; Suen et al., 1991). This is not necessar- vals, which produces a potentially unbiased esti-
ily a limitation of the procedure, however, because mate of duration (Suen et al., 1991).
it reduces the probability of Type I errors. Results of research have typically shown that
momentary time sampling does not systematically
Momentary Time Sampling. Momentary time overestimate or underestimate behavior. Errors
sampling consists of scoring an interval as positive in estimates tend to be close to random, leading
if the target behavior occurs at a predetermined to accurate estimates of the mean (Brookshire,
moment (Harrop & Daniels, 1986). A timer set Nicholas, & Krueger, 1978; Green & Alverson,
to sound every 10 seconds might prompt observ- 1978; Harrop & Daniels, 1986; Mudford, Beale, &
ers to assess whether the target behavior occurred. Singh, 1990; Murphy & Goodall, 1980; Powell et
When the timer goes off, the observers look at the al., 1975, 1977; Repp et al., 1988; Suen et al., 1991;
client and score whether he or she is performing Tyler, 1979). However, Wirth et al. (2014) found
the target behavior at that precise moment (hence that momentary time sampling is biased toward
the term momentary). underestimation when cumulative event durations
Momentary time sampling is useful in provid- are low and overestimation when cumulative event
ing estimates of duration of behavior (Harrop & durations are high. Harrop and Daniels (1993)
Defining and Measuring Behavior 147
caution that although momentary time-sampling actual rates of behavior as partial-interval record-
estimates typically match actual behavior dura- ing (Harrop & Daniels, 1986; Harrop, Daniels,
tions, unbiased estimates are not necessarily ac- & Foulkes, 1990). Tyler (1979) discussed the dif-
curate when based on a single session of observa- ficulty in using momentary time sampling when
tion. Consistent with this notion, researchers have the behavior of interest is not amenable to instant
found that increases in the observation period re- recognition. Despite these reservations, several
duce magnitude and variability of error in momen- researchers have shown that momentary-time-
tary time sampling (Devine et al., 2011; Wirth et sampling measures correlate highly with continu-
al., 2014). Although definitive guidelines for the ous measures (e.g., Brulle & Repp, 1984; Harrop &
use of momentary time sampling do not exist, use Daniels, 1986; Powell et al., 1977).
of this method may be advisable in some situations As in interval recording, shorter interval length
over others, such as when high-rate, long-duration generally will yield more accurate data in momen-
behaviors are of interest and when the observation tary time sampling (Brookshire et al., 1978; Brulle
involves multiple responses or organisms (Hanley, & Repp, 1984; Harrop & Daniels, 1985; Kearns,
Cammilleri, Tiger, & Ingvarsson, 2007; Murphy & Edwards, & Tingstrom, 1990; Mansell, 1985;
Harrop, 1994). We do not recommend momentary Saudargas & Zanolli, 1990). Brulle and Repp
time sampling for responses that have a short dura- (1984) examined different interval lengths and
tion or occur infrequently (Arrington, 1943). found that 10-second, 20-second, 30-second, and
Figures 8.1, 8.2, and 8.3 provide hypothetical 60-second intervals provided accurate estimates of
demonstrations of the sensitivity of momentary the mean duration of the target behavior, but that
time sampling for behavioral streams with response a 120-second interval was accurate only when the
occurrences of various rates and durations. Figure target behavior occurred for more than 10% of the
8.1 shows that momentary time sampling provides session. The 240-second interval was accurate only
a slight underestimate of behavior of varying du- when behavior occurred for more than 20% of the
ration and moderate rate. Figure 8.2 demonstrates session. Harrop and Daniels (1985) cautioned that
how such sampling may grossly underestimate researchers should only use intervals longer than
both frequency and duration of instantaneous re- 30 seconds when the target behavior occurs dur-
sponding of moderate rate. However, momentary ing 25% or more of the total observation period.
time sampling is likely to provide a close estimate Devine et al.’s (2011) results are consistent with
of the overall duration and an overestimate of the the latter recommendation. Thomson, Holmberg,
frequency of relatively long-duration behavior that and Baer (1974) examined interval length and the
occurs for a large portion of the observation pe- rotation of observation intervals when observing
riod, as demonstrated in Figure 8.3. behaviors of multiple organisms. They suggested
Momentary time sampling has several advan- that the smallest error percentage occurs when be-
tages. Cameron, Crosbie, and Crocker (1988) sug- havior is dispersed widely across the observation
gested that it is a good choice because researchers schedule.
can analyze data with inferential statistics, which Some research has compared the accuracy of
can provide a communicative aid when present- momentary time sampling and interval recording.
ing results to psychologists and other professionals Green, McCoy, Burns, and Smith (1982) com-
subscribing to non-behavior-analytic paradigms. pared whole-interval recording, partial-interval
Another key benefit of momentary time sampling recording, and momentary time sampling; they
is the ease of implementation (Brulle & Repp, reported that momentary time sampling provided
1984; Brookshire et al., 1978). Although momen- greater representativeness of the actual behavior
tary time sampling is not an error-free method of and produced fewer observer errors than other
observation, Murphy and Goodall (1980) suggest- interval-recording methods. Wirth et al. (2014)
ed that it may be the best available option when found that momentary time sampling generally
continuous recording is not feasible. resulted in smaller overall absolute magnitude of
There are several potential limitations of mo- error when compared to partial-interval record-
mentary time sampling, including inadequate rep- ing, but that the variability of error was greater
resentation of certain behavioral dimensions and with momentary time sampling (i.e., it sometimes
insensitivity to actual duration, frequency, and underestimated and sometimes overestimated be-
changes in behavior (Repp et al., 1976). Several havior). As stated previously, overall magnitude
authors have reported that momentary time sam- of error tends to decrease with increases in the
pling may not be as sensitive to small changes in length of observation periods.
148 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
Frequency is
of interest
Yes
Can the
behavior No Employ appropriate
indirect measures
be directly
(e.g., product recording)
observed?
Yes
Is extended No Momentary
observation
time sampling
practical?
Yes
No
Yes
Does the
behavior occur at No
a low enough
rate to count?
Yes
Frequency
FIGURE 8.4. Decision flowchart for selecting an observational method when frequency is the behavioral dimension of
interest.
150 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
Duration is
of interest
Yes
Is constant No
observation
possible?
Yes
Is the behavior
Yes Interval
instantaneous? recording
Is the No No
behavior
discrete?
Yes
Momentary
time sampling
Can duration be
easily scored given
No
the number of topographies
and/or participants?
Yes
Duration
FIGURE 8.5. Decision flowchart for selecting an observational method when duration is the behavioral dimension of
interest.
Defining and Measuring Behavior 151
Bailey, J. S., & Burch, M. R. (2002). Research methods Devine, S. L., Rapp, J. T., Testa, J. R., Henrickson, M.
in applied behavior analysis. Thousand Oaks, CA: L., & Schnerch, G. (2011). Detecting changes in
SAGE. simulated events using partial-interval recording and
Bass, R. F. (1987). Computer-assisted observer training. momentary-time sampling: III. Evaluating sensitivity
Journal of Applied Behavior Analysis, 20, 83–88. as a function of session length. Behavioral Interven-
Beaver, B. R., & Busse, R. T. (2000). Informant reports: tions, 26, 103–124.
Conceptual and research bases of interviews with Dixon, M. R. (2003). Creating a portable data-collec-
parents and teachers. In E. S. Shapiro & T. R. Kra- tion system with Microsoft® embedded visual tools
tochwill (Eds.), Behavioral assessment in schools (2nd for the pocket PC. Journal of Applied Behavior Analy-
ed., pp. 257–287). New York: Guilford Press. sis, 36, 271–284.
Bijou, S. W., Peterson, R. F., & Ault, M. H. (1968). A Doll, B., & Elliott, S. N. (1994). Representativeness
method to integrate descriptive and experimental of observed preschool social behaviors: How many
field studies at the level of data and empirical con- data are enough? Journal of Early Intervention, 18,
cepts. Journal of Applied Behavior Analysis, 1, 175–191. 227–238.
Bijou, S. W., Peterson, R. F., Harris, F. R., Allen, K. E., Fisher, W. W., Lindauer, S. E., Alterson, C. J., & Thomp-
& Johnston, M. S. (1969). Methodology for experi- son, R. H. (1998). Assessment and treatment of de-
mental studies of young children in natural settings. structive behavior maintained by stereotypic object
Psychological Record, 19, 177–210. manipulation. Journal of Applied Behavior Analysis,
Bowman, L. G., Fisher, W. W., Thompson, R. H., & Pi- 31, 513–527.
azza, C. C. (1997). On the relation of mands and the Fisher, W. W., Piazza, C. C., Bowman, L. G., & Amari,
function of destructive behavior. Journal of Applied A. (1996). Integrating caregiver report with a system-
Behavior Analysis, 30, 251–265. atic choice assessment to enhance reinforcer identifi-
Brookshire, R. H., Nicholas, L. S., & Krueger, K. (1978). cation. American Journal on Mental Retardation, 101,
Sampling of speech pathology treatment activities: 15–25.
An evaluation of momentary and interval time sam- Fisher, W. W., Rodriguez, N. M., & Owen, T. M. (2013).
pling procedures. Journal of Speech and Hearing Re- Functional assessment and treatment of persevera-
search, 21, 652–667. tive speech about restricted topics in an adolescent
Brulle, A. R., & Repp, A. C. (1984). An investigation with Asperger syndrome. Journal of Applied Behavior
of the accuracy of momentary time sampling proce- Analysis, 46, 307–311.
dures with time series data. British Journal of Psychol- Fritz, J. N., Iwata, B. A., Hammond, J. L., & Bloom, S. E.
ogy, 75, 481–485. (2013). Experimental analysis of precursors to severe
Cameron, D. L., Crosbie, J., & Crocker, A. D. (1988). problem behavior. Journal of Applied Behavior Analy-
A fixed interval momentary sampling method for sis, 46, 101–129.
assessing ongoing behaviors induced by dopamine Fuhrman, A. M., Fisher, W. W., & Greer, B. D. (2016).
receptor agonists. Progress in Neuro-Psychopharmacol- A preliminary investigation on improving functional
ogy and Biological Psychiatry, 12, 595–606. communication training by mitigating resurgence
Cariveau, T., & Kodak, T. (2017). Programming a ran- of destructive behavior. Journal of Applied Behavior
domized dependent group contingency and common Analysis, 49, 884–899.
stimuli to promote durable behavior change. Journal Grace, N. C., Thompson, R., & Fisher, W. W. (1996).
of Applied Behavior Analysis, 50, 121–133. The treatment of covert self-injury through contin-
Connell, J. E., & Witt, J. C. (2004). Applications of gencies on response products. Journal of Applied Be-
computer-based instruction: Using specialized soft- havior Analysis, 29, 239–242.
ware to aid letter-name and letter-sound recognition. Green, S. B., & Alverson, L. G. (1978). A comparison of
Journal of Applied Behavior Analysis, 37, 67–71. indirect measures for long-duration behaviors. Jour-
Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). nal of Applied Behavior Analysis, 11, 530.
Applied behavior analysis (3rd ed.). Upper Saddle Green, S. B., McCoy, J. F., Burns, K. P., & Smith, A. C.
River, NJ: Pearson. (1982). Accuracy of observational data with whole
Cordes, A. K. (1994). The reliability of observational interval, partial interval, and momentary-time-sam-
data: I. Theories and methods for speech language pling recording techniques. Journal of Behavioral As-
pathology. Journal of Speech and Hearing Research, 37, sessment, 4, 103–118.
264–278. Gresham, F. M., Gansle, K. A., & Noell, G. H. (1993).
Critchfield, T. S. (1999). An unexpected effect of re- Treatment integrity in applied behavior analysis with
cording frequency in reactive self-monitoring. Journal children. Journal of Applied Behavior Analysis, 26,
of Applied Behavior Analysis, 32, 389–391. 257–263.
Dempsey, C. M., Iwata, B. A., Fritz, J. N., & Rolider, N. Gutowski, S. J., & Stromer, R. (2003). Delayed match-
U. (2012). Observer training revisited: A comparison ing to two-picture samples by individuals with and
of in vivo and video instruction. Journal of Applied Be- without disabilities: An analysis of the role of nam-
havior Analysis, 45, 827–832. ing. Journal of Applied Behavior Analysis, 36, 487–505.
152 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
Hagopian, L. P., Contrucci Kuhn, S. A., Long, E. S., for collecting real-time observational data. Journal of
& Rush, K. S. (2005). Schedule thinning following Applied Behavior Analysis, 31, 253–261.
communication training: Using competing stimuli to Kazdin, A. E. (1977). Artifact, bias, and complexity of
enhance tolerance to decrements in reinforcer den- assessment: The ABC’s of reliability. Journal of Ap-
sity. Journal of Applied Behavior Analysis, 38, 177–193. plied Behavior Analysis, 10, 141–150.
Hanley, G. P., Cammilleri, A. P., Tiger, J. H., & Ingvars- Kazdin, A. E. (1979). Unobtrusive measures in behav-
son, E. T. (2007). A method for describing preschool- ioral assessment. Journal of Applied Behavior Analysis,
ers’ activity preferences. Journal of Applied Behavior 12, 713–724.
Analysis, 40, 603–618. Kazdin, A. E. (2001). Behavior modification in applied set-
Hanley, G. P., Iwata, B. A., Lindberg, J. S., & Conners, J. tings (6th ed.). Belmont, CA: Wadsworth/Thomson
(2003). Response-restriction analysis: I. Assessment Learning.
of activity preferences. Journal of Applied Behavior Kearns, K., Edwards, R., & Tingstrom, D. H. (1990). Ac-
Analysis, 36, 47–58. curacy of long momentary-time sampling intervals:
Hansen, B. D., Sabey, C. V., Rich, M., Marr, D., Robins, Implications for classroom data collection. Journal of
N., & Barnett, S. (2019). Latency-based functional Psychoeducational Assessment, 8, 74–85.
analysis in schools: Correspondence and differences Kelly, M. B. (1976). A review of academic permanent-
across environments. Behavioral Interventions, 34, product data collection and reliability procedures in
366–376. applied behavior analysis research. Journal of Applied
Harrop, A., & Daniels, M. (1985). Momentary time Behavior Analysis, 9, 211.
sampling with time series data: A commentary on Kent, R. N., Kanowitz, J., O’Leary, K. D., & Cheiken,
the paper by Brulle & Repp. British Journal of Psy- M. (1977). Observer reliability as a function of cir-
chology, 76, 533–537. cumstances of assessment. Journal of Applied Behavior
Harrop, A., & Daniels, M. (1986). Methods of time Analysis, 10, 317–324.
sampling: A reappraisal of momentary-time sampling Kent, R. N., O’Leary, K. D., Dietz, A., & Diament, C.
and partial interval recording. Journal of Applied Be- (1979). Comparison of observational recordings in
havior Analysis, 19, 73–77. vivo, via mirror, and via television. Journal of Applied
Harrop, A., & Daniels, M. (1993). Further reappraisal Behavior Analysis, 12, 517–522.
of momentary-time sampling and partial-interval Lannie, A. L., & Martens, B. K. (2004). Effects of task
recording. Journal of Applied Behavior Analysis, 26, difficulty and type of contingency on students’ allo-
277–278. cation of responding to math worksheets. Journal of
Harrop, A., Daniels, M., & Foulkes, C. (1990). The use Applied Behavior Analysis, 37, 53–65.
of momentary-time sampling and partial interval re- Lepper, T. L., Devine, B., & Petursdottir, A. I. (2016).
cording in behavioral research. Behavioural Psycho- Application of a lag contingency to reduce persevera-
therapy, 18, 121–127. tion on circumscribed interests: A brief report. Devel-
Hartmann, D. P. (1984). Assessment strategies. In D. H. opmental Neurorehabilitation, 20, 313–316.
Barlow & M. Hersen (Eds.), Single case experimental Lerman, D. C., Parten, M., Addison, L. R., Vorndran, C.
designs: Strategies for studying behavior change (2nd M., Volkert, V. M., & Kodak, T. (2005). A methodol-
ed., pp. 107–139). New York: Pergamon Press. ogy for assessing the functions of emerging speech in
Hartmann, D. P., & Wood, D. D. (1982). Observational children with developmental disabilities. Journal of
methods. In A. S. Bellack, M. Hersen, & A. E. Kaz- Applied Behavior Analysis, 38, 303–316.
din (Eds.), International handbook of behavior modifi- Maglieri, K. A., DeLeon, I. G., Rodriguez-Catter, V., &
cation and therapy (pp. 109–133). New York: Plenum Sevin, B. M. (2000). Treatment of covert food steal-
Press. ing in an individual with Prader–Willi syndrome.
Hersen, M., & Barlow, D. H. (1976). Single case experi- Journal of Applied Behavior Analysis, 33, 615–618.
mental designs: Strategies for studying behavior change. Mansell, J. (1985). Time sampling and measurement
Oxford, UK: Pergamon Press. error: the effect of interval length and sampling pat-
Hoch, H., McComas, J. J., Johnson, L., Faranda, N., & tern. Journal of Behavior Therapy and Experimental
Guenther, S. L. (2002). The effects of magnitude and Psychiatry, 16, 245–251.
quality of reinforcement on choice responding dur- McComas, J. J., & Mace, F. C. (2000). Theory and prac-
ing play activities. Journal of Applied Behavior Analy- tice in conducting functional analysis. In E. S. Shap-
sis, 35, 171–181. iro & T. R. Kratochwill (Eds.), Behavioral assessment
Jackson, J., & Dixon, M. R. (2007). A mobile computing in schools (2nd ed., pp. 78–103). New York: Guilford
solution for collecting functional analysis data on a Press.
Pocket PC. Journal of Applied Behavior Analysis, 40, Merrell, K. W. (2000). Informant reports: Theory and
359–384. research in using child behavior rating scales in
Johnston, J. M., & Pennypacker, H. S. (1993). Strategies school settings. In E. S. Shapiro & T. R. Kratoch-
and tactics of behavioral research (2nd ed.). Hillsdale, will (Eds.), Behavioral assessment in schools (2nd ed.,
NJ: Erlbaum. pp. 233–256). New York: Guilford Press.
Kahng, S., & Iwata, B. A. (1998). Computerized systems Miller, D. L., & Kelley, M. L. (1994). The use of goal
Defining and Measuring Behavior 153
setting and contingency contracting for improving Poling, A., Methot, L. L., & LeSage, M. G. (1995). Fun-
children’s homework performance. Journal of Applied damentals of behavior analytic research. New York:
Behavior Analysis, 27, 73–84. Plenum Press.
Miltenberger, R. G., Rapp, J. T., & Long, E. S. (1999). A Powell, J., Martindale, A., & Kulp, S. (1975). An evalu-
low-tech method for conducting real-time recording. ation of time-sample measures of behavior. Journal of
Journal of Applied Behavior Analysis, 32, 119–120. Applied Behavior Analysis, 8, 463–469.
Mudford, O. C., Beale, I. L., & Singh, N. N. (1990). The Powell, J., Martindale, B., Kulp, S., Martindale, A., &
representativeness of observational samples of differ- Bauman, R. (1977). Taking a closer look: Time sam-
ent durations. Journal of Applied Behavior Analysis, pling and measurement error. Journal of Applied Be-
23, 323–331. havior Analysis, 10, 325–332.
Mudford, O. C., Martin, N. T., Hui, J. K. Y., & Taylor, S. Prinz, R. J. (1982). Observing and recording children’s
A. (2009). Assessing observer accuracy in continu- behavior. Behavioral Assessment, 4, 120–121.
ous recording of rate and duration: Three algorithms Rapp, J. T., Colby-Dirksen, A. M., Michalski, D. N.,
compared. Journal of Applied Behavior Analysis, 42, Carroll, R. A., & Lindenberg, A. M. (2008). Detect-
527–539. ing changes in simulated events using partial-interval
Mudford, O. C., Zeleny, J. R., Fisher, W. W., Klum, M. E., recording and momentary-time-sampling. Behavioral
& Owen, T. M. (2011). Calibration of observational Interventions, 23, 237–269.
measurement of rate of responding. Journal of Applied Repp, A. C., Nieminen, G. S., Olinger, E., & Brusca, R.
Behavior Analysis, 44, 571–586. (1988). Direct observation: Factors affecting the ac-
Murphy, G., & Goodall, E. (1980). Case histories and curacy of observers. Exceptional Children, 55, 29–36.
shorter communications: Measurement error in di- Repp, A. C., Roberts, D. M., Slack, D. J., Repp, C. F.,
rect observations: A comparison of common record- & Berkler, M. S. (1976). A comparison of frequency,
ing methods. Behaviour Research and Therapy, 18, interval and time-sampling methods of data collec-
147–150. tion. Journal of Applied Behavior Analysis, 9, 501–508.
Murphy, M. J., & Harrop, A. (1994). Observer effort Rolider, N. U., Iwata, B. A., & Bullock, C. E. (2012).
in the use of momentary-time sampling and partial Influences of response rate and distribution on the
interval recording. British Journal of Psychology, 85, calculation of interobserver reliability scores. Journal
169–180. of Applied Behavior Analysis, 45, 753–762.
Nay, R. N. (1979). Multimethod clinical assessment. New Sanson-Fisher, R. W., Poole, A. D., & Dunn, J. (1980).
York: Gardner Press. An empirical method for determining an appropriate
Nelson, R. O. (1977). Methodological issues in as- interval length for recording behavior. Journal of Ap-
sessment via self-monitoring. In J. D. Cone & R. P. plied Behavior Analysis, 13, 493–500.
Hawkins (Eds.), Behavioral assessment: New directions Sanson-Fisher, R. W., Poole, A. D., Small, G. A., &
in clinical psychology (pp. 217–240). New York: Brun- Fleming, I. R. (1979). Data acquisition in real time:
ner/Mazel. An improved system for naturalistic observations.
North, S. T., & Iwata, B. A. (2005). Motivational in- Behavior Therapy, 10, 543–554.
fluences on performance maintained by food rein- Saudargas, R. A., & Zanolli, K. (1990). Momentary time
forcement. Journal of Applied Behavior Analysis, 38, sampling as an estimate of percentage time: A field
317–333. validation. Journal of Applied Behavior Analysis, 23,
Northup, J. (2000). Further evaluation of the accuracy 533–537.
of reinforcer surveys: A systematic replication. Jour- Schinke, S. P., & Wong, S. E. (1977). Coding group
nal of Applied Behavior Analysis, 33, 335–338. home behavior with a continuous real-time record-
Odom, S. L., & Ogawa, I. (1992). Direct observation of ing device. Behavioral Engineering, 4, 5–9.
young children’s interactions with peers: A review of Sigurdsson, S. O., Aklin, W., Ring, B. M., Needham,
methodology. Behavioral Assessment, 14, 407–441. M., Boscoe, J., & Silverman, K. (2011). Automated
O’Leary, K. D., Kent, R. N., & Kanowitz, J. (1975). Shap- measurement of noise violations in the therapeutic
ing data collection congruent with experimental workplace. Behavior Analysis in Practice, 4, 47–53.
hypotheses. Journal of Applied Behavior Analysis, 8, Slocum, S. K., Grauerholz-Fisher, E., Peters, K. P., &
43–51. Vollmer, T. R. (2018). A multicomponent approach
Page, T. J., & Iwata, B. A. (1986). Interobserver agree- to thinning reinforcer delivery during noncontingent
ment: History, theory, and current practice. In A. reinforcement schedules. Journal of Applied Behavior
Poling & R. W. Fuqua (Eds.), Research methods in Analysis, 51, 61–69.
applied behavior analysis: Issues and advances (pp. 99– Suen, H. K., Ary, D., & Covalt, W. (1991). Reappraisal
126). New York: Plenum Press. of momentary-time sampling and partial-interval
Panos, R., & Freed, T. (2007). The benefits of automatic recording. Journal of Applied Behavior Analysis, 24,
data collection in the fresh produce supply chain. In 803–804.
IEEE International Conference on Automation Science Tapp, J. (2003). Procoder for digital video: User manual.
and Engineering (pp. 1034–1038). Piscataway, NJ: In- Retrieved November 23, 2005, from http://mingus.
stitute of Electrical and Electronics Engineers. kc.vanderbilt.edu/pcdv.
154 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
Tapp, J., & Walden, T. (1993). PROCODER: A pro- Tyler, S. (1979). Time-sampling: Matter of convention.
fessional tape control, coding, and analysis system Animal Behaviour, 27, 801–810.
for behavioral research using videotape. Behavior VanWormer, J. J. (2004). Pedometers and brief e-coun-
Research Methods, Instruments, and Computers, 25, seling: Increasing physical activity for overweight
53–56. adults. Journal of Applied Behavior Analysis, 37, 421–
Taravella, C. C., Lerman, D., Contrucci, S. A., & Roane, 425.
H. S. (2000). Further evaluation of low-ranked items Wasik, B. H., & Loven, M. D. (1980). Classroom ob-
in stimulus-choice preference assessments. Journal of servational data: Sources of inaccuracy and proposed
Applied Behavior Analysis, 33, 105–108. solutions. Behavioral Assessment, 2, 211–227.
Test, D. W., Rose, T. L., & Corum, L. (1990). Applied Wehby, J. H., & Hollahan, M. S. (2000). Effects of
behavior analysis with secondary students: A meth- high-probability requests on the latency to initiate
odological review of research published from 1968 to academic tasks. Journal of Applied Behavior Analysis,
1987. Education and Treatment of Children, 13, 45–62. 33, 259–262.
Thiemann, K. S., & Goldstein, H. (2001). Social stories, Whiting, S. W., & Dixon, M. R. (2012). Creating an
written text cues, and video feedback: Effects on so- iPhone application for collecting continuous ABC
cial communication of children with autism. Journal data. Journal of Applied Behavior Analysis, 45, 643–
of Applied Behavior Analysis, 34, 425–446. 656.
Thomson, C., Holmberg, M., & Baer, D. M. (1974). Wilson, C. W., & Hopkins, B. L. (1973). The effects of
A brief report on a comparison of time-sampling contingent music on the intensity of noise in junior
procedures. Journal of Applied Behavior Analysis, 7, high home economics classes. Journal of Applied Be-
623–626. havior Analysis, 6, 269–275.
Torres-Viso, M., Strohmeier, C. W., & Zarcone, J. R. Wilson, D. M., Iwata, B. A., & Bloom, S. E. (2012).
(2018). Functional analysis and treatment of problem Computer-assisted measurement of wound size asso-
behavior related to mands for rearrangement. Journal ciated with self-injurious behavior. Journal of Applied
of Applied Behavior Analysis, 51, 158–165. Behavior Analysis, 45, 797–808.
Traub, M. R., & Vollmer, T. R. (2019). Response la- Wirth, O., Slaven, J., & Taylor, M. A. (2014). Interval
tency as a measure of behavior in the assessment of sampling methods and measurement error: A com-
elopement. Journal of Applied Behavior Analysis, 52, puter simulation. Journal of Applied Behavior Analysis,
422–438. 47, 83–100.
CHAP TER 9
A primary goal of applied behavior analysis is to individual. Similarly, the use of statistical analysis
understand how specific variables affect socially may obscure behavior change outcomes at the in-
significant behavior or produce behavior change. dividual’s level. For example, suppose an interven-
Single-case experimental designs are ideal for tion produces a statistically significant reduction
this type of analysis, because the focus is on the in the occurrence of self-injurious behavior; these
problem, setting, or individual that is the target findings may lack clinical significance if the re-
for evaluation (Kazdin, 2016). Single-case experi- duced level of self-injury produces tissue damage.
mental designs involve continuous assessment By contrast, single-case experimental designs are
to measure behavior over time. Thus the use of appropriate for tracking continuous, moment-to-
single-case designs allows a behavior analyst to moment changes in the occurrence of the target
determine if an independent variable (e.g., inter- behavior that allow the behavior analyst to deter-
vention) produced the desired change in the tar- mine whether behavior change is clinically signifi-
get dependent variable (e.g., aggression). These cant for the individual.
designs also afford the behavior analyst the flex- The main goal of single-case experimental de-
ibility to modify the measures and method to suit sign is to demonstrate a functional relation—the
the problem of interest. effects of an independent variable on a dependent
Although well suited for use with individuals, variable (Baer, Wolf, & Risley, 1968). An inde-
single-case experimental designs in applied behav- pendent variable is one, such as the intervention,
ior analysis do not exclude group designs or statis- that a behavior analyst can manipulate to affect
tical analysis (Roane, Ringdahl, Kelley, & Glover, a dependent variable, which is generally the be-
2011). However, group designs may not allow a havior targeted for change. A functional relation
behavior analyst to determine individual varia- exists when changes in the dependent variable
tions in the effects of an independent variable occur only during manipulations of the indepen-
on the dependent variable of interest for a given dent variable. Thus the behavior analyst can use
individual. In other words, the behavior analyst a single-case experimental design to systematically
cannot use a group design to determine the ef- identify variables that influence the occurrence
fects (e.g., optimal, minimal, no effect) the inde- of the behavior. In this chapter, we review data
pendent variable has on behavior change for an collection, visual inspection of data, and specific
155
156 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
single-case experimental designs, which we illus- definition should also include details that differen-
trate with hypothetical data and with data from tiate the target behavior from other behaviors, so
our work and that of others. that two independent observers reliably measure
the target behavior. For example, the observer
should record appropriate communication if the
DATA COLLECTION operational definition is contacting a card with
at least one finger on the palm side of the hand,
Effective use of single-case experimental designs and a child contacts the card with one finger on
depends on accurate and reliable data collection. the palm side of the hand, but not when the child
Important steps in developing an accurate and re- contacts the card with other body parts, such as
liable data collection procedure include defining the arm.
the target behavior(s), selecting a data collection The next step in data collection is to select an
system, and collecting baseline data (Roane et al., appropriate measurement system. There are four
2011). In addition, we discuss other variables that categories of behavior measurement: (1) event
influence data collection in strategies involving recording, (2) duration recording, (3) interval
single-case experimental designs. We have focused recording, and (4) permanent-product recording
on how the observation setting and conditions can (Kahng, Ingvarsson, Quigg, Seckinger, & Teich-
affect data collection outcomes, as these are vari- man, 2011; Roane et al., 2011). The operational
ables that behavior analysts often overlook. definition of the target behavior will influence
Development of an operational definition for which measurement system is most appropriate.
the target behavior of interest is an important first Event recording is often appropriate for discrete
step in the data collection process. Operational responses—behaviors that have a clear begin-
definitions should be (1) observable, (2) based ning and end. The observer typically records each
on an action, and (3) descriptive. Measurement occurrence of the target behavior. The recorded
of observable behavior is preferable to measure- events are typically converted into a rate, such as
ment of behaviors that are difficult or impossible frequency over unit of time. In some cases, dura-
to observe, such as internal states. Behavior ana- tion recording, in which the observer records the
lysts measure observable behavior to avoid mak- time a behavior occurs, is appropriate for discrete
ing inferences about the cause of behavior, such behaviors that occur at high rates. Duration re-
as “He was angry,” and to ensure that the mea- cording is also appropriate for continuous behav-
sured behavior is an accurate representation of iors, such as remaining on task, particularly when
the behavior of interest as it occurs in the target the time the behavior occurs is the most impor-
environment. For example, hitting is an observ- tant target for change. The observer might report
able behavior that has a clear beginning and end, a duration record as a duration of time (e.g., “The
whereas being angry is an inferred emotional state. child cried for 5 minutes”) or as a percentage of
An observer might infer that a person is angry if time (e.g., “The child cried for 50% of the ses-
the person hits someone, but hitting could occur sion”). Interval recording is useful if we want to
for other reasons. know whether the behavior occurred or did not
The behavior analyst should consider the ac- occur in a specific interval. The observer would
tion rather than the outcome of the behavior to record the occurrence of the behavior if it oc-
develop an operational definition, unless he or she curred any time during the interval with partial-
is using a permanent-product measure of behav- interval recording, or if the behavior occurred for
ior, such as number of math problems completed. the entire interval with whole-interval recording.
The observer should record an occurrence of the Interval data are typically reported as percentage
behavior if it meets the parameters of the defini- of intervals. Permanent-product recording mea-
tion, regardless of its consequence or perceived in- sures the outcome of a behavior, such as number of
tention. For example, the observer should record completed homework sheets.
a head turn during a mealtime observation if the We have provided a brief overview of opera-
observed behavior matches the operational defini- tional definitions and behavior measurement in
tion (e.g., the head moving 45 degrees away from this chapter; the reader can refer to Kahng et al.
the utensil), independently of whether the behav- (Chapter 8, this volume) for a more in-depth re-
ior occurred after the sound of a door opening and view. Below, we discuss how setting and condition
someone walking into the room. The operational variables can affect data collection outcomes.
Single-Case Experimental Design 157
FIGURE 9.1. Results from Lang et al. (2008) demonstrating correspondence in behavior function despite differences in
setting (top) and variations in behavior function across settings (bottom).
sults (bottom). These results show that functional- tant elements to consider during visual inspection
analytic outcomes may differ, depending on the of single-case experimental design data include
setting in which the observation occurs for some the (1) immediacy of effects; (2) magnitude of
individuals. effects; and (3) proportion of overlapping data
Developing operational definitions and select- points, which should demonstrate clearly that ma-
ing an appropriate measurement system, observa- nipulations of the independent variable produced
tion setting, and observation conditions are in- consistent, meaningful changes in the dependent
tegral steps in single-case experimental designs. variable.
The primary focus of these designs, however, is The level of behavior indicates the extent of
demonstration of functional control. Therefore, in the problem. To put this another way, level in-
the remainder of this chapter, we focus on visual dicates how frequently the target behavior oc-
inspection and on variations in single-case experi- curs and whether the graphed data correspond to
mental designs. other observations and records. Stability refers to
fluctuations in the data over time. Stable levels
of behavior are those that occur within a limited
VISUAL INSPECTION range, so that the behavior analyst can predict the
future level of the target behavior, given that the
Visual inspection of graphed data is the primary independent variable does not change. By con-
method of analyzing single-case experimental trast, unstable data vary across a wider range from
design data. The goal of visual inspection is to data point to data point, so that the future level
make predictions about the future occurrence of of the target behavior is more difficult to predict.
the observed behavior, based on the (1) level, (2) Trend refers to whether the behavior is improving
stability, and (3) trend of the data. Other impor- or worsening over time. An improving trend (e.g.,
Single‑Case Experimental Design 159
FIGURE 9.2. Example of immediate (left) and nonimmediate (right) treatment effects.
160 M e A s u r e M e n t, d e s I g n, A nd M e t h o d o l o g I c A l I s s u e s
FIGURE 9.4. Example of overlapping (left) and nonoverlapping (right) data points across phases.
Visual inspection is the most preferred and most reliable in the absence of the modified structured
often used method for interpreting graphed data criteria, whereas agreement increased when the
from single-case experimental designs (Horner criteria were used. During pretraining, in which
& Kratochwill, 2012), and accurate interpreta- the reviewers had access to the written modified
tion of graphed data is essential for determining criteria for visual inspection, the average agree-
the effects of experimental manipulations. One ments between the reviewers and the expert judge
disadvantage is that interpretations during vi- were .73 and .80, respectively, for the master’s-level
sual inspection of such data may not be reliable and postbaccalaureate interns. Following train-
among untrained individuals. Thus researchers ing, interobserver agreement increased for both
have developed methods to reduce potential bias groups. Specifically, the exact agreement between
and improve agreement during visual inspection the master’s-level reviewers and the expert judge
of single-case experimental design data. Results of increased to an average of .98, whereas the average
these studies have shown that appropriate training exact agreement between the postbaccalaureate
improves interrater agreement to acceptable levels reviewers and the expert judge increased to .95.
(e.g., Fisher, Kelley, & Lomas, 2003; Kahng et al., Roane, Fisher, Kelley, Mevers, and Bouxsein
2010; O’Grady, Reeve, Reeve, Vladescu, & Lake, (2013) modified Hagopian et al.’s (1997) structured
2018; Stewart, Carr, Brandt, & McHenry, 2007; criteria to apply them to results of functional anal-
Wolfe & Slocum, 2015). The generality of these yses with varying lengths. Roane et al. used the
results may be limited, however, because most criteria to train individuals with different levels of
of the studies used data from sequential designs education and clinical experience, which produced
(Hagopian et al., 1997). high levels of interrater agreement and demon-
Hagopian et al. (1997) showed that interrater strated the validity of the modified criteria. Roane
agreement was low (M = .46) when three pred- et al. then used the criteria to train master’s-level
octoral interns used visual inspection to interpret and postbaccalaureate participants to interpret
graphed functional-analytic data. Hagopian et al. graphed functional-analytic data, and agreement
developed structured criteria for interpretation of between the master’s-level and postbaccalaureate
such data, based on the consensus of experts in participants and an expert judge was .98 and .95,
functional analysis. Hagopian et al. used the struc- respectively. Finally, Roane et al. applied the cri-
tured criteria to train the same predoctoral interns teria to the data from 141 functional analyses and
to interpret graphed functional-analytic data, and identified the maintaining reinforcement contin-
mean interrater agreement increased to .81. One gency for problem behavior in a similar percent-
limitation is that Hagopian et al. used functional age of cases relative to those reported previously
analyses that included 10 data points per condi- (e.g., Hagopian, Rooker, Jessel, & DeLeon, 2013;
tion; thus behavior analysts could not use the cri- Hanley, Iwata, & McCord, 2003). Taken together,
teria to interpret functional analyses of different these results suggest that behavior analysts can
lengths. use structured criteria to improve the reliability of
The results of Study 1 suggested that the inter- visual inspection and interpretation of functional-
pretation of functional-analytic outcomes was less analytic data. As Hagopian et al., and Roane et al.
Single-Case Experimental Design 161
note, however, behavior analysts should continue meta-analyses to determine the level of empirical
to use visual inspection as the primary method support for behavioral interventions (see Horner
and use structured criteria as an adjunct for inter- & Kratochwill, 2012, for additional discussion).
pretation of functional-analytic data. Fisher and Lerman noted that recommendations
Researchers and clinicians have recognized from researchers who conduct statistical analyses
the need to evaluate results of single-case experi- for such data have shifted from a philosophy of
mental designs to expand the pool of scientific statistics as the only method of data analysis to
evidence available for review for empirically sup- statistics as an adjunct to visual inspection. This
ported interventions (Kratochwill et al., 2010; shift is likely to become more acceptable, as be-
Pustejovsky & Ferron, 2017). As such, the What havior analysts are not likely to abandon visual
Works Clearinghouse drafted standards for single- inspection completely. One barrier to the addition
case experimental designs, based on input from of statistical analysis to visual inspection is the
national experts. The information from these complexity of statistical-analytic procedures. For
standards overlaps with the contents of this chap- example, Shadish, Hedges, and Pustejovsky (2014)
ter, including (1) an overview of single-case ex- used a mean-difference statistic to analyze single-
perimental designs, (2) the types of questions for case experimental design data that Fisher and Ler-
which a researcher or clinician might use a single- man described as easy to use. By contrast, other
case experimental design, and (3) a discussion of researchers (e.g., Shadish, Zuur, & Sullivan, 2014)
the internal validity of single-case experimental have evaluated statistical-analytic procedures that
designs. The panel proposed standards for these require training and skills that behavior analysts in
designs, which address the internal validity of the practice might not have (Fisher & Lerman, 2014).
designs, and for evidence, which provide guide- Fisher and Lerman concluded that there may be a
lines for determining whether the evidence meets tradeoff between the appropriateness of statistical
the described standards. These guidelines provide methods for single-case experimental design data
another example of the focus on increased objec- and the methods’ user-friendliness. Additionally,
tivity of interpreting single-case experimental de- applications of statistical analysis to data from
sign data. single-case experimental designs are limited, and
Statistical analysis is a well-accepted method replications are needed to determine their ap-
for analyzing group data, but it has not gained propriateness. However, more recent research has
widespread acceptance for analysis of single-case identified the use of randomization techniques as
experimental design data. One reason for the possible alternatives to more traditional statisti-
dearth of statistical analysis in single-case experi- cal analyses. Randomization techniques may be
mental designs is their focus on the individual. more readily appropriate for analyzing single-case
Behavior analysts are generally interested in ro- experimental design data, due to (1) the lack of
bust intervention effects that have a meaningful assumptions regarding the distribution of outcome
impact for the individual. A limitation of statis- variables, (2) the ability to apply the methods to
tical analysis for single-case experimental designs a variety of research designs, and (3) the ability
is that a statistically significant effect may not be to obtain meaningful outcomes when applied to
clinically relevant, or a clinically relevant effect small-N datasets (Craig & Fisher, 2019). Overall,
may not be statistically significant (Kazdin, 2011). increased collaboration between statisticians and
Statistical analysis may be useful in situations in behavior analysts is encouraged to bridge the gap
which clinically relevant effects are difficult to between the use of visual and statistical analysis
determine from visual inspection. Moreover, con- for single-case experimental design data (Fisher &
sensus on and acceptance of statistical analysis to Lerman, 2014).
enhance rather than replace visual analysis are
growing (Cohen, Feinstein, Masuda, & Vowles,
2014). TYPES OF SINGLE‑CASE EXPERIMENTAL DESIGNS
Fisher and Lerman (2014) commented that the
development of a statistical metric for estimating One goal of a single-case experimental design is to
effect sizes for single-case experimental designs show that manipulation of the independent vari-
would represent an important advancement in able is responsible for changes in the dependent
the field. Estimation of effect sizes may increase variable, which we refer to as internal validity. Ex-
the likelihood that researchers would use data ternal validity is the extent to which the results of
from single-case experimental-design research in a study extend beyond the experimental setting.
162 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
Behavior analysts must consider and control for pends on (1) relatively stable levels of responding
threats to internal and external validity during in the A and B phases, or predictably unstable
development and execution of the experimental levels of responding; (2) unambiguous differences
arrangement. in the level of responding in the A and B phases;
Common threats to internal validity in single- (3) trends in responding in the appropriate direc-
case experimental designs include history, matura- tion in the A and B phases; (4) relatively imme-
tion, testing, and instrumentation. These threats diate changes in responding during introduction
could alter, contribute to, or be solely responsible and reversal of A- and B-phase contingencies; (5)
for changes in the dependent variable. History in a clinically acceptable level of responding if the
this context refers to events that occur simultane- goal is to increase or decrease target responding;
ously with the experimental manipulation (Krato- and (6) minimal overlap in the levels of respond-
chwill et al., 2010). Maturation refers to changes ing in the A and B phases. To put all this another
within the experimental participant that occur way, responding reliably turns off and on with the
naturally over time (e.g., aging; Edgington, 1996; introduction and reversal of A- and B-phase con-
Kratochwill et al., 2010). Testing refers to the ef- tingencies. Additional reversals between the two
fects of repeated measurement (e.g., a test score phases, with corresponding changes in the target
improves because the participant took the test response, strengthen the demonstration of a func-
multiple times). Instrumentation is the way observ- tional relation.
ers measure behaviors, and observer drift, which The introduction and reversal of the experi-
is an example of an instrumentation error, is a mental manipulation are essential but also the
change in the way observers record the same tar- most criticized characteristics of the ABAB de-
get behavior over time. sign. This criticism is most relevant when the be-
Threats to external validity may not be appar- havior analyst withdraws or removes an effective
ent until after researchers study and expand on intervention, particularly if the intervention has
the conditions under which the phenomenon oc- produced a clinically significant change in a dan-
curred originally (Kazdin, 2011). External validity gerous behavior. Although this is a valid concern
generally refers to whether the results of a study in this case, the effects of intervention should be
can be applied to other participants, settings, re- robust, and the clinically significant change in re-
sponse measures, and behavior change agents sponding should return when the behavior analyst
(Kazdin, 2011). Reactivity, which is another threat reinstates the intervention. The behavior analyst
to external validity, occurs when a participant must balance the advantages of demonstration of
changes his or her behavior because of knowl- functional control over a reemergence of previous
edge that he or she is involved in an experimental levels of responding. Other limitations include
manipulation or assessment. Pretest sensitization responding that does not return to baseline rates
occurs when preexperimental measurement (e.g., during a reversal, and the potential for responding
baseline) produces changes in participant behav- during the reversal to occur at higher rates than
ior. Finally, multiple-treatment interference occurs during the initial baseline.
when the study includes more than one manipula-
tion, and manipulations that occur earlier in the
Multiple‑Baseline Design
study may affect responding later in the study.
In the remainder of this section, we review the In a concurrent multiple-baseline design, the behav-
characteristics of the most common single-case ior analyst identifies three or four baselines that he
experimental designs. We also review advanced or she can use to evaluate the effects of the experi-
applications of single-case experimental designs, mental manipulation (Byiers, Reichle, & Symons,
such as parametric analysis, component analysis, 2010). The baselines could be behaviors, settings,
and combined designs. or individuals, and each baseline is referred to
as a leg of the multiple-baseline design. The be-
havior analyst implements the experimental ma-
Reversal (ABAB) Design
nipulation on Leg 1 when the level, stability, and
The most basic experimental design is the ABAB trend of the data are appropriate and continues
reversal, in which one condition (the A phase; to implement the baseline contingencies in the
e.g., baseline) is followed by another condition other legs. The behavior analyst implements the
(the B phase; e.g., intervention). Demonstration experimental manipulation in Leg 2 if it produced
of functional control with the ABAB design de- the targeted effect in Leg 1, and implements the
Single-Case Experimental Design 163
experimental manipulation in Leg 3 if it produced at the level of the terminal goal, independently of
the targeted effect in Leg 2. The behavior analyst the current criterion. For example, if the terminal
implements the experimental manipulation at goal of a training program is to teach an individual
different times (i.e., staggered) for each leg. The to sort 50 knives and 50 forks, the behavior analyst
multiple-baseline design demonstrates functional should supply the individual with 50 knives and 50
control when changes in responding occur im- forks on every trial even if the criterion is to sort
mediately after the introduction of the experi- 10 knives and 10 forks, 20 knives and 20 forks, 30
mental manipulation for each leg of the multiple knives and 30 forks, or 40 knives and 40 forks.
baseline. Baseline levels of responding should be A limitation of the changing-criterion design is
maintained for each leg until the behavior analyst that responding may not correspond to the chang-
implements the experimental manipulation, and es in the criterion, which weakens the demonstra-
this feature controls for the passage of time. That tion of functional control. This design may be
is, the behavior analyst can attribute the changes most appropriate for interventions targeting skill
in responding to the experimental manipulation. acquisition or for manipulations that include fad-
Multiple-baseline designs are appropriate when ing and shaping. This design may be less desirable
the experimental manipulation targets behavior when the goal is a large, rapid, or large and rapid
that is not likely to be reversed (i.e., reading skills) change in responding.
or when a reversal to baseline contingencies may
not be desirable (i.e., for dangerous behavior).
Multielement Design
One limitation of the multiple-baseline design
is that effects of the experimental manipulation As the name implies, the multielement design in-
may carry over to the baseline(s) of one or more volves conducting multiple experimental manip-
legs of the design. Carry-over effects decrease con- ulations that the behavior analyst alternates in
fidence that the experimental manipulation was rapid succession, such as from session to session.
responsible for the change in responding. Addi- This rapid alternation is equivalent to completing
tionally, the multiple-baseline design may result in several “mini-reversals.” For example, the behav-
prolonged exposure to baseline contingencies for ior analyst might evaluate the effects of noncon-
some targets, which may potentially delay behav- tingent reinforcement in one condition and time
ior change. out in another condition on rates of disruptive
behavior. The design produces a demonstration of
functional control when there are differences in
Changing‑Criterion Design
responding across the conditions. The ideal dem-
The initial baseline of the changing-criterion de- onstration of functional control with a multiele-
sign is like those of other designs. The changing- ment design involves stable, differentiated levels of
criterion design differs from other designs in that responding across conditions with no or minimal
(1) the experimental manipulation should produce data overlap. The multielement design is appropri-
changes in responding that correspond to a criteri- ate when the goal is to compare the effects of two
on that changes over time; (2) levels of responding or more experimental manipulations while elimi-
should change only when the criterion changes; nating the sequence effects that are problematic
and (3) responding should ultimately correspond with reversal designs. Sequence effects occur when
to the terminal goal of the manipulation. The a history with one condition influences respond-
other designs we have discussed thus far produce ing during a subsequent condition. Behavior ana-
a demonstration of functional control when the lysts may use a multielement design to compare
level and magnitude of responding after the exper- many antecedent and consequent manipulations
imental manipulation are different from those at when the goal is rapid assessment of effects. An
baseline. By contrast, the changing-criterion de- advantage of the multielement design is that it
sign produces a demonstration of functional con- does not require a reversal to or a single prolonged
trol when levels of responding correspond to the baseline phase, given that the behavior analyst
relatively small incremental changes to the crite- can alternate the baseline with the experimental
rion, rather than the large immediate changes we manipulation. Furthermore, implementation of a
generally expect with other designs. An important baseline is not necessarily a requirement of this
component of the changing-criterion design that design. The goal is to evaluate levels of responding
behavior analysts often overlook is that the par- during two or more experimental manipulations,
ticipant should have the opportunity to respond rather than to compare levels of responding during
164 M e a s u r e m e n t, D e s i g n, a nd M e t h o d o l o g i c a l I s s u e s
the experimental manipulation with those during result of carry-over effects (a limitation of the mul-
the baseline. tielement design). We then conducted a reversal
The multielement design has several limitations to baseline, tested the effects of escape extinction,
that may impede a demonstration of functional reversed again to baseline, and tested the effects
control. As with the multiple-baseline design, of avoidance. This example illustrates that single-
carry-over effects are a potential limitation of the case experimental designs are flexible. When we
multielement design. Individuals also may have observed carry-over effects during the multiele-
difficulty discriminating between conditions. Be- ment comparison of the two interventions, we
havior analysts who do not randomize the order added a reversal to strengthen the demonstration
of conditions may observe order effects, when the of functional control.
order of the conditions affects responding. Finally,
demonstration of functional control may be more
Component and Parametric Analyses
difficult with behaviors that change slowly (e.g.,
reading skills), given the rapid, alternating condi- Often behavior analysts implement intervention
tions of the multielement design. in a package that includes multiple components.
For example, consider a child whose disruptive
classroom behavior is maintained by adult atten-
Combined Designs and Other
tion. The classroom teacher implements an inter-
Design Considerations
vention in which he or she delivers attention for
Single-case experimental designs are flexible, such hand raising and withholds attention when the
that the behavior analyst can combine individual child engages in disruptive behavior. The teacher
designs to strengthen the demonstration of func- can be more confident that the intervention has
tional control (Roane et al., 2011). Furthermore, produced increases in hand raising and decreases
including more than one design in an analysis in disruptive behavior if the child’s responding
may mediate several of the disadvantages associ- changes only when the teacher implements the
ated with individual single-case experimental intervention and holds other variables constant.
designs. This section provides a case example to In this section, we discuss how the behavior ana-
demonstrate the flexibility of combined single- lyst can use single-case experimental designs to
case experimental designs. We also discuss design examine the effects of (1) components of a mul-
flexibility relative to assessment of the effects of ticomponent intervention and (2) intervention
(1) various experimental manipulations and (2) parameters.
varying levels or intensity of the experimental ma-
nipulations. Component Analyses. Component analyses
are useful when an intervention consists of mul-
tiple components. Even though a behavior analyst
Combined Designs
may have demonstrated a functional relation be-
The behavior analyst may need additional design tween a multicomponent intervention and levels
elements when a limitation of a common single- of responding, he or she may want to conduct a
case experimental design impedes the demonstra- component analysis to evaluate the effects of indi-
tion of a functional relation. Figure 9.5 shows a vidual intervention components (Cooper, Heron,
clinical example in which we evaluated the dif- & Heward, 2007).
ferential effects of two interventions on inappro- Ward-Horner and Sturmey (2010) have out-
priate mealtime behavior across three participants. lined two methods for component analysis. In a
Our initial plan was to use a multielement design drop-out analysis, the behavior analyst initially
to compare the effects of avoidance and escape implements the intervention package and then
extinction on individual participants, which we systematically removes one component at a time.
embedded in a multiple-baseline design to assess In an add-in analysis, the behavior analyst imple-
intervention effects across participants. Inappro- ments one component of an intervention and
priate mealtime behavior decreased to low levels then introduces individual components or combi-
during the multielement comparison of avoidance nations of components. A change in responding
and escape extinction. We wondered whether both indicates whether the dropped or added compo-
interventions were equally effective, or whether nent or components were responsible for changes
the simultaneous decrease in rates of inappropriate in responding. Component analyses may be useful
mealtime behavior for the two interventions was a when implementation of a package intervention is
Single‑Case Experimental Design 165
FIGURE 9.5. Clinical example for combined designs. The circled data represent examples of potential carry-over effects
in baseline, following introduction of treatment in the first leg of the multiple-baseline design.
indicated, but analysis of the effective components nonremoval of the spoon); and (4) a warm-up
is also important. in which the child had access to toys before the
Cooper et al. (1995) evaluated the components feeder presented food. Food acceptance and con-
of intervention packages to increase the food ac- sumption increased during the multicomponent
ceptance and consumption of four participants intervention. The researchers then conducted a
with intellectual developmental disorder. We component analysis, using a multielement design.
depict the data for one participant in Figure 9.6. Levels of responding did not change when the
The intervention consisted of (1) a choice of food; researchers introduced and withdrew the choice
(2) contingent adult attention for food accep- and warm-up components, suggesting that choice
tance; (3) presentation of the food until the feeder and warm-up did not affect responding. By con-
could deposit the food into the child’s mouth (i.e., trast, levels of responding did change when the
166 M e A s u r e M e n t, d e s I g n, A nd M e t h o d o l o g I c A l I s s u e s
researchers introduced and withdrew nonremoval The researchers compared the effects of FCT plus
of the spoon. extinction to FCT plus punishment in Leg 4 for
Fisher et al. (1993) used an add-in analysis to disruption with the second set of demands.
evaluate an intervention to reduce the problem be-
havior of five participants with intellectual devel- Parametric Analyses. Behavior analysts use
opmental disorder. Figure 9.7 depicts responses per parametric analyses to evaluate the effects of dif-
minute of disruption (top), aggression (second), ferent parameters of an independent variable, such
and self-injury (third) with one set of demands as magnitude, intensity, and integrity, on the de-
(Demand 1), and responses per minute of disrup- pendent variable. For example, a behavior analyst
tion (bottom) during a second set of demands (De- might use a parametric analysis to assess the effects
mand 2) for one participant. The researchers used of 5, 10, 30, or 60 seconds of attention on rates of
reversal and multiple-baseline designs to evaluate problem behavior. Parametric analyses allow the
rates of problem behavior during (1) baseline, (2) behavior analyst to determine the precise level of
extinction, (3) punishment, (4) functional-com- an experimental manipulation needed to produce
munication training (FCT), (5) FCT plus punish- and sustain changes in responding (e.g., Athens
ment, and (6) FCT plus extinction. For example, & Vollmer, 2010; Codding & Lane, 2015; Wilder,
Leg 1 shows rates of disruption during a phase Atwell, & Wine, 2006). We provide several clini-
of extinction followed by phases of punishment cal examples of parametric analyses below, and we
alone, extinction alone, FCT plus punishment, highlight their applied use.
FCT alone, and FCT plus punishment. Rates of In a four-experiment study, Athens and Vollmer
disruption were lowest during FCT plus punish- (2010) conducted parametric analyses to exam-
ment, and the researchers replicated that effect ine the effects of duration, quality, and delay to
in Leg 2 for aggression and Leg 3 for self-injury. reinforcement on mands and problem behavior.
Figure 9.8 shows data for Lana in Experiment 1, and Vollmer replicated the effect in subsequent
in which the researchers used an ABAB design A and B phases. Figure 9.9 shows data for Clark
to examine the effects of reinforcement duration. in Experiment 4, in which the researchers used an
During the A phase, the duration of reinforcement ABAB design to examine the effects of reinforce-
was equal (30 seconds) for mands and problem be- ment duration, quality, and delay. During the A
havior. Lana accessed reinforcement primarily by phase, mands and problem behavior produced 30
engaging in problem behavior when mands and seconds of high-quality reinforcement immediate-
problem behavior produced equal durations of re- ly after the behavior. Clark accessed reinforcement
inforcement. During the B phase, the duration of primarily by engaging in problem behavior when
reinforcement was 30 seconds for mands and 10 mands and problem behavior produced an equal
seconds for problem behavior. Rates of mands in- duration of, quality of, and delay to reinforcement.
creased and rates of problem behavior decreased, During the B phase, mands produced 30 seconds
showing that Lana accessed the longer duration of high-quality reinforcement immediately, and
of reinforcement by engaging in mands. Athens problem behavior produced 5 seconds of low-qual-
168 M e A s u r e M e n t, d e s I g n, A nd M e t h o d o l o g I c A l I s s u e s
FIGURE 9.8. Example of a parametric analysis from Experiment 1 of Athens and Vollmer (2010).
ity reinforcement 10 seconds after the occurrence clean were highest when the feeder presented the
of problem behavior. Rates of mands increased and smoothest texture (pureed), mouth clean did not
rates of problem behavior decreased, showing that increase to clinically acceptable levels. Kadey et
Clark accessed the longer duration of high-quality, al. identified nine foods that were associated with
immediate reinforcement by engaging in mands. lower levels of mouth clean and seven foods that
In their overall findings, Athens and Vollmer were associated with higher levels of it during the
showed that they could bias responding by altering texture analysis. Kadey et al. conducted a second
the various parameters of reinforcement. texture assessment (Figure 9.10, bottom) in which
Kadey, Piazza, Rivas, and Zeleny (2013) provided they compared levels of mouth clean for pureed
another example of a parametric analysis in which and Magic Bullet (which is smoother than pureed)
they evaluated the effects of food texture on mouth textures with the nine foods associated with lower
clean, a product measure of swallowing. Figure levels of mouth clean in the first texture assess-
9.10 (top) depicts the percentage of mouth clean ment. The highest levels of mouth clean were as-
during presentation of chopped food, wet ground sociated with the Magic Bullet texture. This para-
food, and pureed food. Although levels of mouth metric analysis allowed the researchers to identify
FIGURE 9.9. Example of a parametric analysis from Experiment 4 of Athens and Vollmer (2010).
Single‑Case Experimental Design 169
FIGURE 9.10. Example of a parametric analysis from Kadey, Piazza, Rivas, and Zeleny (2013).
SUMMARY REFERENCES
The current chapter has provided a general over- Athens, E. S., & Vollmer, T. R. (2010). An investigation
view of the relevance of single-case experimental of differential reinforcement of alternative behavior
designs in applied behavior analysis. This type of without extinction. Journal of Applied Behavior Anal-
experimental arrangement allows for the system- ysis, 43, 569–589.
Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some
atic evaluation of variables to ensure both clini-
current dimensions of applied behavior analysis.
cally and socially relevant changes in the behav- Journal of Applied Behavior Analysis, 1, 91–97.
ior of interest at the level of the individual and Byiers, B. J., Reichle, J., & Symons, F. J. (2012). Single-
across a range of socially significant challenges. subject experimental design for evidence-based prac-
More specifically, one design or a combination of tice. American Journal of Speech and Language Pathol-
designs may counterbalance the limitations of an- ogy, 21, 397–414.
other design. The flexibility of single-case experi- Carr, J. E., Bailey, J. S., Ecott, C. L., Lucker, K. D., &
mental designs is ideal for demonstrating func- Weil, T. M. (1998). On the effects of non-contingent
tional relations. As such, behavior analysts should delivery of differing magnitudes of reinforcement.
incorporate single-case experimental designs into Journal of Applied Behavior Analysis, 31, 313–321.
Codding, R. S., & Lane, K. L. (2015). A spotlight on
their everyday practice. Integrating these designs
treatment intensity: An important and often over-
into routine clinical practice allows the behavior looked component of intervention inquiry. Journal of
analyst to provide his or her clients, patients, or Behavioral Education, 24, 1–10.
students with the highest quality of care. Single- Cohen, L. L., Feinstein, A., Masuda, A., & Vowles, K.
case experimental designs provide information E. (2014). Single-case research design in pediatric
about the efficacy of intervention and provide the psychology: Considerations regarding data analysis.
clinician with feedback about his or her clinical Journal of Pediatric Psychology, 39, 124–137.
skills. Cooper, J. O., Heron, T. E., & Heward, W. L. (2007).
A behavior analyst must consider several issues Applied behavior analysis (2nd ed.) Upper Saddle
when using single-case experimental designs. De- River, NJ: Pearson/Merrill-Prentice Hall.
Cooper, L. L., Wacker, D. P., McComas, J. J., Brown, K.,
veloping clearly defined operational definitions of
Peck, S. M., Richman, D., et al. (1995). Use of com-
target behavior and selecting an appropriate mea- ponent analyses to identify active variables in treat-
surement system are essential first steps. Historical- ment packages for children with feeding disorders.
ly, visual inspection of data has been the primary Journal of Applied Behavior Analysis, 28, 139–153.
way to analyze data from single-case experimental Craig, A. R., & Fisher, W. W. (2019). Randomization
designs; however, critics have questioned the re- tests as alternative analysis methods for behavior-
liability of visual inspection. Research has shown analytic data. Journal of the Experimental Analysis of
that appropriate training can improve the accu- Behavior, 111, 309–328.
racy and reliability of visual inspection across ob- DeRosa, N. M., Roane, H. S., Bishop, J. R., & Silkowski,
servers. Statistical analysis may become a viable E. L. (2016). The combined effects of noncontingent
reinforcement and punishment on the reduction of
alternative or complement to visual inspection in
rumination. Journal of Applied Behavior Analysis, 49,
the future as these methods improve. However, 680–685.
additional work is needed to establish the role of Edgington, E. S. (1996). Randomized single-subject ex-
statistical analysis in interpretation of single-case perimental designs. Behaviour Research and Therapy,
data. Until then, visual inspection remains the 34, 567–574.
primary means of analysis. Fisher, W. W., Kelley, M. E., & Lomas, J. E. (2003).
Single-case experimental designs are ideal Visual aids and structured criteria for improving
methods for analyzing the effects of manipulations visual inspection and interpretation of single-case
on responding at the individual level. Thus these designs. Journal of Applied Behavior Analysis, 36,
design strategies allow the clinician to evaluate 387–406.
Fisher, W. W., & Lerman, D. C. (2014). It has been said
whether an intervention has produced a mean-
that, “There are three degrees of falsehoods: Lies,
ingful behavior change. Overall, the flexibility of damn lies, and statistics.” Journal of School Psychol-
single-case experimental designs can aid research- ogy, 52, 243–248.
ers and clinicians in identifying the most effective Fisher, W. W., Piazza, C. C., Cataldo, M. F., Harrell, R.,
means for producing socially meaningful changes Jefferson, G., & Conner, R. (1993). Functional com-
in behavior. munication training with and without extinction
Single-Case Experimental Design 171
and punishment. Journal of Applied Behavior Analy- parison of functional analysis results when conduct-
sis, 26, 23–36. ed in contrived versus natural settings. Journal of Ap-
Hagopian, L. P., Fisher, W. W., Thompson, R. H., Owen- plied Behavior Analysis, 41, 441–445.
DeSchryver, J., Iwata, B. A., & Wacker, D. P. (1997). O’Grady, A. C., Reeve, S. A., Reeve, K. F., Vladescu, J.
Toward the development of structured criteria for C., & Lake, C. M. (2018). Evaluation of computer-
interpretation of functional analysis data. Journal of based training to teach adults visual analysis skills of
Applied Behavior Analysis, 30, 313–326. baseline-treatment graphs. Behavior Analysis in Prac-
Hagopian, L. P., Rooker, G. W., Jessel, J., & DeLeon, I. tice, 11, 254–266.
G. (2013). Initial functional analysis outcomes and Pustejovsky, J. E., & Ferron, J. M. (2017). Research syn-
modifications in pursuit of differentiation: A summa- thesis and meta-analysis of single-case designs. In J.
ry of 176 inpatient cases. Journal of Applied Behavior M. Kauffman, D. P. Hallahan, & P. C. Pullen (Eds.),
Analysis, 46, 88–100. Handbook of special education (2nd ed., pp. 168–186).
Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003). New York: Routledge.
Functional analysis of problem behavior: A review. Roane, H. S., Fisher, W. W., Kelley, M. E., Mevers, J. L., &
Journal of Applied Behavior Analysis, 36, 147–185. Bouxsein, K. J. (2013). Using modified visual-inspec-
Horner, R. H., Carr, E. G., Halle, J., McGee, G., Odom, tion criteria to interpret functional analysis outcomes.
S., & Wolery, M. (2005). Research to identify evi- Journal of Applied Behavior Analysis, 46, 130–146.
dence-base practice in special education. Exceptional Roane, H. S., Ringdahl, J. E., Kelley, M. E., & Glover,
Children, 71, 165–179. A. C. (2011). Single-case experimental designs. In W.
Horner, R. H., & Kratochwill, T. R. (2012). Synthesizing W. Fisher, C. C. Piazza, & H. S. Roane (Eds.), Hand-
single-case research to identify evidence-based prac- book of applied behavior analysis (pp. 132–147). New
tices: Some brief reflections. Journal of Behavioral York: Guilford Press.
Education, 21, 266–272. Shadish, W. R., Hedges, L. V., & Pustejovsky, J. E.
Kadey, H. J., Piazza, C. C., Rivas, K. M., & Zeleny, J. (2014). Analysis and meta-analysis of single-case de-
(2013). An evaluation of texture manipulations signs with a standardized mean difference statistic: A
to increase swallowing. Journal of Applied Behavior primer and applications. Journal of School Psychology,
Analysis, 46, 539–543. 52, 15–39.
Kahng, S. W., Chung, K. M., Gutshall, K., Pitts, S. C., Shadish, W. R., Zuur, A. F., & Sullivan, K. J. (2014).
Kao, J., & Girolami, K. (2010). Consistent visual Using generalized additive (mixed) models to analyze
analysis of intra-subject data. Journal of Applied Be- single case designs. Journal of School Psychology, 52,
havior Analysis, 43, 35–45. 41–70.
Kahng, S., Ingvarsson, E. T., Quigg, A. M., Seckinger, K. Stewart, K. K., Carr, J. E., Brandt, C. W., & McHenry,
E., & Teichman, H. M. (2011). Defining and measur- M. M. (2007). An evaluation of the conservative du-
ing behavior. In W. W. Fisher, C. C. Piazza, & H. S. al-criterion method for teaching university students
Roane (Eds.), Handbook of applied behavior analysis to visually inspect AB-design graphs. Journal of Ap-
(pp. 113–131). New York: Guilford Press. plied Behavior Analysis, 40, 713–718.
Kazdin, A. E. (2011). Single-case research design. New Ward-Horner, J., & Sturmey, P. (2010). Component
York: Oxford University Press. analyses using single-subject experimental designs:
Kazdin, A. E. (2016). Single-case experimental designs. A review. Journal of Applied Behavior Analysis, 43,
In A. E. Kazdin (Ed.), Methodological issues and strat- 685–705.
egies in clinical research (pp. 459–483). Washington, Wilder, D. A., Atwell, J., & Wine, B. (2006). The effects
DC: American Psychological Association. of varying levels of treatment integrity on child com-
Kratochwill, T. R., Hitchcock, J., Horner, R. H., Levin, J. pliance during treatment with a three-step prompt-
R., Odom, S. L., Rindskopf, D. M., et al. (2010). Sin- ing procedure. Journal of Applied Behavior Analysis,
gle-case designs technical documentation. Retrieved 39, 369–373.
from https://files.eric.ed.gov/fulltext/ED510743.pdf. Wolfe, K., & Slocum, T. A. (2015). A comparison of two
Lang, R., O’Reilly, M., Machalicek, W., Lancioni, G., approaches to training visual analysis of AB graphs.
Rispoli, M., & Chan, J. (2008). A preliminary com- Journal of Applied Behavior Analysis, 48, 472–477.
PAR T IV
BEHAVIORAL ASSESSMENT
173
174 B e h av i o r a l A s s e s s m e n t
175
176 B e h av i o r a l A s s e s s m e n t
approached the stimulus within 5 seconds of pre- presented stimuli, raising the question of whether
sentation, the therapist gave the stimulus to the some stimuli identified as highly preferred may
participant for 5 seconds. If the participant did not not function as effective reinforcers (Paclawskyj
approach the stimulus within 5 seconds of presen- & Vollmer, 1995). That is, the single-stimulus
tation, the therapist prompted the participant to assessment may produce false positives, or stimuli
touch the stimulus. If the participant did not touch that appear highly preferred but do not function
the stimulus within 5 seconds of the prompt, the as reinforcement.
therapist ended that trial and presented the next
stimulus. Pace et al. labeled stimuli approached on
Paired‑Choice Preference Assessment
at least 80% of presentations preferred, and stimuli
approached on 50% or less of presentations non- Fisher et al. (1992) evaluated a variation of the
preferred. Pace et al. (1985) preference assessment to address
Next, Pace et al. (1985) assessed whether as- its limitations. They used a paired-choice proce-
sessed stimuli functioned as reinforcement. During dure in which they presented stimuli in pairs and
baseline, a simple, free-operant response, such as a prompted each participant, “Pick one.” They pre-
hand raise, produced no differential consequence. sented 16 stimuli and paired each stimulus once
The therapist then delivered either a preferred with every other stimulus. Participant approaches
or a nonpreferred stimulus identified during the toward one of the two stimuli produced access to
preference assessment following the free-operant that stimulus for approximately 5 seconds. The
response. Results indicated that contingent pre- therapist blocked simultaneous approaches toward
sentation of preferred stimuli increased respond- both stimuli, removed the stimulus pair, and re-
ing, compared to baseline and presentation of presented it if 5 seconds elapsed with no response.
nonpreferred stimuli. That is, stimuli approached The therapist prompted the participant to sample
more frequently in the preference assessment func- each stimulus for 5 seconds, re-presented the pair,
tioned as reinforcement more often than stimuli and then presented the next pair if another 5 sec-
approached less frequently. onds elapsed without a response. The choice as-
Although the Pace et al. (1985) single-stimulus sessment identified a hierarchy of preferences for
preference assessment was one of the first objec- participants (see Figure 10.1 for an example).
tive methods of identifying preferred stimuli, it Fisher et al. (1992) compared the results of the
has limitations. The most notable is that partici- paired-choice assessment with the results of the
pants may approach all or most stimuli (Fisher et Pace et al. (1985) single-stimulus assessment. Re-
al., 1992; Mazaleski, Iwata, Vollmer, Zarcone, & sults showed that both assessments identified cer-
Smith, 1993). For example, two of the three partic- tain items as high-preference stimuli. The single-
ipants in Mazaleski et al. (1993) approached most stimulus assessment identified several stimuli as
100
PERCENTAGE SELECTION
80
60
40
20
0
s
s
ys
ad
s
ks
ic
le
ar
ok
us
To
zz
oc
iP
Bo
M
Pu
Bl
al
To
im
An
STIMULUS
FIGURE 10.1. Example of a hierarchy of preferences for one individual, based on the results of a paired-choice prefer-
ence assessment with the names of the presented stimuli on the x-axis and the percentage selection on the y-axis.
The Effectiveness of Positive Reinforcement 177
developmental disorder. The children showed dif- participants. By contrast, Layer, Hanley, Heal, and
ferential preferences for the pictorially depicted ac- Tiger (2008) assessed the accuracy of preference
tivities. Northup et al. (1996) evaluated the accu- assessment for multiple children simultaneously.
racy of a reinforcer survey, a verbal paired-choice First, the researchers identified a preference hi-
questionnaire, and a pictorial paired-choice as- erarchy for each child individually. Next, the re-
sessment for preference identification. The verbal searchers evaluated the preferences of three chil-
and pictorial paired-choice assessments identified dren simultaneously. During the group assessment,
high- and low-preference categories for three of each child privately selected a colored card that
the four participants. By contrast, the survey inac- the experimenters had paired previously with spe-
curately identified preferences. One limitation of cific food reinforcement. After each child selected
pictorial representation is that a participant must a colored card, the researcher placed the three
be able to discriminate that the pictures represent cards in a box. Next, the researcher drew one card
items or activities. For example, Higbee, Carr, and from the box, and each child in the group received
Harrison (1999) found that the picture assessment the food that was associated with that card. Com-
did not consistently identify stimuli that func- parisons of individual and group assessment data
tioned as reinforcement, and suggested that some showed that the two assessments produced similar
individuals may require discrimination training preference rankings, but that the group assessment
before conducting a pictorial preference assess- identified the preferred stimuli more efficiently.
ment. Similarly, Radley, Dart, Battaglia, and Ford (2019)
conducted group preference assessments in a class-
room of 19 students, with all students respond-
Concurrent Chains
ing simultaneously to a prompt (administered via
Most of the preference assessment procedures de- smartphone app) to identify a preferred stimulus.
scribed earlier have focused on identification of They found that the group procedure was a valid
preferred stimuli that behavior analysts could use and rapid method of assessing preference within a
as reinforcement in training programs. Researchers classroom setting.
also have used preference assessments to measure Hanley, Cammilleri, Tiger, and Ingvarsson
participants’ preferences for positive-reinforcement (2007) showed that behavior analysts could use
treatments (Hanley, Piazza, Fisher, Contrucci, & momentary time sampling to assess the activity
Maglieri, 1997), schedules of reinforcement (Luc- preferences of 20 children in a preschool class-
zynski & Hanley, 2009, 2010, 2014), motivational room. They observed less than a 10% error rate
systems (Heal & Hanley, 2007), punishment and when they evaluated preferences with a 120-sec-
extinction components of interventions (Giles, St. ond momentary-time-sampling interval. Subse-
Peter, Pence, & Gibson, 2012; Hanley, Piazza, Fish- quent analyses showed that observers preferred a
er, & Maglieri, 2005), and preferences for choice 90-second interval relative to other intervals, and
and no-choice arrangements (Fisher, Thompson, that this interval duration accurately identified ac-
Piazza, Crosland, & Gotjen, 1997; Tiger, Hanley, tivity preferences for a classroom of children.
& Hernandez, 2006), among others. For example,
Hanley et al. (1997) described a concurrent-chains
procedure for evaluating participant preferences for METHODS FOR EVALUATING
functional-communication training, noncontin- REINFORCEMENT EFFECTS
gent reinforcement, and extinction as treatments
Correlation between Preference
for problem behavior. The concurrent-chains pro-
and Reinforcement Efficacy
cedure consisted of pairing each treatment with a
colored card that participants selected to enter a Researchers have evaluated whether the effec-
room in which they received their chosen treat- tiveness of reinforcement varies positively with
ment. The results of Hanley et al. showed that the the degree of preference (i.e., whether relative
concurrent-chains procedure provided a sensitive preferences demonstrated in preference assess-
measure of participant preferences for treatments ments predict relative reinforcer effectiveness).
for problem behavior. Piazza, Fisher, Hagopian, Bowman, and Toole
(1996) conducted preference assessments with four
individuals with severe behavior problems to iden-
Group Arrangement
tify high-, medium-, and low-preference stimuli.
Most preference assessment studies have focused Subsequent concurrent-operants reinforcement
on shortening administration time for individual assessments showed that the results of the prefer-
The Effectiveness of Positive Reinforcement 181
ence assessment predicted the reinforcing efficacy rent-operants schedule is that the magnitude of
of the high-, medium-, and low-preference stimuli. responding for each operant is a function of the
However, DeLeon, Iwata, and Roscoe (1997) and relative value of each reinforcer, rather than a
Taravella, Lerman, Contrucci, and Roane (2000) function of response competition or interference
showed that lower-ranked stimuli may function as (Catania, 1963; Fisher et al., 1992). The value of
reinforcement under some circumstances. Graff reinforcement is a function of its rate, magnitude,
and Larsen (2011) demonstrated that the reinforc- and quality, and of the immediacy of delivery and
ing efficacy of preferred stimuli may be a product the amount of response effort expended to obtain
of the stimuli used in a preference assessment, in- reinforcement, relative to those of other concur-
dependent of the preference hierarchy that is sub- rently available reinforcers (Fisher & Mazur, 1997).
sequently generated. Thus the rate of each response is a function of the
Lee, Yu, Martin, and Martin (2010) took a value of its reinforcer and the value of other con-
slightly different approach than Piazza et al. (1996) currently available reinforcers. Assume, for exam-
to examine the correspondence between prefer- ple, that Responses A and B concurrently produce
ence and reinforcement effects. Lee et al. identified Reinforcers A and B, respectively. Substantially
stimuli that maintained a range of response rates increasing the reinforcement rate for Response
(i.e., high to low) during a reinforcer assessment. A is likely to increase the rate of Response A.
Next, they conducted stimulus preference assess- The rate of Response B is likely to decrease, even
ments to determine the participants’ preferences though its reinforcement rate remains unchanged.
for those identified reinforcers. They found almost In most natural environments, multiple sources of
perfect correspondence between the preference reinforcement are available simultaneously, and
and reinforcer assessments for one participant, and behavior analysts should assess the value of a given
partial correspondence for the other participant. reinforcer relative to other concurrently available
reinforcers. Concurrent-operants schedules are
ideal for assessing the strength of a reinforcer rela-
Simple versus Complex Responses
tive to other available reinforcers.
Most researchers have used simple, free-operant In some cases, the behavior analyst may want
responses (e.g., hand raise, in-chair behavior) to to assess absolute reinforcement effects (e.g.,
assess the effectiveness of stimuli as reinforce- does Stimulus A function as reinforcement for
ment (e.g., DeLeon & Iwata, 1996; Fisher et al., Response A?). For example, Roscoe, Iwata, and
1992; Pace et al., 1985; Piazza, Fisher, Hagopian, Kahng (1999) showed that the most effective rein-
et al., 1996). Piazza, Fisher, Hagopian, et al. (1996) forcer during a concurrent-operants schedule was
suggested that the use of a simple, free-operant the stimulus identified as highly preferred on both
response during reinforcement assessment has single-stimulus and paired-choice assessments.
several advantages. The goal of the reinforcement However, stimuli identified as highly preferred
assessment is to evaluate whether the stimulus only by the single-stimulus assessment functioned
functions as reinforcement, rather than to teach as reinforcement during the single-operant sched-
a specific response. Simple responses are ideal for ule. Roscoe et al. suggested that concurrent-sched-
these types of evaluations, because individuals ule procedures are useful for the assessment of
with varying functional levels can typically dis- relative reinforcement effects (preference for one
criminate the contingencies rapidly, resulting in a reinforcer over another), and that single-schedule
time-efficient assessment. Failure to emit a more arrangements may be ideal for assessing the abso-
complex response during a reinforcement assess- lute effects of reinforcement.
ment could be due to a skill or a motivational defi-
cit. By contrast, failure to emit a simple response
Progressive‑Ratio Schedules
is less likely to be due to a skill deficit. There may
be situations, however, in which the use of a more Roane, Lerman, and Vorndran (2001) used pro-
complex response (e.g., on-task behavior) is desir- gressive-ratio schedules to assess relative respond-
able (Paramore & Higbee, 2005). ing for two items identified as similarly preferred
during a stimulus preference assessment (Fisher
et al., 1992). In a progressive-ratio schedule, the
Single versus Concurrent Operants
requirement to access reinforcement increases
Fisher et al. (1992) used a concurrent-operants during a single observation (Hodos, 1961). For
schedule to evaluate the reinforcing efficacy of example, the initial response requirement might
preferred stimuli. The advantage of a concur- be completing one math problem to receive a pre-
182 B e h av i o r a l A s s e s s m e n t
ferred toy for 20 seconds. After the first reinforcer assessment under single and concurrent progres-
delivery, the behavior analyst removes the toy and sive-ratio arrangements (Glover, Roane, Kadey, &
increases the response requirement to two math Grow, 2008); (2) assessing the relative effects of
problems to access 20 seconds of reinforcement. highly preferred and less preferred stimuli under
This progression might continue until responding increasing response requirements (Francisco, Bor-
ceases for a specified time. This reinforcer assess- rero, & Sy, 2008; Penrod, Wallace, & Dyer, 2008);
ment can determine how much work a participant and (3) evaluating whether extra-session access to
will complete for a given reinforcer before the par- reinforcement affects rates of academic respond-
ticipant reaches the breakpoint, or the schedule re- ing (Roane, Call, & Falcomata, 2005). Applied
quirement at which the participant does not meet researchers have used progressive-ratio schedules
the criterion for reinforcement. to evaluate (1) preferences for different staff mem-
Roane et al. (2001) showed that concurrent bers in a residential setting (Jerome & Sturmey,
fixed-ratio (FR) schedules failed to differentiate 2008); (2) the relative efficacy of different rein-
higher- from lower-preference stimuli, but that forcement durations (Trosclair-Lasserre, Lerman,
increasing the requirements using progressive- Call, Addison, & Kodak, 2008); and (3) whether
ratio schedules enhanced the differences in re- stimuli that function as differentially effective re-
inforcer effectiveness. Roane et al. also showed inforcers under progressive-ratio schedules are also
that accurate identification of higher-preference differentially effective when incorporated into
stimuli using progressive-ratio schedules was criti- reinforcement-based treatments for problem be-
cal, because the higher-preference stimuli pro- havior (Roane et al., 2001; Smith, Roane, & Ste-
duced greater reductions in problem behavior in phenson, 2009). Despite the potential benefits of
reinforcement-based treatments. Moreover, they progressive-ratio schedules, behavior analysts have
suggested that the within-session increase in re- questioned their utility (e.g., Poling, 2010). Thus
sponse requirements provided a more expeditious researchers should conduct additional studies to
evaluation of relative reinforcer efficacy than did evaluate the relative utility of progressive-ratio
evaluating stimuli across multiple phases of differ- schedules in applied settings.
ent FR requirements (e.g., DeLeon, Iwata, Goh, &
Worsdell, 1997).
Call, Trosclair-Lassare, Findley, Reavis, and ISSUES RELATED TO SPECIFIC STIMULI
Shillingsburg (2012) used progressive-ratio sched- AS REINFORCEMENT
ules to evaluate the accuracy of the paired-stim-
Choice as Reinforcement
ulus and multiple-stimulus-without-replacement
assessments over time. They administered a Researchers have evaluated whether choice func-
paired-choice preference assessment once, fol- tions as reinforcement. Although initial studies on
lowed by daily multiple-stimulus-without-replace- the effects of choice suggested that the opportu-
ment preference assessments and progressive-ratio nity to make choices functioned as reinforcement
reinforcer assessments. They evaluated the corre- (Dunlap et al., 1994; Dyer, Dunlap, & Winterling,
spondence between break points and preferences 1990; Powell & Nelson, 1997), this work was lim-
for the two stimulus preference assessments. They ited because the choice response produced access
found that the highest-ranked stimulus from the to highly preferred stimuli. Thus choice was con-
paired-choice assessment produced the highest founded with the individual’s preferences for the
breakpoints for all seven participants, whereas the chosen items in these investigations.
highest-ranked stimulus from the daily multiple- Fisher et al. (1997) addressed this confound
stimulus-without-replacement assessments cor- by yoking the choice and no-choice conditions.
responded to the highest break point for three of In Experiment 1, participants could choose from
seven participants. two available preferred stimuli as reinforcement in
Applied researchers have used progressive-ratio the choice condition. The researchers yoked the
schedules with increased frequency. These appli- reinforcer they delivered in the no-choice (con-
cations of progressive-ratio schedules have fallen trol) condition to the reinforcer the participant
typically into two categories: bridging basic and chose in the choice condition. For example, if the
applied research, and developing procedures of participant chose Gummy Bears on the first trial,
therapeutic significance (Roane, 2008). Examples Skittles on the second trial, and M&Ms on the
of bridging research with progressive-ratio sched- third trial of the choice condition, then the inves-
ules include (1) evaluating the effects of reinforcer tigator delivered Gummy Bears on the first trial,
The Effectiveness of Positive Reinforcement 183
Skittles on the second trial, and M&Ms on the combination of food and leisure items. The partic-
third trial in the no-choice condition. The results ipants displayed a general tendency to select food
of the study showed that higher levels of respond- over nonfood items in the combined preference as-
ing occurred in the choice than in the no-choice sessment, even though highly preferred food and
condition. In Experiment 2, participants could highly preferred leisure items from the leisure-only
choose from among lower-preference stimuli in assessment functioned as reinforcement. Similarly,
the choice condition or could gain access to high- Fahmie, Iwata, and Jann (2015) found that ed-
er-preference stimuli in the no-choice condition. ible items were more preferred than leisure items
Under these arrangements, participants generally and resulted in higher rates of responding under
allowed the investigator to choose the reinforcer. maintenance conditions in individuals with and
One limitation of yoking is that it does not con- without ASD. Bojak and Carr (1999) found that
trol for momentary fluctuations in preference over preference for edible items persisted even after
time (Tiger et al., 2006). For example, earning mealtimes. However, adults with dementia did not
Skittles in the previous choice condition may re- show a differential preference for edibles over lei-
duce the reinforcing effectiveness of the Skittles in sure items in a study by Ortega, Iwata, Nogales-
the subsequent no-choice condition. An alterna- González, and Frades (2012). These individuals
tive control for examining the effects of choice is often have deficits in sensory perception such as
to offer identical options in choice and no-choice smell and taste, which may reduce the value of ed-
conditions (Tiger et al., 2006). For example, ible items (Shiffman, 1997). This altered sensory
Thompson, Fisher, and Contrucci (1998) provid- perception might account for Ortega et al.’s find-
ed one young boy with autism spectrum disorder ing. Nevertheless, DeLeon et al. suggested that be-
(ASD) the opportunity to choose among three havior analysts should be cautious about including
identical cups of soda in the choice condition or edibles and leisure items in the same preference
gain access to one identical cup of soda that the assessment.
therapist chose in the no-choice condition. Their
results showed that the child preferred the choice
Social Stimuli
arrangement. Moreover, the participant continued
to select the choice condition even when the rein- According to the American Psychiatric Associa-
forcement rate was higher in the no-choice con- tion (2013), deficits in social interaction for chil-
dition. Research has shown that choice functions dren with ASD include but are not limited to a lack
as reinforcement for children with intellectual of social or emotional reciprocity. This diagnostic
developmental disorder and ASD (Fisher et al., feature suggests that social interaction may be less
1997; Thompson et al., 1998; Toussaint, Kodak, likely to function as reinforcement for children
& Vladescu, 2016), preschool-age children (Ack- with ASD. However, Nuernberger, Smith, Czapar,
erlund Brandt, Dozier, Juanico, Laudont, & Mick, and Klatt (2012) demonstrated that children with
2015; Schmidt, Hanley, & Layer, 2009; Tiger et al., ASD preferred some social interactions when the
2006), and individuals with traumatic brain injury researchers presented these stimuli during a mul-
(Tasky, Rudrud, Schulze, & Rapp, 2008). Finally, tiple-stimulus-without-replacement assessment.
Graff and Libby (1999) showed that participants These interactions (e.g., chase, tickles, swinging)
preferred to make choices during the session as op- subsequently functioned as reinforcement. This
posed to before it. was true even for interactions that the multiple-
stimulus-without-replacement assessment identi-
fied as relatively less preferred by the children with
Edible Stimuli
ASD. Clay, Samaha, Bloom, Bogoev, and Boyle
DeLeon, Iwata, and Roscoe (1997) hypothesized (2013) found similar results when they evaluated
that some individuals such as those with intellec- preferences for social interactions and attention
tual developmental disorder or ASD may be more in a paired-choice assessment with children with
likely to select food during preference assessments ASD. Morris and Vollmer (2019) found that pre-
relative to other stimuli. To that end, they assessed ferred social interactions could be identified for
whether edible items were more preferred than five children with ASD. These results suggest that
leisure items and activities during a preference as- social interactions can function as reinforcement
sessment. DeLeon et al. conducted separate mul- for children with ASD, and behavior analysts
tiple-stimulus-without-replacement assessments should include them in preference assessments.
consisting of food only, leisure items only, and a This is an important finding, given that one of
184 B e h av i o r a l A s s e s s m e n t
the primary diagnostic features of ASD is a lack of ing 5 minutes of a video at the end of a session)
typical social behavior and social reciprocity. over distributed consumption of smaller immedi-
ate reinforcers (e.g., watching 30 seconds of the
video after each response). These results suggest
Technology as Reinforcement
that accumulated but delayed reinforcement may
For some individuals, stimuli that rely on ad- be as effective as small, immediate reinforcement,
vanced technology (e.g., tablet computers) may and may be even more preferred for some partici-
have greater preference or reinforcing efficacy pants and for specific reinforcers (Bukala, Hu, Lee,
than stimuli that do not rely on technology. Ward-Horner, & Fienup, 2015; Fienup, Ahlers, &
Hoffmann, Samaha, Bloom, and Boyle (2017) Pace, 2011). Duration of access to a reinforcer (i.e.,
compared the preference and reinforcer efficacy accumulated) also may influence preference hier-
for high- and low-tech stimuli by examining the archies during systematic preference assessments
type and duration of interaction between the two (Steinhilber & Johnson, 2007).
stimuli types. Results suggested that item type and
access duration interacted to influence prefer-
Stimulus Variation
ence and reinforcer efficacy. Participants preferred
high-tech items at longer durations of access. Stimulus variation is one method that research-
However, participants preferred low-tech items at ers have used to enhance the effectiveness of re-
short durations. Moreover, participants engaged in inforcement (Bowman, Piazza, Fisher, Hagopian,
less responding when the high-tech item was pro- & Kogan, 1997; Egel, 1980, 1981; Wine & Wilder,
vided for short durations and when the low-tech 2009). For example, Bowman et al. (1997) found
item was provided for long durations. These results that four of six participants preferred varied ver-
suggest that providing longer access to stimuli that sus constant presentation of preferred stimuli; the
rely on technology when arranged during positive other two participants preferred constant pre-
reinforcement. sentation of preferred stimuli. Wine and Wilder
(2009) extended the work of Bowman et al. by
examining the effects of varied versus constant
FACTORS THAT MAY INFLUENCE reinforcement. Participants could earn access to
THE EFFECTIVENESS OF REINFORCEMENT (1) constant high-preference stimuli; (2) constant
medium-preference stimuli; (3) constant low-pref-
Reinforcement Rate, Quality, Delay,
erence stimuli; or (4) varied stimuli, in which the
and Distribution
experimenter randomly selected a high-, medium-,
Neef and colleagues conducted a series of studies or low-preference stimulus to deliver to the partici-
(Mace, Neef, Shade, & Mauro, 1994; Neef, Mace, pant on each trial. The greatest increases in work
& Shade, 1993; Neef, Mace, Shea, & Shade, 1992; output for both participants occurred in the con-
Neef, Shade, & Miller, 1994) to evaluate how rate, stant high-preference condition, and the varied-
quality, and delay to reinforcement affect respond- presentation condition resulted in work output
ing. Results from these studies suggested that comparable to constant delivery of medium-pref-
participants preferred schedules of reinforcement erence stimuli. Moreover, Keyl-Austin, Samaha,
associated with higher quality (Neef et al., 1992) Bloom, and Boyle (2012) found that presenting
and shorter delays to reinforcement (e.g., Neef et varied medium-preference stimuli produced higher
al., 1993). levels and more sustained responding than did
A longer delay to reinforcement will typically presenting those same stimuli singly. However, a
reduce the effectiveness of a reinforcer than a single highly preferred stimulus resulted in more
shorter delay (Fisher & Mazur, 1997). This is often total responses and a slower decline in within-ses-
true even when an individual is given a choice sion response rate, compared to responding main-
between a delayed larger reinforcer and a small- tained by varied medium-preference stimuli.
er but immediate reinforcer (Madden & Bickel,
2010). However, individuals may prefer larger ac-
Long‑Term Stability of Preferences
cumulated reinforcement over smaller immedi-
ate reinforcement under some circumstances. For Individual preferences for specific stimuli are con-
example, DeLeon et al. (2014) found that three stantly fluctuating, based on establishing opera-
of four participants preferred accumulating and tions and the environmental context in which we
consuming larger delayed reinforcers (e.g., watch- deliver them. For example, water may function
The Effectiveness of Positive Reinforcement 185
as reinforcement following consumption of salty Iwata, and Shore (2002) examined effects of both
pretzels, but may not function as reinforcement in satiation and deprivation of food as reinforce-
other contexts. Moreover, an individual may prefer ment on the pre- and postmeal responding of nine
candy, and candy may function as reinforcement adults with intellectual developmental disorder.
in most circumstances. The value of candy, how- Less than half the participants had higher premeal
ever, may wane if an individual has been consum- than postmeal response rates. The remaining par-
ing the same candy every day for a week. Behav- ticipants had pre- and postmeal response rates that
ior analysts should consider whether preferences were comparable. Satiation and deprivation effects
change over time and should be reevaluated pe- may be somewhat idiosyncratic, depending on the
riodically. Evidence for preference stability across individual and reinforcement type (e.g., Sy & Bor-
time has been somewhat mixed. For example, the rero, 2009).
results of several studies have shown that prefer-
ences vary over time (e.g., Carr, Nicholson, &
Higbee, 2000; Mason, McGee, Farmer-Dougan, & OTHER CONSIDERATIONS IN REINFORCEMENT
Risley, 1989; Zhou, Iwata, Goff, & Shore, 2001). AND STIMULUS SELECTION
By, contrast, Hanley, Iwata, and Roscoe (2006) at-
Teaching New Preferences
tempted to replicate and extend the literature by
evaluating preferences for leisure activities over Ideally, behavior analysts can identify several
3–6 months with 10 adults with intellectual de- highly preferred stimuli to evaluate as reinforcers.
velopmental disorder. Unlike previous researchers, However, some individuals may have few items or
Hanley et al. observed relatively stable preferences activities they prefer. In other cases, alternative
for 80% of participants. Hanley et al. also showed stimuli may be more appropriate to use as reinforc-
that naturally occurring changes in establishing ers in the given environmental context. Research-
operations or conditioning histories disrupted ers have taught individuals to shift their respond-
preference stability. Similarly, the results of Kelley, ing to items or activities previously assessed as less
Shillingsburg, and Bowen (2016) were consistent preferred by rearranging the environment. Stimu-
with Hanley et al.’s, in that preferences tended to lus–stimulus pairing, embedded reinforcement,
be relatively stable across time. Subsequent stud- and manipulation of motivating operations are the
ies have shown that changes in preferences across most common methods to increase the variety of
time does not necessarily affect reinforcer efficacy items or activities an individual prefers or to which
in practice (Verriden & Roscoe, 2016). DeLeon the individual allocates his or her responding.
et al. (2001) showed that when preferences did
change over time, behavior analysts could use
Pairing Procedures
daily brief preference assessments to monitor and
adjust to changes in preferences. Stimulus–stimulus pairing involves presenting
a highly preferred stimulus and a less preferred
stimulus in close temporal proximity, so that the
Satiation versus Deprivation
highly preferred stimulus follows the less preferred
Limiting access to reinforcement outside train- stimulus after a short delay. The less preferred
ing or treatment situations is a commonly recom- stimulus often becomes a conditioned reinforcer
mended strategy to maintain an individual’s moti- through repeated pairing. Researchers have used
vation. Kodak, Lerman, and Call (2007) evaluated this strategy to condition vocal sounds as reinforc-
whether access to postsession reinforcement influ- ers in children with limited vocal repertoires (e.g.,
enced responding. Results of the study showed that Yoon & Bennett, 2000; Esch, Carr, & Michael,
participants engaged in higher levels of responding 2005). Researchers have shown that simultaneous
when the researchers restricted postsession ac- presentation of less preferred with highly preferred
cess to the reinforcer. Hanley, Tiger, Ingvarsson, foods or condiments and highly preferred attention
and Cammilleri (2009) showed that they could or tangibles, in some cases, increases consumption
alter preschoolers’ free-play activity preferences or selection of the less preferred foods (Ahearn,
through satiation manipulations. Preschoolers re- 2003; Piazza et al., 2002; Solberg, Hanley, Layer,
allocated responding to less preferred but impor- & Ingvarsson, 2007). Furthermore, Hanley, Iwata,
tant activities, such as instructional zone, library, Roscoe, Thompson, and Lindberg (2003) pre-
and science, when the researchers used a satiation sented a highly preferred stimulus on a fixed-time
procedure with highly preferred activities. Zhou, schedule and a less preferred activity continuously,
186 B e h av i o r a l A s s e s s m e n t
and weaknesses relative to time efficiency and ac- get response. Stimulus variation may be beneficial
curacy. Single-stimulus assessments may be most under some circumstances. At a minimum, be-
appropriate for individuals who do not make havior analysts should assess preferences periodi-
choices or show approach responses when multiple cally to avoid satiation. They can also minimize
stimuli are available, although the disadvantage satiation by restricting access to reinforcers out-
is that some individuals may show high levels of side learning or treatment contexts. Furthermore,
approach responses to all or most stimuli. Paired- behavior analysta should assess the quality of po-
choice assessments may produce a more differen- tential competing reinforcers in the environment,
tiated hierarchy when individuals demonstrate to ensure that programmed reinforcement is of
choice-making behavior. The multiple-stimulus- higher quality than available alternative reinforce-
without-replacement, multiple-stimulus-with-re- ments. Finally, research has shown that choice in
placement, and free-operant procedures have the and of itself may function as reinforcement.
advantage of time efficiency. Concurrent-chains In summary, the existing literature offers many
procedures are useful for evaluating individuals’ procedures to identify effective positive reinforc-
preferences for procedures such as treatments. ers. Some questions that still arise include these:
Duration assessments may be useful for assessing
levels of inappropriate behavior in the presence 1. Which procedures are most effective for which
of specific stimuli (e.g., Piazza, Fisher, Hanley, individuals?
Hilker, & Derby, 1996; Ringdahl, Vollmer, Mar- 2. What variables or factors should a behavior
cus, & Roane, 1997). Response restriction com- analyst consider when choosing a preference
bines aspects of both trial- and free-operant-based assessment?
assessments and may produce well-differentiated 3. How does a behavior analyst control for vari-
preference hierarchies. Vocal assessments may be ous motivational variables when identifying
most appropriate for individuals who can vocalize preferred stimuli?
their preferences, although the correspondence 4. What is the best approach for progressing
between vocal and observed preferences remains when one or more preference assessments yield
questionable. Pictorial preferences may be an ac- inconclusive results?
curate measure for some individuals, but not oth-
5. Finally, how does a behavior analyst identify
ers. Most research has demonstrated that prefer-
potential negative reinforcers?
ence is a relatively good predictor of reinforcement
efficacy; however, preference assessments may un-
Answers to these questions, among others, will
derestimate the effectiveness of lower-preference
further refine our methods for identifying and en-
stimuli as reinforcement.
hancing positive reinforcers.
The most accurate method of reinforcer iden-
tification is to conduct an assessment to test the
effectiveness of preferred stimuli as reinforcement. REFERENCES
Researchers have used many methods to conduct
reinforcer assessments. Single-operant schedules Ackerlund Brandt, J. A., Dozier, C. L., Juanico, J. F.,
are well suited for evaluating the absolute reinforc- Laudont, C. L., & Mick, B. R. (2015). The value of
ing effects of a stimulus. By contrast, concurrent- choice as a reinforcer for typically developing chil-
operants schedules provide information about dren. Journal of Applied Behavior Analysis, 48, 344–
the relative reinforcing effects of a stimulus. Most 362.
studies have used simple, free-operant responses as Ahearn, W. H. (2003). Using simultaneous presentation
the target in reinforcement assessments for ease to increase vegetable consumption in a mildly selec-
and time efficiency, based on the rationale that tive child with autism. Journal of Applied Behavior
if a stimulus does not function as reinforcement Analysis, 36, 361–365.
American Psychiatric Association. (2013). Diagnostic
for a simple response, then it is not likely to func-
and statistical manual of mental disorders (5th ed.).
tion as reinforcement for a more complex one.
Arlington, VA: Author.
Researchers have used progressive-ratio schedules Bojak, S. L., & Carr, J. E. (1999). On the displacement
to evaluate relative reinforcer efficacy across many of leisure items by food during multiple-stimulus
contexts. preference assessments. Journal of Applied Behavior
Behavior analysts can use several methods to Analysis, 32, 515–518.
maximize the effectiveness of reinforcement, such Bowman, L. G., Piazza, C. C., Fisher, W. W., Hagopian,
as delivering reinforcers immediately after the tar- L. P., & Kogan, J. S. (1997). Assessment of preference
188 B e h av i o r a l A s s e s s m e n t
for varied versus constant reinforcers. Journal of Ap- ence assessment method using activities as stimuli.
plied Behavior Analysis, 30, 451–458. Journal of Applied Behavior Analysis, 42, 563–574.
Bukala, M., Hu, M. Y., Lee, R., Ward-Horner, J. C., & Deci, E. L. (1971). Effects of externally mediated re-
Fienup, D. M. (2015). The effects of work-reinforcer wards on intrinsic motivation. Journal of Personality
schedules on performance and preference in students and Social Psychology, 18, 105–115.
with autism. Journal of Applied Behavior Analysis, 48, DeLeon, I. G., Chase, J. A., Frank-Crawford, M., Car-
215–220. reau-Webster, A., Triggs, M. M., Bullock, C. E., et al.
Call, N. A., Trosclair-Lasserre, N. M., Findley, A. J., (2014). Distributed and accumulated reinforcement
Reavis, A. R., & Shillingsburg, M. A. (2012). Cor- arrangements: Evaluations of efficacy and preference.
respondence between single versus daily preference Journal of Applied Behavior Analysis, 47, 293–313.
assessment outcomes and reinforcer efficacy under DeLeon, I. G., Fisher, W. W., Catter, V. R., Maglieri, K.,
progressive-ratio schedules. Journal of Applied Behav- Herman, K., & Marhefka, J. (2001). Examination of
ior Analysis, 45, 763–777. relative reinforcement effects of stimuli identified
Cameron, J., Banko, K. M., & Pierce, W. D. (2001). Per- through pretreatment and daily brief preference as-
vasive negative effects of rewards on intrinsic moti- sessments. Journal of Applied Behavior Analysis, 34,
vation: The myth continues. Behavior Analyst, 24, 463–473.
1–44. DeLeon, I. G., & Iwata, B. A. (1996). Evaluation of a
Cameron, J., & Pierce, W. D. (1994). Reinforcement, multiple-stimulus presentation format for assessing
reward, and intrinsic motivation: A meta-analysis. reinforcer preferences. Journal of Applied Behavior
Review of Educational Research, 64, 363–423. Analysis, 29, 519–532.
Carr, J. E., Nicolson, A. C., & Higbee, T. S. (2000). DeLeon, I. G., Iwata, B. A., Conners, J., & Wallace, M.
Evaluation of a brief multiple-stimulus preference as- D. (1999). Examination of ambiguous stimulus pref-
sessment in a naturalistic context. Journal of Applied erences with duration-based measures. Journal of Ap-
Behavior Analysis, 33, 353–357. plied Behavior Analysis, 32, 111–114.
Catania, A. C. (1963). Concurrent performances: A DeLeon, I. G., Iwata, B. A., Goh, H. L., & Worsdell,
baseline for the study of reinforcement magnitude. A. S. (1997). Emergence of reinforcer preference as
Journal of Experimental Animal Behavior, 6, 299–300. a function of schedule requirements and stimulus
Ciccone, F. J., Graff, R. B., & Ahearn, W. H. (2015). In- similarity. Journal of Applied Behavior Analysis, 30,
creasing the efficiency of paired-stimulus preference 439–449.
assessments by identifying categories of preference. DeLeon, I. G., Iwata, B. A., & Roscoe, E. M. (1997).
Journal of Applied Behavior Analysis, 48, 221–226. Displacement of leisure reinforcers by food during
Clay, C. J., Samaha, A. L., Bloom, S. E., Bogoev, B. K., preference assessments. Journal of Applied Behavior
& Boyle, M. A. (2013). Assessing preference for so- Analysis, 30, 475–484.
cial interactions. Research in Developmental Disabili- Dunlap, G., dePerczel, M., Clarke, S., Wilson, D.,
ties, 34, 362–371. Wright, S., White, R., & Gomez, A. (1994). Choice
Clements, C., & McKee, J. (1968). Programmed instruc- making to promote adaptive behavior for students
tion for institutionalized offenders: Contingency with emotional and behavioral challenges. Journal of
management and performance contracts. Psychologi- Applied Behavior Analysis, 27, 505–518.
cal Reports, 22, 957–964. Dyer, K., Dunlap, G., & Winterling, V. (1990). Effects
Conyers, C., Doole, A., Vause, T., Harapiak, S., Yu, D. of choice making on the serious problem behaviors
C. T., & Martin, G. L. (2002). Predicting the relative of students with severe handicaps. Journal of Applied
efficacy of three presentation methods for assessing Behavior Analysis, 23, 515–524.
preferences of persons with developmental disabili- Egel, A. L. (1980). The effects of constant vs. varied
ties. Journal of Applied Behavior Analysis, 35, 49–58. reinforcer presentation on responding by autistic
Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). children. Journal of Experimental Child Psychology,
Applied behavior analysis (2nd ed.) Upper Saddle 30, 455–462.
River, NJ: Pearson/Merrill-Prentice Hall. Egel, A. L. (1981). Reinforcer variation: Implications for
Cote, C. A., Thompson, R. H., Hanley, G. P., & McK- motivating developmentally disabled children. Jour-
erchar, P. M. (2007). Teacher report and direct as- nal of Applied Behavior Analysis, 14, 345–350.
sessment of preferences for identifying reinforcers for Esch, B. E., Carr, J. E., & Michael, J. (2005). Evaluating
young children. Journal of Applied Behavior Analysis, stimulus-stimulus pairing and direct reinforcement in
40, 157–166. the establishment of an echoic repertoire of children
Daley, M. F. (1969). The “reinforcement menu”: Find- diagnosed with autism. Analysis of Verbal Behavior,
ing effective reinforcers. In J. D. Krumboltz & C. E. 21, 43–58.
Thorsen (Eds.), Behavioral counseling: Cases and tech- Fahmie, T. A., Iwata, B. A., & Jann, K. E. (2015). Com-
niques. New York: Holt, Rinehart & Winston. parison of edible and leisure reinforcers. Journal of
Daly, E. J., III, Wells, N. J., Swanger-Gagné, M. S., Carr, Applied Behavior Analysis, 48, 331–343.
J. E., Kunz, G. M., & Taylor, A. M. (2009). Evaluation Fantuzzo, J. W., Rohrbeck, C. A., Hightower, A. D.,
of the multiple-stimulus without replacement prefer- & Work, W. C. (1991). Teachers’ use and children’s
The Effectiveness of Positive Reinforcement 189
preferences of rewards in elementary school. Psychol- son, E. T. (2007). A method for describing preschool-
ogy in the Schools, 28, 175–181. ers’ activity preferences. Journal of Applied Behavior
Favell, J. E., & Cannon, P. R. (1976). Evaluation of en- Analysis, 40, 603–618.
tertainment materials for severely retarded persons. Hanley, G. P., Iwata, B. A., & Lindberg, J. S. (1999).
American Journal of Mental Deficiency, 81, 357–361. Analysis of activity preferences as a function of dif-
Fienup, D. M., Ahlers, A. A., & Pace, G. (2011). Prefer- ferential consequences. Journal of Applied Behavior
ence for fluent versus disfluent work schedules. Jour- Analysis, 32(4), 419–435.
nal of Applied Behavior Analysis, 44, 847–858. Hanley, G. P., Iwata, B. A., Lindberg, J. S., & Conners, J.
Fisher, W. W., & Mazur, J. E. (1997). Basic and applied (2003). Response-restriction analysis: I. Assessment
research on choice responding. Journal of Applied Be- of activity preferences. Journal of Applied Behavior
havior Analysis, 30, 387–410. Analysis, 36, 47–58.
Fisher, W. W., Piazza, C. C., Bowman, L. G., & Amari, Hanley, G. P., Iwata, B. A., Roscoe, E. M., Thompson,
A. (1996). Integrating caregiver report with system- R. H., & Lindberg, J. S. (2003). Response-restriction
atic choice assessment to enhance American Journal analysis: II. Alteration of activity preferences. Journal
on Mental Retardation, 101, 15–25. of Applied Behavior Analysis, 36, 59–76.
Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. Hanley, G. P., Iwata, B. A., & Roscoe, E. M. (2006).
P., Owens, J. C., & Slevin, I. (1992). A comparison of Some determinants of changes in preference over
two approaches for identifying reinforcers for persons time. Journal of Applied Behavior Analysis, 39, 189–
with severe and profound disabilities. Journal of Ap- 202.
plied Behavior Analysis, 25, 491–498. Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci,
Fisher, W. W., Thompson, R. H., Piazza, C. C., Crosland, S. A., & Maglieri, K. A. (1997). Evaluation of client
K., & Gotjen, D. (1997). On the relative reinforcing preference for function-based treatment packages.
effects of choice and differential consequences. Jour- Journal of Applied Behavior Analysis, 30, 459–473.
nal of Applied Behavior Analysis, 30, 423–438. Hanley, G. P., Piazza, C. C., Fisher, W. W., & Maglieri,
Fox, R. A., & DeShaw, J. M. (1993). Milestone reinforc- K. A. (2005). On the effectiveness of and preference
er survey. Education and Training in Mental Retarda- for punishment and extinction components of func-
tion, 28, 257–261. tion-based interventions. Journal of Applied Behavior
Francisco, M. T., Borrero, J. C., & Sy, J. R. (2008). Eval- Analysis, 38, 51–65.
uation of absolute and relative reinforcer value using Hanley, G. P., Tiger, J. H., Ingvarsson, E. T., & Cammil-
progressive-ratio schedules. Journal of Applied Behav- leri, A. P. (2009). Influencing preschoolers’ free-play
ior Analysis, 41, 189–202. activity preferences: An evaluation of satiation and
Giles, A. F., St. Peter, C. C., Pence, S. T., & Gibson, embedded reinforcement. Journal of Applied Behavior
A. B. (2012). Preference for blocking or response Analysis, 42, 33–41.
redirection during stereotypy treatment. Research in Heal, N. A., & Hanley, G. P. (2007). Evaluating pre-
Developmental Disabilities, 33, 1691–1700. school children’s preferences for motivational sys-
Glover, A. C., Roane, H. S., Kadey, H. J., & Grow, L. tems during instruction. Journal of Applied Behavior
L. (2008). Preference for reinforcers under progres- Analysis, 40, 249–261.
sive- and fixed-ratio schedules: A comparison of sin- Higbee, T. S., Carr, J. E., & Harrison, C. D. (1999). The
gle- and concurrent-operant arrangements. Journal of effects of pictorial versus tangible stimuli in stimulus-
Applied Behavior Analysis, 41, 163–176. preference assessments. Research in Developmental
Graff, R. B., & Larsen, J. (2011). The relation between Disabilities, 20, 63–72.
obtained preference value and reinforcer potency. Hodos, W. (1961). Progressive-ratio as a measure of re-
Behavioral Interventions, 26, 125–133. ward strength. Science, 134, 943–944.
Graff, R. B., & Libby, M. E. (1999). A comparison of Hoffmann, A. N., Samaha, A. L., Bloom, S. E., & Boyle,
presession and within-session reinforcement choice. M. A. (2017). Preference and reinforcer efficacy of
Journal of Applied Behavior Analysis, 32, 161–173. high- and low-tech items: A comparison of item type
Green, C. W., Reid, D. H., White, L. K., Halford, R. C., and duration of access. Journal of Applied Behavior
Brittain, D. P., & Gardner, S. M. (1988). Identify- Analysis, 50, 222–237.
ing reinforcers for persons with profound handicaps: Jerome, J., & Sturmey, P. (2008). Reinforcing efficacy
Staff opinion versus systematic assessment of prefer- of interactions with preferred and non-preferred staff
ences. Journal of Applied Behavior Analysis, 21, 31–43. under progressive-ratio schedules. Journal of Applied
Greene, D., & Lepper, M. R. (1974). Effects of extrinsic Behavior Analysis, 41, 221–225.
rewards on children’s subsequent intrinsic interest. Kang, S., Lang, R. B., O’Reilly, M. F., Davis, T. N.,
Child Development, 45, 1141–1145. Machalicek, W., Rispoli, M. J., et al. (2010). Problem
Hagopian, L. P., Rush, K. S., Lewin, A. B., & Long, E. S. behavior during preference assessments: An empiri-
(2001). Evaluating the predictive validity of a single cal analysis and practical recommendations. Journal
stimulus engagement preference assessment. Journal of Applied Behavior Analysis, 43, 137–141.
of Applied Behavior Analysis, 34, 475–485. Kang, S., O’Reilly, M. F., Fragale, C., Aguilar, J., Risp-
Hanley, G. P., Cammilleri, A. P., Tiger, J. H., & Ingvars- oli, M., & Lang, R. (2011). Evaluation of the rate of
190 B e h av i o r a l A s s e s s m e n t
problem behavior maintained by different reinforcers ing reinforcer assessment. Journal of Applied Behavior
across preference assessments. Journal of Applied Be- Analysis, 22, 171–179.
havior Analysis, 44, 835–846. Mazaleski, J. L., Iwata, B. A., Vollmer, T. R., Zarcone,
Karsten, A. M., Carr, J. E., & Lepper, T. L. (2011). De- J. R., & Smith, R. G. (1993). Analysis of the rein-
scription of practitioner model for identifying pre- forcement and extinction components in DRO con-
ferred stimuli with individuals with autism spectrum tingencies with self-injury. Journal of Applied Behavior
disorders. Behavior Modification, 35, 347–369. Analysis, 26, 143–156.
Kelley, M. E., Shillingsburg, M. A., & Bowen, C. N. Miguel, C. F., Carr, J. E., & Michael, J. (2002). The ef-
(2016). Stability of daily preference across multiple fects of a stimulus–stimulus pairing procedure on the
individuals. Journal of Applied Behavior Analysis, 49, vocal behavior of children diagnosed with autism.
394–398. Analysis of Verbal Behavior, 18, 3–13.
Keyl-Austin, A. A., Samaha, A. L., Bloom, S. E., & Morris, S. L., & Vollmer, T. R. (2019). Assessing prefer-
Boyle, M. A. (2012). Effects of preference and re- ence for types of social interaction. Journal of Applied
inforcer variation on within-session patterns of re- Behavior Analysis, 52, 1064–1075.
sponding. Journal of Applied Behavior Analysis, 45, Neef, N. A., Mace, F. C., & Shade, D. (1993). Impulsiv-
637–641. ity in students with serious emotional disturbance:
Kodak, T., Lerman, D. C., & Call, N. (2007). Evaluating The interactive effects of reinforcer rate, delay, and
the influence of postsession reinforcement on choice quality. Journal of Applied Behavior Analysis, 26,
of reinforcers. Journal of Applied Behavior Analysis, 37–52.
40, 515–527. Neef, N. A., Mace, F. C., Shea, M. C., & Shade, D.
Layer, S. A., Hanley, G. P., Heal, N. A., & Tiger, J. H. (1992). Effects of reinforcer rate and reinforcer quali-
(2008). Determining individual preschoolers’ prefer- ty on time allocation: Extensions of matching theory
ences in a group arrangement. Journal of Applied Be- to educational settings. Journal of Applied Behavior
havior Analysis, 41, 25–37. Analysis, 25, 691–699.
Lee, M. S. H., Yu, C. T., Martin, T. L., & Martin, G. L. Neef, N. A., Shade, D., & Miller, M. S. (1994). Assess-
(2010). On the relation between reinforcer efficacy ing influential dimensions of reinforcers on choice in
and preference. Journal of Applied Behavior Analysis, students with serious emotional disturbance. Journal
43, 95–100. of Applied Behavior Analysis, 27, 575–583.
Levy, A., DeLeon, I. G., Martinez, C. K., Fernandez, Northup, J., Fusilier, I., Swanson, V., Roane, H., & Bor-
N., Gage, N. A., Sigurdsson, S. Ó., et al. (2017). A rero, J. (1997). An evaluation of methylphenidate as
quantitative review of overjustification effects in per- a potential establishing operation for some common
sons with intellectual and developmental disabilities. classroom reinforcers. Journal of Applied Behavior
Journal of Applied Behavior Analysis, 50, 206–221. Analysis, 30, 615–625.
Luczynski, K. C., & Hanley, G. P. (2009). Do children Northup, J., George, T., Jones, K., Broussard, C., &
prefer contingencies?: An evaluation of the efficacy Vollmer, T. R. (1996). A comparison of reinforcer as-
of and preference for contingent versus noncontin- sessment methods: The utility of verbal and pictorial
gent social reinforcement during play. Journal of Ap- choice procedures. Journal of Applied Behavior Analy-
plied Behavior Analysis, 42, 511–525. sis, 29, 201–212.
Luczynski, K. C., & Hanley, G. P. (2010). Examining Nuernberger, J. E., Smith, C. A., Czapar, K. N., & Klatt,
the generality of children’s preference for contingent K. P. (2012). Assessing preference for social interac-
reinforcement via extension to different responses, tion in children diagnosed with autism. Behavioral
reinforcers, and schedules. Journal of Applied Behavior Interventions, 27, 33–44.
Analysis, 43, 397–409. Ortega, J. V., Iwata, B. A., Nogales-González, C., &
Luczynski, K. C., & Hanley, G. P. (2014). How should Frades, B. (2012). Assessment of preference for edible
periods without social interaction be scheduled?: and leisure items in individuals with dementia. Jour-
Children’s preference for practical schedules of posi- nal of Applied Behavior Analysis, 45, 839–844.
tive reinforcement. Journal of Applied Behavior Analy- Pace, G. M., Ivancic, M. T., Edwards, G. L., Iwata, B. A.,
sis, 47, 500–522. & Page, T. J. (1985). Assessment of stimulus prefer-
Mace, F. C., Neef, N. A., Shade, D., & Mauro, B. C. ence and reinforcer value with profoundly retarded
(1994). Limited matching on concurrent schedule re- individuals. Journal of Applied Behavior Analysis, 18,
inforcement of academic behavior. Journal of Applied 249–255.
Behavior Analysis, 27, 585–596. Paclawskyj, T. R., & Vollmer, T. R. (1995). Reinforcer
Madden, G. J., & Bickel, W. K. (2010). Impulsivity: assessment for children with developmental disabili-
The behavioral and neurological science of discounting. ties and visual impairments. Journal of Applied Behav-
Washington, DC: American Psychological Associa- ior Analysis, 28, 219–224.
tion. Paramore, N. W., & Higbee, T. S. (2005). An evalua-
Mason, S. A., McGee, G. G., Farmer-Dougan, V., & tion of a brief multiple-stimulus preference assess-
Risley, T. R. (1989). A practical strategy for ongo- ment with adolescents with emotional–behavioral
The Effectiveness of Positive Reinforcement 191
disorders in an educational setting. Journal of Applied tions for preference assessments. Journal of Applied
Behavior Analysis, 38, 399–403. Behavior Analysis, 32, 479–493.
Parsons, M. B., & Reid, D. H. (1990). Assessing food Rotatori, A. F., Fox, B., & Switzky, H. (1979). An in-
preferences among persons with profound mental re- direct technique for establishing preferences for cat-
tardation: Providing opportunities to make choices. egories of reinforcement for severely and profoundly
Journal of Applied Behavior Analysis, 23, 183–195. retarded individuals. Perceptual and Motor Skills, 48,
Penrod, B., Wallace, M. D., & Dyer, E. J. (2008). Assess- 1307–1313.
ing reinforcer potency of high-preference and low- Schmidt, A. C., Hanley, G. P., & Layer, S. A. (2009).
preference reinforcers with respect to response rate A further analysis of the value of choice: Control-
and response patterns. Journal of Applied Behavior ling for illusory discriminative stimuli and evaluating
Analysis, 41, 177–188. the effects of less preferred items. Journal of Applied
Piazza, C. C., Fisher, W. W., Hagopian, L. P., Bowman, Behavior Analysis, 42, 711–716.
L. G., & Toole, L. (1996). Using a choice assessment Shiffman, S. S. (1997). Taste and smell losses in normal
to predict reinforcer effectiveness. Journal of Applied aging and disease. Journal of the American Medical
Behavior Analysis, 29, 1–9. Association, 278, 1357–1362.
Piazza, C. C., Fisher, W. W., Hanley, G. P., Hilker, K., Smith, C. J., Roane, H. S., & Stephenson, K. (2009,
& Derby, K. M. (1996). A preliminary procedure May). Evaluation of functional and alternative rein-
for predicting the positive and negative effects of forcers under progressive ratio schedules of reinforce-
reinforcement-based procedures. Journal of Applied ment. In R. G. Smith (Chair), Advances in preference
Behavior Analysis, 29, 137–152. and choice research across multiple applied contexts.
Piazza, C. C., Patel, M. R., Santana, C. M., Goh, H., Symposium conducted at the 35th annual conven-
Delia, M. D., & Lancaster, B. M. (2002). An evalu- tion of the Association for Behavior Analysis, Phoe-
ation of simultaneous and sequential presentation nix, AZ.
of preferred and nonpreferred food to treat food se- Solberg, K. M., Hanley, G. P., Layer, S. A., & Ingvars-
lectivity. Journal of Applied Behavior Analysis, 35, son, E. T. (2007). The effects of reinforcer pairing
259–270. and fading on preschoolers’ snack selection. Journal
Poling, A. (2010). Progressive-ratio schedules and ap- of Applied Behavior Analysis, 40, 633–644.
plied behavior analysis. Journal of Applied Behavior Steinhilber, J., & Johnson, C. (2007). The effects of brief
Analysis, 43, 347–349. and extended stimulus availability on preference.
Powell, S., & Nelson, B. (1997). Effects of choosing aca- Journal of Applied Behavior Analysis, 40, 767–772.
demic assignments on a student with attention defi- Sy, J. R., & Borrero, J. C. (2009). Parametric analysis of
cit hyperactivity disorder. Journal of Applied Behavior presession exposure to edible and nonedible stimuli.
Analysis, 30, 181–183. Journal of Applied Behavior Analysis, 42, 833–837.
Radley, K. C., Dart, E. H., Battaglia, A. A., & Ford, W. Taravella, C. C., Lerman, D. C., Contrucci, S. A., &
B. (2019). A comparison of two procedures for assess- Roane, H. S. (2000). Further evaluation of low-
ing preference in a classroom setting. Behavior Analy- ranked items in stimulus-choice preference assess-
sis in Practice, 12, 95–104. ments. Journal of Applied Behavior Analysis, 33,
Ringdahl, J. E., Vollmer, T. R., Marcus, B. A., & Roane, 105–108.
H. S. (1997). An analogue evaluation of enriched en- Tasky, K. K., Rudrud, E. H., Schulze, K. A., & Rapp, J.
vironment: The role of stimulus preference. Journal of T. (2008). Using choice to increase on-task behavior
Applied Behavior Analysis, 30, 203–216. in individuals with traumatic brain injury. Journal of
Roane, H. S. (2008). On the applied use of progressive- Applied Behavior Analysis, 41, 261–265.
ratio schedules of reinforcement. Journal of Applied Thompson, R. H., Fisher, W. W., & Contrucci, S. A.
Behavior Analysis, 41, 155–161. (1998). Evaluating the reinforcing effects of choice
Roane, H. S., Call, N. A., & Falcomata, T. S. (2005). in comparison to reinforcement rate. Research in De-
A preliminary analysis of adaptive responding under velopmental Disabilities, 19, 181–187.
open and closed economies. Journal of Applied Behav- Tiger, J. H., Hanley, G. P., & Hernandez, E. (2006).
ior Analysis, 38, 335–348. An evaluation of the value of choice with preschool
Roane, H. S., Lerman, D. C., & Vorndran, C. M. (2001). children. Journal of Applied Behavior Analysis, 39,
Assessing reinforcers under progressive schedule re- 1–16.
quirements. Journal of Applied Behavior Analysis, 34, Tourigny Dewhurst, D. L., & Cautela, J. R. (1980). A
145–167. proposed reinforcement survey schedule for special
Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, needs children. Journal of Behavior Therapy and Ex-
B. A. (1998). Evaluation of a brief stimulus preference perimental Psychiatry, 11, 109–112.
assessment. Journal of Applied Behavior Analysis, 31, Toussaint, K. A., Kodak, T., & Vladescu, J. C. (2016).
605–620. An evaluation of choice on instructional efficacy and
Roscoe, E. M., Iwata, B. A., & Kahng, S. (1999). Rela- individual preferences among children with autism.
tive versus absolute reinforcement effects: Implica- Journal of Applied Behavior Analysis, 49, 170–175.
192 B e h av i o r a l A s s e s s m e n t
Trosclair-Lasserre, N. M., Lerman, D. C., Call, N. A., Windsor, J., Piché, L. M., & Locke, P. A. (1994). Pref-
Addison, L. R., & Kodak, T. (2008). Reinforcement erence testing: A comparison of two presentation
magnitude: An evaluation of preference and rein- methods. Research in Developmental Disabilities, 15,
forcer efficacy. Journal of Applied Behavior Analysis, 439–455.
41, 203–220. Wine, B., & Wilder, D. A. (2009). The effects of varied
Tung, S. B., Donaldson, J. M., & Kahng, S. (2017). The versus constant high-, medium-, and low-preference
effects of preference assessment type on problem stimuli on performance. Journal of Applied Behavior
behavior. Journal of Applied Behavior Analysis, 50, Analysis, 42, 321–326.
861–866. Yoon, S., & Bennett, G. M. (2000). Effects of a stimu-
Verriden, A. L., & Roscoe, E. M. (2016). A comparison lus-stimulus pairing procedure on conditioning vocal
of preference-assessment methods. Journal of Applied sounds as reinforcers. Analysis of Verbal Behavior, 17,
Behavior Analysis, 49, 265–285. 75–88.
Vollmer, T. R., Ringdahl, J. E., Roane, H. S., & Marcus, Zhou, L., Iwata, B. A., Goff, G. A., & Shore, B. A.
B. A. (1997). Negative side effects of noncontingent (2001). Longitudinal analysis of leisure-item prefer-
reinforcement. Journal of Applied Behavior Analysis, ences. Journal of Applied Behavior Analysis, 34, 179–
30, 161–164. 184.
Wehman, P. (1976). Selection of play materials for Zhou, L., Iwata, B. A., & Shore, B. A. (2002). Reinforc-
the severely handicapped: A continuing dilemma. ing efficacy of food on performance during pre- and
Education and Training of the Mentally Retarded, 11, postmeal sessions. Journal of Applied Behavior Analy-
46–50. sis, 35, 411–414.
CHAP TER 11
193
194 B e h av i o r a l A s s e s s m e n t
assessment often does not include direct observa- “My child engages in problem behavior at levels
tion of a client or the target behaviors. Thus the far above that which would be normally accepted
assessor depends on caregiver recollections of the in a school.” In this hypothetical rating scale, the
frequency of a target behavior, the settings and levels of agreement range from 1 to 5. Thus the
conditions in which the target behavior is more informant would indicate his or her level of agree-
and less likely to occur, the consequences that ment with this statement as follows: 1—strongly
typically follow the target behavior, and so forth. disagree, 2—somewhat disagree, 3—neither agree
Indirect functional assessment has several other nor disagree, 4—somewhat agree, and 5—strongly
disadvantages. First, investigators have gener- agree. Rating scales often include methods for
ally not evaluated the psychometric properties of summarizing the results, such as adding the scores
some indirect assessment measures. In fact, re- for each question to produce a total score.
sults of some studies have shown that some tools Rating scales are useful because they provide
for indirect functional assessment lack acceptable quantifiable information (Hosp, Howell, & Hosp,
levels of reliability (the stability of measurement 2003). Some rating scales assign numerical values
over time or across stimulus parameters), validity to the informant’s ratings, and behavior analysts
(whether an assessment measures what investiga- can add those values and analyze the results based
tors designed it to measure), or both. For example, on the rating scales’ instructions. An advantage
investigators have reported both high and low is that investigators can assess quantified data for
levels of reliability for the Motivation Assessment reliability across time, between raters, and within
Scale (Durand & Crimmins, 1988), depending on the scale. Investigators also can assess validity by
how investigators calculated reliability and the comparing one scale’s results to those of scales
topography of the assessed target behavior (Bihm, with known psychometric properties that purport
Kienlen, Ness, & Poindexter, 1991; Iwata, Vollmer, to measure the same construct (convergent valid-
& Zarcone, 1990; Paclawskyj, Matson, Rush, ity) or a different construct (discriminant validity).
Smalls, & Vollmer, 2001; Sigafoos, Kerr, & Rob- Second, behavior analysts can compare a rating
erts, 1994; Singh et al., 1993). Saini, Ubdegrove, scale’s results across time, raters, or settings to as-
Biran, and Duncan (2019) showed that that both sess changes in the occurrence of target behavior
the interrater reliability and concurrent validity across those dimensions. Third, rating scales may
of an open-ended functional-analytic interview be free of some biases that affect other measures,
were relatively low. Therapists typically ask par- because a rating scale’s quantifiable responses may
ents and caregivers to provide information about be less open to interpretation.
target individuals during indirect assessments. Rating scales have some noteworthy disadvan-
However, Dracobly, Dozier, Briggs, and Juanico tages that warrant consideration (Hanley, 2010,
(2018) showed that experts were more likely than 2012; Iwata, DeLeon, & Roscoe, 2013). Although
caregivers to agree with each other (interrater reli- generating quantifiable outcomes to compare
ability) and to agree with other objective measures across raters and time may seem appealing, such
(validity). Nevertheless, behavior analysts report assessment often produces unreliable results. That
frequent use of rating scales and structured inter- is, two independent raters may produce different
views to assess the function of problem behavior, outcomes when rating the same individual (New-
probably because they are easy to use (Fryling & ton & Sturmey, 1991; Nicholson, Konstantinidi,
Baires, 2016). In the remainder of the chapter, we & Furniss, 2006; Shogren & Rojahn, 2003; Zar-
provide descriptions and analyses of several of the cone, Rodgers, Iwata, Rourke, & Dorsey, 1991).
most common rating scales and interviews for in- Below, we review some rating scales that behav-
direct behavioral assessment. ior analysts often use to aid in identifying the
reinforcer(s) for problem behavior.
typical episode of problem behavior, using this al. were comparable to those from other studies in
Likert-scale format: Never, 10% of the time, 25% which investigators assessed agreement for a func-
of the time, 50% of the time, 75% of the time, 90% tional behavior assessment rating scale (Conroy,
of the time, or Always. The questionnaire includes Fox, Bucklin, & Good, 1996; Duker & Sigafoos,
a profile that the interviewer completes, based 1998; Sigafoos et al., 1994; Zarcone et al., 1991; see
on the informant’s ratings for problem behavior. Iwata et al., 2013, for a direct comparison of the
The profile includes three major categories: peers, agreement scores across studies).
adults, and settings. Peers and adults each have In Study 2, Iwata et al. (2013) evaluated the
two subcategories: attention and escape. Thus the Functional Analysis Screening Tool’s validity by
questionnaire identifies five potential functions of assessing whether its results predicted the condi-
problem behavior. Lewis et al. (1994) provide spe- tion with the highest rate of problem behavior in
cific scoring criteria for identifying the function each of the analogue functional analyses of 59 in-
of the student’s problem behavior. One limitation dividuals. The investigators determined whether
of the Problem Behavior Questionnaire is that the results of the analogue functional analysis
investigators have not assessed its reliability and matched the results of the Functional Analysis
validity. Thus behavior analysts should not use Screening Tool for both informants, one infor-
the Problem Behavior Questionnaire without ad- mant, or neither informant. The aggregated results
ditional functional assessment. showed the function that one or both informants
identified with the scale matched the function the
analogue functional analysis identified in 64% of
Functional Analysis Screening Tool
cases. Thus the Functional Analysis Screening
The Functional Analysis Screening Tool is a Tool lacked predictive validity for functional-an-
rating scale that provides preliminary informa- alytic outcomes, likely due to its moderate reliabil-
tion about variables that may influence problem ity (Iwata et al., 2013).
behavior (Iwata & DeLeon, 1996). It has several Although the Functional Analysis Screening
open-ended questions about the informant; his Tool was found not to be adequate as a prescriptive
or her relation to the target individual; and the tool, the investigators described ways that it might
topography, frequency, severity, and context for improve the treatment development process. First,
occurrences and nonoccurrences of problem be- it might improve the efficiency and consistency
havior. Sixteen additional questions assess ante- of interviews, because interviewers ask the same
cedents and consequences that might affect prob- questions in every interview. Second, results of
lem behavior, which the informant answers in a the Functional Analysis Screening Tool might
yes–no format. For example, one question is “Does provide information about idiosyncratic behaviors
the person usually engage in the problem behavior or conditions that might not otherwise be appar-
more often when he or she is ill?” The question- ent. Third, behavior analysts might be able to use
naire assesses antecedents and consequences for rater concordance on the Functional Analysis
social positive and social negative reinforcement Screening Tool to rule out reinforcers for problem
and automatic positive and automatic negative re- behavior; this would permit them to conduct fewer
inforcement, and the category with the most yes conditions in the analogue functional analysis,
responses indicates a potential reinforcer for prob- which could potentially reduce its length. Overall,
lem behavior. Iwata and DeLeon (1996) developed the results of Iwata et al. (2013) are consistent with
the Functional Analysis Screening Tool to obtain the existing literature, which suggests that behav-
information from caregivers about potential ante- ior analysts should use the Functional Analysis
cedents and consequences for problem behavior Screening Tool as a supplement to experimental
that could inform an experimental functional functional analyses.
analysis (Iwata et al., 2013).
Iwata et al. (2013) assessed the reliability and
Motivation Assessment Scale
validity of the Functional Analysis Screening Tool
in a two-study experiment. In Study 1, investiga- The Motivation Assessment Scale (Durand &
tors administered the scale to pairs of raters for Crimmins, 1988) has 16 questions about potential
196 problem behaviors. Mean item-to-item agree- antecedent and consequent events for problem be-
ment was 72% (overall range, 29–100%; range for havior. Informants rate how often the individual
individual items, 53–85%; M for highest totals = engages in the behavior in response to each ques-
69%). The agreement coefficients from Iwata et tion on a Likert scale from 1 (never) to 6 (always).
196 B e h av i o r a l A s s e s s m e n t
The assessor transfers the numerical scores for ity. Overall, results of these studies suggest that
each question to a scoring guide organized by the the reliability of the Motivation Assessment Scale
motivation types of attention, escape, sensory, and may not be as high as previously reported. For this
tangible. The highest mean score among the four reason, behavior analysts should not replace direct
motivation types indicates the potential reinforc- observation or experimental manipulation with
er for problem behavior. Durand and Crimmins the Motivation Assessment Scale to determine
(1988, 1992) reported that a reliability assessment the function of problem behavior, to develop treat-
produced acceptable interrater reliability (agree- ment, or both.
ment between raters) and test–retest reliability
(agreement on separate occasions). The validity
Questions about Behavioral Function
assessment compared Motivation Assessment
Scale results with direct observation and indicated The Questions about Behavioral Function scale
that the scale predicted situations in which prob- assesses a broader range of potential variables
lem behavior was more and less likely to occur. than the rating scales we have reviewed thus far. It
Studies by other investigators, however, have includes questions about commonly assessed vari-
produced different results (Bihm et al., 1991; Iwata ables, social attention, escape, tangible reinforce-
et al., 1990; Paclawskyj, Matson, Rush, Smalls, & ment, and nonsocial reinforcement, as well as un-
Vollmer, 2001; Sigafoos et al., 1994; Singh et al., derinvestigated variables such as social avoidance
1993). For example, Zarcone et al. (1991) evalu- and physical discomfort.
ated the reliability of the Motivation Assessment Paclawskyj, Matson, Rush, Smalls, and Vollmer
Scale with one group of individuals with intellec- (2000) evaluated the test–retest, interrater, and
tual developmental disorder in an institution and internal-consistency reliability for Questions
a second group in a school. Informants for each about Behavioral Function. Results suggested that
individual were two staff members who worked assessment data were relatively stable over time,
with the individual regularly in the institution that multiple raters produced similar ratings, and
or school, respectively. Zarcone et al. calculated that subscales were homogeneous. Simo-Pinatella
the reliability coefficients Durand and Crimmins et al. (2013) assessed the psychometric properties
(1988) reported, to directly compare the results of Questions about Behavioral Function in Span-
from the two studies. The coefficients included (1) ish and found that it had good test–retest reli-
a Pearson product–moment correlation that com- ability. Matson, Bamburg, Cherry, and Paclawskyj
pared the two raters’ raw scores across items on the (1999) used Questions about Behavioral Function
questionnaire, (2) a Pearson product–moment cor- to identify the function of participants’ problem
relation that compared the two raters’ mean scores behavior. The scale identified a clear function—
for the four motivation categories, and (3) a Spear- defined as a minimum score of four of five pos-
man rank-order correlation for each pair of raters sible endorsements on a subscale, with no other
on their rank orders for each motivation category. subscales containing significant endorsements
Zarcone et al. also calculated exact agreement (see Matson & Vollmer, 1995)—for 84% of par-
(whether the two raters selected the same response ticipants. Next, the investigators assessed whether
on each item of the Motivation Assessment Scale) they could prescribe effective treatment for partici-
and agreement for adjacent scores (whether one pants whose Questions about Behavioral Function
rater’s score for an item was plus or minus 1 from results identified a function for problem behavior.
the other rater’s score on the item). The five reli- Treatments based on these results were more effec-
ability methods produced coefficients of .27, .41, tive for decreasing occurrences of problem behav-
.41, .20, and .40, respectively. ior than those not based on these results. By con-
In another study, the reliability coefficients for trast, Questions about Behavioral Function results
the Motivation Assessment Scale when aggression matched results of an analogue functional analysis
was the target behavior were lower (Sigafoos et al., in 56% of cases in Paclawskyj et al. (2001).
1994) than those Durand and Crimmins (1988) Although investigators have evaluated Ques-
reported for self-injurious behavior. Finally, Duker tions about Behavioral Function more extensively
and Sigafoos (1998) evaluated the reliability, in- than most other indirect functional behavioral
ternal consistency, and construct validity of the assessments, additional research is warranted. Re-
Motivation Assessment Scale across three topog- sults thus far are mixed about whether the scale
raphies of problem behavior with two calculation accurately identifies the function of problem be-
methods. Results suggested that the scale had low havior and whether behavior analysts can use it to
levels of reliability and ambiguous construct valid- develop effective function-based treatment.
Indirect Behavioral Assessments 197
with severe and profound disabilities. Journal of Ap- ing a functional assessment of problem behavior in
plied Behavior Analysis, 25, 491–498. applied settings. Journal of the Association for Persons
Fryling, M. J., & Baires, N. A. (2016). The practical with Severe Handicaps, 14, 304–311.
importance of the distinction between open and Lewis, T. J., Scott, T. M., & Sugai, G. (1994). The Prob-
closed-ended indirect assessments. Behavior Analysis lem Behavior Questionnaire: A teacher-based instru-
in Practice, 9, 146–151. ment to develop functional hypotheses of problem
Green, G., & Striefel, S. (1988). Response restriction behavior in general education classrooms. Diagnos-
and substitution with autistic children. Journal of the tique, 19, 103–115.
Experimental Analysis of Behavior, 50, 21–32. Lloyd, B. P., & Kennedy, C. H. (2014). Assessment and
Hanley, G. P. (2010). Prevention and treatment of se- treatment of challenging behaviour for individuals
vere problem behavior. In E. Mayville & J. Mulick with intellectual disability: A research review. Jour-
(Eds.), Behavioral foundations of autism intervention nal of Applied Research in Intellectual Disabilities, 27,
(pp. 233–256). New York: Sloman. 187–199.
Hanley, G. P. (2012). Functional assessment of problem Mace, F. C., Neef, N. A., Shade, D., & Mauro, B. C.
behavior: Dispelling myths, overcoming implemen- (1996). Effects of problem difficulty and reinforcer
tation obstacles, and developing new lore. Behavior quality on time allocated to concurrent arithmetic
Analysis in Practice, 5, 54–72. problems. Journal of Applied Behavior Analysis, 29,
Horner, R. H., & Day, H. M. (1991). The effects of re- 11–24.
sponse efficiency on functionally equivalent compet- Matson, J. L., Bamburg, J. W., Cherry, K. E., & Pa-
ing behaviors. Journal of Applied Behavior Analysis, clawskyj, T. R. (1999). A validity study on the Ques-
24, 719–732. tions about Behavioral Function (QABF) scale: Pre-
Hosp, J. L., Howell, K. W., & Hosp, M. K. (2003). Char- dicting treatment success for self-injury, aggression,
acteristics of behavior rating scales: Implications for and stereotypies. Research in Developmental Disabili-
practice in assessment and behavioral support. Jour- ties, 20, 142–160.
nal of Positive Behavior Interventions, 5, 201–208. Matson, J. L., & Vollmer, T. R. (1995). User’s guide:
Individuals with Disabilities Education Improvement Questions about Behavioral Function (QABF). Baton
Act of 2004, 20 U.S.C. 1400 et seq. (2004). Rouge, LA: Scientific.
Iwata, B. A., & DeLeon, I. G. (1996). Functional Analy- Michael, J. (1993). Establishing operations. Behavior
sis Screening Tool (FAST). Gainesville: Florida Cen- Analyst, 16, 191–206.
ter on Self-Injury, University of Florida. Michael, J. (2000). Implications and refinements of the
Iwata, B. A., DeLeon, I. G., & Roscoe, E. M. (2013). establishing operation concept. Journal of Applied Be-
Reliability and validity of the functional analysis havior Analysis, 33, 401–410.
screening tool. Journal of Applied Behavioral Analysis, Neef, N. A., & Lutz, M. N. (2001). A brief computer-
46(1), 271–284. based assessment of reinforcer dimensions affect-
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & ing choice. Journal of Applied Behavior Analysis, 34,
Richman, G. S. (1994). Toward a functional analysis 57–60.
of self-injury. Journal of Applied Behavior Analysis, 27, Neef, N. A., Mace, F. C., & Shade, D. (1993). Impulsivity
197–209. (Reprinted from Analysis and Intervention in in students with serious emotional disturbance: The
Developmental Disabilities, 2, 3–20, 1982) interactive effects of reinforcer rate, delay, and qual-
Iwata, B. A., Vollmer, T. R., & Zarcone, J. R. (1990). The ity. Journal of Applied Behavior Analysis, 26, 37–52.
experimental (functional) analysis of behavior disor- Neef, N. A., Shade, D., & Miller, M. S. (1994). Assess-
ders: Methodology, applications, and limitations. In ing influential dimensions of reinforcers on choice in
A. C. Repp & N. N. Singh (Eds.), Perspectives on the students with serious emotional disturbance. Journal
use of nonaversive and aversive interventions for persons of Applied Behavior Analysis, 27, 575–583.
with developmental disabilities (pp. 301–330). Syca- Newton, J. T., & Sturmey, P. (1991). The Motivation
more, IL: Sycamore. Assessment Scale: Inter-rater reliability and internal
Jin, C. S., Hanley, G. P., & Beaulieu, L. (2013). An in- consistency in a British sample. Journal of Mental De-
dividualized and comprehensive approach to treating ficiency Research, 35, 472–474.
sleep problems in young children. Journal of Applied Nicholson, J., Konstantinidi, E., & Furniss, F. (2006).
Behavior Analysis, 46, 161–180. On some psychometric properties of the Questions
Kelley, M. E., Fisher, W. W., Lomas, J. E., & Sanders, R. about Behavioral Function (QABF) scale. Research in
Q. (2006). Some effects of stimulant medication on Developmental Disabilities, 27, 337–352.
response allocation: A double-blind analysis. Journal O’Neill, R. E., Horner, R. H., Albin, R. W., Sprague, J.
of Applied Behavior Analysis, 39, 243–247. R., Storey, K., & Newton, J. (1997). Functional assess-
Kennedy, C. H., & Meyer, K. A. (1996). Sleep depri- ment and program development for problem behavior: A
vation, allergy symptoms, and negatively reinforced practical handbook. Pacific Grove, CA: Brooks/Cole.
problem behavior. Journal of Applied Behavior Analy- O’Reilly, M. F. (1997). Functional analysis of episodic
sis, 29, 133–135. self-injury correlated with recurrent otitis media.
Lennox, D. B., & Miltenberger, R. G. (1989). Conduct- Journal of Applied Behavior Analysis, 30, 165–167.
Indirect Behavioral Assessments 201
Pace, G. M., Ivancic, M. T., Edwards, G. L., Iwata, B. A., Shogren, K. A., & Rojahn, J. (2003). Convergent reli-
& Page, T. J. (1985). Assessment of stimulus prefer- ability and validity of the Questions about Behavior-
ence and reinforcer value with profoundly retarded al Function and the Motivation Assessment Scale: A
individuals. Journal of Applied Behavior Analysis, 18, replication study. Journal of Developmental and Physi-
249–255. cal Disabilities, 15, 367–375.
Paclawskyj, T. R., Matson, J. L., Rush, K. S., Smalls, Y., Sigafoos, J., Kerr, M., & Roberts, D. (1994). Interrater
& Vollmer, T. R. (2000). Questions about Behavioral reliability of the Motivation Assessment Scale: Fail-
Function (QABF): A behavioral checklist for func- ure to replicate with aggressive behavior. Research in
tional assessment of aberrant behavior. Research in Developmental Disabilities, 15, 333–342.
Developmental Disabilities, 21, 223–229. Simo-Pinatella, D., Alomar-Kurz, E., Font-Roura, J.,
Paclawskyj, T. R., Matson, J. L., Rush, K. S., Smalls, Gine, C., Matson, J. L., & Cifre, I. (2013). Questions
Y., & Vollmer, T. R. (2001). Assessment of the con- about Behavioral Function (QABF): Adaptation and
vergent validity of the Questions about Behavioral validation of the Spanish version. Research in Devel-
Function scale with analogue functional analysis and opmental Disabilities, 34, 1248–1255.
the Motivation Assessment Scale. Journal of Intellec- Singh, N. N., Donatelli, L. S., Best, A., Williams, D. E.,
tual Disability Research, 45, 484–494. Barrera, F. J., Lenz, M. W., et al. (1993). Factor struc-
Piazza, C. C., & Fisher, W. W. (1991). A faded bedtime ture of the Motivation Assessment Scale. Journal of
with response cost protocol for treatment of multiple Intellectual Disability Research, 37, 65–74.
sleep problems in children. Journal of Applied Behav- Steege, M. W., & Watson, T. S. (2009). Conducting
ior Analysis, 24, 129–140. school-based functional behavioral assessments (2nd
Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, ed.). New York: Guilford Press.
B. A. (1998). Evaluation of a brief stimulus preference Umbreit, J. (1996). Functional analysis of disruptive
assessment. Journal of Applied Behavior Analysis, 31, behavior in an inclusive classroom. Journal of Early
605–620. Intervention, 20, 18–29.
Saini, V., Ubdegrove, K., Biran, S., & Duncan, R. (2019). Zarcone, J. R., Rodgers, T. A., Iwata, B. A., Rourke, D.
A preliminary evaluation of interrater reliability and A., & Dorsey, M. F. (1991). Reliability analysis of the
concurrent validity of open-ended indirect assess- Motivation Assessment Scale: A failure to replicate.
ment. Behavior Analysis in Practice, 13, 114–125. Research in Developmental Disabilities, 12, 349–360.
CHAP TER 12
Direct Observation
Rachel H. Thompson and John C. Borrero
This chapter focuses on methods of assessment the key components of a descriptive analysis. They
involving the direct observation of behavior. Re- described the classroom, the routine, the number
searchers and behavior analysts typically refer of children and adults present, and precise op-
to these as descriptive methods, in that the as- erational definitions of child and adult behavior.
sessments describe a series of naturally occurring They conducted observations during 3-hour blocks
events, but do not demonstrate a functional re- across 28 days during one of several classroom ac-
lation between those events (i.e., there is no ex- tivities such as art or snacktime. They assessed ob-
perimental manipulation). As with indirect as- server reliability and presented data graphically to
sessment, the goal of the descriptive analysis1 is depict levels of various forms of child and teacher
to identify naturally occurring behavior–environ- behavior. In short, the researchers developed a
ment relations. Unlike indirect methods, descrip- model for descriptive analyses and provided a case
tive analysis involves the measurement of behavior illustration of the method in application.
and various environmental events through repeat- Bijou et al. (1968) noted that experimental
ed direct observation. studies are essential for understanding behav-
Researchers have used descriptive-analysis ior, in that experimental manipulation uncov-
methods widely in the behavioral sciences for de- ers functional relations between behavior and
cades. In fact, one of the defining features of ethol- environment. However, as Baer (1973) pointed
ogy, a field that grew out of the biological tradition, out, an experimental demonstration that a given
is direct observation of naturally occurring behav- variable produces a particular behavior change
ior (Hine, 1982; Tinbergen, 1951). A seminal paper demonstrates only that the relation is possible.
by Bijou, Peterson, and Ault (1968) introduced this That those same circumstances influence behav-
approach to the field of applied behavior analysis. ior under naturally occurring conditions does not
This paper highlighted the importance of descrip- necessarily follow. For example, a behavior ana-
tive studies of behavior, described an ideal interre- lyst might demonstrate language acquisition ex-
lation between descriptive and experimental stud- perimentally, using a shaping procedure (Ghaem-
ies, and recommended specific descriptive-analysis maghami, Hanley, Jessel, & Landa, 2018). These
procedures to improve the objectivity and ease of results would indicate only that the behavior ana-
interpretation of descriptive data. lyst can shape language, but they do not provide
Bijou et al. (1968) presented a case study of a direct evidence that language is shaped through
young boy in a nursery school setting to illustrate typical parent–child interactions. The addition
202
Direct Observation 203
of descriptive data showing that parents provide puters to collect, organize, and analyze data; and
some potentially reinforcing event following suc- (3) the availability of a trained professional to in-
cessive approximations to language would support terpret and use results to make intervention de-
the contention that shaping is responsible for lan- cisions. When such resources are not available,
guage acquisition outside the laboratory. Thus, as the only reasonable option may be to implement
Bijou et al. and Baer point out, descriptive analysis descriptive-analysis procedures that are relatively
is a vital tool in understanding naturally occurring easy and inexpensive to carry out. For example,
behavior–environment relations (e.g., Glodowski, observers might use paper and pencil during brief
Thompson, & Martel, 2019). intervals (e.g., momentary time sampling) to col-
lect data throughout the day. Importantly, behav-
ior analysts should consider how to use the avail-
CHOOSING AN APPROPRIATE PROCEDURE able resources to incorporate descriptive analysis
into their everyday clinical practice.
A behavior analyst may consider several factors
when selecting a method for a descriptive analysis.
Perhaps the most obvious factor is the overall pur- DEVELOPING HYPOTHESES
pose of the descriptive analysis; that is, what is the ABOUT BEHAVIORAL FUNCTION
behavior analyst attempting to accomplish by con-
ducting the analysis? Although behavior analysts Behavior analysts often use descriptive analysis in
conduct descriptive analyses most commonly to practice to develop hypotheses about the function
develop hypotheses regarding behavioral function, of behavior. Oliver, Pratt, and Normand (2015)
there may be other purposes as well. Such purpos- surveyed 682 Board Certified Behavior Analysts
es may include (1) identifying common contingen- and reported that 94% always or almost always
cies in naturalistic settings (e.g., Borrero, Woods, conducted descriptive analyses. Similarly, Roscoe,
Borrero, Masler, & Lesser, 2010; McComas et al., Phillips, Kelly, Farber and Dube (2015) surveyed
2009; McKerchar & Thompson, 2004; Rodriguez, 205 behavior analysts in the state of Massachu-
Thompson, Stocco, & Schlichenmeyer, 2013; setts, and 84% of respondents reported using de-
Thompson & Iwata, 2001; Simmons, Akers, & scriptive analysis most frequently. A concerning
Fisher, 2019); (2) establishing a baseline by which finding of both studies is that behavior analysts
to assess the efficacy of intervention (e.g., Rapp, reported frequently using descriptive analyses to
Vollmer, St. Peter, Dozier, & Cotnoir, 2004); (3) generate hypotheses about behavioral function,
studying basic behavioral processes (e.g., reinforce- without using a functional analysis to verify that
ment, punishment, extinction) under naturally a functional relation existed. Oliver et al. reported
occurring circumstances (Addison & Lerman, that 63% of behavior analysts reported never or
2009; Borrero, Vollmer, Borrero, & Bourret, 2005; almost never conducting functional analyses, and
Sloman et al., 2005); and (4) studying quantita- the Roscoe et al. survey found that 62% of behav-
tive models of behavior, such as the matching law ior analysts reported using a descriptive analysis in
(e.g., Borrero & Vollmer, 2002; Cero & Falligant, the absence of a functional analysis.
2020; Oliver, Hall, & Nixon, 1999) and behav- The practice of prescribing interventions based
ioral momentum (e.g., Strand, Wahler, & Herring, on the results of a descriptive analysis alone is not
2000). Even though the focus of this chapter is on empirically supported. Although there are isolated
applications of descriptive-analysis procedures to reports of successful intervention based only on
develop hypotheses about behavioral function, be- descriptive data (e.g., VanDerHeyden, Witt, &
havior analysts should recognize that descriptive Gatti, 2001), there is strong evidence of poor cor-
analysis is a highly flexible and widely used means respondence between descriptive and functional
of studying naturally occurring behavior that they analyses (e.g., Camp, Iwata, Hammond, & Bloom,
can adapt easily for many purposes. 2009; Hall, 2005; Mace & Lalli, 1991; Martens
A second consideration in the selection of et al., 2019; Pence, Roscoe, Bourret, & Ahearn,
descriptive-analysis methods involves available 2009; Piazza et al., 2003; Thompson & Iwata,
resources. The most sophisticated methods of 2007). Descriptive analysis may identify events
gathering and analyzing descriptive-analysis data correlated with but not functionally related to the
involve (1) direct observation by trained observers target behavior, or may fail to identify sources of
who are free from other responsibilities (e.g., pa- reinforcement that caregivers deliver only inter-
tient care) during the observation period; (2) com- mittently or under circumscribed conditions that
204 B e h av i o r a l A s s e s s m e n t
the descriptive analysis did not sample (see Ler- delivery, to identify potential reinforcers for food
man & Iwata, 1993, for a discussion). When this refusal. Behavior analysts may refer to Hanley,
occurs, the prescribed intervention may include ir- Iwata, and McCord (2001), Hagopian et al. (2013),
relevant components, may lack essential features, and Schlichenmeyer et al. (2015) for guidance on
or both. This lack of correspondence between how to identify a broader range of variables to
descriptive and functional analyses raises consid- measure during the descriptive analysis. Narrative
erable concern about the practice of prescribing recording is also a useful starting point for identi-
interventions based solely on the results of descrip- fying idiosyncratic variables correlated with prob-
tive analyses. lem behavior (see Rodriguez et al., 2013).
Behavior analysts should use the results of de-
scriptive analyses to generate hypotheses about
behavioral function and to inform the develop- VARIATIONS IN MEASUREMENT
ment of functional-analysis conditions. Results of DURING DIRECT OBSERVATION
the functional analysis are then used to evaluate
whether events correlated with problem behavior Generally, direct observation may involve peri-
under natural conditions are functionally related odic sampling or continuous recording of behavior
to problem behavior. Hagopian, Rooker, Jessel, and other environmental events. Considerations
and DeLeon (2013) described a clinical model for using either sampling or continuous recording
whereby they initially used standard functional- include the effort of each method and the amount
analysis conditions to assess problem behavior, and and quality of the resultant data. Sampling meth-
they modified the functional analysis as necessary ods involve collecting data at the end of speci-
to obtain clear results. The standardized analysis fied intervals. Harding et al. (1999) used a time-
yielded an interpretable outcome in only 47% of sampling procedure to assess child behavior at the
176 consecutive clinical cases. The functional end of each 10-second interval, while allocating
analysis ultimately produced clear results in 87% of the remainder of the interval recording to teacher
cases when the researchers modified antecedents, behavior. Sampling procedures require somewhat
consequences, the experimental design, or a com- less effort, depending on how frequently observers
bination of these variables. These results suggest collect data, but provide data that are less compre-
the value of individualizing functional analyses. hensive than those of continuous methods. Con-
Hagopian et al. individualized functional analyses tinuous methods involve collecting data through-
only after a standard analysis produced unclear re- out the observation period on each instance of the
sults (see also Roscoe, Schlichenmeyer, & Dube, target response, and in some cases environmental
2015). An alternative approach is to individual- events occurring in close temporal proximity. For
ize functional-analysis conditions from the start, example, Moss et al. (2005) used continuous re-
rather than waiting for unclear functional-analysis cording to collect data on the self-injurious behav-
results (see Hanley, 2012, for a discussion). Data ior (SIB) of eight participants diagnosed with Cor-
from descriptive analyses may be a useful source of nelia de Lange syndrome and potential evocative
information for individualizing functional-analysis events contiguous with SIB. Continuous methods
conditions (see Schlichenmeyer, Roscoe, Rooker, are more labor-intensive than sampling, but also
Wheeler, & Dube, 2015). provide richer samples of behavioral data.
The behavior analyst must structure the descrip- Event recording (i.e., recording the number of
tive analysis to identify idiosyncratic variables for times a response occurs), and recording the oc-
inclusion in a functional analysis. Behavior ana- currence or nonoccurrence of behavior in relatively
lysts often structure analyses to identify correla- small intervals (e.g., 10–20 seconds), are contin-
tions between problem behavior and the anteced- uous-observation methods. A study by Anderson
ents and consequences described by Iwata, Dorsey, and Long (2002) illustrates both methods. Observ-
Slifer, Bauman, and Richman (1982/1994). The ers recorded each time problem behavior occurred
behavior analyst should expand the descriptive (i.e., event recording) and scored the occurrence
analyses to include more events and to describe or nonoccurrence of potentially evocative envi-
specific interactions beyond those that Iwata et al. ronmental events (e.g., periods of low attention)
assessed. For example, Borrero et al. (2010) record- during 5-second intervals.
ed specific caregiver responses such as coaxing, Partial-interval and whole-interval recording
threats, and statements of concern, rather than are options for recording the occurrence or non-
simply recording the general category of attention occurrence of events. Partial-interval recording
Direct Observation 205
involves scoring an event if it occurs during any correlate with the problematic times. Further iso-
portion of a specified interval, which may overes- lating the environmental conditions associated
timate the occurrence of events (Gardenier, Mac- with problem behavior may require additional
Donald, & Green, 2004). By contrast, whole-inter- assessment if multiple variables occur during the
val recording involves scoring an event if it occurs problematic times. For example, Arndorfer, Milt-
for the duration of the specified interval, which enberger, Woster, Rortvedt, and Gaffaney (1994)
may underestimate the occurrence of events (see used a parent-completed scatterplot to identify
Cooper, Heron, & Heward, 2007). Selection of the times during which problem behavior was likely.
appropriate measurement procedure will depend Observers then recorded antecedents and conse-
in part on whether the response is high- or low- quences associated with problem behavior during
frequency, and whether the response is targeted for times identified through the scatterplot.
increase or decrease. Touchette et al. (1985) illustrated the use of
Below, we describe methods of data collection the scatterplot, presenting data on the temporal
and analysis that behavior analysts commonly patterns of problem behavior displayed by three
use. We progress from methods that examine be- participants. The scatterplot showed that problem
havior–environment relations with relatively low behavior correlated reliably with particular times
resolution to those with relatively greater resolu- for two participants, and this information led to
tion. Each method of data collection involves ei- environmental modifications aimed at reducing
ther continuous observation or periodic sampling problem behavior. Scatterplot data were uninter-
of behavior. In this section, we highlight methods pretable for the third participant; problem behav-
of summarizing data and make specific recommen- ior was not correlated reliably with any particular
dations regarding appropriate applications of each time. Results of a study by Kahng et al. (1998)
method. suggested that the third case may be more repre-
sentative. These researchers examined scatterplots
depicting the frequency of problem behavior for 15
The Scatterplot
participants, and found that none of the datasets
The scatterplot is a form of descriptive analysis showed a predictable temporal pattern. However,
used to identify temporal patterns of a target be- the researchers identified temporal patterns of be-
havior under naturally occurring conditions. Typi- havior for 12 participants, using statistical-control
cally, this form of assessment involves continuous charts of the same data (Pfadt & Wheeler, 1995);
observation of behavior and recording of the tar- these findings suggest that the main limitation
get behavior and the time during which the be- of the scatterplot may be the data depiction and
havior occurred. The scatterplot differs from other analysis rather than the measurement. Neverthe-
forms of descriptive analysis, because the behav- less, the practical utility of the scatterplot is lim-
ior analyst does not record environmental events ited severely if the user must construct statistical-
surrounding the target behavior. In addition, the control charts for data interpretation.
manner of data depiction is different from that of Behavior analysts report using scatterplots fre-
other methods. In most cases, the behavior analyst quently, despite the method’s limitations (Elling-
uses a code to indicate target behavior frequency son, Miltenberger, & Long, 1999), perhaps because
on a grid in which each cell indicates a time inter- they are easy to use. The published literature de-
val (e.g., 30 minutes), and each column represents scribes scatterplot assessment infrequently, but
a day. For example, an empty cell may indicate further evaluation of this method may be valuable.
that the target behavior did not occur during the For example, the scatterplot may be more useful
interval; a cell with a slash may indicate a low fre- for temporally organized responses. Ashbaugh and
quency of the target behavior; and a darkened cell Peck (1998) used a scatterplot to evaluate parent-
may indicate a high frequency of the target behav- collected data on disturbed sleep exhibited by a
ior (Touchette, MacDonald, & Langer, 1985). typically developing 2-year-old girl. Data revealed
The behavior analyst analyzes data through many intervals of sleep during scheduled awake
visual inspection of the grid to identify times as- hours, and many intervals awake during sched-
sociated with zero, low, and high frequencies of uled sleep hours. Bedtime fading and response
problem behavior. If the analyst identifies a tem- cost modified this pattern. More recently, Maas,
poral pattern of behavior, he or she might use this Didden, Bouts, Smits, and Curfs (2009) used scat-
information to modify features of the environ- terplot data to assess signs of sleepiness and dis-
ment, such as staffing patterns or activities, that ruptive behavior for individuals diagnosed with
206 B e h av i o r a l A s s e s s m e n t
METHODS OF DATA ANALYSIS: ample, it is likely that any observed target behav-
PROBABILITY ANALYSES ior will be followed by teacher attention if teacher
attention is available (independently of behavior)
Methods of analysis for direct-observation data nearly continuously during an observation. In
range from simple (e.g., tallying the frequency of this case, the conditional probability of attention
antecedents, behavior, and consequences) to com- would be very high, although the target behavior
plex (e.g., computing the Yule’s Q statistic; Rah- does not increase the probability of attention. To
man, Oliver, & Alderman, 2010). Typical methods address this weakness, Vollmer et al. (2001) rec-
of data analysis provide gross descriptions of be- ommend comparing conditional probabilities to
havior and related environmental events for many unconditional or background probabilities of the
of the descriptive-analysis procedures described same event, to determine whether the probability
earlier. However, more complex methods of data of an event (e.g., attention) changes due to the tar-
analysis provide more fine-grained descriptions. get behavior (i.e., to detect a possible contingency)
In this section, we focus on two methods that (see also Herscovitch, Roscoe, Libby, Bourret, &
researchers have studied extensively both within Ahearn, 2009). A conceptually similar alterna-
and outside applied behavior analysis: compara- tive to comparing the conditional probability of
tive-probability analyses and lag-sequential analy- an environmental event, given target behavior, to
sis. We can conduct both types of analyses using the unconditional probability of that event is to
hand data-collection procedures; however, com- compare the conditional probability to the prob-
puterized data collection and analysis programs ability of the environmental event, given that no
may facilitate these analyses. Notably, however, behavior occurred (e.g., Reed et al., 2010). Though
comparative-probability and lag-sequential anal- conceptually similar, this approach may prove
yses do not represent the end of the complexity more challenging, because it requires a somewhat
continuum. Readers interested in statistics that arbitrary determination of what defines a period
capture the degree of sequential associations be- with no behavior (e.g., Hagopian, Paclawskyj, &
tween events may start with Yoder and Symons Contrucci-Kuhn, 2005).
(2010) and Lloyd, Kennedy, and Yoder (2013).
Dynamic‑Probability Analyses
Static‑Probability Analyses
Closely related to the concept of comparative-
Many researchers have analyzed descriptive-anal- probability analysis is the method of lag-sequential
ysis data by calculating conditional probabilities analysis (Bakeman & Gottman, 1997). Typically,
to determine whether relations exist between evaluations of comparative probabilities produce
behavior and environmental events. Conditional one conditional and one unconditional probability
probabilities evaluate the likelihood of one event value. For example, observers may record data for 4
(e.g., attention), given that another event occurred hours and report the probability of attention and
(e.g., aggression). Frequently, evaluations of condi- the probability of attention, given an instance of
tional probabilities involve analyses of the target the target response, for the entire 4-hour observa-
response (e.g., aggression) and a potential reinforc- tion. This information provides a rather static de-
er (e.g., attention). In many cases, researchers cal- piction of what is likely a very dynamic exchange,
culate and compare the conditional probabilities even though it has proven useful in evaluations of
of various events (e.g., attention, escape, material behavior–environment relations. Lag-sequential
presentation) to determine the event(s) with the analysis, on the other hand, can provide a more
highest conditional probability, given the target refined description of exchanges in the natural en-
behavior (e.g., Reed, Luiselli, Morizio, & Child, vironment by depicting comparative probabilities
2010). This event or events, then, are considered before and after an instance of the target response
the likely maintaining variable(s) (e.g., Anderson on a second-by-second basis.
& Long, 2002; Noell, VanDerHeyden, Gatti, & The term lag conveys that researchers may eval-
Whitmarsh, 2001). uate behavior–environment relations or behavior–
One limitation of this approach is that the behavior relations several seconds before or after
analysis may identify variables that occur at a high the response occurs (e.g., Emerson, Thompson,
frequency following the target behavior only be- Reeves, Henderson, & Robertson, 1995; Samaha
cause the caregiver presents the event at a high et al., 2009). For example, a lag +1 indicates that
frequency, independently of responding. For ex- the probability of an event is calculated 1 unit
Direct Observation 209
(e.g., 1 second) after the occurrence of the target tween environmental events and target responses.
response. A lag –1 indicates that the probability of Furthermore, relative to the method Borrero and
an environmental event is calculated 1 unit (e.g., Borrero described (which involved analyses of ±
1 second) before the occurrence of the target re- 50 seconds), the windows Vollmer et al. evaluated
sponse. Borrero and Borrero (2008) demonstrated may be more practical for those working in nonre-
the use of lag-sequential analysis. The researchers search settings.
hypothesized that occurrences of severe problem Relative rather than absolute probabilities are
behavior, such as head hitting, were preceded re- important for both static- and dynamic-probability
liably by less severe forms of problem behavior, analyses (e.g., Rooker, DeLeon, Borrero, Frank-
such as screaming, for two students with autism Crawford, & Roscoe, 2015). For example, a positive
spectrum disorder. Borrero and Borrero described contingency between aggression and attention is
the less severe forms of problem behavior as poten- unlikely if the probability of attention, given prob-
tial precursors to more severe forms, to the extent lem behavior, is .20. By contrast, problem behavior
that occurrences of one might be reliable predic- increases the probability of attention if the prob-
tors of occurrences of the other. They evaluated ability of attention, given problem behavior, is .10.
the probability of severe problem behavior, given Researchers have been the primary users of the
that it was preceded by an instance of less severe data-analysis techniques just described, and they
problem behavior (i.e., conditional probability), have used them primarily to evaluate descriptive-
and compared that to the probability of the less analysis data for research (e.g., Anderson & Long,
severe problem behavior (i.e., the unconditional 2002; Borrero & Borrero, 2008; Doggett, Edwards,
probability) to quantify this potential relation. Moore, Tingstrom, & Wilczynski, 2001; Forman,
The window for the lag-sequential analysis was Hall, & Oliver, 2002; Marion, Touchette, & Sand-
+50 seconds (i.e., 50 seconds after each instance man, 2003; Moss et al., 2005; Noell et al., 2001;
of problem behavior) and –50 seconds (i.e., 50 sec- Woods, Borrero, Laud, & Borrero, 2010). Perhaps
onds before each instance of problem behavior). behavior analysts do not use these methods be-
Results for both students showed a sharp increase cause of their complexity. In addition, although
in the probability of a potential precursor in the these methods provide a finer-grained analysis of
1-second intervals immediately before an instance naturally occurring behavior–environment rela-
of problem behavior. tions than other descriptive-analysis methods pro-
Vollmer et al. (2001) also conducted compar- duce, they are limited because results suggest cor-
ative-probability analyses and used a variation of relations between two events (e.g., attention and
the lag-sequential analysis procedure described by aggression) but do not identify functional relations
Borrero and Borrero (2008). The researchers con- (e.g., St. Peter et al., 2005).
ducted a descriptive analysis of interactions be-
tween participants referred for the assessment and
intervention of severe problem behavior and their RECOMMENDATIONS
primary caregivers. Next, the researchers com-
pared the probability of an environmental event to The measurement and analysis of naturally oc-
the probability of an environmental event, given curring behavior–environment relations are nec-
problem behavior in the context of various poten- essary for a complete understanding of behavior.
tial establishing operations (e.g., low attention). However, the role of descriptive analysis in uncov-
Researchers evaluated probabilities with lags of +5, ering potential controlling variables for problem
+10, +15, and +20, using a variant of the lag-se- behavior is less clear. The results of two recent
quential-analysis procedures; that is, the research- surveys show that behavior analysts frequently de-
ers calculated the probability of an event within velop interventions for problem behavior based on
5, 10, 15, and 20 seconds of a particular time for the results of descriptive analysis alone. Yet results
the unconditional probability or an instance of of several studies showing poor correspondence
the target response for the conditional probabil- between descriptive- and experimental-analysis
ity. Even though the procedure does not provide outcomes raise significant concerns about this
the same level of analysis as the method Borrero practice (Camp et al., 2009; Hall, 2005; Mace &
and Borrero described, the method does provide Lalli, 1991; Pence et al., 2009; Thompson & Iwata,
four intervals to evaluate probabilistic changes, 2007).
and Vollmer et al. reported that the method was There are inherent limitations to the informa-
useful in identifying potential contingencies be- tion that we can gain through descriptive analyses,
210 B e h av i o r a l A s s e s s m e n t
because this method is limited to correlational and of recording (Lerman et al., 2009). Mayer and Di-
not functional descriptions of naturally occurring Gennaro Reed (2013) found that narrative data
behavior–environment relations. Thus we may collected by direct care staff did not identify the
gain the most complete understanding of these antecedents of or the consequences for problem
relations by combining descriptive and experi- behavior accurately. In the same study, researchers
mental methods. Descriptive-analysis data provide demonstrated immediate improvement in accura-
information regarding the environmental events cy after brief training and the provision of detailed
that are correlated with behavior under naturally written instructions. Future research should focus
occurring conditions, and the experimental analy- on identifying efficient training methods to ensure
sis identifies those events that are related func- accuracy of descriptive analyses.
tionally to the behavior of interest. We have focused in this chapter on the use of
We have described several descriptive-analysis descriptive analyses to generate hypotheses regard-
methods that vary with respect to the level of de- ing naturally occurring reinforcers for problem be-
tail provided, as well as the level of expertise and havior. However, there may be many other uses
amount of resources necessary to conduct the for descriptive analysis in behavioral assessment.
analysis. Of these methods, the scatterplot, narra- For example, researchers and behavior analysts
tive recording, and the A-B-C checklist are gener- might use descriptive analysis to identify general
ally easy to use, and those directly responsible for practices that appear to promote or interfere with
the participant can implement them with mini- desirable behavior. For example, Austin, Carr, and
mal training. A more detailed descriptive analysis Agnew (1999) suggested that descriptive analyses
involves recording a target behavior and its sur- might identify the form of instruction that produc-
rounding events via frequency, interval, or time- es the most accurate and efficient performance in
sample recording and analyzing relations among organizations. Behavior analysts also might use de-
the variables. However, this method is resource- scriptive analysis to identify socially valid stimuli,
intensive and likely to require special training; responses, or levels of responding to inform social
thus only trained personnel who are dedicated skills training (e.g., Minkin et al., 1976).
specifically to behavioral assessment are likely to
use this type of descriptive analysis. Behavior ana-
lysts who allocate resources toward various forms REFERENCES
of assessment should remember that even the most
sophisticated descriptive analyses yield only cor- Addison, L., & Lerman, D. C. (2009). Descriptive anal-
relational data. In many cases, descriptive analyses ysis of teachers’ responses to problem behavior fol-
uncover correlations between problem behavior lowing training. Journal of Applied Behavior Analysis,
and events that do not exert control over problem 42, 485–490.
behavior (e.g., Thompson & Iwata, 2007). There- Anderson, C. M., & Long, E. S. (2002). Use of a struc-
tured descriptive analysis methodology to identify
fore, we recommend conducting a brief descriptive
variables affecting problem behavior. Journal of Ap-
analysis to inform the development of functional-
plied Behavior Analysis, 35, 137–154.
analysis conditions in most cases. Arndorfer, R. E., Miltenberger, R. G., Woster, S. H.,
Rortvedt, A. K., & Gaffaney, T. (1994). Home-based
descriptive and experimental analysis of problem be-
FUTURE RESEARCH haviors in children. Topics in Early Childhood Special
Education, 14, 64–87.
Despite the limitations of descriptive analysis, the Ashbaugh, R., & Peck, S. M. (1998). Treatment of sleep
apparent ubiquity of its implementation (Oliver problems in a toddler: A replication of the faded bed-
et al. 2015; Roscoe, Phillips, et al., 2015) and the time with response cost protocol. Journal of Applied
importance of understanding the conditions natu- Behavior Analysis, 31, 127–129.
rally surrounding behavior necessitate additional Austin, J., Carr, J. E., & Agnew, J. L. (1999). The need
for assessment of maintaining variables in OBM.
research. One important use for descriptive analy-
Journal of Organizational Behavior Management, 19,
sis is to identify variables for inclusion in a func- 59–87.
tional analysis. The flexibility of narrative record- Baer, D. M. (1973). The control of the developmen-
ing is appealing when the behavior analyst’s goal tal process: Why wait? In J. R. Nesselroade & H.
is to identify idiosyncratic variables for inclusion W. Reese (Eds.), Lifespan developmental psychology:
in a functional analysis. However, the open-ended Methodological issues (pp. 187–193). New York: Aca-
nature of this assessment may reduce the accuracy demic Press.
Direct Observation 211
Bakeman, R., & Gottman, J. M. (1997). Observing in- & Robertson, J. (1995). Descriptive analysis of mul-
teraction: An introduction to sequential analyses (2nd tiple response topographies of challenging behavior
ed.). New York: Cambridge University Press. across two settings. Research in Developmental Dis-
Barker, R., & Wright, H. (1955). Midwest and its chil- abilities, 16, 301–329.
dren: The psychological ecology of an American town. Forman, D., Hall, S., & Oliver, C. (2002). Descriptive
Oxford, UK: Peterson. analysis of self-injurious behavior and self-restraint.
Bijou, S. W., Peterson, R. F., & Ault, M. H. (1968). A Journal of Applied Research in Intellectual Disabilities,
method to integrate descriptive and experimental 15, 1–7.
field studies at the levels of data and empirical con- Galiatsatos, G. T., & Graff, R. B. (2003). Combining de-
cepts. Journal of Applied Behavior Analysis, 1, 175–191. scriptive and functional analyses to assess and treat
Bijou, S. W., Peterson, R. F., Harris, F. R., Allen, K. E., screaming. Behavioral Interventions, 18, 123–138.
& Johnston, M. S. (1969). Methodology for experi- Gardenier, N. C., MacDonald, R., & Green, G. (2004).
mental studies of young children in natural settings. Comparison of direct observational methods for
Psychological Record, 19, 177–210. measuring stereotypical behavior in children with
Borrero, C. S. W., & Borrero, J. C. (2008). Descriptive autism spectrum disorders. Research in Developmental
and experimental analyses of potential precursors to Disabilities, 25, 99–118.
problem behavior. Journal of Applied Behavior Analy- Ghaemmaghami, M., Hanley, G. P., Jessel, J., & Landa,
sis, 41, 83–96. R. (2018). Shaping complex functional communica-
Borrero, C. S. W., Vollmer, T. R., & Borrero, J. C. tion responses. Journal of Applied Behavior Analysis,
(2004). Combining descriptive and functional analy- 51, 502–520.
sis logic to evaluate idiosyncratic variables maintain- Glodowski, K. R., Thompson, R. H., & Martel. L.
ing aggression. Behavioral Interventions, 19, 247–262. (2019). The rooting reflex as an infant feeding cue.
Borrero, C. S. W., Vollmer, T. R., Borrero, J. C., & Bour- Journal of Applied Behavior Analysis, 52, 17–27.
ret, J. (2005). A method of evaluating parameters of Groden, G. (1989). A guide for conducting a compre-
reinforcement during parent–child interactions. Re- hensive behavioral analysis of a target behavior. Jour-
search in Developmental Disabilities, 26, 577–592. nal of Behavior Therapy and Experimental Psychiatry,
Borrero, C. S. W., Woods, J. N., Borrero, J. C., Masler, 20, 163–170.
E. A., & Lesser, A. D. (2010). Descriptive analyses of Groden, G., & Lantz, S. (2001). The reliability of the
pediatric food refusal and acceptance. Journal of Ap- Detailed Behavior Report (DBR) in documenting
plied Behavior Analysis, 43, 71–88. functional assessment observations. Behavioral Inter-
Borrero, J. C., & Vollmer, T. R. (2002). An application ventions, 16, 15–25.
of the matching law to severe problem behavior. Jour- Hall, S. S. (2005). Comparing descriptive, experimen-
nal of Applied Behavior Analysis, 35, 13–27. tal and informant-based assessments of problem be-
Camp, E. M., Iwata, B. A., Hammond, J. L., & Bloom, haviors. Research in Developmental Disabilities, 26,
S. E. (2009). Antecedent versus consequent events 514–526.
as predictors of problem behavior. Journal of Applied Hagopian, L. P., Paclawskyj, T. R., & Contrucci-Kuhn,
Behavior Analysis, 42, 469–483. S. (2005). The use of conditional probability analysis
Carr, E. G., Smith, C. E., Giacin, T. A., Whelan, B. to identify a response chain leading to the occur-
M., & Pancari, J. (2003). Menstrual discomfort as a rence of eye poking. Research in Developmental Dis-
biological setting event for severe problem behavior: abilities, 26, 393–397.
Assessment and intervention. American Journal of Hagopian, L. P., Rooker, G. W., Jessel, J., & DeLeon, I.
Mental Retardation, 108, 117–133. G. (2013). Initial functional analysis outcomes and
Cero, I., & Falligant, J. M. (2020). Application of the modifications in pursuit of differentiation: A summa-
generalized matching law to chess openings: A gam- ry of 176 inpatient cases. Journal of Applied Behavior
bit analysis. Journal of Applied Behavior Analysis, 53, Analysis, 46, 88–100.
835–845. Hamo, M., Blum-Kulka, S., & Hacohen, G. (2004).
Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). From observation to transcription and back: Theory,
Applied behavior analysis (2nd ed.). Upper Saddle practice, and interpretation in the analysis of chil-
River, NJ: Pearson/Merrill-Prentice Hall. dren’s naturally occurring discourse. Research on Lan-
Doggett, R. A., Edwards, R. P., Moore, J. W., Tingstrom, guage and Social Interaction, 37, 71–92.
D. H., & Wilczynski, S. M. (2001). An approach to Hanley, G. P. (2012). Functional assessment of problem
functional assessment in general education class- behavior: Dispelling myths, overcoming implemen-
room settings. School Psychology Review, 30, 313–328. tation obstacles, and developing new lore. Behavior
Ellingson, S. A., Miltenberger, R. G., & Long, E. S. Analysis in Practice, 5, 54–72.
(1999). A survey of the use of functional assessment Hanley, G. P., Iwata, B. A., & McCord, B. E. (2001).
procedures in agencies serving individuals with de- Functional analysis of problem behavior: A review.
velopmental disabilities. Behavioral Interventions, 14, Journal of Applied Behavior Analysis, 36, 147–185.
187–198. Harding, J., Wacker, D. P., Cooper, L. J., Asmus, J., Jen-
Emerson, E., Thompson, S., Reeves, D., Henderson, D., sen-Kovalan, P., & Grisolano, L. (1999). Combining
212 B e h av i o r a l A s s e s s m e n t
descriptive and experimental analyses of young chil- McComas, J. J., Moore, T., Dahl, N., Hartman, E., Hoch,
dren with behavior problems in preschool settings. J., & Symons, F. (2009). Calculating contingencies in
Behavior Modification, 23, 316–333. natural environments: Issues in the application of se-
Hart, B., & Risley, T. (1995). Meaningful differences in quential analysis. Journal of Applied Behavior Analy-
the everyday experience of young American children. sis, 42, 413–423.
Baltimore: Brookes. McKerchar, P. M., & Thompson, R. H. (2004). A de-
Herscovitch, B., Roscoe, E. M., Libby, M. E., Bourret, J. scriptive analysis of potential reinforcement contin-
C., & Ahearn, W. H. (2009). A procedure for identi- gencies in the preschool classroom. Journal of Applied
fying precursors to problem behavior. Journal of Ap- Behavior Analysis, 37, 431–444.
plied Behavior Analysis, 42, 697–702. Minkin, N., Braukmann, C. J., Minkin, B. L., Timbers,
Hine, R. A. (1982). Ethology: Its nature and relations with G. D., Timbers, B. J., Fixsen, D. L., et al. (1976). The
other sciences. New York: Oxford University Press. social validation and training of conversation skills.
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Journal of Applied Behavior Analysis, 9, 127–139.
Richman, G. S. (1994). Toward a functional analysis Moss, J., Oliver, C., Hall, S., Arron, K., Sloneem, J., &
of self-injury. Journal of Applied Behavior Analysis, 27, Petty, J. (2005). The association between environ-
197–209. (Reprinted from Analysis and Intervention in mental events and self-injurious behaviour in Cor-
Developmental Disabilities, 2, 3–20, 1982) nelia de Lange syndrome. Journal of Intellectual Dis-
Kahng, S. W., Iwata, B. A., Fischer, S. M., Page, T. J., ability Research, 49, 269–277.
Treadwell, K. R., Williams, D. E., et al. (1998). Tem- Noell, G. H., VanDerHeyden, A. M., Gatti, S. L., &
poral distributions of problem behavior based on Whitmarsh, E. L. (2001). Functional assessment of
scatterplot analysis. Journal of Applied Behavior Anal- the effects of escape and attention on students’ com-
ysis, 31, 593–604. pliance during instruction. School Psychology Quar-
Lerman, D. C., Hovanetz, A., Strobel, M., & Tetreault, terly, 16, 253–269.
A. (2009). Accuracy of teacher-collected descriptive Oliver, A. C., Pratt, L. A., & Normand, M. P. (2015). A
analysis data: A comparison of narrative and struc- survey of functional behavioral assessment methods
tured recording formats. Journal of Behavioral Educa- used by behavior analysts in practice. Journal of Ap-
tion, 18, 157–172. plied Behavior Analysis, 48, 817–829.
Lerman, D. C., & Iwata, B. A. (1993). Descriptive and Oliver, C., Hall, S., & Nixon, J. (1999). A molecular to
experimental analyses of variables maintaining self- molar analysis of communicative and problem behav-
injurious behavior. Journal of Applied Behavior Analy- ior. Research in Developmental Disabilities, 20, 197–213.
sis, 26, 293–319. Pence, S. T., Roscoe, E. M., Bourret, J. C., & Ahearn, W.
Lloyd, B. P., Kennedy, C. H., & Yoder, P. J. (2013). H. (2009). Relative contributions of three descriptive
Quantifying contingent relations from direct obser- methods: Implications for behavioral assessment.
vation data: Transitional probability comparisons Journal of Applied Behavior Analysis, 42, 425–446.
versus Yule’s Q. Journal of Applied Behavior Analysis, Pfadt, A., & Wheeler, D. J. (1995). Using statistical pro-
46, 479–497. cess control to make data-based clinical decisions.
Maas, A. P. H. M., Didden, R., Bouts, L., Smits, M. G., Journal of Applied Behavior Analysis, 28, 349–370.
& Curfs, L. M. G. (2009). Scatterplot analysis of Piazza, C. C., Fisher, W. W., Brown, K. A., Shore, B. A.,
excessive daytime sleepiness and severe disruptive Patel, M. R., Katz, R. M., et al. (2003). Functional
behavior in adults with Prader–Willi syndrome: A analysis of inappropriate mealtime behaviors. Journal
pilot study. Research in Developmental Disabilities, 30, of Applied Behavior Analysis, 36, 187–204.
529–537. Rahman, B., Oliver, C., & Alderman, N. (2010). De-
Mace, F. C., & Lalli, J. S. (1991). Linking descriptive scriptive functional analysis of challenging behav-
and experimental analyses in the intervention of bi- iours shown by adults with acquired brain injury.
zarre speech. Journal of Applied Behavior Analysis, 24, Neuropsychological Rehabilitation, 20, 212–238.
553–562. Rapp, J. T., Vollmer, T. R., St. Peter, C., Dozier, C. L.,
Marion, S. D., Touchette, P. E., & Sandman, C. A. & Cotnoir, N. M. (2004). Analysis of response al-
(2003). Sequential analysis reveals a unique struc- location in individuals with multiple forms of stereo-
ture for self-injurious behavior. American Journal on typed behavior. Journal of Applied Behavior Analysis,
Mental Retardation, 108, 301–313. 37, 481–501.
Martens, B. K., Baxter, E. L., McComas, J. J., Sallade, S. Reed, D. D., Luiselli, J. K., Morizio, L. C., & Child, S.
J., Kester, J. S., Caamano, M., et al. (2019). Agreement N. (2010). Sequential modification and the identi-
between structured descriptive assessments and func- fication of instructional components occasioning
tional analyses conducted over a telehealth system. self-injurious behavior. Child and Family Behavior
Behavior Analysis: Research and Practice, 19, 343–356. Therapy, 32, 1–16.
Mayer, K. L., & DiGennaro Reed, F. D. (2013). Effects Repp, A. C., & Karsh, K. G. (1994). Hypothesis-based
of a training package to improve the accuracy of de- interventions for tantrum behaviors of persons with
scriptive analysis data recording. Journal of Organiza- developmental disabilities in school settings. Journal
tional Behavior Management, 33, 226–243. of Applied Behavior Analysis, 27, 21–31.
Direct Observation 213
Richman, D. M., & Hagopian, L. P. (1999). On the ef- S. W., Sloman, K. N., & Rapp, J. T. (2005). On the
fects of “quality” of attention in the functional analy- role of attention in naturally occurring matching
sis of destructive behavior. Research in Developmental relations. Journal of Applied Behavior Analysis, 38,
Disabilities, 20, 51–62. 429–433.
Rodriguez, N. M., Thompson, R. H., Schlichenmeyer, Strand, P. S., Wahler, R. G., & Herring, M. (2000). Mo-
K. S., & Stocco, C. S. (2012). Functional analysis of mentum in child compliance and opposition. Journal
arranging and ordering by individuals with an autism of Child and Family Studies, 9, 363–375.
spectrum disorder. Journal of Applied Behavior Analy- Thompson, R. H., & Iwata, B. A. (2001). A descriptive
sis, 45, 1–22. analysis of social consequences following problem
Rodriguez, N. M., Thompson, R. H., Stocco, C. S., & behavior. Journal of Applied Behavior Analysis, 34,
Schlichenmeyer, K. S. (2013). Arranging and order- 169–178.
ing in autism spectrum disorder: Characteristics, Thompson, R. H., & Iwata, B. A. (2007). A comparison
severity, and environmental correlates. Journal of In- of outcomes from descriptive and functional analy-
tellectual and Developmental Disabilities, 38, 242–255. ses of problem behavior. Journal of Applied Behavior
Rooker, G. W., DeLeon, I. G., Borrero, C. S., Frank- Analysis, 40, 333–338.
Crawford, M. A., & Roscoe, E. M. (2015). Reducing Thompson, T., Symons, F. J., & Felce, D. (2000). Prin-
ambiguity in the functional assessment of problem ciples of behavioral observation: Assumptions and
behavior. Behavioral Interventions, 30, 1–35. strategies. In T. Thompson, D. Felce, & F. J. Symons
Roscoe, E. M., Phillips, K. M., Kelly, M. A., Farber, R., (Eds.), Behavioral observation: Technology and appli-
& Dube, W. V. (2015). A statewide survey of prac- cations in developmental disabilities (pp. 3–16). Balti-
titioners’ use and perceived utility of functional as- more: Brookes.
sessment. Journal of Applied Behavior Analysis, 48, Tinbergen, N. (1951). The study of instinct. Oxford, UK:
830–844. Clarendon Press.
Roscoe, E. M., Schlichenmeyer, K. J., & Dube, W. V. Touchette, P. E., MacDonald, R. F., & Langer, S. N.
(2015). Functional analysis of problem behavior: (1985). A scatterplot for identifying stimulus con-
A systematic approach to identifying idiosyncratic trol of problem behavior. Journal of Applied Behavior
variables. Journal of Applied Behavior Analysis, 48, Analysis, 18, 343–351.
289–314. VanDerHeyden, A. M., Witt, J. C., & Gatti, S. (2001).
Samaha, A. L., Vollmer, T. R., Borrero, C., Sloman, K., Descriptive assessment method to reduce overall
St. Peter Pipkin, C., & Bourret, J. (2009). Analyses of disruptive behavior in a preschool classroom. School
response–stimulus sequences in descriptive observa- Psychology Review, 30, 548–567.
tions. Journal of Applied Behavior Analysis, 42, 447–468. Vollmer, T. R., Borrero, J. C., Wright, C. S., Van Camp,
Schlichenmeyer, K. J., Roscoe, E. M., Rooker, G. W., C., & Lalli, J. S. (2001). Identifying possible contin-
Wheeler, E. E., & Dube, W. V. (2013). Idiosyncratic gencies during descriptive analyses of severe behav-
variables affecting functional analysis outcomes: A ior disorders. Journal of Applied Behavior Analysis, 34,
review (2001–2010). Journal of Applied Behavior Anal- 269–287.
ysis, 46, 339–348. Wahler, R. G., Winkel, G. H., Peterson, R. F., & Morri-
Simmons, C. A., Akers, J. S., & Fisher, W. W. (2019). son, D. C. (1965). Mothers as behavior therapists for
Functional analysis and treatment of covert food their own children. Behaviour Research and Therapy,
stealing in an outpatient setting. Behavioral Develop- 3, 113–124.
ment, 24, 42–57. Woods, J. N., Borrero, J. C., Laud, R. B., & Borrero, C.
Sloman, K. N., Vollmer, T. R., Cotnoir, N., Borrero, C. S. W. (2010). Descriptive analyses of pediatric food
S. W., Borrero, J. C., Samaha, A. L., et al. (2005). refusal: The structure of parental attention. Behavior
Descriptive analyses of parent reprimands. Journal of Modification, 34, 35–36.
Applied Behavior Analysis, 38, 373–383. Yoder, P., & Symons, F. (2010). Observational measure-
St. Peter, C., Vollmer, T. R., Bourret, J. C., Borrero, C. ment of behavior. New York: Springer.
CHAP TER 13
Functional Analysis
History and Methods
Traditional methods of classifying behavior dis- havior. Although Skinner described several pro-
orders (e.g., American Psychiatric Association, cedures for conducting a functional analysis of
2013) rely primarily on correlations among symp- behavior, we focus in the current chapter on an
toms. For example, an easily distracted child who assessment in which the behavior analyst directly
often fidgets and squirms while seated might re- manipulates the consequences hypothesized to re-
ceive a diagnosis of attention-deficit/hyperactivity inforce problem behavior. We use the term func-
disorder (ADHD). This approach focuses on the tional analysis to describe this assessment.
structural characteristics of responses and on the
extent to which responses covary. A behavior-
Analyzing Behavioral Function
analytic alternative to the structural approach is
to categorize problem behavior according to en- Researchers in the late 1960s began systematically
vironmental variables of which the behavior is a manipulating environmental variables to study
function. their effects on self-injurious behavior (SIB) and
other problem behavior. Lovaas, Freitag, Gold, and
Kassorla (1965), for example, observed that the
HISTORY AND CONCEPTUAL FOUNDATION frequency of a participant’s SIB increased when
OF FUNCTIONAL ANALYSIS the therapist provided attention in the form of
sympathetic statements after occurrences of SIB.
Skinner (1953) adapted the mathematical term Similarly, Lovaas and Simmons (1969) established
functional analysis and introduced it to the field of that attention was a reinforcer for one participant’s
behavior analysis. According to Skinner, a func- SIB, and that SIB decreased when attention did
tional relation exists when an independent variable not follow its occurrence.
produces an orderly and predictable change in a Carr and colleagues showed that SIB (Carr,
dependent variable. Among Skinner’s primary in- Newsom, & Binkoff, 1976) and aggression (Carr,
terests were the effects of environmental events on Newsom, & Binkoff, 1980) occurred more fre-
human behavior, and he used the term functional quently during a condition in which the researcher
analysis to describe a process to identify indepen- presented demands than during conditions with-
dent variables functionally related to human be- out demands. These data suggested that escape
214
Functional Analysis 215
reprimands) in the natural environment. Thus the the participant to complete the task if the partici-
therapist simulates these conditions by diverting pant did not complete it after the vocal or model
his or her attention initially, providing attention prompt. The therapist removed the task materials,
following occurrences of problem behavior, and stopped prompting, and turned away from the par-
providing no differential consequence for other ticipant for 30 seconds if the participant engaged
participant behavior, such as communication. in SIB during the prompting sequence.
Some behavior analysts include a brief preatten- The rationale for this arrangement is that learn-
tion interaction between the therapist and par- ing tasks are aversive to some individuals. Problem
ticipant, so that the participant samples therapist behavior may postpone, prevent, or remove these
attention immediately before the session. aversive events (Iwata, 1987). For example, a care-
Results of several studies have demonstrated giver may stop prompting a boy to brush his teeth
that qualitatively different forms of attention may if the boy becomes aggressive during tooth brush-
vary in their reinforcing effects on problem behav- ing. Thus the boy may be more likely to engage in
ior (e.g., Fisher et al., 1996; Kodak, Northup, & aggression again in the future during self-care or
Kelley, 2007; Piazza et al., 1999). In our clinic, we other tasks if he learns that it ends the aversive
generally match the type and quality of attention activity. Beavers, Iwata, and Lerman (2013) identi-
the participant’s caregiver provides to our direct fied negative reinforcement as the most common
observations described above. The therapist will reinforcer for problem behavior.
use the same vocal statements, prosody, tone, and The presentation of learning trials is a defin-
volume the caregiver uses to provide attention fol- ing feature of the demand condition. Iwata et al.
lowing problem behavior. (1982/1994) chose tasks the participants were
In the Iwata et al. (1982/1994) study, toys were unlikely to complete to increase the participant’s
within arm’s reach of the participant. Several stud- motivation to escape. We select tasks based on our
ies have shown that response-independent avail- pre-functional-analysis observations, caregiver or
ability of preferred items, such as toys, can reduce teacher reports of tasks that correlate with prob-
attention-reinforced problem behavior to near-zero lem behavior, or a combination of these methods
levels (e.g., Fisher, O’Connor, Kurtz, DeLeon, & (Fisher, Adelinis, Thompson, Worsdell, & Zar-
Gotjen, 2000). The behavior analyst may consider cone, 1998; McComas, Hoch, Paone, & El-Roy,
using low-preference items in the attention con- 2000). The tasks that evoke problem behavior may
dition; otherwise, the functional analysis may not differ for individual participants, depending on
identify an attention function that is responsible variables such as amount of movement required,
for the maintenance of problem behavior in the difficulty level, and presentation rate (McCord,
natural environment. Thompson, & Iwata, 2001; Smith, Iwata, Goh,
Divided attention is a modification of the at- & Shore, 1995). Thus selection and use of tasks
tention condition in which the therapist provides that evoke problem behavior in the natural envi-
attention to another child or adult until the par- ronment may increase the probability of accurate
ticipant engages in problem behavior, which then functional-analysis results.
produces the therapist’s attention (Fahmie, Iwata,
Harper, & Querim, 2013). The rationale for this
Test for Automatic Reinforcement
modification is that an adult providing attention
to another child or adult may signal attention Iwata et al. (1982/1994) described an alone condi-
availability or increase the value of attention as tion, in which the participant was alone in a room
reinforcement. with no toys or materials that would function as
external stimulation. The rationale for this ar-
rangement is that some problem behaviors occur
Test for Social Negative Reinforcement, Escape
in the absence of social reinforcement—reinforce-
Iwata et al. (1982/1994) also described an academic ment that another person mediates or controls.
demand condition, in which the therapist present- Rather, the stimulation that the problem behavior
ed demands to complete nonpreferred academic automatically produces reinforces problem behav-
tasks. The therapist used sequential vocal, model, ior. The absence of stimulation in the alone con-
and physical prompts to encourage the individual dition increases the probability that no other re-
to complete each task. The therapist provided inforcement is available that would compete with
praise if the participant completed the task after the automatic reinforcement the problem behavior
the vocal or model prompt, or physically guided produces.
Functional Analysis 217
The risk of harm for some participants may stereotypical object flipping and spinning. He as-
outweigh the benefits of an alone condition. An sessed this hypothesis by placing a carpet on the
ignore condition is an alternative in which a thera- table to attenuate the sound of the objects while
pist in the room blocks the participant’s problem flipping and spinning. Decreases and increases in
behavior to reduce injury risk, but provides no flipping and spinning occurred with the introduc-
additional programmed consequence for problem tion and withdrawal of the carpet, respectively.
behavior. An ignore condition also may be ap-
propriate if the therapist suspects that automatic
Control Condition
reinforcement maintains problem behavior, such
as aggression or property destruction. These cases Iwata et al. (1982/1994) used an unstructured play
require the therapist in the room for the partici- condition as the control condition of the function-
pant to exhibit aggression or materials in the room al analysis. The participant had toys; the therapist
for the participant to exhibit property destruction. interacted with the participant about once every
Another variation is a baited environment, which 30 seconds in the absence of SIB, but provided no
provides the participant with the materials to en- differential consequence after occurrences of SIB.
gage in the problem behavior. Piazza et al. (1998) The rationale for this arrangement is that non-
described a procedure for a baited environment to contingent attention, the absence of demands, the
assess pica (consumption of unsafe items such as presence of toys, and the absence of a contingency
lead paint chips). The researchers identified each for problem behavior should decrease the partici-
participant’s pica items and bought or made items pant’s motivation to engage in problem behavior
that would serve as proxies. For example, Piazza to access attention, escape, or automatic reinforce-
et al. made simulated paint chips from flour and ment, respectively.
water for participants who consumed paint chips
(Finney, Russo, & Cataldo, 1982). This prepara-
Test for Social Positive Reinforcement, Tangible
tion allowed Piazza et al. to observe and measure
proxies of pica without exposing these participants Children may engage in problem behavior to ac-
to the risks associated with consumption of the cess a preferred item or activity, such as when
pica materials in the natural environment (e.g., Child B hits Child A to get the truck with which
paint chips containing lead). Child A is playing. Researchers have systematical-
Querim et al. (2013) used the alone condition ly tested whether access to tangible items functions
as a screening tool to predict whether automatic as reinforcement for problem behavior (cf. Fisher
or social reinforcement would maintain problem et al., 1993), which is an addition to the original
behavior. After conducting a series of consecutive Iwata et al. (1982/1994) procedure. The therapist
5-minute sessions, they hypothesized that auto- gives the participant the tangible item briefly be-
matic reinforcement likely maintained problem fore the session, starts the session by removing the
behavior that persisted across sessions, and that tangible item, and then provides the tangible item
social reinforcement likely maintained problem for about 30 seconds after occurrences of problem
behavior that extinguished across sessions. This behavior. The rationale for this arrangement is
screening tool predicted whether automatic or so- that removal of the tangible item should increase
cial reinforcement maintained problem behavior the participant’s motivation to engage in problem
in 28 of 30 cases. behavior to access the item if the item functions as
Withdrawing and reinstating reinforcement reinforcement for problem behavior.
consequences constitute another method for con- We select the tangible item based on our direct
firming the reinforcer for problem behavior. This observations of participant behavior or caregiver
method is more difficult when problem behavior report. We conduct a stimulus preference assess-
produces automatic reinforcement. For example, a ment (Fisher et al., 1992) and use the most pre-
therapist may not be able to block or prevent SIB if ferred item(s) from the assessment if the caregiver
a participant has multiple topographies that occur reports that the participant engages in problem
simultaneously at a high rate. In addition, the behavior to access several tangible items.
behavior analyst may not be able to control the We only conduct a tangible condition if we ob-
onset or offset of automatic reinforcement. One serve or a caregiver reports a relationship between
exception in the literature was a study by Rincover a problem behavior and a tangible item, because of
(1978). He hypothesized that auditory stimulation the risk of teaching the participant a response–re-
was the automatic reinforcement for a participant’s inforcement relation that does not exist currently
218 B e h av i o r a l A s s e s s m e n t
in the individual’s natural environment (Rooker, conduct functional analyses should also wear pro-
Iwata, Harper, Fahmie, & Camp, 2011; Shirley, tective equipment, such as arm guards, if problem
Iwata, & Kahng, 1999). For example, Shirley et al. behavior (e.g., biting) might cause injury to them.
(1999) showed that occurrences of automatically Functional analyses should have session termina-
reinforced hand mouthing increased when the tion criteria when problem behavior causes or has
participant received a tangible item after occur- the potential to cause injury, as well as criteria for
rences of hand mouthing. Shirley et al. concluded resuming sessions, such as when the injuries heal
that the tangible item became a reinforcer for hand or after consulting with a medical care provider.
mouthing, because the researchers provided it In general, behavior analysts should consider the
after occurrences of hand mouthing. risks carefully and only conduct a functional anal-
ysis when the benefits clearly outweigh the risks.
Kahng et al. (2015) evaluated whether partici-
MITIGATING THE RISKS OF CONDUCTING pation in a functional analysis increased injury
A FUNCTIONAL ANALYSIS risk relative to when the participants were not in
the functional analysis, based on records from 99
The occurrence of problem behavior during a func- patients with SIB treated in an inpatient hospital.
tional analysis is necessary to identify response–re- The frequency of injury was comparable during
inforcement relations. The possibility exists, there- and outside of the functional analyses. Injury rate
fore, that the functional analysis will increase the adjusted for time tended to be higher during the
risk of harm to the participant, the caregiver or functional analysis, but the injuries that occurred
therapist, or the environment, due to occurrences were relatively infrequent and of low severity.
of problem behavior. Iwata et al. (1982/1994) im-
plemented safeguards to minimize risks to partici-
pants in their study. First, each participant received INTERPRETING FUNCTIONAL‑ANALYSIS DATA
a medical examination and appropriate diagnostic
consultations to assess the participant’s physical The goal of a functional analysis is to identify the
status and to rule out organic causes for the partici- reinforcer(s) for problem behavior. Observers col-
pant’s SIB, such as headaches. Second, the partici- lect data on the target problem behavior via direct
pants’ physician recommended criteria for termi- observation during the test and control condi-
nating functional-analysis sessions based on degree tions. The behavior analyst converts the data to
of injury or level of responding, and intermittently a measure that is comparable across conditions,
observed sessions and modified termination crite- such as responses per minute; plots the data on a
ria if needed. Third, the therapist stopped the ses- line graph; and visually inspects the line graph to
sion if a participant met the termination criterion, determine whether levels of problem behavior are
and a physician or nurse examined the participant higher in one or more test conditions than in the
and determined whether sessions could continue. control condition. Generally, when levels of prob-
Fourth, a nurse examined participants after every lem behavior are higher in one or more test condi-
four sessions and noted any issues from SIB. Iwata tions than in the control condition, the behavior
et al. concluded that the risk of study participation analyst concludes that the functional analysis has
was no greater than the risk the participants ex- identified the reinforcer(s) for problem behavior,
perienced from occurrences of SIB in the natural or says that the analysis is differentiated. Iwata et
environment, due to their study’s safeguards. al. (1982/1994) found that higher levels of SIB oc-
Standard functional-analysis practice should curred in specific test conditions for six of the nine
include safeguards to minimize risk to participants participants. The researchers concluded that the
(Weeden, Mahoney, & Poling, 2010). These might “data provide information regarding the specific
include interviewing the caregiver(s) to determine conditions that may affect self-injury” (p. 203).
how often, when, and where problem behavior has Visual inspection is a reasonable method of
produced harm to the participant, others, or the data analysis when functional-analysis results are
environment. The behavior analyst should have relatively clear, but it may not be the most reliable
a plan to block or prevent SIB that might cause method of interpreting such results (Hagopian et
significant damage, such as using arm splints to al., 1997). Researchers have developed and used
prevent eye gouging or a helmet if head banging structured criteria to interpret functional-analysis
might cause a concussion (e.g., Fisher, Piazza, Bow- data, to address the problem of visual inspection’s
man, Hanley, & Adelinis, 1997). Therapists who poor interrater reliability.
Functional Analysis 219
Attention
4
Ignore
2
Play Demand
0
5 10 15
SESSION
FIGURE 13.1. An example of a functional analysis in which contingent attention functions as social positive reinforce-
ment for Mike’s problem behavior.
Multiply Controlled Problem Behavior other test and control conditions, and that rates of
problem behavior are higher in the attention con-
When more than one reinforcer maintains prob-
dition than in the control condition. These data
lem behavior, behavior analysts use the term suggest that the tangible item and attention func-
multiply controlled to describe the behavior. In tion as social positive reinforcement for problem
other words, levels of multiply controlled problem behavior.
behavior are higher in more than one test condi-
tion of a functional analysis relative to the control
condition. The functional analysis in Figure 13.3 Undifferentiated Functional Analysis
shows that rates of Willie’s problem behavior are Interpretation of functional-analysis data may be
highest in the tangible condition relative to the more difficult when problem behavior occurs at
Ignore
STEREOTYPE (PERCENTAGE OF INTERVALS)
Attention
100
Demand
75
Tangible
50
Play
25
0
5 10 15 20
SESSION
FIGURE 13.2. An example of a functional analysis in which automatic reinforcement likely maintains Hank’s problem
behavior.
Functional Analysis 221
Tangible
Demand
1
Play
Ignore
0
5 10 15
SESSION
FIGURE 13.3. An example of a functional analysis in which multiple consequences function as reinforcement for Wil-
lie’s problem behavior.
relatively equivalent rates across conditions, rates Variables That May Affect Interpretation
of problem behavior are variable across conditions,
Although we expect the programmed anteced-
or problem behavior seldom or never occurs during
the functional analysis. In this case, the behavior ents and consequences of a functional analysis to
analyst might conclude that problem behavior is control responding, research has shown that other
undifferentiated; that is, the functional analysis did procedural components of the functional analy-
not identify the reinforcer for problem behavior. sis may affect the results. We review these issues
Figure 13.4 depicts undifferentiated responding, below. We then review procedures for clarifying
in which rates of Lynn’s problem behavior are low data on undifferentiated problem behavior in the
and variable across conditions. subsequent section.
4
AGGRESSION PER MINUTE
2
Play Demand
Attention
1 Ignore
Tangible
0
5 10 15 20 25
SESSION
conditions, rates of problem behavior are variable shows data from an undifferentiated functional
across conditions, or problem behavior seldom or analysis followed by extended alone sessions in
never occurs during the functional analysis. How- the final phase. Problem behavior was maintained
ever, an alternative conclusion is that some other during extended alone sessions, suggesting auto-
variable(s)—such as carry-over effects, multiple- matic reinforcement as the function.
treatment interference, or a participant’s failure to
discriminate the condition contingencies—have
Re‑Presenting Social‑Reinforcement
contributed to the unclear results. In these cases,
Test Conditions
researchers have described and tested several pro-
cedures to clarify the results. Hagopian, Rooker, When problem behavior does not persist during ex-
Jessel, and DeLeon (2013) modified the functional tended alone or ignore sessions, re-presenting the
analysis for 176 cases with initially undifferenti- social-reinforcement conditions of the functional
ated results. The modifications included (1) con- analysis in a reversal design may clarify results. Fig-
ducting sessions in a different location, (2) imple- ure 13.6 shows that problem behavior did not per-
menting extinction for certain problem behaviors, sist during an extended ignore condition. The be-
or (3) increasing session duration. These manipu- havior analyst then conducted a reversal between
lations identified the function of problem behav- a contingent-escape condition and an ignore con-
ior in 153 of 176 cases. Below, we review specific dition. Rates of behavior were higher in the con-
procedures researchers have used to clarify undif- tingent-escape conditions, suggesting that escape
ferentiated functional-analysis results. functioned as negative reinforcement for problem
behavior. A potentially more time-efficient alter-
native is to conduct a pairwise comparison of one
Providing Extended Alone Sessions
test condition and the control condition (Iwata,
Vollmer, Marcus, Ringdahl, and Roane (1995) sug- Duncan, Zarcone, Lerman, & Shore, 1994).
gested conducting consecutive or extended alone
or ignore sessions when problem behavior occurs
Including Additional Discriminative Stimuli
during a functional analysis, but the results are un-
differentiated (Querim et al., 2013; Vollmer et al., Some functional analyses may produce undiffer-
1995). Automatic reinforcement is the likely func- entiated results because the discriminative stimuli
tion of problem behavior that persists during ex- associated with the conditions do not control the
tended alone sessions, because no other conditions participant’s problem behavior. Adding different,
are present to cause multiple-treatment interfer- salient discriminative stimuli in each condition
ence, and alternative sources of reinforcement are may address this problem (Conners et al., 2000).
not available in the alone condition. Figure 13.5 For example, a therapist might wear a red shirt
12
Alone
10
PICA PER MINUTE
4 Demand Attention
2 Play
0
5 10 15 20
SESSION
0
5 10 15 20 25 30 35 40 45 50
SESSION
during the attention condition, a purple shirt dur- Analyzing Within‑Session Response Patterns
ing the demand condition, and a yellow shirt dur-
Vollmer, Iwata, Zarcone, Smith, and Mazaleski
ing the control condition.
(1993) analyzed within-session data for a partici-
pant whose rates of SIB were high in the alone and
Altering Motivating Operations, Reinforcement control conditions. The analysis showed that ex-
Contingencies, or Both tinction bursts occurred at the beginning of alone
Another reason for undifferentiated functional- and control sessions that immediately followed an
analysis results is that the motivating operation for attention session. These results supported the con-
problem behavior may not be present in the func- clusion that attention reinforced SIB.
tional analysis. Altering one or several of the moti- Roane, Lerman, Kelley, and Van Camp (1999)
vating operations may produce more differentiated suggested analyzing within-session data to de-
or conclusive results. Finally, a functional analy- termine the effects of momentary changes in
sis may produce undifferentiated or inconclusive the establishing operation on problem behavior.
results because the consequences that maintain Roane et al. used within-session data to determine
problem behavior may not be present in the func- whether problem behavior occurred when the con-
tional analysis. Therefore, adding or altering the dition’s motivating operation was present versus
consequences for problem behavior may produce absent. Results of the within-session analysis gen-
more differentiated or conclusive results (Hago- erally agreed with the results of the multielement
pian et al., 2013). functional analysis: Problem behavior tended to
occur when the motivating operation for problem
behavior identified by the functional analysis was
Extending Session Duration present versus when it was absent.
Extending the duration of sessions for a functional
analysis may add clarity to the results. For exam- Identifying Idiosyncratic Consequences
ple, the participant in Kahng, Abt, and Schon-
bachler (2001) never displayed aggression during Still another possible explanation for undifferenti-
the 10-minute functional-analysis sessions. Con- ated results is that the reinforcer for problem be-
ducting one condition per day from 9:00 A.M. to havior is not present in the functional analysis. For
4:00 P.M. produced data suggesting that attention example, Van Camp et al. (2000) observed a corre-
functioned as positive reinforcement for problem lation between the occurrence of problem behav-
behavior, and a treatment analysis validated this ior and the presence of a Bumble Ball® during an
conclusion. undifferentiated functional analysis. Researchers
Functional Analysis 225
evaluated this relationship in conditions in which Currently, we use the following steps to iden-
(1) the participant had the ball, and the therapist tify idiosyncratic reinforcers for problem behavior.
removed the ball for 20 seconds after occurrences First, we interview caregivers to establish (1) the
of problem behavior; (2) the therapist gave the participant’s daily routine, (2) the times when
ball to the participant for 20 seconds after occur- problem behavior is most likely to occur during
rences of problem behavior; and (3) the partici- the routine, and (3) which activities correlate
pant had access to toys but not social interaction. with the occurrence of problem behavior. Next,
Results suggested that rates of problem behavior we ask caregivers to collect descriptive anteced-
were higher when the ball was present, but that ent–behavior–consequence data (Sulzer-Azaroff
automatic reinforcement maintained problem be- & Mayer, 1977), and therapists conduct observa-
havior. Schlichenmeyer, Roscoe, Rooker, Wheeler tions based on caregiver interview and descriptive
and Dube (2013) identified more than 30 idiosyn- data to identify the antecedents that evoke or oc-
cratic variables that influenced problem behavior casion problem behavior and the consequences
during functional analyses published between that reinforce it. A therapist then conducts several
2001 and 2010. Thus behavior analysts should as- 1- to 2-minute sessions in which he or she presents
sess whether idiosyncratic consequences reinforce the identified antecedents and consequences to re-
problem behavior. create the scene (see Van Houten & Rolider, 1988).
Our research group has used informal descrip- We develop hypotheses and test and control con-
tive assessments to generate hypotheses about id- ditions if problem behavior reliably occurs when
iosyncratic reinforcers for problem behavior, and we re-create the scene (e.g., Bowman et al., 1997).
these hypotheses inform our subsequent test and Roscoe, Schlichenmeyer, and Dube (2015)
control conditions (Bowman, Fisher, Thompson, developed a systematic method to generate and
& Piazza, 1997; Fisher, Adelinis, et al., 1998; Fisher, test hypotheses about idiosyncratic functions of
Lindauer, Alterson, & Thompson, 1998; Thomp- problem behavior when initial functional-analysis
son, Fisher, Piazza, & Kuhn, 1998). For example, results were undifferentiated. They first adminis-
Bowman et al. (1997) conducted informal obser- tered questionnaires to two informants familiar
vations of participants with undifferentiated func- with each participant, to identify potential idio-
tional analyses and their caregivers. Each caregiver syncratic antecedents and consequences of prob-
frequently acquiesced to the participant’s requests lem behavior. Next, Roscoe et al. video-recorded
or mands, even ones that were unreasonable, such and observed participants when problem behavior
as telling the adult to hop on one foot while play- reportedly occurred. This method identified the
ing a card game. Bowman et al. hypothesized that reinforcer(s) for problem behavior for five of six
a precurrent relation had developed in which the participants.
precurrent response (problem behavior) increased
the probability of reinforcement of the current
response (the mand). In this case, because prob- VARIATIONS OF FUNCTIONAL‑ANALYSIS
lem behavior occurred when the caregiver did not PROCEDURES
acquiesce to the participant’s mand, the caregiver
Trial‑Based Functional Analysis
was likely to deliver reinforcement for the mand to
avoid the occurrence of problem behavior. A trial-based functional analysis capitalizes on
Bowman et al. (1997) conducted a mand analysis naturally occurring events, but also allows the be-
in which the therapist complied with the partici- havior analyst to manipulate variables that poten-
pant’s mands for 1–2 minutes before the test session tially reinforce participant problem behavior in a
started. The therapist stopped complying with the discrete-trial format (Bloom, Iwata, Fritz, Roscoe,
participant’s mands when the session started, and & Carreau, 2011; Sigafoos & Saggers, 1995). Each
occurrences of problem behavior produced 30 sec- trial consists of presentation and removal of the
onds of therapist compliance with the mand. The motivating operation for the putative reinforcer.
therapist complied with the participant’s mands For example, the attention condition of a trial-
in the control condition, and problem behavior based functional analysis in a classroom might
produced no differential consequence. Problem be- consist of sequential 2-minute intervals in which
havior occurred at high rates in the test condition the teacher presents attention in one interval and
and at near-zero rates in the control condition. The removes it in the next interval. Conceptually, the
mand analysis informed a treatment that reduced presentation and removal of teacher attention are
problem behavior to near-zero levels. the test and control conditions, respectively. Ob-
226 B e h av i o r a l A s s e s s m e n t
servers collect data on the occurrence of problem logue functional analyses and functional analyses
behavior per trial, and the behavior analyst con- conducted in natural environments, such as in the
verts occurrences to a percentage of trials. Figure home with the caregiver as therapist. Results of
13.7 shows the results of a trial-based functional the two types of functional analyses corresponded
analysis for David, which suggests that attention for four of five participants. Additional research is
functions as positive reinforcement for aggression. needed to assess the correspondence of analogue
Conducting the assessment in the environ- and naturalistic functional analyses (Hanley et al.,
ment where the behavior occurs is an advantage 2003).
of the trial-based functional analysis. Researchers
have conducted trial-based functional analyses
Latency‑Based Functional Analysis
in classrooms for typically developing children
(Austin, Groves, Raynish, & Francis, 2015) and The occurrence of problem behavior may be nec-
children with autism spectrum disorder (Kodak, essary for a functional analysis to identify the re-
Fisher, Paden, & Dickes, 2013). Moreover, class- inforcer for problem behavior. Some participants,
room teachers and residential staff working with however, exhibit problem behavior that is a dan-
adults with developmental disabilities have imple- ger to themselves, caregivers and staff, property,
mented trial-based functional analyses with good or a combination of these. Measuring latency to
procedural integrity (Bloom, Lambert, Dayton, & the occurrence of problem behavior is one way
Samaha, 2013; Lambert, Bloom, Kunnavatana, to minimize the risk of problem behavior. As in
Collins, & Clay, 2013). other functional-analysis procedures, the behav-
Trial duration may be one limitation of trial- ior analyst presents the motivating operation in
based functional analyses. Exposure to the moti- test condition sessions; however, the session ends
vating operation is short and may be insufficient when the participant engages in problem behav-
to evoke problem behavior for some participants. ior. Shorter latencies to the occurrence of problem
Research thus far, however, has shown that trial- behavior in a test condition relative to the con-
based functional analysis is a relatively simple, ef- trol condition suggest that the tested consequence
ficient, and effective method for identifying the functions as reinforcement for problem behavior.
function of problem behavior in natural environ- Thomason-Sassi, Iwata, Neidert, and Roscoe
ments. (2011) examined response rate and response la-
tency in 38 functional analyses, and found that
rate and latency identified the same reinforcer for
Functional Analysis in the Natural Environment
problem behavior in 33 of 38 cases. Latency-based
The trial-based method is not the only procedure functional analysis may be time-efficient if prob-
for conducting a functional analysis in the natural lem behavior occurs reliably and lawfully because
environment (e.g., Broussard & Northup, 1995; the session ends after the first occurrence of prob-
Ringdahl & Sellers, 2000; Sigafoos & Saggers, lem behavior or after a specific time has elapsed.
1995; Umbreit, 1996). Thomason-Sassi, Iwata, and Figure 13.8 shows the results of a latency-based
Fritz (2013) directly compared the results of ana- functional analysis for Brian. The short latencies
AGGRESSION (% OF TRIALS)
100 Test
80
60
40
Control
20
0
Attention Demand Ignore Tangible
CONDITION
250
200 Tangible
150
100
50
0
5 10 15 20 25
SESSION
to problem behavior in the tangible condition sug- sor behavior is a reasonable approach, particularly
gest that access to tangible items reinforces prob- when the risks outweigh the benefits of conducting
lem behavior. a functional analysis of problem behavior.
lem behavior. For example, in a synthesized con- limitation, in which they trained a function (e.g.,
tingency of contingent escape and access to toys, escape) for a surrogate problem response (e.g., hit-
the behavior analyst cannot determine whether ting a cushioned pad) that had no prior history of
problem behavior is maintained by (1) negative re- reinforcement. They then conducted a traditional
inforcement as an individual contingency, (2) pos- functional analysis, followed by a synthesized con-
itive reinforcement as an individual contingency, tingency analysis based on the IISCA, and then
or (3) the combination of negative and positive re- another traditional functional analysis for this
inforcers as a synthesized contingency. Moreover, newly established response. The traditional func-
researchers use an open-ended interview and in- tional analysis only identified the trained func-
formal observations to identify the contingencies tion of the surrogate problem behavior for all six
they include in an IISCA. The open-ended inter- cases, thus providing support for the validity of
view is of unknown reliability, and informal ob- the traditional functional analysis. By contrast,
servations often identify events that correlate with after the surrogate problem behavior was exposed
but do not reinforce the target behavior (Thomp- to the synthesized contingency analysis, three of
son & Borrero, 2011). the six participants showed a new function of the
Fisher, Greer, Romani, Zangrillo, and Owen surrogate problem behavior during the second tra-
(2016) conducted within-participant comparisons ditional functional analysis. These results suggest
of the results of an IISCA and a traditional func- that synthesizing contingencies in a manner such
tional analysis. An IISCA is based on the assump- as that of the IISCA has the potential to produce
tion that the multiple, synthesized contingencies new functions of problem behavior. Behavior ana-
interact to reinforce problem behavior, rather than lysts should consider this potential side effect of
operating on the behavior independently, but an IISCA before using such synthesized analyses.
it tests neither the interactive nor the indepen-
dent effects of those contingencies. By contrast,
a traditional functional analysis assumes that the CONCLUSION
putative reinforcement contingencies (e.g., atten-
tion, escape) operate independently on problem Functional analysis has emerged as the predomi-
behavior, and it tests this assumption by testing nant method of prescribing effective behavioral
each contingency in isolation. Fisher et al. found treatments for persons with intellectual devel-
that both the IISCA and the functional analysis opmental disorder who display problem behavior
produced clear and differentiated results for four (Repp, 1994). Several investigations have com-
of five participants. In each of these four cases, pared behavioral interventions that are and are
results supported the assumption of a traditional not based on a functional analysis, and the re-
functional analysis that the tested contingen- sults have favored the function-based approach
cies operated in isolation, but not the assump- consistently (Kuhn, DeLeon, Fisher, & Wilke,
tion that the putative reinforcers included in the 1999; Smith, Iwata, Vollmer, & Zarcone, 1992).
IISCA interacted to produce a differential effect In addition, the results across studies included in
on problem behavior. In addition, 6 (55%) of the large-scale meta-analyses have indicated that be-
11 contingencies the IISCA included across these havioral interventions tend to be more effective
four participants appeared to be irrelevant, in that than pharmacological interventions, and that
they did not increase problem behavior relative to function-based behavioral interventions tend to
the individual contingency identified by the tradi- be more effective than non-function-based inter-
tional functional analysis. Greer, Mitteer, Briggs, ventions (Didden, Duker, & Korzilius, 1997; Iwata,
Fisher, and Sodawasser (2020) replicated and ex- Pace, Dorsey, et al., 1994). Although Iwata et al.
tended the findings of Fisher et al. with a larger (1982/1994) originally developed their procedure
cohort and produced equivalent findings. to assess the SIB of individuals with intellectual
One limitation of studies that compare func- developmental disorder, researchers have adapted
tional-analysis procedures, such as an IISCA and this procedure and applied it to ADHD (Northup
a traditional functional analysis, is that the be- et al., 1997), autism spectrum disorder (Fisher,
havior analyst cannot determine the true function Piazza, Alterson, & Kuhn, 1999), breath holding
of problem behavior in the participant’s natural (Kern, Mauk, Marder, & Mace, 1995), bruxism
environment before he or she conducts the analy- (Armstrong, Knapp, & McAdam, 2014; Lang et
ses. Retzlaff, Fisher, Akers, and Greer (2020) con- al., 2013), disruptive behavior (Asmus et al., 1999),
ducted a translational investigation to address this drug ingestion (Chapman, Fisher, Piazza, & Kurtz,
Functional Analysis 229
1993), elopement (Lang et al., 2010; Lehardy, Ler- Asmus, J. M., Wacker, D. P., Harding, J., Berg, W. K.,
man, Evans, O’Connor, & LeSage, 2013; Piazza et Derby, K. M., & Kocis, E. (1999). Evaluation of an-
al., 1997), feeding problems (Piazza et al., 2003), tecedent stimulus parameters for the treatment of
behavior in nonhuman animals (Dorey, Rosales- escape-maintained aberrant behavior. Journal of Ap-
plied Behavior Analysis, 32, 495–513.
Ruiz, Smith, & Lovelace, 2009; Martin, Bloom-
Austin, J. L., Groves, E. A., Reynish, L. C., & Francis, L.
smith, Kelley, Marr, & Maple, 2011), noncompli- L. (2015). Validating trial-based functional analyses
ance (Wilder, Harris, Reagan, & Rasey, 2007), in mainstream primary school classrooms. Journal of
pica and coprophagia (Ing, Roane, & Veenstra, Applied Behavior Analysis, 48, 274–288.
2011; Piazza et al., 1998), physical exercise (Lar- Beavers, G. A., Iwata, B. A., & Lerman, D. C. (2013).
son, Normand, Morley, & Miller, 2013), psychotic Thirty years of research on the functional analysis of
speech (Fisher, Piazza, & Page, 1989), rumination problem behavior. Journal of Applied Behavior Analy-
(Woods, Luiselli, & Tomassone, 2013), and tan- sis, 46, 1–21.
trums (Repp & Karsh, 1994). Researchers also Berkson, G., & Mason, W. A. (1963). Stereotyped move-
have modified functional-analysis procedures to ments of mental defectives: III. Situation effects.
American Journal of Mental Deficiency, 68, 409–412.
improve their accuracy (e.g., Vollmer et al., 1995),
Berkson, G., & Mason, W. A. (1964). Stereotyped
applicability to different populations (e.g., Coo- movements of mental defectives: IV. The effects of
per et al., 1992), applicability to telehealth (e.g., toys and the character of the acts. American Journal
Wacker et al., 2013), efficiency (e.g., Derby et al., of Mental Deficiency, 68, 511–524.
1992; Kahng & Iwata, 1999; Wallace & Iwata, Bloom, S. E., Iwata, B. A., Fritz, J. N., Roscoe, E. M., &
1999), generality to natural environments (e.g., Carreau, A. B. (2011). Classroom application of trial-
Mace & Lalli, 1991; Thomason-Sassi, Iwata, & based functional analysis. Journal of Applied Behavior
Fritz, 2013), and use for identifying idiosyncratic Analysis, 44, 19–31.
reinforcers (e.g., Fisher, Lindauer, et al., 1998; Ros- Bloom, S. E., Lambert, J. M., Dayton, E., & Samaha, A.
coe et al., 2015; Saini, Greer, & Fisher, 2015). L. (2013). Teacher-conducted trial-based functional
analyses as the basis for intervention. Journal of Ap-
In summary, the functional-analysis procedure
plied Behavior Analysis, 46, 208–218.
Iwata et al. (1982/1994) developed arguably rep- Bowman, L. G., Fisher, W. W., Thompson, R. H., &
resents the most important advance in applied Piazza, C. C. (1997). On the relation of mands and
behavior analysis in the last half century. It is a the function of problem behavior. Journal of Applied
prescriptive assessment, in that its results directly Behavior Analysis, 30, 251–265.
inform intervention. Functional analysis generally Broussard, C. D., & Northup, J. (1995). An approach
leads to more effective intervention (Didden et al., to functional assessment and analysis of disruptive
1997; Iwata, Pace, Dorsey, et al., 1994). Functional behavior in regular education classrooms. School Psy-
analysis has led to the development of many inno- chology Quarterly, 10, 151–164.
vative interventions (Bowman et al., 1997; Iwata, Carr, E. G. (1977). The motivation of self-injurious be-
havior: A review of some hypotheses. Psychological
Pace, Cowdery, et al., 1994). Finally, functional
Bulletin, 84, 800–816.
analysis provides a controlled method for conduct-
Carr, E. G., Newsom, C. D., & Binkoff, J. A. (1976).
ing large-scale epidemiological investigations to Stimulus control of self-destructive behavior in a psy-
study environmental influences on problem be- chotic child. Journal of Abnormal Child Psychology,
havior (Iwata, Pace, Dorsey, et al., 1994). 4, 139–153.
Carr, E. G., Newsom, C. D., & Binkoff, J. A. (1980).
Escape as a factor in the aggressive behavior of two
REFERENCES retarded children. Journal of Applied Behavior Analy-
sis, 13, 101–117.
American Psychiatric Association. (2013). Diagnostic Chapman, S., Fisher, W. W., Piazza, C. C., & Kurtz, P.
and statistical manual of mental disorders (5th ed.). F. (1993). Functional assessment and treatment of
Arlington, VA: Author. life-threatening drug ingestion in a dually diagnosed
Armstrong, A., Knapp, V. M., & McAdam, D. B. (2014). youth. Journal of Applied Behavior Analysis, 26, 255–
Functional analysis and treatment of the diurnal 256.
bruxism of a 16-year-old girl with autism. Journal of Conners, J., Iwata, B. A., Kahng, S., Hanley, G. P., Wors-
Applied Behavior Analysis, 47, 415–419. dell, A. S., & Thompson, R. H. (2000). Differential
Asmus, J. M., Franzese, J. C., Conroy, M. A., & Dozier, responding in the presence and absence of discrimi-
C. L. (2003). Clarifying functional analysis outcomes native stimuli during multielement functional analy-
for disruptive behaviors by controlling consequence ses. Journal of Applied Behavior Analysis, 33, 299–308.
delivery for stereotypy. School Psychology Review, 32, Cooper, L. J., Wacker, D. P., Sasso, G. M., Reimers, T.
624–630. M., & Donn, L. K. (1990). Using parents as therapists
230 B e h av i o r a l A s s e s s m e n t
to evaluate appropriate behavior of their children: Fisher, W. W., Ninness, H. A. C., Piazza, C. C., & Owen-
Application to a tertiary diagnostic clinic. Journal of DeSchryver, J. S. (1996). On the reinforcing effects
Applied Behavior Analysis, 23, 285–296. of the content of verbal attention. Journal of Applied
Cooper, L. J., Wacker, D. P., Thursby, D., Plagmann, L. Behavior Analysis, 29, 235–238.
A., Harding, J., Millard, T., et al. (1992). Analysis of Fisher, W. W., O’Connor, J. T., Kurtz, P. F., DeLeon, I.
the effects of task preferences, task demands, and G., & Gotjen, D. L. (2000). The effects of noncon-
adult attention on child behavior in outpatient and tingent delivery of high- and low-preference stimuli
classroom settings. Journal of Applied Behavior Analy- on attention-maintained problem behavior. Journal
sis, 25, 823–840. of Applied Behavior Analysis, 33, 79–83.
Derby, K. M., Hagopian, L., Fisher, W. W., Richman, Fisher, W. W., Piazza, C. C., Alterson, C. J., & Kuhn,
D., Augustine, M., Fahs, A., et al. (2000). Functional D. E. (1999). Interresponse relations among aberrant
analysis of aberrant behavior through measurement behaviors displayed by persons with autism and de-
of separate response topographies. Journal of Applied velopmental disabilities. In P. M. Ghezzi, W. L. Wil-
Behavior Analysis, 33, 113–117. liams, & J. E. Carr (Eds.), Autism: Behavior-analytic
Derby, K. M., Wacker, D. P., Peck, S., Sasso, G., DeRaad, perspectives (pp. 113–135). Reno, NV: Context Press.
A., Berg, W., et al. (1994). Functional analysis of Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L.
separate topographies of aberrant behavior. Journal P., Owens, J. C., & Slevin, I. (1992). A comparison of
of Applied Behavior Analysis, 27, 267–278. two approaches for identifying reinforcers for persons
Derby, K. M., Wacker, D. P., Sasso, G., Steege, M., Nor- with severe and profound disabilities. Journal of Ap-
thup, J., Sigrand, K., et al. (1992). Brief functional as- plied Behavior Analysis, 25, 491–498.
sessment techniques to evaluate aberrant behavior in Fisher, W. W., Piazza, C. C., Bowman, L. G., Hanley, G.
an outpatient setting: A summary of 79 cases. Journal P., & Adelinis, J. D. (1997). Direct and collateral ef-
of Applied Behavior Analysis, 25, 713–721. fects of restraints and restraint fading. Journal of Ap-
Didden, R., Duker, P. C., & Korzilius, H. (1997). Meta- plied Behavior Analysis, 30, 105–120.
analytic study on treatment effectiveness for problem Fisher, W. W., Piazza, C. C., Cataldo, M., Harrell, R., Jef-
behaviors with individuals who have mental retarda- ferson, G., & Conner, R. (1993). Functional commu-
tion. American Journal on Mental Retardation, 101, nication training with and without extinction and
387–399. punishment. Journal of Applied Behavior Analysis, 26,
Dorey, N. R., Rosales-Ruiz, J., Smith, R., & Lovelace, 23–36.
B. (2009). Functional analysis and treatment of self- Fisher, W. W., Piazza, C. C., & Page, T. J. (1989). Assess-
injury in a captive olive baboon. Journal of Applied ing independent and interactive effects of behavioral
Behavior Analysis, 42, 785–794. and pharmacologic interventions for a client with
Dracobly, J. D., & Smith, R. G. (2012). Progressing from dual diagnoses. Journal of Behavior Therapy and Ex-
identification and functional analysis of precursor perimental Psychiatry, 20, 241–250.
behavior to treatment of self-injurious behavior. Jour- Fritz, J. N., Iwata, B. A., Hammond, J. L., & Bloom, S. E.
nal of Applied Behavior Analysis, 45, 361–374. (2013). Experimental analysis of precursors to severe
Fahmie, T. A., Iwata, B. A., Harper, J. M., & Querim, A. problem behavior. Journal of Applied Behavior Analy-
C. (2013). Evaluation of the divided attention con- sis, 46, 101–129.
dition during functional analyses. Journal of Applied Ghaemmaghami, M., Hanley, G. P., Jin, S. C., & Van-
Behavior Analysis, 46, 71–78. selow, N. R. (2016). Affirming control by multiple
Finney, J. W., Russo, D. C., & Cataldo, M. F. (1982). Re- reinforcers via progressive treatment analysis. Behav-
duction of pica in young children with lead poison- ioral Interventions, 31, 70–86.
ing. Journal of Pediatric Psychology, 7, 197–207. Greer, B. D., Mitteer, D. R., Briggs, A. M., Fisher, W.
Fisher, W. W., Adelinis, J. D., Thompson, R. H., Wors- W., & Sodawasser, A. J. (2020). Comparisons of
dell, A. S., & Zarcone, J. R. (1998). Functional anal- standardized and interview-informed synthesized
ysis and treatment of problem behavior maintained reinforcement contingencies relative to functional
by termination of “don’t” (and symmetrical “do”) analysis. Journal of Applied Behavior Analysis, 53,
requests. Journal of Applied Behavior Analysis, 31, 82–101.
339–356. Hammond, J. L., Iwata, B. A., Rooker, G. W., Fritz, J. N.,
Fisher, W. W., Greer, B. D., Romani, P. W., Zangrillo, A. & Bloom, S. E. (2013). Effects of fixed versus random
N., & Owen, T. M. (2016). Comparisons of synthe- condition sequencing during multielement function-
sized- and individual-reinforcement contingencies al analyses. Journal of Applied Behavior Analysis, 46,
during functional analysis. Journal of Applied Behav- 22–30.
ior Analysis, 49, 596–616. Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003).
Fisher, W. W., Lindauer, S. E., Alterson, C. J., & Functional analysis of problem behavior: A review.
Thompson, R. H. (1998). Assessment and treatment Journal of Applied Behavior Analysis, 36, 147–185.
of problem behavior maintained by stereotypic object Hanley, G. P., Jin, C. S., Vanselow, N. R., & Hanratty,
manipulation. Journal of Applied Behavior Analysis, L. A. (2014). Producing meaningful improvements
31, 513–527. in problem behavior of children with autism via syn-
Functional Analysis 231
thesized analyses and treatments. Journal of Applied analysis in early intervention classrooms. Journal of
Behavior Analysis, 47, 16–36. Applied Behavior Analysis, 46, 301–306.
Hagopian, L. P., Fisher, W. W., Thompson, R. H., Owen- Kodak, T., Northup, J., & Kelley, M. E. (2007). An
DeSchryver, J., Iwata, B. A., & Wacker, D. P. (1997). evaluation of the types of attention that maintain
Toward the development of structured criteria for problem behavior. Journal of Applied Behavior Analy-
interpretation of functional analysis data. Journal of sis, 40, 167–171.
Applied Behavior Analysis, 30, 313–326. Kuhn, D. E., DeLeon, I. G., Fisher, W. W., & Wilke,
Hagopian, L. P., Rooker, G. W., Jessel, J., & DeLeon, I. A. E. (1999). Clarifying an ambiguous functional
G. (2013). Initial functional analysis outcomes and analysis with matched and mismatched extinction
modifications in pursuit of differentiation: A summa- procedures. Journal of Applied Behavior Analysis, 32,
ry of 176 inpatient cases. Journal of Applied Behavior 99–102.
Analysis, 46, 88–100. Lambert, J. M., Bloom, S. E., Kunnavatana, S. S., Col-
Ing, A. D., Roane, H. S., & Veenstra, R. A. (2011). lins, S. D., & Clay, C. J. (2013). Training residential
Functional analysis and treatment of coprophagia. staff to conduct trial-based functional analyses. Jour-
Journal of Applied Behavior Analysis, 44, 151–155. nal of Applied Behavior Analysis, 46, 296–300.
Iwata, B. A. (1987). Negative reinforcement in applied Lang, R., Davenport, K., Britt, C., Ninci, J., Garner, J.,
behavior analysis: An emerging technology. Journal & Moore, M. (2013). Functional analysis and treat-
of Applied Behavior Analysis, 20, 361–378. ment of diurnal bruxism. Journal of Applied Behavior
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Analysis, 46, 322–327.
Richmand, G. S. (1994). Toward a functional analy- Lang, R., Davis, T., O’Reilly, M., Machalicek, W., Risp-
sis of self-injury. Journal of Applied Behavior Analysis, oli, M., Sigafoos, J., et al. (2010). Functional analy-
27, 197–209. (Reprinted from Analysis and Interven- sis and treatment of elopement across two school
tion in Developmental Disabilities, 2, 3–20, 1982) settings. Journal of Applied Behavior Analysis, 43,
Iwata, B. A., Duncan, B. A., Zarcone, J. R., Lerman, D. 113–118.
C., & Shore, B. A. (1994). A sequential, test-control Laraway, S., Snycerski, S., Michael, J., & Poling, A.
methodology for conducting functional analyses of (2003). Motivating operations and terms to describe
self-injurious behavior. Behavior Modification, 18, them: Some further refinements. Journal of Applied
289–306. Behavior Analysis, 36, 407–414.
Iwata, B. A., Pace, G. M., Cowdery, G. E., & Milten- Larson, T. A., Normand, M. P., Morley, A. J., & Miller,
berger, R. G. (1994). What makes extinction work: B. G. (2013). A functional analysis of moderate-to-
An analysis of procedural form and function. Journal vigorous physical activity in young children. Journal
of Applied Behavior Analysis, 27, 131–144. of Applied Behavior Analysis, 46, 199–207.
Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R., Lehardy, R. K., Lerman, D. C., Evans, L. M., O’Connor,
Vollmer, T. R., Smith, R. G., et al. (1994). The func- A., & LeSage, D. L. (2013). A simplified methodol-
tions of self-injurious behavior: An experimental– ogy for identifying the function of elopement. Journal
epidemiological analysis. Journal of Applied Behavior of Applied Behavior Analysis, 46, 256–270.
Analysis, 27, 215–240. Lovaas, O. I., Freitag, G., Gold, V. J., & Kassorla, I. C.
Jessel, J., Hanley, G. P., & Ghaemmaghami, M. (2016). (1965). Experimental studies in childhood schizo-
Interview-informed synthesized contingency analy- phrenia: Analysis of self-destructive behavior. Jour-
ses: Thirty replications and reanalysis. Journal of Ap- nal of Experimental Child Psychology, 2, 67–84.
plied Behavior Analysis, 49, 576–595. Lovaas, O. I., & Simmons, J. Q. (1969). Manipulation of
Kahng, S., Abt, K. A., & Schonbachler, H. E. (2001). self-destruction in three retarded children. Journal of
Assessment and treatment of low-rate high intensity Applied Behavior Analysis, 2, 143–157.
problem behavior. Journal of Applied Behavior Analy- Mace, F. C., & Lalli, J. S. (1991). Linking descriptive
sis, 34, 225–228. and experimental analyses in the treatment of bi-
Kahng, S., Hausman, N. L., Fisher, A. B., Donaldson, J. zarre speech. Journal of Applied Behavior Analysis, 24,
M., Cox, J. R., Lugo, M., et al. (2015). The safety of 553–562.
functional analyses of self-injurious behavior. Journal Martin, A. L., Bloomsmith, M. A., Kelley, M. E., Marr,
of Applied Behavior Analysis, 48, 107–114. M. J., & Maple, T. L. (2011). Functional analysis and
Kahng, S., & Iwata, B. A. (1999). Correspondence be- treatment of human-directed undesirable behavior
tween outcomes of brief and extended functional exhibited by a captive chimpanzee. Journal of Applied
analyses. Journal of Applied Behavior Analysis, 32, Behavior Analysis, 44, 139–143.
149–159. McComas, J., Hoch, H., Paone, D., & El-Roy, D.
Kern, L., Mauk, J. E., Marder, T. J., & Mace, F. C. (2000). Escape behavior during academic tasks: A
(1995). Functional analysis and intervention for preliminary analysis of idiosyncratic establishing
breath-holding. Journal of Applied Behavior Analysis, operations. Journal of Applied Behavior Analysis, 33,
28, 339–340. 479–493.
Kodak, T., Fisher, W. W., Paden, A., & Dickes, N. (2013). Northup, J., Jones, K., Broussard, C., DiGiovanni, G.,
Evaluation of the utility of a discrete-trial functional Herring, M., Fusilier, I., et al. (1997). A preliminary
232 B e h av i o r a l A s s e s s m e n t
analysis of interactive effects between common class- Roscoe, E. M., Schlichenmeyer, K. J., & Dube, W. V.
room contingencies and methylphenidate. Journal of (2015). Functional analysis of problem behavior: A
Applied Behavior Analysis, 30, 121–125. systematic approach for identifying idiosyncratic
Piazza, C. C., Bowman, L. G., Contrucci, S. A., Delia, variables. Journal of Applied Behavior Analysis, 48,
M. D., Adelinis, J. D., & Goh, H. (1999). An evalu- 289–314.
ation of the properties of attention as reinforcement Saini, V., Fisher, W. W., & Retzlaff, B. J. (2018). Pre-
for problem and appropriate behavior. Journal of Ap- dictive validity and efficiency of ongoing visual-in-
plied Behavior Analysis, 32, 437–449. spection criteria for interpreting functional analyses.
Piazza, C. C., Fisher, W. W., Brown, K. A., Shore, B. A., Journal of Applied Behavior Analysis, 51, 303–320.
Patel, M. R., Katz, R. M., et al. (2003). Functional Saini, V., Fisher, W. W., Retzlaff, B. J., & Keevy, M. D.
analysis of inappropriate mealtime behaviors. Journal (2020). Efficiency in functional analysis of problem
of Applied Behavior Analysis, 36, 187–204. behavior: A quantitative and qualitative review. Jour-
Piazza, C. C., Fisher, W. W., Hanley, G. P., LeBlanc, nal of Applied Behavior Analysis, 53, 44–66.
L. A., Worsdell, A. S., Lindauer, S. E., et al. (1998). Saini, V., Greer, B. D., & Fisher, W. W. (2015). Clari-
Treatment of pica through multiple analyses of its fying inconclusive functional analysis results: As-
reinforcing functions. Journal of Applied Behavior sessment and treatment of automatically reinforced
Analysis, 31, 165–189. aggression. Journal of Applied Behavior Analysis, 48,
Piazza, C. C., Hanley, G. P., Bowman, L. G., Ruyter, J. 315–330.
M., Lindauer, S. E., & Saiontz, D. M. (1997). Func- Schlichenmeyer, K. J., Roscoe, E. M., Rooker, G. W.,
tional analysis and treatment of elopement. Journal Wheeler, E. E., & Dube, W. V. (2013). Idiosyncratic
of Applied Behavior Analysis, 30, 653–672. variables that affect functional analysis outcomes:
Querim, A. C., Iwata, B. A., Roscoe, E. M., Schlichen- A review (2001–2010). Journal of Applied Behavior
meyer, K. J., Ortega, J. V., & Hurl, K. E. (2013). Func- Analysis, 46, 339–348.
tional analysis screening for problem behavior main- Shirley, M. J., Iwata, B. A., & Kahng, S. (1999). False-
tained by automatic reinforcement. Journal of Applied positive maintenance of self-injurious behavior by
Behavior Analysis, 46, 47–60. access to tangible reinforcers. Journal of Applied Be-
Repp, A. (1994). Comments on functional analysis pro- havior Analysis, 32, 201–204.
cedures for school-based behavior problems. Journal Sigafoos, J., & Saggers, E. A. (1995). Discrete-trial ap-
of Applied Behavior Analysis, 27, 409–411. proach to the functional analysis of aggressive be-
Repp, A. C., & Karsh, K. G. (1994). Hypothesis-based haviour in two boys with autism. Australian and New
interventions for tantrum behaviors of persons with Zealand Journal of Developmental Disabilities, 20,
developmental disabilities in school settings. Journal 287–297.
of Applied Behavior Analysis, 27, 21–31. Skinner, B. F. (1953). Science and human behavior. New
Retzlaff, B. J., Fisher, W. W., Akers, J. S., & Greer, B. D. York: Free Press.
(2020). A translational evaluation of potential iatro- Smith, R. G., Iwata, B. A., Goh, H. L., & Shore, B. A.
genic effects of single and combined contingencies (1995). Analysis of establishing operations for self-
during functional analysis of target responses. Journal injury maintained by escape. Journal of Applied Be-
of Applied Behavior Analysis, 53, 67–81. havior Analysis, 28, 515–535.
Rincover, A. (1978). Sensory extinction: A procedure Smith, R. G., Iwata, B. A., Vollmer, T. R., & Zarcone,
for eliminating self-stimulatory behavior in devel- J. R. (1992). Experimental analysis and treatment of
opmentally disabled children. Journal of Abnormal multiply controlled self-injury. Journal of Applied Be-
Child Psychology, 6, 299–310. havior Analysis, 26, 183–196.
Ringdahl, J. E., & Sellers, J. A. (2000). The effects of Sulzer-Azaroff, B., & Mayer, G. (1977). Applying behav-
different adults as therapists during functional analy- ior analysis procedures with children and youth. Austin,
ses. Journal of Applied Behavior Analysis, 33, 247–250. TX: Holt, Rinehart & Winston.
Roane, H. S., Fisher, W. W., Kelley, M. E., Mevers, J. Thomason-Sassi, J. L., Iwata, B. A., & Fritz, J. N. (2013).
L., & Bouxsein, K. J. (2013). Using modified visual- Therapist and setting influences on functional anal-
inspection criteria to interpret functional analysis ysis outcomes. Journal of Applied Behavior Analysis,
outcomes. Journal of Applied Behavior Analysis, 46, 46, 79–87.
130–146. Thomason-Sassi, J. L., Iwata, B. A., Neidert, P. L., &
Roane, H. S., Lerman, D. C., Kelley, M. E., & Van Camp, Roscoe, E. M. (2011). Response latency as an index
C. M. (1999). Within-session patterns of responding of response strength during functional analyses of
during functional analyses: The role of establishing problem behavior. Journal of Applied Behavior Analy-
operations in clarifying behavioral function. Re- sis, 44, 51–67.
search in Developmental Disabilities, 20, 73–89. Thompson, R. H., & Borrero, J. C. (2011). Direct obser-
Rooker, G. W., Iwata, B. A., Harper, J. M., Fahmie, T. vation. In W. W. Fisher, C. C. Piazza, & H. S. Roane
A., & Camp, E. M. (2011). False-positive tangible (Eds.), Handbook of applied behavior analysis (pp. 191–
outcomes of functional analyses. Journal of Applied 205). New York: Guilford Press.
Behavior Analysis, 44, 737–745. Thompson, R. H., Fisher, W. W., Piazza, C. C., & Kuhn,
Functional Analysis 233
D. E. (1998). The evaluation and treatment of ag- Vollmer, T. R., Marcus, B. A., Ringdahl, J. E., & Roane,
gression maintained by attention and automatic re- H. S. (1995). Progressing from brief assessments to
inforcement. Journal of Applied Behavior Analysis, 31, extended experimental analyses in the evaluation of
103–116. aberrant behavior. Journal of Applied Behavior Analy-
Thompson, R. J., & Iwata, B. A. (2001). Functional sis, 28, 561–576.
analysis and treatment of self-injury associated with Wacker, D. P., Lee, J. F., Dalmau, Y. C. P., Kopelman,
transitions. Journal of Applied Behavior Analysis, 34, T. G., Lindgren, S. D., Kuhle, J., et al. (2013). Con-
195–210. ducting functional analyses of problem behavior via
Umbreit, J. (1996). Functional analysis of disruptive telehealth. Journal of Applied Behavior Analysis, 46,
behavior in an inclusive classroom. Journal of Early 31–46.
Intervention, 20, 18–29. Wallace, M. D., & Iwata, B. A. (1999). Effects of session
Van Camp, C. M., Lerman, D. C., Kelley, M. E., Roane, duration on functional analysis outcomes. Journal of
H. S., Contrucci, S. A., & Vorndran, C. M. (2000). Applied Behavior Analysis, 32, 175–183.
Further analysis of idiosyncratic antecedent influ- Weeden, M., Mahoney, A., & Poling, A. (2010). Self-in-
ences during the assessment and treatment of prob- jurious behavior and functional analysis: Where are
lem behavior. Journal of Applied Behavior Analysis, the descriptions of participant protections? Research
33, 207–221. in Developmental Disabilities, 31, 299–303.
Van Houten, R., & Rolider, A. (1988). Recreating the Wilder, D. A., Harris, C., Reagan, R., & Rasey, A.
scene: An effective way to provide delayed punish- (2007). Functional analysis and treatment of non-
ment for inappropriate motor behavior. Journal of Ap- compliance by preschool children. Journal of Applied
plied Behavior Analysis, 21, 187–192. Behavior Analysis, 40, 173–177.
Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. G., Woods, K. E., Luiselli, J. K., & Tomassone, S. (2013).
& Mazaleski, J. L. (1993). Within-session patterns of Functional analysis and intervention for chronic ru-
self-injury as indicators of behavioral function. Re- mination. Journal of Applied Behavior Analysis, 46,
search in Developmental Disabilities, 14, 479–492. 328–332.
PAR T V
INTERVENTIONS FOR INCREASING
DESIRABLE BEHAVIOR
A pplied behavior analysis has a rich history and an array of procedures to pro-
mote skill development and increase desirable behavior across a broad range of
learners. The topics covered in Part V reflect the breadth of these procedures and
progress from teaching alternatives to problem behavior, to teaching novel verbal
behavior, to teaching complex, multistep behavior repertoires, to providing staff
training.
In Chapter 14, Tiger and Hanley discuss the topic of differential reinforce-
ment and the many ways in which differential-reinforcement paradigms may be
used to promote desirable behavior and decrease maladaptive behavior. Attention
is given to the subtypes of differential-reinforcement procedures, their various
applications, and the conditions under which some procedures might be war-
ranted. For their current chapter, the authors have expanded their discussion in
the first edition to summarize recent research on differential reinforcement and
describe tools to facilitate programming for maintenance and increasing delays to
reinforcement.
It is common for behavior analysts to focus on promoting extended behav-
ioral sequences or chains (e.g., assembling a bicycle). In Chapter 15, Noell, Call,
Ardoin, and Miller have updated their first-edition chapter on building complex
behaviors through stimulus control, chaining, and strategic behavior. New for this
edition is an increased focus on the conceptualization of complex behavior rep-
ertoires, as well as new information on promoting maintenance, generalization,
self-management, and response variability.
Since the first edition of this handbook appeared, there have been several
advances in the area of verbal behavior. Chapter 16, by Tincani, Miguel, Bondy,
and Crozier, summarizes much of this recent research. The chapter begins by
reviewing Skinner’s formulation of verbal behavior before reviewing applications
for developing verbal behavior repertoires. It then provides a summary of recent
235
236 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
Differential-Reinforcement Procedures
Jeffrey H. Tiger and Gregory P. Hanley
The term reinforcement describes the process in ner, particularly when reinforcers for engaging in
which a behavior strengthens when its occurrence important behavior are delayed or intermittent, or
is followed by some improvement in the environ- when a chain of behavior is necessary to produce
ment. By strengthened, we mean that the behavior reinforcement. Imagine trying to learn to drive
is more likely to occur in the future in similar en- a manual-transmission car based solely on the
vironmental conditions. The process of reinforce- natural consequences of that behavior. Two dis-
ment is fundamental to the way people interact tinguishing capacities of humans are the abilities
with and learn from their environment. For in- to relay personal learning histories to other people
stance, children repeat phrases that made their through verbal behavior, such as speech and writ-
parents laugh; teenagers wear the same clothes ing, and to arrange contingencies to develop and
that made their friends take notice; and adults refine important behaviors in others. Thus we can
swing a golf club with a particular form when increase the speed at which important behavior
doing so has produced long and accurate drives. develops and eventually contacts natural rein-
We can understand much of early human learn- forcement contingencies. In this regard, differential
ing by acknowledging the regular, natural, and reinforcement is applicable as a procedural term to
often accidental reinforcement and punishment describe the act of increasing the occurrence of a
contingencies that infants experience (Bijou, desirable behavior in others by arranging for im-
1996; Schlinger, 1995). For example, an infant girl provements to follow such behavior.
may experience reinforcers for grasping her food By arranging for reinforcers to occur more
only when she applies the appropriate amount of often following one behavior than following an-
grip strength. Grasping too hard will squash the other, differential reinforcement has two effects:
food or cause it to slip from her hands. Grasping It strengthens the target behavior and weakens
too softly will not capture the food. Such gradual other behavior that is functionally similar. Given
and natural reinforcement processes may at least this latter effect, investigators have used differ-
partially account for learning to reach, grasp, and ential reinforcement to reduce problem behavior
then chew, and other important behaviors such as (see Vollmer, Athens, & Fernand, Chapter 19,
babbling, standing, and walking. this volume). By many accounts, differential-
Although natural contingencies may account reinforcement procedures have revolutionized the
for much of human learning, alone they may educational and care practices for young children,
change behavior in a slow and inefficient man- especially children with intellectual developmen-
237
238 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
tal disorder and severe problem behaviors (Risley, This simple manipulation produced increased ap-
2005). However, the accelerative effects of differ- propriate peer interactions and decreased occur-
ential reinforcement are also valuable for design- rences of aggression.
ing teaching and habilitative environments, and In differential reinforcement of other behavior
our chapter focuses primarily on the use of dif- (DRO), by contrast, reinforcement is arranged for
ferential reinforcement to develop and refine new periods in which target behavior does not occur,
behavior and to maintain this behavior in many and this may produce shifts from one topography
settings. to another. For instance, Protopopova, Kisten, and
Differential reinforcement as a procedure is Wynne (2016) delivered food remotely to dogs that
deceptively simple: Identify a behavior you would historically engaged in high rates of nuisance bark-
like to occur more often, arrange reinforcers to fol- ing after periods in which no barking occurred.
low the occurrence of the behavior or features of This DRO schedule eliminated nuisance barking
the behavior, and do not present these same rein- for four of five dogs. DRO is not as precise as DRA
forcers following occurrences of other behaviors. for strengthening target behavior, and response
Socially important behavior change, however, is topographies that the omission contingency did
often not that simple. Behavior analysts have de- not target may emerge and be strengthened (Jes-
veloped a comprehensive technology for increas- sel, Borrero, & Becraft, 2015; Jessel & Ingvarsson,
ing desirable behavior through differential rein- 2016).
forcement and have used this technology since the
inception of the field in the early 1960s. We review
Rate
those technological developments in this chapter.
Specifically, we provide descriptions and examples Rate is the number of responses emitted in a cer-
of features of behavior that behavior analysts may tain period. Some responses must occur repeatedly
strengthen through differential reinforcement and in a period to be useful or functional (e.g., typing
highlight considerations for analysts designing speed, answering math facts). Much differential-
differential-reinforcement-based interventions. In reinforcement research focuses on increasing the
addition, we highlight the diverse array of applica- rate of various socially important behaviors. In a
tions with differential reinforcement at their core. recent creative example, Stasolla et al. (2017) used
automated differential reinforcement to increase
the ambulation rate of two girls with multiple dis-
FEATURES OF BEHAVIOR TO TARGET abilities. When optic sensors detected a forward
WITH DIFFERENTIAL REINFORCEMENT step, the automated device provided brief access
to music, lights, or tactile vibration, and this ar-
In this section, we define features of behavior that rangement produced large increases in ambulation
are sensitive to differential reinforcement and pro- rate. Furthermore, these girls showed higher indi-
vide illustrative examples of how differential rein- ces of happiness when ambulation produced rein-
forcement has modified these features. forcement than when the same reinforcers were
available noncontingently. These findings are like
those of studies in which children demonstrated
Topography
a preference for differential over noncontingent
Common uses of differential reinforcement in- reinforcement during concurrent-chain schedules
volve reinforcement of appropriate behavior in (Hanley, Piazza, Fisher, Contrucci, & Maglieri,
lieu of problem behavior. We often refer to this 1997; Luczynski & Hanley, 2009, 2010).
procedure as differential reinforcement of alternative When the base rate of a behavior is insuffi-
behavior (DRA). Pinkston, Reese, LeBlanc, and cient, differential reinforcement of high-rate behavior
Baer (1973) provided an example of DRA for ap- (DRH) can produce acceleration in the behavior’s
propriate peer interactions in lieu of aggression. In base rate. A DRH schedule arranges reinforce-
baseline, teachers typically responded to instances ment delivery if the participant emits a minimum
of peer aggression with reprimands (e.g., “You can’t number of responses before the end of a specified
do that here!”) and responded infrequently to ap- interval. Ingham and Andrews (1973) used a pro-
propriate social interaction, resulting in relatively cedure to treat stuttering that we can conceptual-
high rates of aggression. The investigators then ize as a DRH schedule. The investigators treated
taught the teachers to withhold attention follow- participants for stuttering with auditory feedback
ing aggression and to provide attention when the in which a tone sounded when the participant
children engaged in desirable peer interactions. stuttered. This treatment produced stutter-free
Differential-Reinforcement Procedures 239
speech, but the speech was slow and unnatural, iors, but behavior analysts can also use differential
according to the investigators. Ingham and An- reinforcement to sustain responding. Miller and
drews then delivered token reinforcement for pro- Kelley (1994) taught parents to use differential re-
gressively higher rates of spoken words. This DRH inforcement to sustain the homework engagement
maintained stutter-free speech and increased the of four school-age children. After the parent and
rate and naturalness of the spoken words. child set a goal for such engagement, the parent
In other cases, certain behaviors are socially ac- provided access to preferred activities when the
ceptable only when they occur at moderate to low child met or exceeded the goal. Investigators also
rates. For instance, recruiting teacher attention is have used DRO schedules to increase the duration
a common and desirable behavior of young chil- of other important behavior. For instance, Cox,
dren and is a common target for children who do Virues-Ortega, Julio, and Martin (2017) arranged
not demonstrate this skill (e.g., Stokes, Fowler, & DRO schedules to reduce the excessive motion
Baer, 1978). However, children who make frequent of children with autism spectrum disorder during
bids for attention can be disruptive to typical class- magnetic resonance imaging (MRI). The DRO
room environments. In differential reinforcement of schedule arranged reinforcement for movement-
low-rate behavior (DRL), reinforcement is arranged free intervals, and participants learned to lie still
when a behavior occurs below a certain threshold. up to 5 minutes during mock MRI sessions. These
Investigators have used DRL schedules to main- findings are important because excessive move-
tain moderate or low rates of behavior, frequently ment produces unusable MRI results, with the sub-
as an initial treatment for problem behavior. For sequent need to repeat this expensive procedure to
instance, Austin and Bevan (2011) described what obtain usable results.
they called a full-session DRL procedure with three
elementary school students. The classroom teacher
Intensity
set a maximum-response criterion for each student,
such as nine responses in a 20-minute session, and Intensity is the physical force or magnitude of
each student who made fewer requests than their the target response. For instance, the volume at
individualized maximum received a point that he which an individual emits speech is integral to a
or she could use in the classroom behavior man- conversation partner’s ability to respond. An in-
agement system. This procedure reduced request- dividual who speaks too softly may not be heard,
ing behavior to levels more appropriate for the and excessively loud speech may be aversive to the
classroom. Unlike forms of differential reinforce- listener. Fleece et al. (1981) demonstrated the use
ment that target zero or near-zero levels of a be- of differential reinforcement of response intensity
havior, DRL schedules may maintain behavior at to increase the speech volume of two preschool
low rates, but see Jessel and Borrero (2014) and Be- children with intellectual developmental disorder.
craft, Borrero, Davis, Mendres-Smith, and Castillo Investigators used a sound-sensitive apparatus that
(2018) for laboratory-based studies including varia- they calibrated to respond to participant vocaliza-
tions of DRL schedules that produced response tions that exceeded a minimum threshold by pro-
maintenance relative to response suppression. ducing red- and green-colored lights in the shape
of a Christmas tree, which was a presumed rein-
forcer. The investigators increased the minimum
Duration
threshold for reinforcement as the children suc-
Duration is the amount of time a participant per- cessfully activated the device. The speech volume
forms a behavior. For behaviors such as completing of the participants increased but did not exceed
homework, exercising, or reading, the number of the speech volume of their peers.
instances of behavior is less informative than the
amount of time a participant performs the behav-
Latency
ior. For instance, knowing that a student studied
for 3 hours in the past week is likely more infor- Latency is the amount of time that passes between
mative than knowing that the student studied on the occurrence of some event and the completion
three occasions, particularly if the three occasions of behavior. For instance, we might define latency
lasted only 30 seconds each. Thus response dura- to awakening as the time between an alarm clock’s
tion may be a more important target than response sounding and a person’s getting out of bed. Tiger,
frequency in these cases. Bouxsein, and Fisher (2007) used differential re-
Our previous examples of differential reinforce- inforcement of response latencies with an adult
ment target increased frequency or speed of behav- with Asperger syndrome who displayed delayed
240 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
responding to questions. The investigators asked Sprague, O’Brien, and Heathfield (1990) showed
the participant for information, such as his sib- the importance of response effort when teaching
lings’ names and their addresses, during baseline. alternative communicative responses to two par-
The participant responded accurately, but mean ticipants who engaged in socially mediated prob-
response latency was 24 seconds. The investigators lem behavior. Acquisition was slow and incom-
then arranged a differential-reinforcement contin- plete, and problem behavior persisted when the
gency in which the participant earned tokens ex- investigators required participants to type a full
changeable for access to a movie for each question sentence on an augmentative-communication de-
he answered within an identified latency. By the vice to access reinforcement. Participants learned
third differential-reinforcement session, the par- and maintained a less effortful alternative (press-
ticipant’s mean latency to respond was 5 seconds. ing a key to generate the same sentence) more
quickly, and problem behavior decreased and re-
mained low. When speed of acquisition is critical,
Interresponse Time
decreasing response effort is an important tactic
Interresponse time is the time between two instanc- to consider. The behavior analyst may still teach
es of a response. Differential reinforcement of short more effortful and complex responses by first ar-
interresponse times produces rapid responding ranging differential-reinforcement contingencies
(i.e., short pausing between similar responses) and for less effortful or more simple responses, and
differential reinforcement of longer interresponse then gradually increasing the response effort and
times produces slow responding (i.e., greater paus- response complexity required to access reinforce-
ing between similar responses). For instance, Len- ment (see Hernandez, Hanley, Ingvarsson, &
nox, Miltenberger, and Donnelly (1987) reduced Tiger, 2007, for an example of this strategy).
the rapid eating of three adults with profound in-
tellectual developmental disorder by differentially
Immediacy of Reinforcers
reinforcing long interresponse times between con-
sumption of bites of food. The investigators used Reinforcer immediacy or reinforcer contiguity is the
baseline data to set a target interresponse time of time between an instance of behavior and re-
15 seconds. They blocked participants’ attempts inforcement delivery (Vollmer & Hackenberg,
to place food in the mouth more often than once 2001). Individuals may acquire responses when
every 15 seconds. Lennox et al. also prompted the considerable time expires between the response
participant to engage in an incompatible response and a reinforcing event (i.e., acquisition under de-
during the 15-second interval, and participants’ layed-reinforcement conditions; Gleeson & Lattal,
rate of bite consumption decreased. 1987), and short delays may sometimes increase
response persistence for primary reinforcers (Leon,
Borrero, & DeLeon, 2016). The acquisition pro-
CONSIDERATIONS FOR cess is usually substantially longer or incomplete,
DIFFERENTIAL‑REINFORCEMENT PROCEDURES however, even with brief delays (Carroll, Kodak,
& Adolf, 2016; Gleeson & Lattal, 1987). The con-
Behavior analysts can implement differential rein- tingency-weakening effects of delayed reinforce-
forcement in many ways. Several parameters of the ment are well documented (Fisher, Thompson,
response–reinforcer relation affect the likelihood Hagopian, Bowman, & Krug, 2000; Hanley, Iwata,
of differential reinforcement’s effectiveness. These & Thompson, 2001), and sometimes a single in-
include the effort of the target response and the stance of immediate reinforcement will strength-
immediacy, schedule, magnitude, type, and quality en a response (Skinner, 1948). Thus ensuring the
of reinforcement. We discuss each of these param- immediate delivery of reinforcement following a
eters below. target behavior is critical for rapidly increasing
the behavior through differential reinforcement
(Hanley et al., 2001).
Response Effort
Delays to social and tangible reinforcement are
The response effort is likely to affect the rate at inevitable outside of highly resourced teaching
which an individual learns a response. Individuals conditions, however. Differential reinforcement
acquire responses with lower effort more quickly is still essential for generating and maintaining
than those with higher effort, and simple respons- important behavior under these conditions. In an
es more quickly than complex responses. Horner, early example, Lalli, Casey, and Kates (1995) used
Differential-Reinforcement Procedures 241
differential reinforcement of progressively increas- ery; the organism then pauses for a period before
ing chains of responses to strengthen task comple- high-rate responding resumes (Ferster & Skinner,
tion and maintain functional communication, 1957; Orlando & Bijou, 1960).
despite consistent delay to reinforcement. Ghaem- Variable-ratio (VR) schedules arrange reinforce-
maghami, Hanley, and Jessel (2016) extended this ment delivery around a mean number of responses
work by showing that socially important behavior that changes from trial to trial. For instance, rein-
such as functional communication, tolerance, and forcement delivery would occur after a mean of five
compliance with instructions maintained de- responses in a VR 5 schedule. Thus, reinforcement
spite long delays to reinforcement by (1) provid- delivery might occur after one, three, five, seven,
ing immediate reinforcement for each behavior or nine responses. VR schedules tend to produce
type at least intermittently, and (2) progressively high response rates without postreinforcement
strengthening chains of appropriate behavior with pauses, and behavior analysts often use them for
the contingent termination of the delay versus response maintenance (Ferster & Skinner, 1957;
time-based termination of the delay. This pro- Schlinger, Derenne, & Baron, 2008).
cess strengthens initial behaviors in the response Progressive-ratio schedules arrange reinforce-
chain, even though initial behaviors do not con- ment delivery on a schedule that changes across
tact much immediate reinforcement. The process reinforcer deliveries. These schedules progress ei-
also mitigates resurgence of problem behavior dur- ther by the addition of a fixed number of responses
ing delays by strengthening appropriate behavior (arithmetic increases) or by multiplying each pro-
during the delay. Thus the appropriate behavior gressive-schedule value by a constant (geometric
that occurs during the delay is available for rein- increases). For instance, an investigator might use
forcement when the delay ends. Investigators also a geometric progressive-ratio schedule in which
have shown that procedures that develop behavior the number of responses required to produce re-
chains mitigate the untoward effects of delays to inforcement doubles after each reinforcer deliv-
automatic reinforcement. For instance, Slaton and ery. Investigators use progressive-ratio schedules
Hanley (2016) showed that chained schedules pro- to compare the strength of two or more stimuli
duced more consistent item engagement and lower as reinforcers (see DeLeon et al., Chapter 7, this
levels of stereotypy. volume). For instance, Roane, Lerman, and Vorn-
dran (2001) demonstrated that progressive-ratio
schedules were more sensitive to differences in
Reinforcement Schedules
the reinforcing efficacy of stimuli than traditional
Reinforcement schedules specify the number and preference assessments.
type of responses required to produce reinforce-
ment or the time that must elapse before rein-
Interval Schedules
forcement is available. The reinforcement sched-
ule specifies the rules for reinforcement delivery. Interval schedules arrange for reinforcement deliv-
Because Mace, Pritchard, and Penney (Chapter ery for the first response occurring after a specified
4, this volume) describe reinforcement schedules interval and may be either fixed or variable. In a
more fully, we only briefly review them here. fixed-interval (FI) schedule, reinforcement delivery
occurs for the first response that occurs after in-
terval lengths that remain constant. For instance,
Ratio Schedules
the first response after 60 seconds will produce a
Ratio schedules arrange reinforcement delivery reinforcer in an FI 60-second schedule. FI sched-
based on number of responses, which may be ules may generate high rates of responding, espe-
constant, variable, or progressive. In a fixed-ratio cially with low-effort responses and relatively small
(FR) schedule, the number of responses required schedule values (e.g., Hanley et al., 2001). These
to produce reinforcement remains constant. For schedules tend to produce a scalloped behavior
instance, every response produces a reinforcer in pattern in which little responding occurs early in
an FR 1 schedule; every fifth response produces a the interval, but responding gradually accelerates
reinforcer in an FR 5 schedule. Behavior analysts as the interval progresses (Ferster & Skinner, 1957;
use FR 1 schedules commonly to establish and Weiner, 1969).
strengthen behavior. FR schedules may produce a A variable-interval (VI) schedule arranges re-
pause-and-run pattern in which responding occurs inforcement for the first response occurring after
at consistent high rates until reinforcement deliv- a specified interval that varies around a defined
242 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
mean. For instance, reinforcement delivery might less often (Iwata, 1987), examples include studies
occur for the first response after 10, 30, 80, or 90 by Piazza et al. (1997) and Lalli et al. (1999). They
seconds in a VI 60-second schedule. VI schedules provided negative reinforcement in the form of a
tend to produce steady response rates with little break when the participant complied with a task
pausing (Orlando & Bijou, 1960). demand.
Error correction is a common differential-
negative-reinforcement procedure that behavior
Reinforcer Magnitude
analysts incorporate into teaching programs. Error
Reinforcer magnitude is the amount or duration of correction involves prompting additional respond-
a reinforcer. Social or practical constraints often ing when a learner makes an error. For instance,
influence a reinforcer’s magnitude, such as when the teacher might point to the correct picture,
a teacher is available only for 5 minutes, or when say, “That is the elephant,” and then prompt the
someone wants to limit the amount of candy a child to “Point to the elephant” in an error cor-
child consumes. These constraints, however, may rection trial for receptive identification. Research
influence the efficacy of differential-reinforce- has shown that learners will acquire novel skills
ment procedures. For instance, Trosclair-Lasserre, to avoid these additional prompts (Kodak et al.,
Lerman, Call, Addison, and Kodak (2008) showed 2016; McGhan & Lerman, 2013; Rodgers & Iwata,
that larger amounts of attention and toys main- 1991).
tained responding at higher schedule values than
did smaller amounts for three children diagnosed
Automatic Reinforcers
with autism spectrum disorder. Like most func-
tional relations, there are relevant boundary con- Most studies of differential reinforcement in the
ditions. For instance, delivering copious amounts literature have used social reinforcers. Neverthe-
of reinforcement may produce reinforcer satiation less, programming nonsocial or automatic rein-
and limit the effectiveness of the differential-re- forcers following the occurrence of target behav-
inforcement procedure. Therefore, behavior ana- ior is possible and may be useful. For instance,
lysts should base their selection of reinforcement Linscheid, Iwata, Ricketts, Williams, and Griffin
amount or magnitude on practicality, social ac- (1990) described a device to treat severe self-in-
ceptability, and effectiveness. jurious behavior that could detect occurrences of
head banging and deliver a preferred event, such
as music or visual stimulation, when head banging
Types of Reinforcers
did not occur for a specified period.
Behavior analysts generally distinguish between Behavior analysts also can arrange differen-
positive and negative reinforcement and social and tial automatic negative reinforcement for target
nonsocial reinforcement. A positive reinforcer is responses. For instance, Azrin, Ruben, O’Brien,
one that a behavior analyst presents contingent Ayllon, and Roll (1968) engineered a device that
on a response, which increases the future prob- emitted a quiet tone, followed in 3 seconds by a
ability of the response. A negative reinforcer is loud tone when participants engaged in slouching.
one the behavior analyst removes contingent on a Participants could correct their posture after the
response, which increases the future probability of quiet tone and avoid the loud tone, or could re-
a response. Social reinforcers are ones that we can main erect and avoid both tones.
control, such as saying, “Nice work!” or giving a Investigators have shown that providing access
child a cookie. By contrast, nonsocial or automatic to automatically reinforced stereotypical behavior
reinforcers are events that occur as a direct result can function as reinforcement for other target re-
of the behavior (e.g., obtaining a cookie from a sponses, such as academic discriminations and play
vending machine; Vaughan & Michael, 1982). skills (Charlop-Christy & Haymes, 1996; Char-
lop, Kurtz, & Casey, 1990; Wolery, Kirk, & Gast,
1985). For instance, Hanley, Iwata, Thompson,
Positive versus Negative Reinforcers
and Lindberg (2000) showed that participants’ ste-
Most reported applied-behavior-analytic studies reotypic behavior persisted in the absence of social
of differential reinforcement have used positive consequences during an experimental functional
reinforcers, such as vocal and physical attention, analysis, suggesting that the consequences pro-
edible items, or leisure activities. Although inves- duced by the behavior functioned as automatic re-
tigators use differential negative reinforcement inforcement. Hanley et al. provided access to ste-
Differential-Reinforcement Procedures 243
reotypic behaviors contingent on play with leisure that temporarily diminish the value of a reinforcer,
materials, which increased participants’ play with called abolishing operations. The most common es-
leisure materials. Potter, Hanley, Augustine, Clay, tablishing operation is deprivation, and the most
and Phelps (2013) used a similar arrangement to common abolishing operation is satiation. Con-
teach complex, multistep play to adolescents with trol and manipulation of establishing operations
autism spectrum disorder. The participants could can increase the effectiveness of differential-rein-
engage in stereotypy by completing progressively forcement procedures. For instance, Goh, Iwata,
complicated play routines. Slaton and Hanley and DeLeon (2000) showed that no participants
(2016) taught participants to inhibit stereotypy acquired a novel mand when the reinforcer for the
and engage appropriately with items, using access mand was available on a dense schedule of non-
to stereotypy as reinforcement. A chained sched- contingent reinforcement (NCR). When the in-
ule of reinforcement produced higher levels of item vestigators made the NCR schedule progressively
engagement and stimulus control of stereotypy leaner, participants acquired the novel mand, pre-
than a schedule in which access to stereotypy was sumably because the relevant establishing opera-
time-based. tion increased with decreases in time-based rein-
Using automatic reinforcers in differential- forcer deliveries.
reinforcement contingencies may be desirable for Satiation is a serious challenge when a behav-
several additional reasons. First, the delivery of so- ior analyst is arranging reinforcement contingen-
cial reinforcers commonly requires a caregiver to cies, because each reinforcer delivery serves as an
continuously monitor and document participant abolishing operation for the reinforcer that sub-
behavior. The procedures described by Linscheid sequent responses produce. For instance, each sip
et al. (1990) and Azrin et al. (1968) require nei- of water decreases the establishing operation for
ther, which may increase their utility. Second, au- subsequent sips of water over the near term. Using
tomated delivery of reinforcers is likely to be more the smallest amount of reinforcement necessary to
precise and immediate than delivery of reinforcers maintain responding is one way to mitigate satia-
by humans if the device functions properly. Third, tion. Another is to restrict the reinforcer to the en-
individuals may acquire skills more readily when vironment for the contingency arrangement. For
the consequence of responding results directly instance, Roane, Call, and Falcomata (2005) dem-
from the behavior (Thompson & Iwata, 2000). onstrated that responding persisted more when
they restricted the reinforcer to the progressive-
ratio-schedule arrangement than when the rein-
Reinforcer Quality
forcer was available outside the progressive-ratio-
Reinforcer quality is a participant’s subjective valu- schedule arrangement.
ation of a reinforcing stimulus. Results of multi- In some cases, a behavior analyst cannot ethi-
ple studies have shown that attention to quality cally or legally restrict a potential reinforcer. In
improves the efficacy of reinforcement programs these cases, the behavior analyst can schedule
(e.g., Johnson, Vladescu, Kodak, & Sidener, 2017; training in ways that maximize the effectiveness
Mace, Neef, Shade, & Mauro, 1996) and their ac- of reinforcers. For instance, the analyst might
ceptability to the participants (e.g., Johnson et al., schedule a training session before the partici-
2017). Presumably, effective procedures rely on re- pant’s regularly scheduled lunch and use food as a
inforcers of sufficient quality, and reinforcer value reinforcer (e.g., North & Iwata, 2005; Vollmer &
is idiosyncratic and may change over time. Iwata, 1991).
Reinforcer Variation
STRATEGIES TO INCREASE OR MAINTAIN
THE EFFECTIVENESS OF REINFORCERS Varying reinforcers for responding may delay satia-
tion and prolong the effectiveness of differential
Motivating Operations
reinforcement (Bowman, Piazza, Fisher, Hagopian,
A motivating operation is an event that alters the & Kogan, 1997; Egel, 1981; Koehler, Iwata, Roscoe,
effectiveness of a stimulus as reinforcement. There Rolider, & O’Steen, 2005). For instance, Bowman
are two broad categories of motivating operations et al. (1997) showed that five of seven participants
(Laraway, Snycerski, Michael, & Poling, 2003): preferred varied delivery of three lesser preferred
those that temporarily increase the value of a re- items to constant delivery of a more preferred item,
inforcer, called establishing operations, and those and Egel (1981) showed that varying reinforcers
244 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
produced more stable levels of correct responding din & Bootzin, 1972). Token systems allow care-
and on-task behavior for several children diag- givers to deliver multiple reinforcers contingent
nosed with autism spectrum disorder. on desirable behavior without adversely affecting
the value of the primary or backup reinforcers,
and without interrupting learning tasks or com-
Reinforcer Choice
plex behaviors for reinforcer consumption. For in-
Providing a choice of reinforcers may be a simple stance, Krentz, Miltenberger, and Valbuena (2016)
yet highly effective means of improving the ef- used token reinforcement to increase the distance
ficacy of differential-reinforcement procedures walked by overweight and obese adults with in-
(Ackerlund Brandt, Dozier, Juanico, Laudont, & tellectual developmental disorder at an adult day
Mick, 2015; Dunlap et al., 1994; Dyer, Dunlap, & training center.
Winterling, 1990; Fisher, Thompson, Piazza, Cro-
sland, & Gotjen, 1997; Sellers et al., 2013; Thomp-
son, Fisher, & Contrucci, 1998; Tiger, Hanley, & COMPLEMENTARY PROCEDURES TO DEVELOP
Hernandez, 2006; Toussaint, Kodak, & Vladescu, NEW BEHAVIOR
2016). Choice making may be effective because it
produces reinforcer variation, which minimizes Although differential reinforcement alone can
satiation secondary to the repeated delivery of produce new behavior, combining it with other
the same reinforcer. In addition, choice making procedures when teaching new behavior is often
capitalizes on establishing operations that produce more effective. This section describes procedures
momentary fluctuations in the value of reinforc- to complement differential reinforcement to de-
ers, because the participant can choose the rein- velop new behavior.
forcer he or she prefers at that moment.
There also is evidence that the opportunity to
Prompting
choose adds value to differential reinforcement
beyond the value of obtaining the most preferred Behavior analysts often pair prompting with differ-
item. For instance, we (Tiger & Hanley, 2006) ential reinforcement to teach new behavior. The
showed that six of seven preschoolers preferred general sequence involves prompting the individu-
to engage in academic seatwork when they could al to engage in a response, providing reinforcement
choose a single edible from an identical edible for the prompted response, and gradually eliminat-
array for correct responding, rather than when ing the prompt over time. The behavior analyst
the teacher provided the same amount and type can provide prompts in many forms (such as vocal,
of edible from the same type of array for correct visual, or physical-response prompts; within-stim-
responding. Note that reinforcer amount, qual- ulus prompts; or extrastimulus prompts) and can
ity, and type were identical in the two conditions; choose the prompt based on the modality of the
the only difference was the choice component. target behavior and the individual’s capabilities.
We also showed that children engaged in 12 to 16 For instance, Thompson, McKerchar, and Dancho
times more academic work in the choice condi- (2004) used delayed physical prompts and differ-
tion. These data show that programming opportu- ential reinforcement to teach three infants to emit
nities to make choices may enhance the efficacy of the manual signs Please and More with food as the
differential reinforcement. reinforcer.
By contrast, behavior analysts cannot prompt
nonmotor target behavior, such as vocalizations;
Token Reinforcement Systems
therefore, they must pair alternative prompting
Using conditioned reinforcers that an individ- procedures with differential reinforcement. Bour-
ual can trade later for preferred items, known as ret, Vollmer, and Rapp (2004) used vocal and
backup reinforcers, is another strategy to decrease model prompts to teach vocalizations to two chil-
satiation. Token economies, for instance, involve dren with autism spectrum disorder. The therapist
providing arbitrary items, such as tickets, tokens, vocally prompted the participant to emit the target
stickers, or points, following the occurrence of tar- vocalization (e.g., “Say tunes”). Correct vocaliza-
get behaviors. Later the individual can exchange tions produced access to music. If the participant
the tokens for preferred items (see Reitman, Boer- did not emit the target vocalization, the therapist
ke, & Vassilopoulos, Chapter 22, this volume, or modeled progressively shorter vocalizations (e.g.,
reviews by Hackenberg, 2018; Kazdin, 1982; Kaz- changing “Say tunes” to “Say tuh”). As the partici-
Differential-Reinforcement Procedures 245
pant successfully emitted the modeled vocaliza- most complex to arrange percentile schedules for
tion, the therapist required the participant to emit complex behavior. The behavior analyst keeps a
a vocalization that more closely approximated the running stream of the temporal order and form of
target vocalization before receiving reinforcement. the behavior, with a focus on the most recent re-
One disadvantage is that prompting may pro- sponses. The behavior analyst delivers a reinforcer
duce prompt dependence. The behavior analyst for a response if it exceeds the formal qualities of
can pair differential reinforcement with various the most recent subset of responses.
tactics to fade and to eliminate prompts eventu- Behavior analysts may use percentile schedules
ally (see Halle, 1987, for a discussion of spontane- to shape higher rates or durations of responding.
ity). Thompson et al. (2004) and Bourret et al. For instance, Athens et al. (2007) used percentile
(2004) eliminated prompts by increasing the delay schedules to increase the academic-task engage-
between the presentation of the evocative event ment of four students with intellectual develop-
(such as placing a toy in a participant’s reach) and mental disorder. The investigators measured the
the prompts, so that reinforcement was more im- duration of task engagement for each participant.
mediate for independent responses. Other tactics During the percentile-schedule phase, engagement
include withholding reinforcement for prompted produced a token exchangeable for food if engage-
responses (Touchette & Howard, 1984) or decreas- ment duration exceeded the median duration of
ing the physical intensity of the prompts (see Wol- the previous 5, 10, or 20 engagement durations, de-
ery & Gast, 1984). pending on the experimental condition. Thus the
reinforcement criterion constantly shifted, given
the participant’s recent engagement duration.
Shaping and Percentile Schedules
The percentile schedule produced increased en-
When prompting is not appropriate to increase gagement durations, with the biggest increases in
responding, we recommend shaping as an alterna- conditions in which the participant’s previous be-
tive tool. Shaping involves differential reinforce- havior determined the momentary criterion for re-
ment of successive approximations of a behavior. inforcement. For instance, Athens et al. observed
To initiate a shaping procedure, a behavior analyst higher engagement durations when they used the
must (1) identify a behavior the individual cur- previous 20 versus the previous 5 engagement du-
rently emits that approximates the target behav- rations to determine the reinforcement criterion.
ior; (2) provide reinforcement for that behavior;
and (3) require closer approximations to the termi-
Response Chaining and Task Analysis
nal behavior, such as more complex forms or dif-
ferent rates or durations of behavior, for reinforcer Commonly taught behaviors are often not single,
delivery. Investigators have used shaping to teach unitary responses; instead, they include a series of
many complex behaviors, including eye contact in topographically distinct behaviors that a partici-
children with autism spectrum disorder (e.g., Mc- pant must complete in sequence. Behavior analysts
Connell, 1967), vocal speech in mute adults diag- often refer to these behaviors as response chains and
nosed with psychosis (Sherman, 1965), and limb to each component behavior as a link in the chain.
use in patients after strokes (Taub et al., 1994). Al- Providing reinforcement for a single response in
though shaping is among behavior analysts’ oldest a chain or for the entire response chain may not
and most celebrated tools, there are few formalized be an efficient or effective way to shape behavior.
rules for shaping. Therefore, behavior analysts typically use prompt-
Galbicka (1994) described a formalized shaping ing and differential reinforcement or shaping to
system using percentile schedules, and investigators establish individual links of the response chain,
have published studies in which percentile sched- then differentially reinforce sequences of links
ules are the cornerstones of their behavior change until a participant produces an entire response
procedures (Athens, Vollmer, & St. Peter Pipkin, chain. Behavior analysts use one of two general
2007; Lamb, Kirby, Morral, Galbicka, & Iguchi, procedures, called forward chaining and backward
2004; Lamb, Morral, Kirby, Iguchi, & Galbicka, chaining, to teach response chains. Forward chain-
2004). Percentile schedules dictate rules for the ing involves teaching the response chain in the
timing of reinforcement delivery, and these rules same order in which the participant will ultimately
can be adjusted based on recent or local rates, du- perform it. That is, the behavior analyst differen-
rations, or types of responding. The behavior ana- tially reinforces emission of the first behavior in
lyst rank-orders responses from the simplest to the the chain, then the first and second behaviors, and
246 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
so forth. By contrast, the behavior analyst provides & Greer, 2016). Other methods for introducing
differential reinforcement for the last behavior in delays successfully include introducing brief de-
the chain and adds behaviors of the chain to the lays that gradually increase (Schweitzer & Sulzer-
differential-reinforcement contingency in reverse Azaroff, 1988), providing a signal when the delay
order, in backward chaining. begins (Vollmer, Borrero, Lalli, & Daniel, 1999),
Task analysis, or identifying individual behav- and providing alternative activities during the
iors in the response chain, is necessary before delay (Austin & Tiger, 2015; Fisher et al., 2000;
teaching a complex behavior. For instance, Neef, Hagopian, Contrucci Kuhn, Long, & Rush, 2005).
Parrish, Hannigan, Page, and Iwata (1989) dem- A multiple schedule is another effective method
onstrated the importance of task analysis. They for thinning reinforcement schedules (Hanley et
taught self-catheterization skills to two young girls al., 2001; Tiger & Hanley, 2004). During a mul-
with spina bifida by identifying each step of self- tiple schedule, the behavior analyst correlates
catheterization and then partitioning the task into reinforcement and extinction periods with dis-
6–11 component steps. They taught each step to tinct discriminative signals. Hanley et al. (2001)
each participant, using prompting and differential showed that multiple schedules maintained newly
reinforcement, until the two girls could indepen- acquired social manding, even though reinforce-
dently self-catheterize (see also Noell, Call, Ar- ment was available only one-fifth of the time. In
doin, & Miller, Chapter 15, this volume). addition, we (Tiger & Hanley, 2006) showed that
children preferred conditions with signaled re-
inforcement and extinction versus ones without
RESPONSE MAINTENANCE signals. Luczynski and Hanley (2009) showed that
AND SCHEDULE THINNING children preferred multiple schedules to briefly sig-
naled delays.
Although immediate, dense schedules of reinforce- Slaton and Hanley (2016) showed that students
ment are important for establishing responses, preferred chained schedules of reinforcement in
caregivers may have difficulty implementing such which the reinforcement component included
schedules with high integrity over the long term. differential versus noncontingent reinforcement.
Therefore, thinning of a reinforcement schedule is Ghaemmaghami et al. (2016) used intermittent
an important part of response maintenance. and unpredictable reinforcement of several alter-
One method of reinforcement schedule thin- native responses to thin the reinforcement sched-
ning is to deliver reinforcement intermittently by ule. Ghaemmaghami et al. suggested that the
progressively increasing response requirements schedule they used was more like those in natural
for reinforcement. For instance, Van Houten and environments, where reinforcement contingencies
Nau (1980) used FR- and VR-like reinforcement are often ambiguous.
schedules to increase the attending behaviors of
elementary school students. The fixed schedule
arranged reinforcement for every eight intervals DIFFERENTIAL REINFORCEMENT
with attending behavior. Children could reach OF DIVERSE RESPONDING
into a grab bag with a one-eighth probability of
payoff after 5 minutes of continuous attending be- Behavior analysts use differential reinforcement
havior in the VR schedule. Procedures like these to increase the occurrence of a target behavior
may allow caregivers to miss a few instances of an and to increase the diversity of behavior. For in-
important behavior without inadvertently weak- stance, Goetz and Baer (1973) provided descrip-
ening the behavior. tive praise (e.g., “Wow, a tower; that is new”) to
A second technique for making differential re- preschoolers for creating a block structure that
inforcement more practical is to include delays to they had not built that day, and new forms of
reinforcement, which allows extra time for care- building increased. Investigators have used similar
givers to provide the reinforcer. However, delays to procedures to increase diverse verbal responses to
reinforcement often result in extinction of newly questions (Lee, McComas, & Jawor, 2002), activ-
acquired behaviors (Fisher et al., 2000; Hanley et ity selections during free-play periods (Cammilleri
al., 2001), and exposure to extinction may produce & Hanley, 2005), and martial-arts performances
resurgence (i.e., the reemergence of previously (Harding, Wacker, Berg, Rick, & Lee, 2004). A
reinforced behavior) and increase the likelihood renewed emphasis on using differential-reinforce-
of problem behavior returning (Fuhrman, Fisher, ment schedules to promote diverse verbal behavior
Differential-Reinforcement Procedures 247
also is apparent in the applied literature. Inves- group contingencies to be effective in reducing
tigators have used lag-differential-reinforcement students’ verbal disruptions, inappropriate sitting,
schedules to increase preschoolers’ diversity of and off-task behaviors during the Good Behavior
selections during free-play periods (Cammilleri & Game, but students preferred the interdependent-
Hanley, 2005), mand variability in young children group contingency.
with autism spectrum disorder (Brodhead, Higbee,
Gerencser, & Akers, 2016), varied intraverbal re-
sponses of children with autism spectrum disorder LOTTERY CONTINGENCIES
(Contreras & Betz, 2016), and varied item naming
of typically developing children (Wiskow & Don- Lottery contingencies are useful for increasing the
aldson, 2016). time- and cost-effectiveness of differential rein-
forcement. For instance, Petry et al. (1998) used
a lottery-type reinforcement procedure to decrease
GROUP CONTINGENCIES inappropriate verbal behavior in a heroin treat-
ment center. Patients earned stickers for engaging
When the behavior of many individuals is the in desirable verbal behavior. Investigators entered
target of intervention, dependent- or interdepen- the names of sticker recipients into a lottery and
dent-group contingencies may be useful. The con- drew one name from the lottery at the end of each
sequences for the group depend on the behavior week. The winner received $25. This procedure
of one or some members of the group with a de- produced increased compliments and pleasantries,
pendent-group contingency. For instance, an entire and decreased profanity and discussion of evading
classroom may earn extra recess time if one child the police.
scores well on an exam (Litow & Pumroy, 1975).
Dependent contingencies may motivate students
to aid each other, such as helping each other pre- CONCLUSION
pare for an exam (Speltz, Shimamura, & McReyn-
olds, 1982) or to help one child behave more ap- Differential reinforcement is one of the most
propriately. Poorly designed group contingencies well-researched procedures available to behavior
may cause unwelcome peer pressure, however, and analysts. A mature understanding of differential
some children may not prefer reinforcement con- reinforcement involves knowledge of (1) the range
tingencies that depend on someone else’s behavior and complexity of behavior it affects, (2) the pa-
(e.g., Speltz et al., 1982). One way to mitigate the rameters that are responsible for or increase its
negative effects of a dependent-group contingency effectiveness, and (3) the procedures that may
is to randomly select the student on whose behav- complement it to produce significant and lasting
ior the contingencies will depend (Cariveau & changes in socially important behavior. This sort
Kodak, 2017). of understanding, combined with supervised and
In an interdependent-group contingency, the dedicated practice in implementing differential re-
group’s behavior determines whether the group inforcement, is critical to the development of an
receives reinforcement (Litow & Pumroy, 1975). effective behavior analyst.
For instance, the group receives reinforcement if
the students score above 80% on an exam. Hirsch,
Healy, Judge, and Lloyd (2016) used an interdepen- REFERENCES
dent-group contingency to increase second-grade
students’ engagement in physical education activi- Ackerlund Brandt, J. A., Dozier, C. L., Juanico, J. F.,
ties; the students rated this group contingency fa- Laudont, C. L., & Mick, B. R. (2015). The value of
vorably. Interdependent-group contingencies may choice as a reinforcer for typically developing chil-
involve competition between two or more groups, dren. Journal of Applied Behavior Analysis, 48, 344–
362.
with the highest-scoring group receiving the rein-
Athens, E. S., Vollmer, T. R., & St. Peter Pipkin, C. C.
forcer, such as the Good Behavior Game for stu- (2007). Shaping academic task engagement with per-
dents in classrooms (Barrish, Saunders, & Wolf, centile schedules. Journal of Applied Behavior Analy-
1969) or the Good Productivity Game for staff sis, 40, 475–488.
members in hospitals (Lutzker & White-Black- Austin, J. E., & Tiger, J. H. (2015). Providing alterna-
burn, 1979). Groves and Austin (2017) recently tive reinforcers to facilitate tolerance to delayed re-
found both independent- and interdependent- inforcement following functional communication
248 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
training. Journal of Applied Behavior Analysis, 48, dren with autism. Journal of Applied Behavior Analy-
663–668. sis, 49, 3–16.
Austin, J. L., & Bevan, D. (2011). Using differential re- Cox, A. D., Virues-Ortega, J., Julio, F., & Martin, T. L.
inforcement of low rates to reduce children’s requests (2017). Establishing motion control in children with
for teacher attention. Journal of Applied Behavior autism and intellectual disability: Applications for
Analysis, 44, 451–461. anatomical and functional MRI. Journal of Applied
Azrin, N., Rubin, H., O’Brien, F., Ayllon, T., & Roll, Behavior Analysis, 50, 8–26.
D. (1968). Behavioral engineering: Postural control Dunlap, G., dePerczel, M., Clarke, S., Wilson, D.,
by a portable operant apparatus. Journal of Applied Wright, S., White, R., et al. (1994). Choice making
Behavior Analysis, 1, 99–108. to promote adaptive behavior for students with emo-
Barrish, H. H., Saunders, M., & Wolf, M. M. (1969). tional and behavioral challenges. Journal of Applied
Good Behavior Game: Effects of individual contin- Behavior Analysis, 27, 505–518.
gencies for group consequences on disruptive behav- Dyer, K., Dunlap, G., & Winterling, V. (1990). Effects
ior in a classroom. Journal of Applied Behavior Analy- of choice making on the serious problem behaviors
sis, 2, 119–124. of students with severe handicaps. Journal of Applied
Becraft, J. L., Borrero, J. C., Davis, B. J., Mendres-Smith, Behavior Analysis, 23, 515–524.
A. E., & Castillo, M. I. (2018). The role of signals in Egel, A. L. (1981). Reinforcer variation: Implications for
two variations of differential-reinforcement-of-low- motivating developmentally disabled children. Jour-
rate procedures. Journal of Applied Behavior Analysis, nal of Applied Behavior Analysis, 14, 345–350.
51, 3–24. Ferster, C. B., & Skinner, B. F. (1957). Schedules of rein-
Bijou, S. W. (1996). New directions in behavior develop- forcement. New York: Appleton-Century-Crofts.
ment. Reno, NV: Context Press. Fisher, W. W., Thompson, R. H., Hagopian, L. P., Bow-
Bourret, J., Vollmer, T. R., & Rapp, J. T. (2004). Evalu- man, L. G., & Krug, A. (2000). Facilitating tolerance
ation of a vocal mand assessment and vocal mand to delayed reinforcement during functional commu-
training procedures. Journal of Applied Behavior nication training. Behavior Modification, 24, 3–29.
Analysis, 37, 129–144. Fisher, W. W., Thompson, R. H., Piazza, C. C., Crosland,
Bowman, L. G., Piazza, C. C., Fisher, W. W., Hagopian, K., & Gotjen, D. (1997). On the relative reinforcing
L. P., & Kogan, J. S. (1997). Assessment of preference effects of choice and differential consequences. Jour-
for varied versus constant reinforcers. Journal of Ap- nal of Applied Behavior Analysis, 30, 423–438.
plied Behavior Analysis, 30, 451–458. Fleece, L., Gross, A., O’Brien, T., Kistner, J., Rothblum,
Brodhead, M. T., Higbee, T. S., Gerencser, K. R., & E., & Drabman, R. (1981). Elevation of voice volume
Akers, J. S. (2016). The use of a discrimination-train- in young developmentally delayed children via an
ing procedure to teach mand variability to children operant shaping procedure. Journal of Applied Behav-
with autism. Journal of Applied Behavior Analysis, 49, ior Analysis, 14, 351–355.
34–48. Fuhrman, A. M., Fisher, W. W., & Greer, B. D. (2016).
Cammilleri, A. P., & Hanley, G. P. (2005). Use of a lag A preliminary investigation on improving functional
differential reinforcement contingency to increase communication training by mitigating resurgence
varied selections of classroom activities. Journal of of destructive behavior. Journal of Applied Behavior
Applied Behavior Analysis, 38, 111–115. Analysis, 49, 884–899.
Cariveau, T., & Kodak, T. (2017). Programming a ran- Galbicka, G. (1994). Shaping in the 21st century: Mov-
domized dependent group contingency and common ing percentile schedules into applied settings. Journal
stimuli to promote durable behavior change. Journal of Applied Behavior Analysis, 27, 739–760.
of Applied Behavior Analysis, 50, 121–133. Ghaemmaghami, M., Hanley, G. P., & Jessel, J. (2016).
Carroll, R. A., Kodak, T., & Adolf, K. J. (2016). Effect Contingencies promote delay tolerance. Journal of
of delayed reinforcement on skill acquisition during Applied Behavior Analysis, 49, 548–575.
discrete-trial instruction: Implications for treatment- Gleeson, S., & Lattal, K. A. (1987). Response–reinforc-
integrity errors in academic settings. Journal of Ap- er relations and the maintenance of behavior. Journal
plied Behavior Analysis, 49, 176–181. of the Experimental Analysis of Behavior, 48, 383–393.
Charlop, M. H., Kurtz, P. F., & Casey, F. G. (1990). Using Goetz, E. M., & Baer, D. M. (1973). Social control of
aberrant behaviors as reinforcers for autistic children. form diversity and the emergence of new forms in
Journal of Applied Behavior Analysis, 23, 163–181. children’s block building. Journal of Applied Behavior
Charlop-Christy, M. H., & Haymes, L. K. (1996). Using Analysis, 6, 209–217.
obsessions as reinforcers with and without mild re- Goh, H., Iwata, B. A., & DeLeon, I. G. (2000). Com-
ductive procedures to decrease inappropriate behav- petition between noncontingent and contingent re-
iors of children with autism. Journal of Autism and inforcement schedules during response acquisition.
Developmental Disorders, 26, 527–546. Journal of Applied Behavior Analysis, 33, 195–205.
Contreras, B. P., & Betz, A. M. (2016). Using lag sched- Groves, E. A., & Austin, J. L. (2017). An evaluation of
ules to strengthen the intraverbal repertoires of chil- interdependent and independent group contingen-
Differential-Reinforcement Procedures 249
cies during the Good Behavior Game. Journal of Ap- applied research on DRO procedures. Journal of Ap-
plied Behavior Analysis, 50, 552–566. plied Behavior Analysis, 49, 991–995.
Hackenberg, T. D. (2018). Token reinforcement: Trans- Johnson, K. A., Vladescu, J. C., Kodak, T., & Sidener,
lational research and application. Journal of Applied T. M. (2017). An assessment of differential reinforce-
Behavior Analysis, 51, 393–435. ment procedures for learners with autism spectrum
Hagopian, L. P., Contrucci Kuhn, S. A., Long, E. S., disorder. Journal of Applied Behavior Analysis, 50,
& Rush, K. S. (2005). Schedule thinning following 290–303.
communication training: Using competing stimuli to Kazdin, A. E. (1982). The token economy: A decade
enhance tolerance to decrements in reinforcer den- later. Journal of Applied Behavior Analysis, 15, 431–
sity. Journal of Applied Behavior Analysis, 38, 177–193. 445.
Halle, J. W. (1987). Teaching language in the natural Kazdin, A. E., & Bootzin, R. R. (1972). The token econ-
environment: An analysis of spontaneity. Journal of omy: An evaluative review. Journal of Applied Behav-
the Association for Persons with Severe Handicaps, 12, ior Analysis, 5, 343–372.
28–37. Kodak, T., Campbell, V., Bergmann, S., LeBlanc, B.,
Hanley, G. P., Iwata, B. A., & Thompson, R. H. (2001). Kurtz-Nelson, E., Cariveau, T., et al. (2016). Ex-
Reinforcement schedule thinning following treat- amination of efficacious, efficient, and socially valid
ment with functional communication training. Jour- error-correction procedures to teach sight words and
nal of Applied Behavior Analysis, 34, 17–38. prepositions to children with autism spectrum disor-
Hanley, G. P., Iwata, B. A., Thompson, R. H., & Lind- der. Journal of Applied Behavior Analysis, 49, 532–547.
berg, J. S. (2000). A component analysis of “stereoty- Koehler, L. J., Iwata, B. A., Roscoe, E. M., Rolider, N. U.,
py as reinforcement” for alternative behavior. Journal & O’Steen, L. E. (2005). Effects of stimulus variation
of Applied Behavior Analysis, 33, 285–297. on the reinforcing capability of nonpreferred stimuli.
Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci, Journal of Applied Behavior Analysis, 38, 469–484.
S. A., & Maglieri, K. A. (1997). Evaluation of client Krentz, H., Miltenberger, R., & Valbuena, D. (2016).
preference for function-based treatments. Journal of Using token reinforcement to increase walking for
Applied Behavior Analysis, 30, 459–473. adults with intellectual disabilities. Journal of Applied
Harding, J. W., Wacker, D. P., Berg, W. K., Rick, G., & Behavior Analysis, 49, 745–750.
Lee, J. F. (2004). Promoting response variability and Lalli, J. S., Casey, S., & Kates, K. (1995). Reducing es-
stimulus generalization in martial arts training. Jour- cape behavior and increasing task completion with
nal of Applied Behavior Analysis, 37, 185–195. functional communication training, extinction, and
Hernandez, E., Hanley, G. P., Ingvarsson, E. T., & Tiger, response chaining. Journal of Applied Behavior Analy-
J. H. (2007). An evaluation of the emergence of novel sis, 28, 261–268.
mand forms. Journal of Applied Behavior Analysis, 40, Lalli, J. S., Vollmer, T. R., Progar, P. R., Wright, C.,
137–156. Borrero, J., Daniel, D., et al. (1999). Competition
Hirsch, S. E., Healy, S., Judge, J. P., & Lloyd, J. W. (2016). between positive and negative reinforcement in the
Effects of an interdependent group contingency on treatment of escape behavior. Journal of Applied Be-
engagement in physical education. Journal of Applied havior Analysis, 32, 285–296.
Behavior Analysis, 49, 975–979. Lamb, R. J., Kirby, K. C., Morral, A. R., Galbicka, G., &
Horner, R. H., Sprague, J. R., O’Brien, M., & Heathfield, Iguchi, M. Y. (2004). Improving contingency man-
L. T. (1990). The role of response efficiency in the agement programs for addiction. Addictive Behaviors,
reduction of problem behaviors through functional 29, 507–523.
equivalence training: A case study. Journal of the As- Lamb, R. J., Morral, A. R., Kirby, K. C., Iguchi, M. Y.,
sociation for Persons with Severe Handicaps, 15, 91–97. & Galbicka, G. (2004). Shaping smoking cessation
Ingham, R. J., & Andrews, G. (1973). An analysis of a using percentile schedules. Drug and Alcohol Depen-
token economy in stuttering therapy. Journal of Ap- dence, 76, 247–259.
plied Behavior Analysis, 6, 219–229. Laraway, S., Snycerski, S., Michael, J., & Poling, A.
Iwata, B. A. (1987). Negative reinforcement in applied (2003). Motivating operations and terms to describe
behavior analysis: An emerging technology. Journal them: Some further refinements. Journal of Applied
of Applied Behavior Analysis, 20, 361–378. Behavior Analysis, 36, 407–414.
Jessel, J., & Borrero, J. C. (2014). A laboratory compari- Lee, R., McComas, J. J., & Jawor, J. (2002). The effects
son of two variations of differential-reinforcement- of differential and lag reinforcement schedules on
of-low-rate procedures. Journal of Applied Behavior varied verbal responding by individuals with autism.
Analysis, 47, 314–324. Journal of Applied Behavior Analysis, 35, 391–402.
Jessel, J., Borrero, J. C., & Becraft, J. L. (2015). Differen- Lennox, D. B., Miltenberger, R. G., & Donnelly, D. R.
tial reinforcement of other behavior increases untar- (1987). Response interruption and DRL for the re-
geted behavior. Journal of Applied Behavior Analysis, duction of rapid eating. Journal of Applied Behavior
48, 402–416. Analysis, 20, 279–284.
Jessel, J., & Ingvarsson, E. T. (2016). Recent advances in Leon, Y., Borrero, J. C., & DeLeon, I. G. (2016). Para-
250 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
metric analysis of delayed primary and conditioned Piazza, C. C., Fisher, W. W., Hanley, G. P., Remick, M.
reinforcers. Journal of Applied Behavior Analysis, 49, A., Contrucci, S. A., & Aitken, T. (1997). The use of
639–655. positive and negative reinforcement in the treatment
Linscheid, T. R., Iwata, B. A., Ricketts, R. W., Williams, of escape-maintained destructive behavior. Journal of
D. E., & Griffin, J. C. (1990). Clinical evaluation of Applied Behavior Analysis, 30, 279–297.
the self-injurious behavior inhibiting system. Journal Pinkston, E. M., Reese, N. M., LeBlanc, J. M., & Baer,
of Applied Behavior Analysis, 23, 53–78. D. M. (1973). Independent control of a preschool
Litow, L., & Pumroy, D. K. (1975). Brief technical re- child’s aggression and peer interaction by contingent
port: A brief review of classroom group-oriented teacher attention. Journal of Applied Behavior Analy-
contingencies. Journal of Applied Behavior Analysis, sis, 6, 115–124.
8, 341–347. Potter, J. N., Hanley, G. P., Augustine, M., Clay, C. J.,
Luczynski, K. C., & Hanley, G. P. (2009). Do children & Phelps, M. C. (2013). Treating stereotypy in ado-
prefer contingencies?: An evaluation of the efficacy lescents diagnosed with autism by refining the tactic
of and preference for contingent versus noncontin- of “using stereotypy as reinforcement.” Journal of Ap-
gent social reinforcement during play. Journal of Ap- plied Behavior Analysis, 46, 407–423.
plied Behavior Analysis, 42, 511–525. Protopopova, A., Kisten, D., & Wynne, C. (2016).
Luczynski, K. C., & Hanley, G. P. (2010). Examining Evaluating a humane alternative to the bark collar:
the generality of children’s preference for contingent Automated differential reinforcement of not barking
reinforcement via extension to different responses, in a home-alone setting. Journal of Applied Behavior
reinforcers, and schedules. Journal of Applied Behavior Analysis, 49, 735–744.
Analysis, 43, 397–409. Risley, T. R. (2005). Montrose M. Wolf (1935–2004).
Lutzker, J. R., & White-Blackburn, G. (1979). The Good Journal of Applied Behavior Analysis, 38, 279–287.
Productivity Game: Increasing work performance in Roane, H. S., Call, N. A., & Falcomata, T. S. (2005).
a rehabilitation setting. Journal of Applied Behavior A preliminary analysis of adaptive responding under
Analysis, 12, 488. open and closed economies. Journal of Applied Behav-
Mace, F. C., Neef, N. A., Shade, D., & Mauro, B. C. ior Analysis, 38, 335–348.
(1996). Effects of problem difficulty and reinforcer Roane, H. S., Lerman, D. C., & Vorndran, C. M. (2001).
quality on time allocated to concurrent arithmetic Assessing reinforcers under progressive schedule re-
problems. Journal of Applied Behavior Analysis, 29, quirements. Journal of Applied Behavior Analysis, 34,
11–24. 145–167.
McConnell, O. L. (1967). Control of eye contact in an Rodgers, T. A., & Iwata, B. A. (1991). An analysis of
autistic child. Journal of Child Psychology and Psychia- error-correction procedures during discrimination
try, 8, 249–255. training. Journal of Applied Behavior Analysis, 24,
McGhan, A. C., & Lerman, D. C. (2013). An assess- 775–781.
ment of error-correction procedures for learners with Schlinger, H. D. (1995). A behavior analytic view of child
autism. Journal of Applied Behavior Analysis, 46, development. New York: Plenum Press.
626–639. Schlinger, H. D., Derenne, A., & Baron, A. (2008).
Miller, D. L., & Kelley, M. L. (1994). The use of goal What 50 years of research tell us about pausing under
setting and contingency contracting for improving ratio schedules of reinforcement. Behavior Analyst,
children’s homework performance. Journal of Applied 31, 39–60.
Behavior Analysis, 27, 73–84. Schweitzer, J. B., & Sulzer-Azaroff, B. (1988). Self-con-
Neef, N. A., Parrish, J. M., Hannigan, K. F., Page, T. J., trol: Teaching tolerance for delay in impulsive chil-
& Iwata, B. A. (1989). Teaching self-catheterization dren. Journal of the Experimental Analysis of Behavior,
skills to children with neurogenic bladder complica- 50, 173–186.
tions. Journal of Applied Behavior Analysis, 22, 237– Sellers, T. P., Bloom, S. E., Samaha, A. L., Dayton, E.,
243. Lambert, J. M., & Keyl-Austin, A. A. (2013). Evalu-
North, S. T., & Iwata, B. A. (2005). Motivational in- ation of some components of choice making. Journal
fluences on performance maintained by food rein- of Applied Behavior Analysis, 46, 455–464.
forcement. Journal of Applied Behavior Analysis, 38, Sherman, J. A. (1965). Use of reinforcement and imita-
317–333. tion to reinstate verbal behavior in mute psychotics.
Orlando, R., & Bijou, S. W. (1960). Single and multiple Journal of Abnormal Psychology, 70, 155–164.
schedules of reinforcement in developmentally re- Skinner, B. F. (1948). “Superstition” in the pigeon. Jour-
tarded children. Journal of the Experimental Analysis nal of Experimental Psychology, 38, 168–272.
of Behavior, 3, 339–348. Slaton, J., & Hanley, G. P. (2016). Effects of multiple
Petry, N. M., Bickel, W. K., Tzanis, E., Taylor, R., Kubik, versus chained schedules on stereotypy and function-
E., Foster, M., et al. (1998). A behavioral interven- al engagement. Journal of Applied Behavior Analysis,
tion for improving verbal behaviors of heroin addicts 49, 927–946.
in a treatment clinic. Journal of Applied Behavior Speltz, M. L., Shimamura, J. W., & McReynolds, W. T.
Analysis, 31, 291–297. (1982). Procedural variations in group contingencies:
Differential-Reinforcement Procedures 251
Effects on children’s academic and social behaviors. ing: Reinforcement contingencies and stimulus con-
Journal of Applied Behavior Analysis, 15, 533–544. trol transfer in delayed prompting. Journal of Applied
Stasolla, F., Caffò, A. O., Perilli, V., Boccasini, A., Behavior Analysis, 17, 175–188.
Stella, A., Damiani, R., et al. (2017). A microswitch- Toussaint, K. A., Kodak, T., & Vladescu, J. (2016). An
based program for promoting initial ambulation re- evaluation of choice on instruction efficacy and indi-
sponses: An evaluation with two girls with multiple vidual preference for children with autism. Journal of
disabilities. Journal of Applied Behavior Analysis, 50, Applied Behavior Analysis, 40, 170–175.
345–356. Trosclair-Lasserre, N. M., Lerman, D. C., Call, N. A.,
Stokes, T. F., Fowler, S. A., & Baer, D. M. (1978). Train- Addison, L. R., & Kodak, T. (2008). Reinforcement
ing preschool children to recruit natural communi- magnitude: An evaluation of preference and rein-
ties of reinforcement. Journal of Applied Behavior forcer efficacy. Journal of Applied Behavior Analysis,
Analysis, 11, 285–303. 41, 203–220.
Taub, E., Crago, J. E., Burgio, L. D., Groomes, T. E., Van Houten, R., & Nau, P. A. (1980). A comparison of
Cook, E. W., DeLuca, S. C., et al. (1994). An operant the effects of fixed and variable ratio schedules of re-
approach to rehabilitation medicine: Overcoming inforcement on the behavior of deaf children. Journal
learned nonuse by shaping. Journal of the Experimen- of Applied Behavior Analysis, 13, 13–21.
tal Analysis of Behavior, 61, 281–293. Vaughan, M. E., & Michael, J. L. (1982). Automatic
Thompson, R. H., Fisher, W. W., & Contrucci, S. A. reinforcement: An important but ignored concept.
(1998). Evaluating the reinforcing effects of choice Behaviorism, 10, 101–112.
in comparison to reinforcement rate. Research in De- Vollmer, T. R., Borrero, J. C., Lalli, J. S., & Daniel, D.
velopmental Disabilities, 19, 181–187. (1999). Evaluating self-control and impulsivity in
Thompson, R. H., & Iwata, B. A. (2000). Response ac- children with severe behavior disorders. Journal of
quisition under direct and indirect contingencies of Applied Behavior Analysis, 32, 451–466.
reinforcement. Journal of Applied Behavior Analysis, Vollmer, T. R., & Hackenberg, T. D. (2001). Reinforce-
33, 1–11. ment contingencies and social reinforcement: Some
Thompson, R. H., McKerchar, P. M., & Dancho, K. A. reciprocal relations between basic and applied re-
(2004). The effects of delayed physical prompts and search. Journal of Applied Behavior Analysis, 34,
reinforcement on infant sign language acquisition. 241–253.
Journal of Applied Behavior Analysis, 37, 379–383. Vollmer, T. R., & Iwata, B. A. (1991). Establishing op-
Tiger, J. H., Bouxsein, K. J., & Fisher, W. W. (2007). erations and reinforcement effects. Journal of Applied
Treating excessively slow responding of a young man Behavior Analysis, 24, 279–291.
with Asperger syndrome using differential reinforce- Weiner, H. (1969). Controlling human fixed-interval
ment of short response latencies. Journal of Applied performance. Journal of the Experimental Analysis of
Behavior Analysis, 40, 559–563. Behavior, 12, 349–373.
Tiger, J. H., & Hanley, G. P. (2004). Developing stimulus Wiskow, K. M., & Donaldson, J. M. (2016). Evaluation
control of preschooler mands: An analysis of sched- of a lag schedule of reinforcement in a group contin-
ule-correlated and contingency-specifying stimuli. gency to promote varied naming of categories items
Journal of Applied Behavior Analysis, 37, 517–521. with children. Journal of Applied Behavior Analysis,
Tiger, J. H., & Hanley, G. P. (2006). The effectiveness 49, 472–484.
of and preschoolers’ preferences for variations of Wolery, M., & Gast, D. L. (1984). Effective and efficient
multiple-schedule arrangements. Journal of Applied procedures for the transfer of stimulus control. Topics
Behavior Analysis, 39, 475–488. in Early Childhood Special Education, 4, 52–77.
Tiger, J. H., Hanley, G. P., & Hernandez, E. (2006). A Wolery, M., Kirk, K., & Gast, D. L. (1985). Stereotyp-
further evaluation of the reinforcing value of choice. ic behavior as a reinforcer: Effects and side-effects.
Journal of Applied Behavior Analysis, 39, 1–16. Journal of Autism and Developmental Disorders, 15,
Touchette, P. E., & Howard, J. S. (1984). Errorless learn- 149–161.
CHAP TER 15
Complexity is a cardinal feature of human be- havior analyst reduces complex behavior to small
havior. Humans routinely exhibit behavior that is teachable units. For some consumers, this approach
varied; is subtly discriminated; requires long-term initially appears reductive and far removed from
planning; is maintained by delayed, ambiguous the goal of teaching subtle, complex, and adaptive
contingencies; is initiated by verbal rules; and is behavior. When behavior analysts have sought to
recursive. Human behavior’s complexity, subtle ca- establish more complex behavior, they have typi-
pacity of adaptation, and incorporation of rich lan- cally achieved it by combining simpler behaviors
guage largely accounts for our evolutionary domi- to form a more complex chain, or by elaborating
nance over the earth (Ehrlich & Ehrlich, 2008). on a simpler behavior to produce a more complex
This wondrous capacity has also created enormous behavior (e.g., Sauttfr, LeBlanc, Jay, Goldsmith, &
challenges for individual human beings, as societ- Carr, 2011). Due to different conceptual bases and
ies have developed increasingly complex behav- emphasis on different outcomes (e.g., understand-
ioral requirements for successful adaptation. As the ing vs. behaving), cognitivist and behavior-analyt-
complexity of behavioral expectations increases, ic approaches are commonly viewed as conflicting
the demand for efficient teaching increases. or competing approaches, but the two approaches
Cognitive-constructivist approaches to teach- have much in common. For example, they both
ing have achieved dominance in many areas, identify some of the same procedures as effective,
because they explicitly emphasize the complex, such as practice with feedback, but provide differ-
varied nature of the target material and its accom- ent explanations for how and why the procedures
panying behavior (e.g., Haywood, 2004; Martens are effective (Carroll, Kodak, & Adolf, 2016, vs.
& Daly, 1999; Richardson, 2003). This approach Trapman, van Gelderen, van Steensel, van Schoo-
to teaching makes sense to many consumers and is ten, & Hulstijn, 2014). Both approaches seek to
sufficient for many learners. By contrast, behavior- help individuals develop complex behavioral rep-
analytic teaching procedures have commonly em- ertoires that include flexible and generalized re-
phasized an elemental approach, in which the be- sponding to diverse stimuli.
252
Building Complex Repertoires from Discrete Behaviors 253
A strength of the systematic, elemental building Behaviors themselves are the most intuitive
approach that behavior-analytic teaching adopts is source of complexity. No generally accepted defi-
that simplified behaviors can be taught to individ- nition of behavioral complexity exists in the be-
uals who have difficulty acquiring new or complex havior-analytic literature. We suggest that the fol-
behavior (Luiselli & Hurley, 2005). By contrast, lowing five dimensions are important to consider
some have criticized behavior-analytic approaches in establishing new behavior: (1) subordinate and
for failing to capture the symbolic meaning or superordinate skills, (2) sequencing, (3) promoting
underlying structure of complex behaviors and variability in responding, (4) ambiguity in natu-
for fostering dependence on instructors (Hickey, ral criteria, and (5) establishing self-management
Moore, & Pellegrino, 2001; Kroesbergen, Van Luit, skills. Most complex behaviors include or require
& Maas, 2004). This criticism of behavior-analytic several subordinate or prerequisite skills. For ex-
teaching appears quite reasonable when we exam- ample, fluent decoding skills are precursors to text
ine studies for which the goal was to establish a search and reading comprehension skills (Nation-
specific response (e.g., Swain, Lane, & Gast, 2015). al Reading Panel, 2000).
It appears less tenable in the context of broader Another considerable challenge is identify-
and more systematic behavior-analytic approaches ing the required proficiency level for prerequisite
to teaching, such as direct instruction, that clearly behaviors before teaching the target skill. That
emphasize meaning, structure, and behaviors that a student must be able to complete addition and
represent understanding (Liem & Martin, 2013). subtraction operations correctly before learning
to balance a checkbook may be obvious. However,
defining which operations and what accuracy and
COMPLEXITY: FLEXIBLE, DIVERSE BEHAVIOR fluency levels to require before teaching the target
skill may pose a considerable challenge (see Kelley,
Behavioral complexity is difficult to define, be- Reitman, & Noell, 2002, for a discussion of accu-
cause it exists as a relative comparison of behaviors racy and fluency criteria in subordinate skills).
within and across individuals. Driving to the cor- A second source of complexity arises from be-
ner market to buy a gallon of milk can be insur- havioral chains, a series of behaviors that occur
mountably complex for some individuals or quite sequentially and produce a consistent end state
simple for others. Behaviors also vary in complex- (Cooper, Heron, & Heward, 2007). Sequences of
ity. For example, talking to a friend who is pres- a different order of behaviors or variations within
ent is less complex than e-mailing the friend. Be- the chain may produce the same end state. For
haviors that were complex will become simple as example, many successful variations of the hand-
an individual develops greater skill in a domain. washing chain are possible, dependent on personal
Decoding a single word can be a complex process preferences and environmental context. Each be-
that includes many discriminations (Snow, Burns, havior in an established chain produces the con-
& Griffin, 1998). Later in the process of becom- ditioned reinforcer that serves as the discrimina-
ing literate, reading that same word can become a tive stimulus for the next behavior in the chain
simple behavior in which the reader perceives the (Cooper et al., 2007).
word as a single stimulus. A common goal of teach- The requirement that competence includes
ing is to help learners master simple behaviors that varied behavior, which is necessary for response
will become elements of more complex behaviors. generalization, creates a third source of complex-
Complexity emerges in all parts of the anteced- ity. For example, initiating play with peers requires
ent–behavior–consequence (A-B-C) chain. For variety across occasions and available materials
example, a behavior is more complex when its an- (Ledbetter-Cho et al., 2015). Variable behavior
tecedent is ambiguous (Harding, Wacker, Cooper, that maintains contact with reinforcement is a
Millard, & Jensen-Kovalan, 1994); when a delay common feature of social, vocational, and aca-
occurs between the onset of the discriminative demic behaviors. For example, the person who
stimulus and the target behavior; or when the same tells the same joke over and over is not likely to
stimulus is discriminative for different behavioral receive continued reinforcement for joke telling,
repertoires, depending on the context (Hughes even if the joke initially occasioned laughter. Re-
& Barnes-Holmes, 2014). Complex stimuli whose sponse generalization and behavioral flexibility
functional properties change across contexts are are very challenging for some learners. Individuals
common aspects of social interactions, academic who require many trials with carefully controlled
activities, and vocational activities. antecedent and consequent stimuli to learn also
254 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
have difficulty learning to respond to variability of those behaviors to produce more flexible, adap-
in natural contexts, and this is one of the central tive repertoires. We first discuss selected issues in
features of some developmental disabilities (Reit- the assessment of behaviors and individuals before
zel et al., 2013). teaching begins. We then discuss many specific
A fourth source of complexity is evident for be- procedures behavior analysts might use to estab-
haviors that are so varied that correct and incor- lish new behavior. We present the procedures as
rect responses are difficult to define. These behav- they might arise in practice as an analyst moves
iors are so common in human interaction that we from establishing an initial response, to elaborat-
might describe them as normative. For example, ing on that response, to creating a more complex
a response to a simple greeting may have a nearly repertoire. Thus these sections progress from shap-
infinite number of acceptable responses. ing and prompting, to chaining, to strategy in-
A fifth source of complexity arises when behav- struction, and finally to generalization.
ior requires substantial planning, progress moni-
toring, and plan revision, also known as executive
control (Mahy, Moses, & Kliegel, 2014) or self- ASSESSING BEHAVIORS AND INDIVIDUALS
management (Gureasko-Moore, DuPaul, & White,
2006). Executive control or self-management in Any program to establish new behavior should
this context refers to organizing and evaluating begin with an assessment of the individual’s cur-
responses necessary to complete complex behav- rent skills, behaviors, goals, and preferences. We
iors. Although the term executive control typically can conceptualize establishing new behavior as
describes internal unobservable processes, plan- the answers to three questions. First, what do we
ning, progress monitoring, and plan revision can expect the learner to do? Second, what does the
be observable behaviors. learner know how to do? Third, what procedures
The consequences of behavior can create an ad- can we use to build on what the learner does now,
ditional source of complexity in establishing and so that he or she can meet the new expectations?
maintaining behaviors. Delayed consequences, Although these questions are intuitive, complex-
thin reinforcement schedules, and small-magni- ity arises in the details. For example, most indi-
tude, cumulative consequences are extraordinarily viduals will need to learn many new behaviors
common in human endeavor (Malott, 1989) and that likely overlap in function and topography.
are frequently problematic in establishing and sus- We often must prioritize the behaviors that we
taining behavior. Natural consequences can be will teach. Space limitations preclude an exten-
sufficiently delayed, on such thin schedules, and sive consideration of prioritizing strategies in this
so ambiguous in their presentation that they are chapter. Generally, we should target those be-
insufficient to teach behavior or maintain existing haviors that have the broadest possible adaptive
behavior (Malott, 1989). importance and those that are prerequisites of
Behavior-analytic teaching often focuses on these broad and important behaviors. Research-
specific responses, using tightly controlled pro- ers have used the term behavioral cusps to describe
cedures that can produce inflexible, tightly con- behaviors that make broad contributions to an in-
trolled responding. We should conceptualize this dividual’s adaptive success (Rosales-Ruiz & Baer,
outcome, however, as a beginning rather than an 1997). Obvious examples include spoken language
end. This approach arose in part from demonstra- in social contexts, and reading in educational and
tions that individuals regarded as “unteachable” or vocational contexts.
“disabled” could learn far more and far faster than Instructional planning should begin by identi-
anyone thought was possible. The striking success fying criteria for competence that indicate when
in using principles of applied behavior analysis we should terminate instruction or shift to new
(ABA) led to ABA’s successful application to typi- targets. The end point might be age-appropriate
cally developing individuals (Daly, Persampieri, oral-language skills or reading text and correctly
McCurdy, & Gortmaker, 2005; Koscinski & Hoy, answering comprehension questions with an in-
1993). Moving from discrete teachable behaviors termediate stage to prepare the student to learn
to elaborate flexible repertoires is a fundamental additional material. For example, teaching a child
goal of teaching represented in the ABA literature with autism spectrum disorder (ASD) to articulate
(e.g., Reid, Lienemann, & Hagaman, 2013). In the targeted words in response to appropriate anteced-
balance of this chapter, we focus on critical issues ent stimuli is less likely to be an end than a stage in
for establishing new behaviors and the elaboration the process of building oral-language skills.
Building Complex Repertoires from Discrete Behaviors 255
Once we identify the immediate goal of in- portant to behavior assessment as it is to behavior
struction, we can identify the behavior’s critical treatment. A standardized assessment may suggest
components. We often identify relevant approxi- that a learner lacks a skill, when in fact the as-
mations or create a task analysis of the behavior. sessment environment simply lacked the supports
Relevant approximations are behaviors that are necessary to elicit the learner’s performance. The
topographically like the target behavior, and that absence of contingencies for competent respond-
we can potentially shape through differential rein- ing and reinforcement for competing responses
forcement to produce the target behavior (Isaacs, or distractions can yield negatively biased assess-
Thomas, & Goldiamond, 1960). For example, a ments (Noell, Freeland, Witt, & Gansle, 2001).
student may already read grade-level words, but Behavior analysts should assess behavior under
with insufficient fluency to be competent. In this varied conditions that test consequences.
case, we may be able to use reading slowly to shape The next stage is to identify an instructional
fluent reading (e.g., Noell et al., 1998). procedure that is appropriate for establishing or
A task analysis is the process of breaking down refining the target behavior. We describe specific
a complex behavior into a series of discrete, mea- ABA teaching procedures below. Behavior ana-
surable, and teachable components. For example, lysts should note that the assessment–treatment or
students must read numbers and operation signs, assessment–teaching process is recursive and con-
identify correct answers, and write the correct tinuous. Once teaching begins, ongoing assess-
answers to complete multiplication problems. The ment data should guide decisions that will change
three strategies researchers most commonly rec- the instructional plan as necessary.
ommend for completing task analyses are as follow.
First, we can watch the performance and record
the steps that competent individuals use to com- SHAPING
plete the task. Second, we can consult an expert
(an individual with specialized content expertise Shaping is an instructional approach that is par-
or a specialized published resource). Third is sim- ticularly important for learners who have a low
ply to complete the target behavior ourselves and probability of exhibiting the target behavior even
record the steps that were necessary to complete with prompting, but engage in some related behav-
the task. ior that we can use to begin instruction. Shaping
Once we identify the component steps of the involves increasing the probability of a behavior’s
target task, assessment will shift to the learner. occurrence through the gradual transformation
We will ask the learner to complete the task and of some property of responding. Differential rein-
record which steps the learner completes indepen- forcement of successive approximations of a tar-
dently. We will compare the learner’s performance geted operant class produces this transformation.
to a performance standard to identify which steps Shaping across topographies modifies the topogra-
the learner preformed competently and which phy of a response, and researchers have demon-
we need to teach, including instructional targets. strated it in several classic studies (Horner, 1971;
Unfortunately, evidence-based standards are not Isaacs et al., 1960; Skinner, 1938). For example,
available for many tasks, but we can apply ratio- Isaacs et al. (1960) shaped the eye movements of
nal and local, normative standards for some tasks. an individual diagnosed with comatose schizo-
Adequate standards typically require attention to phrenia into lip movements, then speech sounds,
quality, accuracy, and fluency (see Kelley et al., and eventually recognizable words. Shaping within
2002). For example, the assessment may show that topography modifies the rate, magnitude, or some
the student’s reading fluency is too low for him or other property of the target operant. Research-
her to answer comprehension questions accurately. ers have used this type of shaping to increase the
Assessment of reading performance may suggest arm extension of an athlete during a critical step
that the student has sufficient fundamental skills in pole vaulting (Rea & Williams, 2002) and to
but needs fluency building, that some critical sight increase the duration individuals held their breath
words are missing, that phonics/decoding skills are before measuring exhaled carbon monoxide levels
insufficient for reading novel words, or a combina- during treatment for smoking cessation (Scott,
tion. Each of these outcomes suggests a different Scott, & Goldwater, 1997).
initial focus for instruction. Learners typically emit a distribution of be-
One of the central tenets of ABA is that the haviors relevant to the targeted response dimen-
environment changes behavior. This fact is as im- sion. Shaping uses extinction and reinforcement
256 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
to shift this distribution, so that the proportion servations has the potential to decrease efficiency,
of responses that contain the desired response because we must postpone reinforcement until the
property increases (Galbicka, 1994). Continued learner has emitted a sufficiently large number of
differential reinforcement of responses above a cri- responses to establish the criterion.
terion value produces differentiation in which an
increasing proportion of behavior is at or nearer
to the target behavior. Factors affecting the prob- PROMPTING
ability of successful shaping include properties of
the initial response and the way we establish the When a response is not in a learner’s repertoire or
criterion for reinforcement. The response targeted is not under appropriate stimulus control, prompt-
early in the process must occur at a sufficient level ing may be necessary to evoke the response so that
or rate to permit initial reinforcement. It must we can reinforce it. Prompts are antecedent stimuli
also approximate the target behavior, so that we that increase the probability of a desired response.
can differentially reinforce the relevant response Prompting can help a stimulus become discrimi-
dimension. Finally, the initial response must have native by increasing opportunities to provide dif-
enough variability that we can provide differen- ferential reinforcement in its presence (Alberto &
tial reinforcement for responses that exceed an Troutman, 1986; Cooper et al., 2007; Demchak,
established criterion, thus shaping the response 1990; Miltenberger, 2001). Researchers have used
toward the target. Note that increased variability prompts to teach communication (Matson, Sevin,
is a predictable side effect of both reinforcement Fridley, & Love, 1990; Williams, Donley, & Keller,
(Skinner, 1938) and extinction (Lerman & Iwata, 2000), academic skills (Noell, Connell, & Duhon,
1996), so initial responses that seem relatively in- 2006; Stevens, Blackhurst, & Slaton, 1991), leisure
variant may be amenable to shaping. skills (DiCarlo, Reid, & Stricklin, 2003; Schleien,
Determining which responses to reinforce and Wehman, & Kiernan, 1981; Oppenheim-Leaf,
which responses to extinguish can be challenging. Leaf, & Call, 2012), social skills (Krantz & Mc-
If the criterion for reinforcement is too low, we will Clannahan, 1993; Garcia-Albea, Reeve, Brothers,
reinforce a large proportion of responses and shap- & Reeve, 2014), self-help skills (Pierce & Sch-
ing will proceed slowly. By contrast, if the criterion reibman, 1994; Taylor, Hughes, Richard, Hoch,
is too high, we will reinforce a small proportion & Coello, 2004), and vocational skills (Bennett,
of responses, and responding may be extinguished. Ramasamy, & Honsberger, 2013).
Galbicka (1994) recommended that instructors use Behavior analysts have categorized prompts as
percentile schedules to empirically determine the stimulus and response prompts (Schoen, 1986;
criterion for reinforcement during shaping. All re- Wolery & Gast, 1984). Stimulus prompts are those
sponses that exceed a preestablished rank-ordered in which we alter or present some property of the
response from a sample of the previous responses criterion stimulus (Etzel & LeBlanc, 1979). For ex-
produce reinforcement (e.g., the third highest ample, a teacher uses a stimulus prompt when he
from the last 10). Researchers have successfully or she places two pictures in front of the learner,
used this approach to shape behaviors such as eye and the correct picture is larger than the incorrect
contact (Hall, Maynes, & Reiss, 2009) and aca- one. By contrast, response prompts involve teacher
demic-task engagement (Athens, Vollmer, & St. behavior to evoke the desired learner behavior. For
Peter Pipkin, 2007; Clark, Schmidt, Mezhoudi, & example, a teacher uses a response prompt when
Kahng, 2016). he or she points to the correct picture after the
The advantage of a percentile schedule is that it learner did not respond when the teacher said,
constantly updates the distribution of responses it “Point to [correct picture].”
uses to establish the criterion for reinforcement as A script is a stimulus prompt that we can use
responding varies, which keeps the proportion of to facilitate complex behavior, especially conver-
reinforced responses constant. To date, researchers sational skills. Subtlety is a distinct advantage of
have conducted few studies to determine the opti- prompting social behaviors such as conversations
mal rank order and number of responses to sample with a script, because more contrived prompts
to establish a percentile schedule. In a notable ex- (such as vocal instructions) might be off-putting
ception, Athens et al. (2007) found that percentile to conversational partners. We can fade the script
schedule procedures were more effective when they length from a complete text to a single word or a
based the criterion for reinforcement on a larger symbol (e.g., Krantz & McClannahan, 1993), or
number of observations. However, using more ob- eliminate the script entirely (e.g., Garcia-Albea
Building Complex Repertoires from Discrete Behaviors 257
et al., 2014). Researchers have used scripts as a prompting, such as evaluating the effects of model
caregiver-mediated intervention for children with type and of recording perspective (e.g., Domire &
ASD to promote verbal interactions during play Wolfe, 2014).
(Reagon & Higbee, 2009). However, scripts are
limited to learners who can read, and script fading
can be lengthy with some learners. PROMPT FADING
Response prompts exist along a continuum
of intrusiveness in the amount of assistance re- Prompt dependence occurs when the prompt over-
quired to evoke the desired behavior from the shadows the criterion stimulus to such an extent
least intrusive verbal prompts, to the moderately that it never takes on discriminative properties in
intrusive gestural or model prompts, to the more the absence of the prompt. Thus we must transfer
intrusive physical prompts (Cooper et al., 2007; stimulus control from the prompt to the criterion
Miltenberger, 2001). We can deliver most prompts stimulus. Fading is a method of gradually removing
at different intrusiveness levels, such as a partial a prompt so that the behavior eventually comes
verbal prompt instead of a complete instruction or under control of the criterion stimulus in the ab-
a physical prompt, to guide the learner to perform sence of prompts. For example, Wichnick-Gillis,
the first portion instead of the entire behavior. Vener, and Poulson (2019) used textual prompts
Modeling, in which the antecedent stimulus is embedded within instructional stimuli (i.e.,
topographically identical to the target behavior, scripts) to teach three children with ASD to en-
is a prompt that can be especially effective for gage in social interactions. During generalization,
teaching complex behaviors (Bandura, Ross, & the scripts were gradually faded by removing one
Ross, 1963). Video modeling and prompting are word at a time from the end of a given script until
methods to demonstrate target behavior without all words were faded.
requiring an instructor to model the behavior each
time the learner receives the prompt. The prompt
Fading Stimulus Prompts
in video modeling is a video of an individual com-
pleting the behavior, which provides the learner There are two primary methods of fading stimu-
with a visual overview of the steps in the behavior lus prompts: stimulus shaping and stimulus fading.
sequence. The prompt in video prompting is a clip In stimulus shaping, we alter the property of the
of an individual completing one step of a behavior criterion stimulus that is critical to the intended
sequence and a video clip of the next step after discrimination, so that the learner can initially
the learner completes the previous step (Domire & make the discrimination. For example, when a
Wolfe, 2014). Researchers have used both methods chef is teaching a sous-chef to make subtle dis-
to teach complex chains of behavior, including so- criminations of saltiness, initial training may in-
cial skills and daily living tasks (Ayres & Langone, clude samples with very distinct differences in salt
2005). There is some evidence that video prompt- content. Once the learner is reliably making the
ing may facilitate acquisition of chained responses discrimination, we gradually diminish the altered
more effectively than video modeling (Cannella- stimulus property until the stimulus is representa-
Malone et al., 2006). tive of the criterion stimulus (Etzel & LeBlanc,
Several factors can influence the effective- 1979). Thus the difference in the amount of salt in
ness of modeling, such as the learner’s observing the two samples during stimulus shaping may be-
the modeled behavior produce reinforcement, as come smaller until the sous-chef can detect even
well as the similarity between the model and the subtle differences.
learner or between discriminative stimuli pre- In stimulus fading, we alter a property of the
sented to the model and the learner (Bandura criterion stimulus other than the dimension criti-
et al., 1963). Video self-modeling, in which the cal for the discrimination (Etzel & LeBlanc, 1979;
learner is also the model in the video, is one way Wichnick-Gillis et al., 2019). For example, we can
to maximize similarity between model and learner alter position prompts or the size of a target letter
(Buggey & Ogle, 2012). Although modeling is ef- when teaching letter identification. In this case,
fective, two of its limitations are that the learner neither position nor size is the critical dimension
must have generalized imitation skills and attend for discriminating the target letter from other let-
to the model during instruction. Further research ters. Rather, the form of the letter is the critical
is needed to identify the components that influ- property. During stimulus fading, we bring the size
ence the effectiveness of video modeling and video or position of the target letter closer to the position
258 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
or size of the alternative letter until the learner learns to emit the target behavior independently
can discriminate the letter by its form. (Billingsley & Romer, 1983).
A review of studies comparing stimulus shaping During graduated guidance, the instructor grad-
and stimulus fading found that both were effective ually eliminates the controlling prompt by only
instructional approaches (Ault, Wolery, Doyle, & presenting the level of prompt necessary to evoke
Gast, 1989). However, stimulus shaping appears the target behavior. A controlling prompt is one
to be more effective than stimulus fading, perhaps that consistently results in the learner’s exhibiting
because stimulus fading requires the learner to the target behavior (Wolery et al., 1992). We can
shift the discrimination from an irrelevant stimu- use least-to-most prompting to identify controlling
lus dimension to the relevant dimension. Making prompts, which often include physical guidance.
such a shift may be difficult for some learners—es- For example, parents taught yoga poses to their
pecially those who selectively attend to certain di- children by gradually fading the amount of physi-
mensions of stimuli, such as some individuals with cal guidance from a firm hold to shadowing their
ASD (Wolery & Gast, 1984). children (Gruber & Poulson, 2015). We can use
graduated guidance to fade physical guidance and
to transfer stimulus control to other types of con-
Fading Response Prompts
trolling prompts, such as verbal prompts (Schoen,
The five main fading procedures researchers have 1986).
studied for transferring stimulus control from a An advantage of graduated guidance is that the
response prompt to a criterion stimulus are least- learner can be as independent as possible, because
to-most, graduated guidance, most-to-least, time- the instructor only provides the minimum amount
delay, and simultaneous prompting. Least-to-most of guidance necessary. A disadvantage is that fad-
prompting, or the system of least prompts, is adapt- ing is not systematic. Fading relies on subjective
able to teaching behavior chains such as a series judgments about the required prompting level
of motor responses (e.g., folding laundry) and (Wolery & Gast, 1984), which the instructor must
discrete behavior such as object labeling. During make rapidly, based on the learner’s responses.
least-to-most prompting, the instructor presents This can affect implementation integrity and can
the criterion stimulus so the learner can emit the be difficult to evaluate in the absence of systematic
correct response independently. If the learner research.
does not emit the correct response, the instructor Most-to-least fading begins with the delivery of a
presents increasingly intrusive prompts until the controlling prompt, and the amount of assistance
learner emits the target behavior. For example, necessary for the individual to complete the behav-
an instructor may use least-to-most prompting to ior correctly varies across trials instead of within a
teach a learner to correctly identify sight words by trial. The intrusiveness of the prompt decreases or
first presenting the criterion stimulus, a flash card. increases on subsequent trials, based on whether
The instructor provide a more intrusive prompt, the learner meets a mastery or failure criterion, re-
such as the first syllable, if the learner does not spectively, for the current prompt level. Note that
make the target response before an interval ex- if the first prompting level is a controlling prompt,
pired (e.g., 5 seconds), and an even more intrusive the learner should always meet the mastery cri-
prompt, the word, if the learner does not make the terion, because the response should always occur
target response after the previous prompt. Wol- following this most intrusive prompt. Graduated
ery and Gast (1984) suggested that as instructors, guidance and most-to-least prompt fading are well
we should present the criterion stimulus at each suited for teaching chained motor responses and
prompt level and use a constant response interval for learners who require many response–reinforc-
after each prompt. We should also consider wheth- er pairings to achieve independence (Wolery &
er successive prompts increase the probability of Gast, 1984).
the target behavior. If not, we should use the least Libby, Weiss, Bancroft, and Ahearn (2008)
intrusive prompt that is likely to occasion the be- modified most-to-least prompt fading by insert-
havior. ing a delay between controlling prompts to allow
An advantage of least-to-most prompt fading learners an opportunity for independence. This
is that the learner can emit the correct behavior time-delay fading procedure produced mastery al-
without prompts. It also may be easier to imple- most as rapidly as least-to-most prompting, but
ment than other strategies, because more intrusive with fewer errors. The two types of time-delay fad-
prompts become unnecessary as the individual ing, constant and progressive (O’Neill, McDowell,
Building Complex Repertoires from Discrete Behaviors 259
& Leslie, 2018; Snell & Gast, 1981), begin with may be more appropriate for individuals who lack
0-second-delay trials in which the criterion stimu- these skills. In general, research has shown that
lus and prompt occur simultaneously. We typically the prompt-fading methods described above can
use more 0-second-delay trials for difficult tasks or effectively transfer stimulus control to a criterion
lower-functioning learners. In progressive time-de- stimulus for at least some tasks and participants.
lay fading, the time between the criterion stimulus We cannot draw further conclusions about prompt
presentation and the prompt gradually increases fading beyond the idiosyncratic variables evalu-
after each trial, several trials, or each instructional ated, such as participant characteristics, tasks, and
session (Heckaman, Alber, Hooper, & Heward, prompting variations.
1998). With constant time-delay fading, the in-
structor delays the prompt for a specified time after
presentation of the criterion stimulus, and this la- CHAINING
tency remains fixed during instructional sessions
(Snell & Gast, 1981). Each step or component of a behavior chain has
Some advantages of time-delay fading include its own conditioned reinforcers and discrimina-
its low error rate and its simplicity. Constant time- tive stimuli (Kelleher, 1966; Skinner, 1938). That
delay fading is especially simple to use, which may is, the consequence following completion of each
produce better treatment integrity (Wolery et al., component of the behavior chain may function as
1992). Simplicity may be especially important both a conditioned reinforcer for the previous be-
when caregivers or supervisees implement inter- havior and a discriminative stimulus for the next
ventions. For example, DiPipi-Hoy and Jitendra one. We typically complete a task analysis for the
(2004) showed that parents could implement con- chain of target behaviors and teach the chain by
stant time-delay procedures with good fidelity. using forward chaining, backward chaining, or
During simultaneous prompting, the instructor total-task presentation.
delivers the controlling prompt immediately after During backward chaining and forward chain-
presenting the demand (Morse & Schuster, 2004). ing, the instructor teaches one component of the
That is, trials have a 0-second delay, and the in- behavior chain at a time. The instructor teaches
structor does not introduce a delay. The instructor each additional component as the learner meets
conducts an assessment before instructional ses- the mastery criterion for the previous components.
sions to determine whether the learner can emit The difference is that in backward chaining, the
the target response without the prompt. Research- instructor teaches the behavior chain in reverse
ers have used simultaneous prompting to teach order, starting with the last component; in forward
discrete behaviors (Tekin-Iftar, Acar, & Kurt, chaining, behaviors are taught in the order they
2003), chained tasks (Parrott, Schuster, Collins, occur in the chain, starting with the first compo-
& Gassaway, 2000), and vocational tasks (Fetko, nent. In total- or whole-task chaining, the learner
Schuster, Harley, & Collins, 1999). The advan- performs the entire behavior chain on every in-
tages of simultaneous prompting are low error structional trial.
rate and simplicity, but it may be less sensitive to During reverse chaining, the instructor physi-
detecting the moment at which mastery occurs. cally guides the learner to perform the behavior
However, instructors can conduct more trials per chain’s components until the last one, which the
time than they can with prompt-delay procedures, learner performs independently (Sternberg &
and simultaneous prompting produces more rapid Adams, 1982). The instructor teaches components
mastery than other prompting strategies do (Akm- in reverse order by physically guiding progressively
anoglu, Kurt, & Kapan, 2015; Schuster, Griffen, & fewer components (e.g., last two, last three) as
Wolery, 1992; Swain et al., 2015). the learner masters each component. In backward
Studies comparing prompt-fading methods have chaining with leap ahead (Spooner, Spooner, &
produced conflicting results. Limitations of these Ulicny, 1986), the instructor does not teach every
comparative studies is that most participants dis- component directly, to increase time efficiency.
played generalized imitation, attended well, waited Rather, the instructor conducts ongoing assess-
for teacher assistance, and demonstrated clear pref- ment to determine whether the learner can per-
erences for potential reinforcers. Wolery and Gast form some components without training.
(1984) suggested that constant time-delay may A potential advantage of backward chaining is
be more efficient for students who exhibit these that the learner produces natural reinforcement
behaviors, but that other prompt-fading strategies by completing each component of the behavior
260 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
chain, whereas completing initial components ability in responding can be important, because
during forward-chaining produces conditioned it often determines whether a response will pro-
reinforcement (Spooner et al., 1986). However, duce natural reinforcers. For example, peers may
the natural consequences produced in backward perceive a student who always asks for help using
chaining may not function as reinforcement for a single phrase in the same tone of voice as odd,
some individuals. The advantage of total-task and may ignore or shun the student. McClanna-
presentation is that the learner has increased op- han and Krantz (2005) used scripts to teach sev-
portunities for conditioned reinforcement by prac- eral mand frames (i.e., I want, I need, or I would
ticing every component in the behavior chain on like) to participants and observed increases in
every trial. However, total-task presentation may spontaneous, novel mand frames. Similarly, Betz,
be less time-efficient. Forward chaining may be Higbee, Kelley, Sellers, and Pollard (2011) taught
easiest to use, because the instructor teaches com- varied responses to individuals with ASD by using
ponents in the order in which they occur in the stimulus prompts, each associated with a unique
chain, and this may produce the best long-term color. They faded the stimulus prompts and used
performance (Watters, 1992). the color prompts, and faded the color prompts for
Smith (1999) showed that fewer errors occurred two of three participants who continued to emit
at the beginning of the chain during forward the varied responses.
chaining and at the end of the chain during back- We can promote response variability by manip-
ward chaining. Thus an instructor should use for- ulating the consequences of responding. Extinc-
ward chaining if a learner is unlikely to complete tion-induced variability is one example in which
the chain after an error. Direct comparisons of the extinction of a previously reinforced response in-
different chaining methods have shown mixed creases the likelihood that various other responses
results (Ash & Holding, 1990; Hur & Osborne, will occur (Goetz & Baer, 1973; Sullivan et al.,
1993; Slocum & Tiger, 2011; Spooner & Spooner, 2020). Extinction-induced variability has the
1984; Spooner, Weber, & Spooner, 1983; Watters, unique benefit of eliminating instructional time
1992; Wightman & Sistrunk, 1987). As with other for behaviors that are in the learner’s repertoire
methods described in this chapter, acquisition of but occur infrequently. For example, Valentino,
skills via different chaining methods is likely to be Shillingsburg, Call, Burton, and Bowen (2011)
idiosyncratic across populations and highly influ- implemented extinction for the signed mands of
enced by the target skill. children with limited vocalizations, and observed
increases in vocalizations. The chances for extinc-
tion of the original target behavior and emergence
PROMOTING RESPONSE GENERALIZATION of problem behavior are disadvantages of extinc-
AND VARIETY tion-induced variability.
Another method to produce response variability
We direct readers to Chapters 5 (Spradlin, Simon, is a lag schedule of reinforcement. The instructor
& Fisher) and 6 (Podlesnik, Jimenez-Gomez, & reinforces a response in a lag schedule if it differs
Kelley) of this volume for discussions of generaliza- from a designated number of previously emitted re-
tion and methods useful for its facilitation, includ- sponses (Falcomata et al., 2018; Page & Neuringer,
ing descriptions of stimulus control, equivalence 1985). For example, if an individual had previously
classes, and recombinative generalization. We emitted the response hi followed by hello, neither
review some issues about generalization relevant of these greetings would be eligible for reinforce-
to establishing complex behaviors. For example, ment as the third response on a Lag 2 schedule.
multiple-exemplar training, in which the instruc- Only a novel greeting (e.g., good morning) would
tor prompts and reinforces responding to several produce reinforcement, but a novel response or hi
members of a stimulus class, can produce gener- for the fourth greeting would produce reinforce-
alization and promote spontaneous or varied re- ment, because neither was one of the previous two
sponding (Stokes & Baer, 1977). For example, responses. Researchers have used lag schedules to
Krantz and McClannahan (1993) used scripts establish variety for verbal behavior (Esch, Esch,
with varied content that prompted comments & Love, 2009; Falcomata et al., 2018; Lee, McCo-
about activities to teach children with ASD to mas, & Jawor, 2002; Silbaugh, Falcomata, & Fer-
initiate social interactions. Scripted comments guson, 2018; Wiskow, Matter, & Donaldson, 2018)
increased, and spontaneous, unscripted comments and building block structures (Napolitano, Smith,
also increased for several participants. Such vari- Zarcone, Goodkin, & McAdam, 2010), and to
Building Complex Repertoires from Discrete Behaviors 261
maintain variable responding after discontinua- the benefits of using the strategy. Teachers then
tion of the lag schedule (Heldt & Schlinger, 2012). model the strategy for students by writing essays
A disadvantage is that patterned responding may while asking themselves questions aloud, followed
emerge. Using variable-lag schedules may mitigate by modeling self-instruction procedures (e.g.,
this problem (Lee et al., 2002). self-evaluation and self-reinforcement). A second
common component of strategy instruction is
providing students with opportunities to practice
STRATEGIC INSTRUCTION problems in a programmed sequence from sim-
pler to more complex problems. Teachers provide
Strategies are more complex sequences of behav- students with corrective feedback, reinforcement,
ior that include assessment, planning, execution, and many models, allowing the students to prac-
and evaluation of a course of action. Importantly, tice skills collaboratively. Collaborative practice,
a strategy is a behavioral process by which an in- when possible, enables teachers to support stu-
dividual chooses, orders, and evaluates behav- dents while gradually providing them with greater
ior toward solving diverse problems, rather than independence. Teachers provide students with
a fixed behavioral sequence. We can distinguish mnemonics for remembering the strategy’s steps
skills and strategies by the unique roles that they and teach them to use self-monitoring, prompt
play in learning and achievement. Alexander and cards, or a combination to foster independence
Murphy (1999) describe skills as procedural knowl- and generalization. The students practice these
edge that students develop, which enables them to steps verbally until they memorize the strategy’s
perform tasks effectively with speed and accuracy. steps. Providing students with opportunities to
Students who achieve automaticity or fluency of practice newly learned strategies across multiple
skills can attend to more complex task dimensions exemplars promotes generalization and adaptation
(e.g., comprehending text after mastering decod- of strategies (Alber-Morgan, Hessler, & Konrad,
ing and sight words). For example, Wagner et al. 2007). For example, teachers should provide stu-
(2011) reported that fluency in writing individual dents with opportunities to practice reading and
letters contributed to the quality and complexity writing strategies and continued support across a
of first- and fourth-grade students’ writing. One variety of text types and writing tasks, so that the
critical goal of effective education is to provide students effectively use these strategies.
students with sufficient opportunities to practice Teaching students to use self-regulatory strate-
fundamental skills (e.g., decoding, computing gies, such as self-monitoring, self-recording, self-
basic math facts), so that they can develop ade- assessing, and self-reinforcing, will promote strat-
quate fluency to use those basic skills in a strategic egy use (Alexander et al., 1998; Reid et al., 2013).
manner (Ardoin & Daly, 2007). Teachers meet with students individually, review
As they do with basic skills, students learn and students’ work to establish a baseline, explain the
perform strategies better when we teach these ex- benefits of each self-regulatory behavior, help stu-
plicitly, when the environment supports effective dents to establish self-monitoring goals, and ver-
skill use, and when the environment naturally re- bally model the strategy. Research has shown that
wards strategy use (Duffy, Roehler, Sivan, & Rack- teaching students self-regulatory behaviors in-
liffe, 1987; Manset-Williamson & Nelson, 2005). creases their understanding of their academic abil-
An extensive literature exists describing models of ities, which enables them to connect strategy use
strategy instruction in reading, writing, and math- and successful performance. Students who have
ematics for students with and without learning been taught explicitly to use self-regulatory behav-
difficulties (Alexander, Graham, & Harris, 1998; iors experience greater acquisition, maintenance,
Reid et al., 2013). and generalization of self-regulatory strategy devel-
Instructors who use effective strategy instruc- opment (De La Paz, 1999; Pressley & Levin, 1987).
tion ensure that students have the prerequisite
background skills, explain the strategy to them,
model the strategy, and explain why they should DIRECT INSTRUCTION
use the strategy. For example, Self-Regulated
Strategy Development, an empirically validated Direct instruction focuses on teaching skills and
curriculum for teaching writing strategies, involves strategies to a level of mastery via explicit teacher-
explaining to students how good writers might use directed instruction (Grossen, 2004). A guiding
a strategy (e.g., planning, listing main ideas) and principle of direct instruction is that students can
262 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
learn, and failure to learn is viewed as the function they learn to the trained examples and to the wid-
of inappropriate teaching, curriculum, or both est array of new items and situations. Engelmann
rather than as a student characteristic. Siegfried and Becker (1978) referred to this as general-case
Englemann and Carl Bereiter developed direct programming. Teachers show students examples
instruction in the 1960s to enable students who and nonexamples in a specific sequence that al-
many believed could not learn to read to learn at lows students to easily recognize differences and
a pace that would allow them to catch up to their generalize what they have learned (Watkins &
nondisabled peers. Over 30 years of research sug- Slocum, 2004).
gests that the Corrective Reading Program (En- A misperception is that direct instruction fo-
glemann, 1999) improves reading skills at two to cuses only on rote learning and promotes passive
three times the typical rate if implemented con- learning (Adams & Engelmann, 1996; Leontov-
sistently and with integrity, allowing struggling ich, 1999). Results of studies suggest that direct
students to catch up to their peers (see Grossen, instruction promotes more generalization than
1998, for a review of the research). Researchers alternative programs supported by constructivists,
have broadened direct instruction’s application who commonly argue that teachers should only
to include critical reading (Darch & Kame’enui, guide students as the students discover rules and
1987), chemistry (Carnine, 1989), earth science strategies on their own. To examine this issue,
(BFA Educational Media, 1991), expressive writ- Klahr and Nigam (2004) assigned 112 third- and
ing (Walker, Shippen, Alberto, Houchins, & fourth-grade students to a direct-instruction or a
Cihak, 2005), U.S. history (Carnine, Crawford, discovery-learning condition to teach the control-
Harniss, & Hollenbeck, 1994), and problem solv- of-variables strategy, an elementary science objec-
ing (BFA Educational Media, 1991). tive. Teachers provided students in the direct-in-
Direct-instruction curricula share several com- struction condition with information on how and
mon characteristics to ensure that students with why control of variables works, showed examples
various needs will succeed. For example, assess- and nonexamples of the strategy, and explained
ment of a student’s instructional needs drives the differences between the examples. In the
placement in a direct-instruction curriculum. discovery-learning condition, teachers provided
Ongoing assessment provides information about students with identical materials, and students
the student’s skill development and teacher effec- developed their own experiments. Researchers
tiveness. Students are commonly placed in homo- conducted acquisition assessments in which stu-
geneous skill groups to ensure that students in a dents developed four of their own experiments,
group need the same level of instruction and guid- and a generalization assessment in which students
ance. It also decreases the probability that some evaluated two science fair projects. Results repli-
students either have mastered the targeted skills or cated previous studies indicating that acquisition
do not have the prerequisite skills to benefit from of the control-of-variables strategy was greater for
instruction. Groupings are temporary, however, students in the direct-instruction condition (Chen
because teachers use ongoing individualized per- & Klahr, 1999). In addition, results extended the
formance assessment to alter group membership literature by showing that students in the direct-
appropriately. instruction condition demonstrated more gener-
Direct-instruction materials streamline lessons alization, based on their science fair evaluations,
when this is practical. Englemann recommends than students in the discovery-learning condition
not teaching information students do not need, did. Results of a meta-analytic study showed that
so that struggling students can catch up (Engel- explicit instructions produced a mean effect size of
mann & Becker, 1978). Direct-instruction lessons d = –0.38, suggesting that explicit instruction pro-
are organized as logical developmental sequences, duced greater learning than unassisted discovery
so that students know the rules, concepts, opera- instruction (Alfieri, Brooks, Aldrich, & Tenen-
tions, and strategies necessary to learn the target baum, 2011).
skill (Kim & Axelrod, 2005; Kozloff, LaNuiziata, Teachers teach direct-instruction lessons at a
Cowardin, & Bessellieu, 2000). For example, brisk pace, to provide students with more learning
teachers only teach letter sounds, not names, to and response opportunities and to maintain active
students learning to sound out words, because the student engagement. Teachers and students use
letter names are not needed to sound out words. consistent language to define concepts, state rules,
The teacher teaches generalizable skills, concepts, and employ strategies, which prevents confusion
and strategies, so that students can apply what from variations in language. Teachers also use con-
Building Complex Repertoires from Discrete Behaviors 263
sistent verbal signals, nonverbal signals, or both behaviors in the teaching context and across con-
for frequent group unison responding. Students texts, persons, and times, and to exhibit a range
write their answers on dry-erase boards; this facili- of related behaviors that were never explicitly in-
tates academic engagement and practice, as well structed. The likelihood of students’ retaining and
as assessment of student performance (Watkins generalizing learning increases substantially when
& Slocum, 2004). Haydon and Hunter (2011) re- generalization is part of the curriculum rather than
ported that unison responding decreased teachers’ an afterthought (Daly, Martens, Barnett, Witt, &
redirective statements and increased their praise Olson, 2007).
statements, and that it increased students’ on-task One technique to promote generalization is to
behavior, correct responses, and test scores. provide students with numerous opportunities to
Teachers consistently use a model–lead–test– practice skills and strategies to promote mastery.
delayed-test procedure in which teachers first Evidence exists that teaching students to a high
model the target content, and then the teachers level of fluency, rather than just accuracy, pro-
and students work through skills and operations motes maintenance, generalization, and adapta-
together (Kozloff et al., 2000). As students become tion (Binder, 1996). There are also data suggesting
more proficient, teachers provide less information, that overlearning further increases retention and
decrease their prompts, and fade their corrective generalization. Overlearning is achieved by bring-
feedback from immediate to delayed. Problems ing students to levels of mastery and then provid-
increase in complexity. These shifts ensure ini- ing additional practice opportunities (Driskell,
tial and ongoing student success, thus maintain- Copper, & Willis, 1992). For example, Ardoin,
ing their motivation to learn, to master skills, and Williams, Klubnik, and McCall (2009) found that
to become more independent. Teaching skills to overlearning through increased opportunities to
mastery increases the likelihood that students respond enhanced maintenance effects, but did
will maintain their learning and will general- not substantially increase students’ initial reading
ize and adapt it to new situations (Binder, 1996). fluency.
After modeling, teachers test students’ acquisition Frequent reinforcement for accurate and flu-
through unison responding and provide immedi- ent responding also promotes generalization.
ate corrective feedback for mistakes. Teachers con- Teacher assistance in the form of modeling and
duct group and individual delayed tests later in the performance feedback ensures success. Teachers
same lesson and during subsequent lessons, to en- should graduate their assistance such that students
sure maintenance and to promote generalization experience success, but they should also require
(Kozloff et al., 2000). students to apply the skills and strategies inde-
Adhering to the precise sequence of teaching pendently across tasks that systematically increase
skills and strategies, and maintaining the clarity in difficulty. Teachers should systematically plan
and consistency of instructions, are the keys to the what academic responses to reinforce, which will
effectiveness of direct instruction. Teachers use increase the probability of skill generalization and
lesson scripts to increase the integrity and plan- allow for natural contingencies to serve eventually
fulness of instruction (Watkins & Slocum, 2004). as reinforcers (Daly et al., 2007).
A direct-instruction approach or curriculum is a Teachers should model and provide practice op-
relatively comprehensive instructional method portunities, using various examples of when stu-
that incorporates explicit strategy instruction into dents should use the target skill or strategy (Troia,
teaching basic skills and more complex behavioral 2002), and should provide students with examples
chains. Direct instruction includes ongoing assess- and nonexamples. In a study by Hicks, Bethune,
ment of student progress and needs, with recursive Wood, Cooke, and Mims (2011), teachers taught
plan revision, shaping, prompting, chaining, and correct preposition use to students with intellectu-
strategy instruction. al disabilities. For example, the teacher first placed
a ball on a box and said, “This is on” (example),
and then placed the ball at least 0.3 meters from
GENERALIZATION OF COMPLEX SKILLS the box and said, “This is not on” (nonexample).
The teacher next decreased the distance between
Both strategy instruction and direct instruction the ball and the box. After this, the teacher re-
devote considerable attention to generalization placed the ball with common classroom objects to
from the outset of teaching. Typically, the goal of demonstrate examples and nonexamples of on. Fi-
teaching is for the student to demonstrate the new nally, students participated in a scavenger hunt to
264 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
Daly, E. J., III, Persampieri, M., McCurdy, M., & Gort- Adami, S. Hoffman, K., Shpall, C., et al. (2018). Lag
maker, V. (2005). Generating reading interventions schedules and functional communication training:
through experimental analysis of academic skills: Persistence of mands and relapse of problem behav-
Demonstration and empirical evaluation. School Psy- ior. Behavior Modification, 42, 314–334.
chology Review, 34(3), 395–414. Fetko, K. S., Schuster, J. W., Harley, D. A., & Collins,
Darch, C. B., & Kame’enui, E. J. (1987). Teaching LD B. C. (1999). Using simultaneous prompting to teach
students critical reading skills: A systematic replica- a chained vocational task to young adults with se-
tion. Learning Disability Quarterly, 10, 82–91. vere intellectual disabilities. Education and Training
De La Paz, S. (1999). Self-regulated strategy instruction in Mental Retardation and Developmental Disabilities,
in regular education settings: Improving outcomes 34, 318–329.
for students with and without learning disabilities. Galbicka, G. (1994). Shaping in the 21st century: Mov-
Learning Disabilities Research and Practice, 14, 92– ing percentile schedules into applied settings. Journal
106. of Applied Behavior Analysis, 27, 739–760.
Delano, M. E. (2007). Improving written language per- Garcia-Albea, E., Reeve, S. A., Brothers, K. J., & Reeve,
formance of adolescents with Asperger syndrome. K. F. (2014). Using audio script fading and multiple-
Journal of Applied Behavior Analysis, 40, 345–351. exemplar training to increase vocal interactions in
Demchak, M. (1990). Response prompting and fading children with autism. Journal of Applied Behavior
methods: A review. American Journal on Mental Re- Analysis, 47, 325–343.
tardation, 94, 603–615. Goetz, E. M., & Baer, D. M. (1973). Social control of
DiCarlo, C. F., Reid, D. H., & Stricklin, S. B. (2003). form diversity and the emergence of new forms in
Increasing toy play among toddlers with multiple dis- children’s blockbuilding. Journal of Applied Behavior
abilities in an inclusive classroom: A more-to-less, Analysis, 62, 209–217.
child-directed intervention continuum. Research in Grossen, B. (1998). The research base for Corrective Read-
Developmental Disabilities, 24, 195–209. ing. Blacklick, OH: Science Research Associates.
DiPipi-Hoy, C., & Jitendra, A. (2004). A parent-deliv- Grossen, B. (2004). Success of a direct instruction
ered intervention to teach purchasing skills to young model at a secondary level school with high risk stu-
adults with disabilities. Journal of Special Education, dents. Reading and Writing Quarterly, 20, 161–178.
38(3), 144–157. Gruber, D. J., & Poulson, C. L. (2015). Graduated guid-
Domire, S. C., & Wolfe, P. (2014). Effects of video ance delivered by parents to teach yoga to children
prompting techniques on teaching daily living skills with developmental delays. Journal of Applied Behav-
to children with autism spectrum disorder: A review. ior Analysis, 49, 1–6.
Research and Practice for Persons with Severe Disabili- Gureasko-Moore, S., DuPaul, G. J., & White, G. P.
ties, 39(3), 211–226. (2006). The effects of self-management in general
Driskell, J. E., Copper, C., & Willis, R. P. (1992). Effect education classrooms on the organizational skills of
of overlearning on retention. Journal of Applied Psy- adolescents with ADHD. Behavior Modification, 30,
chology, 77, 615–622. 159–183.
Duffy, G. G., Roehler, L. R., Sivan, E., & Rackliffe, G. Hall, S. S., Maynes, N. P., & Reiss, A. L. (2009). Using
(1987). Effects of explaining the reasoning associated percentile schedules to increase eye contact in chil-
with using reading strategies. Reading Research Quar- dren with fragile X syndrome. Journal of Applied Be-
terly, 22(3), 347–368. havior Analysis, 42, 171–176.
Ehrlich, P. R., & Ehrlich, A. H. (2008). The dominant Harding, J., Wacker, D. P., Cooper, L. J., Millard, T.,
animal. Washington, DC: Island Press. & Jensen-Kovalan, P. (1994). Brief hierarchical as-
Englemann, S. (1999). Corrective reading program. sessment of potential treatment components with
Blacklick, OH: Science Research Associates. children in an outpatient clinic. Journal of Applied
Englemann, S., & Becker, W. S. (1978). Systems for Behavior Analysis, 27, 291–300.
basic instruction: Theory and applications. In A. C. Haydon, T., & Hunter, W. (2011). The effects of two
Catania & T. A. Bringham (Eds.), Handbook of ap- types of teacher questioning on teacher behavior and
plied behavior analysis (pp. 325–377). New York: Ir- student performance: A case study. Education and
vington. Treatment of Children, 34(2), 229–245.
Esch, J. W., Esch, B. E., & Love, J. R. (2009). Increasing Haywood, H. C. (2004). Thinking in, around, and
vocal variability in children with autism using a lag about the curriculum: The role of cognitive educa-
schedule of reinforcement. Analysis of Verbal Behav- tion. International Journal of Disability, Development
ior, 25(1), 73–78. and Education, 51, 231–252.
Etzel, B. C., & LeBlanc, J. M. (1979). The simplest treat- Heckaman, K. A., Alber, S., Hooper, S., & Heward, W.
ment alternative: Appropriate instructional control L. (1998). A comparison of least-to-most prompts
and errorless learning procedures for the difficult- and progressive time delay on the disruptive behavior
to-teach child. Journal of Autism and Developmental of students with autism. Journal of Behavioral Educa-
Disorders, 9, 361–382. tion, 8(2), 171–201.
Falcomata, T. S., Muething, C. S., Silbaugh, B. C., Heldt, J., & Schlinger, H. D. (2012). Increased variabil-
Building Complex Repertoires from Discrete Behaviors 267
ity in tacting under a lag 3 schedule of reinforcement. in the Netherlands. Elementary School Journal, 104,
Analysis of Verbal Behavior, 28(1), 131–136. 233–251.
Hickey, D. T., Moore, A. L., & Pellegrino, J. W. (2001). Ledbetter-Cho, K., Lang, R., Davenport, K., Moore, M.,
The motivational and academic consequences of el- Lee, A., Howell, A., et al. (2015). Effects of script
ementary mathematics environments: Do construc- training on the peer-to-peer communication of chil-
tivist innovations and reforms make a difference? dren with autism spectrum disorder. Journal of Ap-
American Educational Research Journal, 38, 611–652. plied Behavior Analysis, 48, 785–799.
Hicks, S. C., Bethune, K. S., Wood, C. L., Cooke, N. Lee, R., McComas, J. J., & Jawor, J. (2002). The effects
L., & Mims, P. J. (2011). Effects of direct instruction of differential and lag reinforcement schedules on
on the acquisition of prepositions by students with varied verbal responding by individuals with autism.
intellectual disabilities. Journal of Applied Behavior Journal of Applied Behavior Analysis, 35, 391–402.
Analysis, 44, 675–679. Leontovich, M. (1999, August). Direct controversial:
Horner, R. D. (1971). Establishing use of crutches by a Direct instruction makes enemies, converts. Title I
mentally retarded spina bifida child. Journal of Ap- Report. Retrieved July 2004, from www.ncld.org/sum-
plied Behavior Analysis, 4, 183–189. mit99/keys99-nichd.htm.
Hughes, S., & Barnes-Holmes, D. (2014). Associative Lerman, D. C., & Iwata, B. A. (1996). Developing a
concept learning, stimulus equivalence, and rela- technology for the use of operant extinction in clini-
tional frame theory: Working out the similarities and cal settings: An examination of basic and applied
differences between human and nonhuman behav- research. Journal of Applied Behavior Analysis, 29,
ior. Journal of the Experimental Analysis of Behavior, 345–382.
101(1), 156–160. Libby, M. E., Weiss, J. S., Bancroft, S., & Ahearn, W.
Hur, J., & Osborne, S. (1993). A comparison of forward H. (2008). A comparison of most-to-least and least-
and backward chaining methods used in teaching to-most prompting on the acquisition of solitary play
corsage making skills to mentally retarded adults. skills. Behavior Analysis in Practice, 1(1), 37–43.
British Journal of Developmental Disabilities, 39(77), Liem, G. D., & Martin, A. J. (2013). Direct instruction.
108–117. In J. Hattie & E. M. Anderman (Eds.), Internation-
Isaacs, W., Thomas, J., & Goldiamond, I. (1960). Ap- al guide to student achievement (pp. 366–368). New
plication of operant conditioning to reinstate verbal York: Routledge.
behavior in psychotics. Journal of Speech and Hearing Luiselli, J. K., & Hurley, A. D. (2005). The significance
Disorders, 25, 8–12. of applied behavior analysis in the treatment of in-
Kelleher, R. T. (1966). Chaining and conditioned rein- dividuals with autism spectrum disorders (ASD).
forcement. In W. K. Honig (Ed.), Operant behavior: Mental Health Aspects of Developmental Disabilities,
Areas of research and application (pp. 160–212). Engle- 8, 128–130.
wood Cliffs, NJ: Prentice-Hall. Mahy, C. V., Moses, L. J., & Kliegel, M. (2014). The
Kelley, M. L., Reitman, D., & Noell, G. H. (2002). Prac- development of prospective memory in children: An
titioner’s guide to empirically based measures of school executive framework. Developmental Review, 34(4),
behavior. New York: Kluwer Academic/Plenum. 305–326.
Kim, T., & Axelrod, S. (2005). Direct instruction: An Malott, R. W. (1989). The achievement of evasive goals.
educators’ guide and a plea for action. Behavior Ana- In S. C. Hayes (Ed.), Rule-governed behavior: Cogni-
lyst Today, 6(2), 111–120. tion, contingencies, and instructional control (pp. 269–
Klahr, D., & Nigam, M. (2004). The equivalence of 322). New York: Plenum Press.
learning paths in early science instruction: Effects of Manset-Williamson, G., & Nelson, J. M. (2005). Bal-
direct instruction and discovery learning. Psychologi- anced, strategic reading instruction for upper-ele-
cal Science, 15, 661–667. mentary and middle school students with reading
Koscinski, S. T., & Hoy, C. (1993). Teaching multiplica- disabilities: A comparative study of two approaches.
tion facts to students with learning disabilities: The Learning Disability Quarterly, 28, 59–74.
promise of constant time delay procedures. Learning Martens, B. K., & Daly, E. J. (1999). Discovering the al-
Disabilities Research and Practice, 8(4), 260–263. phabetic principle: A lost opportunity for education-
Kozloff, M. A., LaNuiziata, L., Cowardin, J., & Bessel- al reform. Journal of Educational Psychology, 9, 35–43.
lieu, F. B. (2000). Direct instruction: Its contribu- Matson, J. L., Sevin, J. A., Fridley, D., & Love, S. R.
tions to high school achievement. High School Jour- (1990). Increasing spontaneous language in three au-
nal, 84, 54–77. tistic children. Journal of Applied Behavior Analysis,
Krantz, P. J., & McClannahan, L. E. (1993). Teaching 23, 227–233.
children with autism to initiate to peers: Effects of McClannahan, L. E., & Krantz, P. J. (2005). Teaching
a script-fading procedure. Journal of Applied Behavior conversation to children with autism: Scripts and script
Analysis, 26, 121–132. fading. Bethesda, MD: Woodbine House.
Kroesbergen, E. H., Van Luit, J. E. H., & Maas, C. J. M. Miltenberger, R. G. (2001). Behavior modification: Prin-
(2004). Effectiveness of explicit and constructivist ciples and procedures (2nd ed.). Belmont, CA: Wad-
mathematics instruction for low-achieving students sworth/Thomson Learning.
268 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
Morse, T. E., & Schuster, J. W. (2004). Simultaneous Rea, J., & Williams, D. (2002). Shaping exhale dura-
prompting: A review of the literature. Education and tions for breath CO detection for men with mild
Training in Developmental Disabilities, 39(2), 153–168. mental retardation. Journal of Applied Behavior Anal-
Napolitano, D. A., Smith, T., Zarcone, J. R., Goodkin, ysis, 35, 415–418.
K., & McAdam, D. B. (2010). Increasing response Reagon, K. A., & Higbee, T. S. (2009). Parent-imple-
diversity in children with autism. Journal of Applied mented script fading to promote play-based verbal
Behavior Analysis, 43, 265–271. initiations in children with autism. Journal of Applied
National Reading Panel. (2000). Report of the National Behavior Analysis, 42, 659–664.
Reading Panel. Teaching children to read: An evidence- Reid, R., Lienemann, T. O., & Hagaman, J. L. (2013).
based assessment of the scientific research literature Strategy instruction for students with learning disabili-
on reading and its implications for reading instruction ties (2nd ed.). New York: Guilford Press.
(NIH Publication No. 00-4769). Washington, DC: Reitzel, J., Summers, J., Lorv, B., Szatmari, P., Zwaigen-
National Institute of Child Health and Human De- baum, L., Georgiades, S., et al. (2013). Pilot random-
velopment. ized controlled trial of a functional behavior skills
Nguyen, K.-V., Binder, K. S., Nemier, C., & Ardoin, S. training program for young children with autism
P. (2014). Gotcha!: Catching kids during mindless spectrum disorder who have significant early learn-
reading. Scientific Studies of Reading, 18(4), 274–290. ing skill impairments and their families. Research in
Noell, G. H., Connell, J. M., & Duhon, G. J. (2006). Autism Spectrum Disorders, 7(11), 1418–1432.
Spontaneous response generalization during whole Richardson, V. (2003). Constructivist pedagogy. Teach-
word instruction: Reading to spell and spelling to ers College Record, 105(9), 1623–1640.
read. Journal of Behavioral Education, 15(3), 121–130. Rosales-Ruiz, J., & Baer, D. M. (1997). Behavioral cusps:
Noell, G. H., Freeland, J. T., Witt, J. C., & Gansle, K. A developmental and pragmatic concept for behav-
A. (2001). Using brief assessments to identify effec- ior analysis. Journal of Applied Behavior Analysis, 30,
tive interventions for individual students. Journal of 533–544.
School Psychology, 39(4), 335–355. Sauttfr, R. A., LeBlanc, L. A., Jay, A. A., Goldsmith, T.
Noell, G. H., Gansle, K. A., Witt, J. C., Whitmarsh, E. R., & Carr, J. E. (2011). The role of problem solving
L., Freeland, J. T., LaFleur, L. H., et al. (1998). Effects in complex intraverbal repertoires. Journal of Applied
of contingent reward and instruction on oral reading Behavior Analysis, 44, 227–244.
performance at differing levels of passage difficulty. Schleien, S. J., Wehman, P., & Kiernan, J. (1981). Teach-
Journal of Applied Behavior Analysis, 31, 659–664. ing leisure skills to severely handicapped adults: An
O’Neill, S. J., McDowell, C., & Leslie, J. (2018). A com- age-appropriate darts game. Journal of Applied Behav-
parison of prompt delays with trial-and-error instruc- ior Analysis, 14, 513–519.
tion in conditional discrimination training. Behavior Schoen, S. F. (1986). Assistance procedures to facilitate
Analysis in Practice, 11, 370–380. the transfer of stimulus control: Review and analy-
Oppenheim-Leaf, M. L., Leaf, J. B., & Call, N. A. sis. Education and Training of the Mentally Retarded,
(2012). Teaching board games to two children with 21(1), 62–74.
autism spectrum disorder. Journal of Developmental Schuster, J. W., Griffin, A. K., & Wolery, M. (1992).
and Physical Disabilities, 24(4), 347–358. Comparisons of simultaneous prompting and con-
Page, S., & Neuringer, A. (1985). Variability is an oper- stant time delay procedures in teaching sight words to
ant. Journal of Experimental Psychology: Animal Be- elementary students with moderate mental retarda-
havior Processes, 11(3), 429–452. tion. Journal of Behavioral Education, 2(3), 302–325.
Parrott, K. A., Schuster, J. W., Collins, B. C., & Gas- Scott, D., Scott, L. M., & Goldwater, B. (1997). A per-
saway, L. J. (2000). Simultaneous prompting and formance improvement program for an international-
instructive feedback when teaching chained tasks. level track and field athlete. Journal of Applied Behav-
Journal of Behavioral Education, 10(1), 3–19. ior Analysis, 30, 573–575.
Pierce, K. L., & Schreibman, L. (1994). Teaching daily Silbaugh, B. C., Falcomata, T. S., & Ferguson, R. H.
living skills to children with autism in unsupervised (2018). Effects of a lag schedule of reinforcement
settings through pictorial self-management. Journal with progressive time delay on topographical mand
of Applied Behavior Analysis, 27, 471–481. variability in children with autism. Developmental
Plavnick, J. B., Ferreri, S. J., & Maupin, A. N. (2010). Neurorehabilitation, 21, 166–177.
The effects of self-monitoring on the procedural in- Silla-Zaleski, V. A., & Vesloski, M. J. (2010). Using
tegrity of a behavioral intervention for young chil- DRO, behavioral momentum, and self-regulation to
dren with developmental disabilities. Journal of Ap- reduce scripting by an adolescent with autism. Jour-
plied Behavior Analysis, 43, 315–320. nal of Speech and Language Pathology–Applied Behav-
Pressley, M., & Levin, J. R. (1987). Elaborative learn- ior Analysis, 5(1), 80–87.
ing strategies for the inefficient learner. In S. J. Ceci Skinner, B. F. (1938). The behavior of organisms: An exper-
(Ed.), Handbook of cognitive, social, and neuropsycho- imental analysis. New York: Appleton-Century-Crofts.
logical aspects of learning disabilities (pp. 175–212). Slocum, S. K., & Tiger, J. H. (2011). An assessment of
Hillsdale, NJ: Erlbaum. the efficiency of and child preference for forward
Building Complex Repertoires from Discrete Behaviors 269
and backward chaining. Journal of Applied Behavior Trapman, M., van Gelderen, A., van Steensel, R., van
Analysis, 44, 793–805. Schooten, E., & Hulstijn, J. (2014). Linguistic knowl-
Smith, G. J. (1999). Teaching a long sequence of a be- edge, fluency and meta-cognitive knowledge as com-
havior using whole task training, forward chaining, ponents of reading comprehension in adolescent low
and backward chaining. Perceptual and Motor Skills, achievers: Differences between monolinguals and
89, 951–965. bilinguals. Journal of Research in Reading, 37, S3–S21.
Snell, M. E., & Gast, D. L. (1981). Applying time delay Troia, G. A. (2002). Teaching writing strategies to chil-
procedure to the instruction of the severely handi- dren with disabilities: Setting generalization as the
capped. Journal of the Association for the Severely goal. Exceptionality, 10, 249–269.
Handicapped, 6(3), 3–14. Valentino, A. L., Shillingsburg, M. A., Call, N. A., Bur-
Snow, C. E., Burns, M. S., & Griffin, P. (1998). Prevent- ton, B., & Bowen, C. N. (2011). An investigation of
ing reading difficulties in young children. Washington, extinction-induced vocalizations. Behavior Modifica-
DC: National Academy Press. tion, 35(3), 284–298.
Spooner, F., & Spooner, D. (1984). A review of chain- Wagner, R. K., Puranik, C. S., Foorman, B., Foster, E.,
ing techniques: Implications for future research and Wilson, L. G., Tschinkel, E., et al. (2011). Modeling
practice. Education and Training of the Mentally Re- the development of written language. Reading and
tarded, 19, 114–124. Writing, 24, 203–220.
Spooner, F., Spooner, D., & Ulicny, G. (1986). Com- Walker, B., Shippen, M. E., Alberto, P., Houchins, D. E.,
parisons of modified backward chaining: Backward & Cihak, D. F. (2005). Using the expressive writing
chaining with leap-aheads and reverse chaining with program to improve the writing skills of high school
leap-aheads. Education and Treatment of Children, students with learning disabilities. Learning Disabili-
9(2), 122–134. ties Research and Practice, 20, 175–183.
Spooner, F., Weber, L. H., & Spooner, D. (1983). The Watkins, C. L., & Slocum, T. A. (2004). The com-
effects of backward chaining and total task presenta- ponents of direct instruction. In N. E. Marchand-
tion on the acquisition of complex tasks by severely Martella, T. A. Slocum, & R. C. Martella (Eds.),
retarded adolescents and adults. Education and Treat- Introduction to direct instruction (pp. 28–65). Boston:
ment of Children, 6, 401–420. Allyn & Bacon.
Sternberg, L., & Adams, G. L. (1982). Educating severely Watters, J. K. (1992). Retention of human sequenced be-
and profoundly handicapped students. Rockville MD: havior following forward chaining, backward chain-
Aspen Systems. ing, and whole task training procedures. Journal of
Stevens, K. B., Blackhurst, A. E., & Slaton, D. B. (1991). Human Movement Studies, 22(3), 117–129.
Teaching memorized spelling with a microcomputer: Wichnick-Gillis, A. M., Vener, S. M., & Poulson, C. L.
Time delay and computer-assisted instruction. Jour- (2019). Script fading for children with autism: Gen-
nal of Applied Behavior Analysis, 24, 153–160. eralization of social initiation skills from school to
Stokes, T. F., & Baer, D. M. (1977). An implicit tech- home. Journal of Applied Behavior Analysis, 52, 451–
nology of generalization. Journal of Applied Behavior 466.
Analysis, 10, 349–367. Wightman, D. C., & Sistrunk, F. (1987). Part-task train-
Sullivan, W. E., Saini, V., DeRosa, N. M., Craig, A. R., ing strategies in simulated carrier landing final-ap-
Ringdahl, J. E., & Roane, H. S. (2020). Measurement proach training. Human Factors, 29(3), 245–254.
of nontargeted problem behavior during investiga- Williams, G., Donley, C. R., & Keller, J. W. (2000).
tions of resurgence. Journal of Applied Behavior Anal- Teaching children with autism to ask questions about
ysis, 53, 249–264. hidden objects. Journal of Applied Behavior Analysis,
Swain, R., Lane, J. D., & Gast, D. L. (2015). Comparison 33, 627–630.
of constant time delay and simultaneous prompting Wiskow, K. M., Matter, A. L., & Donaldson, J. M. (2018).
procedures: Teaching functional sight words to stu- An evaluation of log schedules and prompting meth-
dents with intellectual disabilities and autism spec- ods to increase variability of naming category items
trum disorder. Journal of Behavioral Education, 24(2), in children with autism spectrum disorder. Analysis
210–229. of Verbal Behavior, 34, 100–123.
Taylor, B. A., Hughes, C. E., Richard, E., Hoch, H., & Wolery, M., & Gast, D. L. (1984). Effective and efficient
Coello, A. R. (2004). Teaching teenagers with autism procedures for the transfer of stimulus control. Topics
to seek assistance when lost. Journal of Applied Be- in Early Childhood Special Education, 4(3), 52–77.
havior Analysis, 37, 79–82. Wolery, M., Holcombe, A., Cybriwsky, C., Doyle, P. M.,
Tekin-Iftar, E., Acar, G., & Kurt, O. (2003). The effects Schuster, J. W., Ault, M. J., et al. (1992). Constant
of simultaneous prompting on teaching expressive time delay with discrete responses: A review of effec-
identification of objects: An instructive feedback tiveness and demographic, procedural, and method-
study. International Journal of Disability, Development, ological parameters. Research in Developmental Dis-
and Education, 50(2), 149–167. abilities, 13(3), 239–266.
CHAP TER 16
Lori is eating popcorn in a classroom while she teaching communication and language to those
prepares a lesson for her preschool students, who with limited or no verbal repertoire.
are on the playground. Lily walks into the class-
room, approaches Lori, and says, “Popcorn!” Lori
smiles and gives Lily popcorn. Jack walks into the SKINNER’S ANALYSIS OF VERBAL BEHAVIOR
classroom and sees the popcorn. He grabs for the
bag without saying anything. Lori looks at him In his seminal book Verbal Behavior, Skinner (1957)
and says, “What do you want?” Jack says, “Pop- showed how we can analyze language by using the
corn!” Lori smiles and gives Jack popcorn. Char principles of operant and respondent condition-
walks into the classroom, sees the popcorn, and ing. Skinner chose the term verbal behavior rather
grabs for the bag while remaining silent. Lori says, than speech, because he did not want to restrict his
“What do you want?”, but Char continues to grab analysis to vocal behavior. Additionally, tradition-
for the bag while remaining silent. Lori then says, alists had already used the term language to refer to
“Say popcorn.” Char immediately says, “Popcorn!” an ability or system in the organism responsible for
Lori smiles and gives Char popcorn. generating speech. The source of control for lan-
If Lori were to complete a checklist based on guage or verbal behavior, respectively, was a major
these interactions that included the question “Can difference between the traditional approach and
each child say popcorn?”, the correct answer would Skinner’s approach. According to the traditional-
be yes. However, if we were to ask Lori, “Does each ists, control for language originates within the or-
child do the same thing?”, the correct answer ganism. According to Skinner, by contrast, control
would be no. Although the form of the response for verbal behavior originates from contingencies of
popcorn is the same, each example involves a dif- reinforcement acting upon the organism’s behavior.
ferent controlling relation. The difference in per- Skinner’s terminology emphasized that verbal be-
spective between the first and second questions is havior is like any other operant behavior. The term
at the heart of a functional analysis of verbal be- verbal behavior, which at that time was “relatively
havior. Although knowing the form of verbal be- unfamiliar in traditional modes of explanation”
havior is useful, understanding its functional con- (Skinner, 1957, p. 1), set the stage for a departure
trol is more important. This chapter reviews how a from traditional explanations of language.
functional analysis of verbal behavior can provide Skinner (1957) defined verbal behavior as “be-
guidance about the complex issues associated with havior reinforced through the mediation of other
270
Teaching Verbal Behavior 271
persons” (p. 2). In other words, a speaker’s behavior We contrast the behavioral approach Skinner
is reinforced through or mediated by the behav- (1957) advocated with the traditional approach to
ior of a listener.1 The topography of the speaker’s language development and intervention of Chom-
behavior, such as vocal, gestural, or visual, is ir- sky (1965), Brown (1973), and Piaget (1951), who
relevant within this definition. Skinner stated his conceptualized language by the form or topogra-
unique orientation to verbal behavior explicitly: phy of the learner’s verbal repertoire (i.e., vocabu-
“In defining verbal behavior as behavior rein- lary, grammar, syntax), with little regard for func-
forced through the mediation of other persons, we tion. From the traditional perspective, language
do not, and cannot, specify any one form, mode, development is the function of hypothesized in-
or medium” (p. 14). Sometimes our behavior influ- nate developmental, neurological, and cognitive
ences the environment in direct ways. For exam- structures. We use the term language deficit or delay
ple, a little boy ties his own shoelaces and can run when a learner’s verbal repertoire is deficient in
around immediately. At other times, our behavior comparison to the verbal repertoire of a same-age,
influences other people whose actions lead to re- typically developing learner. Proponents of a tra-
inforcement. For example, the boy asks his brother ditional-language approach often relate language
to tie his shoes and runs around after his brother deficits to genetic or neurological abnormalities
does so. Although the first example does not meet (Lord, Cook, Leventhal, & Amaral, 2000). Inter-
the definition of verbal behavior, the second one vention tends to focus on the acquisition of forms,
does. from sounds to words to larger structures, with less
The verbal community selects specific forms attention to the behavioral function of such forms
of verbal behavior to function in certain ways. (American Speech–Language–Hearing Associa-
The behavior forms may be vocal, such as spo- tion, 2016).
ken words; graphic lines, such as writing; or hand By contrast, the behavioral approach focuses on
postures and movements, such as sign language. contingency or functional analysis of language or
These behavior forms produce an effect on the verbal behavior. Specific environment–behavior
environment. Skinner added an important refine- relations or contingencies of reinforcement are
ment to the definition of verbal behavior when he responsible for language development, according
wrote, “The listener must be responding in ways to the behavioral approach. Although Skinner
that have been conditioned precisely to reinforce the (1957) did not review language deficits extensively
behavior of the speaker” (p. 225, emphasis in origi- in his book, researchers have applied his analysis
nal). An individual learns listening behavior as a to teaching language to children and adults with
member of a verbal community. Thus the listener language difficulties (Barbera & Rasmussen, 2007;
learns to react to the speaker’s verbal behavior, Carr & Miguel, 2013; Frost & Bondy, 2002; LaF-
and this is a requirement for behavior to be ver- rance & Miguel, 2014; Sundberg, 2008; Sundberg
bal. In the example above, when the little boy asks & Partington, 1998). However, some behaviorally
his brother, “Will you tie my shoes?”, the auditory oriented language training programs (e.g., Lovaas,
product of this behavior, the sound pattern, serves 2003; Maurice, Green, & Luce, 1996) have not
as a stimulus that evokes his brother’s shoe tying. used Skinner’s analysis as a framework for teach-
But notice that the brother can react appropriately ing verbal behavior. In our view, Skinner’s analy-
as a listener by performing the task only if he has sis is essential to developing successful programs
learned how to respond to this request previously. for training verbal behavior. In the sections that
In lay terms, the listener must understand the follow, we illustrate the basic tenets of Skinner’s
speaker.2 approach to verbal behavior and illustrate how we
may apply it in specific training protocols.
1 Even though we commonly use the terms speaker and lis-
tener to refer to vocal behavior (i.e., speech), in Skinner’s PRIMARY VERBAL OPERANTS
terms, the speaker is the one behaving (e.g., talking, sign-
ing, using pictures), and the listener is the one being af-
fected by the response products of the speaker’s behavior. Skinner (1957) identified and named six types of
2 During
functional relations between controlling variables
any given verbal episode, we behave as both speak-
ers and listeners, as when we make a request or fulfill one,
and verbal responses. These consist of the mand,
respectively. Additionally, we can react as listeners to our tact, intraverbal, textual, echoic, and audience rela-
own verbal behavior, in that we can understand and react tions. He named two more in the section on tran-
to the things we say (Miguel, 2016). scription: copying a text and taking dictation. Skin-
272 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
ner referred to these relations as verbal operants, tacts such as vehicle, red, and fast. Environmental
because he classified them by the antecedents and stimuli are likely to control many verbal responses,
consequences that control their form (i.e., operant and we discuss this issue more fully below.
behavior). Skinner (1957) also illustrated that tacts can
A mand is a verbal operant that a characteristic occur in the presence of novel objects or events
consequence reinforces, and the relevant condi- to which the speaker has not been exposed previ-
tions of deprivation or aversive stimulation control ously if the novel object shares physical properties
the response. Deprivation or aversive stimulation with the original stimulus. For example, a child
are motivating operations, or events that alter the may give the verbal response car in the presence of
value of a reinforcer. Establishing operations in- a novel exemplar because it shares common physi-
crease the value of a stimulus as a reinforcer. Abol- cal properties with the stimulus that was present
ishing operations decrease the value of a stimulus as when the child learned to say, sign, or write car. In
a reinforcer3 (Michael & Miguel, 2020; Michael, his book, Skinner referred to this type of stimulus
1993). For example, behavior that produces water, generalization as generic tact extension. Other types
such as touching a communication card that sig- of extensions include metaphorical and metonymic
nals the therapist to deliver water, is likely to in- extensions. A metaphoric extension occurs when
crease after consumption of salty snacks, which the novel stimulus shares some, but not all, charac-
is the motivating operation in this example. The teristics with the original stimulus. The word surf-
learner is likely to emit behavior that has produced ing when referring to the Internet is an example.
water in the past under these conditions. Newborn A metonymical extension occurs when the novel
babies and crying provide another example. Al- stimulus does not share any physical similarity
though newborns cry reflexively, they learn to cry with the original one. Rather, they just happened
when hungry when crying produces food. In this to appear together during the acquisition of the
example, food deprivation is the motivating opera- tact. For example, the sentence The White House
tion, which will control crying if crying produces released a statement is equivalent to The President
food. These examples show that mands develop released a statement, because the President and the
when specific response forms, such as touching a White House usually appear together.
communication card or crying, produce specific Skinner (1957) identified other verbal operants
consequences, such as water or food, respectively. whose forms, such as what a learner says or writes,
Mands are unique among verbal operants, be- are all evoked by verbal discriminative stimuli (i.e.,
cause mands are controlled by relevant motiving the products of someone else’s verbal behav-
operations. By contrast, discriminative stimuli ior). The echoic relation is one of these operants
control other verbal operants, such as tacts and in which the speaker’s behavior is controlled by
intraverbals. Another difference between mands the auditory stimulus arising from someone else’s
and other verbal operants is that specific stimulus vocal behavior. In the case of the echoic, the re-
forms reinforce mands. By contrast, nonspecific, sponse bears formal similarity to the stimulus. For
generalized stimuli reinforce other verbal operants. instance, a girl says, “Ball,” after her teacher says,
The tact is a verbal operant in which a response “Ball,” when there is no ball in view. Copying a
of a given form, such as vocal, sign, or writing, is text is like the echoic in that the response form
controlled by a nonverbal stimulus or “a particular also bears formal similarity with the stimulus, and
object or event or property of an object or event” reinforcement may also depend on close corre-
(Skinner, 1957, p. 82). The presence of a car, for spondence between the stimulus and response. For
example, increases the likelihood of the learner’s instance, a girl writes the word ball after seeing the
emitting the vocal, signed, or written response printed word ball. Taking dictation is writing what
car. The object evokes the response because the someone says. Writing ball when hearing some-
English-speaking verbal community reinforced one say, “Ball,” is an example. There is still cor-
this specific verbalization in the presence of this respondence between the stimulus and response,
object. An object may serve as a discriminative because specific sound patterns control specific
stimulus for various response forms. A toy car can hand movements. There is no physical similarity
evoke not only the response car, but also other between them, however, because the stimulus is
auditory, and the response produces a printed word
3 Motivating operations may also modulate the effective- or visual stimulus. Like taking dictation, the textual
ness of consequences as punishers. relation consists of reading printed words.
Teaching Verbal Behavior 273
The intraverbal relation is a response form tion from inattentive listeners. In applied settings,
evoked by a verbal discriminative stimulus. The increasing the consistency of the communicative
stimulus and response are different words and partner’s attention and responses to the speaker’s
do not resemble each other. For example, a par- communicative attempts is an important inter-
ent asks, “What day is it today?”, and the child vention goal (Goldstein, Kaczmarek, Pennington,
responds, “Thursday.” Or the teacher says, “Two, & Shafer, 1992). This is necessary to support the
four, six . . . ,” and the student responds, “Eight.” efforts of the speaker in his or her communicative
In these cases, there are no parts or subdivisions attempts.
of the stimulus controlling parts or subdivisions of Importantly, effective communicators can act as
the response. both speakers and listeners, which allows them to
Even though all verbal operants except the understand (i.e., react to) their own verbal behav-
mand are maintained by generalized conditioned ior. Unfortunately, young children or those with
reinforcement, Skinner suggested that “the ac- developmental disabilities may not speak readily
tion that a listener takes with respect to a verbal after learning to react to words as listeners, or vice
response is often more important to the speaker versa (Petursdottir & Carr, 2011). This suggests
than generalized reinforcement” (p. 151). In other that when attempting to establish functional ver-
words, verbal behavior is sensitive to listener be- bal skills, we must teach speaker and listener be-
havior, such as a listener’s nodding as a speaker havior simultaneously (e.g., Fiorile & Greer, 2007).
provides directions to a restaurant. This observa-
tion suggests that if we want to establish an effec-
tive speaker repertoire, we must establish listener MULTIPLE CONTROL
reactions as effective forms of reinforcement (e.g.,
Maffei, Singer-Dudek, & Keohane, 2014). The verbal operants discussed thus far involve
control by a single variable, such as a motivat-
ing operation or a verbal or nonverbal stimulus.
THE LISTENER However, most verbal behavior involves different
topographies under control of multiple variables
Emphasizing the role of the listener in these prima- (Michael, Palmer, & Sundberg, 2011). Multiple
ry verbal relations, including the sources of control control occurs when a single response is controlled
for the listener’s behavior, is important. According by more than one variable, or a single variable
to Skinner (1957), an analysis of both speaker and controls more than one response (Bondy, Tincani,
listener behavior is necessary to understand the & Frost, 2004; Skinner, 1957). The first type of
“total verbal episode” (p. 36). The mand primarily multiple control, convergent control, occurs when
benefits the speaker, such as when a girl requests the verbal community arranges reinforcement
and receives water. By contrast, other primary ver- for a response form in the presence of more than
bal operants, such as the tact, largely benefit the one stimulus. For example, a girl receives rein-
listener. For example, if a speaker says, “It’s going forcement for saying, “Ball,” in response to the
to rain,” in response to a dark sky, the listener printed word ball, a picture of a ball, and the ques-
may contact reinforcement that he or she would tion “What do you throw?” The second type of
not have contacted otherwise. The stimulus may multiple control, divergent control, occurs when a
evoke listener behavior such as carrying an um- given variable strengthens multiple responses. An
brella or delaying a walk outside. The reinforcer example is when liquid deprivation strengthens
for the listener is avoidance of an aversive stimu- verbal responses such as Water please, May I have a
lus: getting wet. Thus competent speakers extend drink?, and I’m thirsty.
a listener’s contact with the stimulating environ- Multiple control that produces impure verbal
ment, and this effect on the listener may serve as operants, (i.e., those that more than one variable
an important yet subtle source of reinforcement for strengthens simultaneously) may be present when
the speaker’s verbal behavior. teaching verbal responses (Skinner, 1957, p. 151).
Additionally, effective programs for train- For example, a boy who is liquid-deprived is more
ing language must explicitly teach the speaker likely to say, “Juice, please,” if we present him with
to engage the listener. For example, after begin- a cup of juice than when deprivation or the cup
ning speakers learn to make requests to attentive of juice is presented alone. The motivating opera-
listeners, they may need to learn to recruit atten- tion and nonverbal stimulus have a combined ef-
274 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
fect, producing an impure verbal operant, which is the verbal stimuli eat and breakfast. Although the
the mand–tact. Several variables may combine to stimulus eat may evoke several responses (diver-
produce a response. When a mother asks, “Would gent control), including, pasta, pizza, toast, and
you like some juice?” with a cup of juice present, broccoli, the addition of the stimulus breakfast will
and her daughter says, “Yes, please,” we may call serve to strengthen the response toast (convergent
the girl’s response a mand–tact–intraverbal if it control). In the absence of specific instructions
is multiply controlled by a motivating operation to distinguish between both features of this com-
(deprivation), nonverbal stimulus (cup of juice), pound verbal stimulus, eat and breakfast, a learner
and verbal stimulus (mother’s question). In this may attend to only one component (i.e., eat) and
manner, we can identify the potential sources of continue to respond with toast to any question that
control and establish multiply controlled operants. includes the stimulus eat (e.g., Q: “What do rab-
The relation among the different verbal oper- bits eat?” A: toast). Behavior analysts can prevent
ants is important in teaching verbal skills. For ex- this kind of failure, or rote responding, by teaching
ample, does acquisition of a response form under learners to respond to both features of the complex
one set of variables lead to emission of the same verbal stimulus at the onset of instruction (Axe,
response form under a different set of variables? 2008; DeSouza, Fisher, & Rodriguez, 2019; Sund-
There is considerable evidence that verbal oper- berg & Sundberg, 2011)
ants are functionally independent under certain The determination of controlling variables
conditions (Kelley, Shillingsburg, Castro, Addi- helps direct teaching protocols for several relevant
son, & LaRue, 2007; LaFrance, Wilder, Normand, situations. For example, a seemingly simple verbal
& Squires, 2009; LaMarre & Holland, 1985; Pe- skill, such as Martin’s learning to say, “Swing,” may
tursdottir, Carr, & Michael, 2005; Twyman, 1996). have many possible sources of control:
That is, a response topography (i.e., word) learned
with one set of controlling variables will not nec- 1. Martin learns an echoic. He says, “Swing,”
essarily occur in the presence of different variables when the teacher says, “Swing,” and the teach-
unless it is explicitly taught. A response topogra- er provides praise.
phy taught as a mand, therefore, will not occur 2. Martin learns an echoic–mand. When his
automatically as a tact, or vice versa. For example, teacher says, “Swing,” he says, “Swing,” and his
a boy who is taught to say, sign, or write tree in re- teacher puts him on the swing.
sponse to the question “What is it?” and a picture 3. Martin learns a mand. Martin says, “Swing,”
of a tree may not reply with tree in the presence of to his teacher, and his teacher puts him on the
the picture of the tree by itself (tact), because the swing.
learned response involves a different controlling 4. Martin learns a mand–tact. Martin sees a
relation—the picture of the tree plus the question. swing and says, “Swing,” to his teacher. His
The analysis of multiply controlled verbal oper- teacher puts him on the swing.
ants becomes particularly relevant in the design of 5. Martin learns an intraverbal–mand. His teach-
communication training programs. Even though er asks, “What do you want to do?”, without a
many behaviorally oriented language training pro- swing in sight. Martin says, “Swing,” and his
grams seek to establish complex verbal operants teacher puts him on the swing.
(e.g., Leaf & McEachin, 1999; Lovaas, 2003; Mau- 6. Martin learns a pure tact. Martin sees a swing,
rice et al., 1996), they do not describe procedures and he says, “Swing.” The teacher says, “Yes, I
to transfer control from the question plus object to see it too,” but does not put him on the swing.
the object itself. Without explicit procedures for 7. Martin learns an intraverbal–tact. When his
transfer of stimulus control, the learner is likely to teacher points to a swing and asks, “What is
develop a highly selective repertoire in which tacts that?”, Martin says, “Swing.” He teacher pro-
will occur only in the presence of objects accom- vides praise, but does not put him on the swing.
panied by questions. By contrast, a more function- 8. Martin learns an intraverbal–mand–tact. Mar-
al tacting repertoire involves response topogra- tin sees a swing, and his teacher asks, “What
phies that occur in several stimulus combinations, do you want to do?” Martin says, “Swing,” and
including presentation of the object alone. his teacher puts him on the swing.
Intraverbals are usually under the control of
multiple verbal stimuli. For example, the response Martin’s teacher must determine which variables
toast in the presence of the question “What do you are relevant and explicitly arrange each variation
eat for breakfast?” must be under control of both to teach Martin to say, “Swing.”
Teaching Verbal Behavior 275
tation skills. Aided systems include picture-based Progressive time-delay prompting or errorless teach-
systems, such as PECS (Frost & Bondy, 2002), ing (Karsten & Carr, 2009; Touchette & Howard,
and speech-generating devices (Lorah, Parnell, 1984) is a variation of the time-delay procedure, in
Whitby, & Hantula, 2015). Relevant factors for se- which the teacher presents the cue simultaneously
lecting an aided system may include (1) the avail- with the prompt and gradually increases the dura-
ability of a device or system; (2) the ease of use for tion between the cue and prompt as the learner
the primary listeners, such as parents, siblings, or makes independent responses. Such time-delay
teachers; (3) the potential for the device or system procedures use establishing operations (Michael,
to accommodate several communication symbols 1993) to promote mand acquisition. The interrupt-
and vocabulary; and (4) the capability of the de- ed-chain procedure (Hall & Sundberg, 1997) also
vice or system to produce a repertoire of indepen- uses establishing operations, in which the teacher
dent, functional verbal behavior for the listener prevents the learner from completing a behavior
(Tincani, 2007; Tincani & Boutot, 2005). chain until the learner performs an appropriate
mand. For example, a teacher might hide a step
stool from a learner who typically uses the stool
TEACHING THE MAND to access a game on a shelf. The teacher would use
this situation to teach the learner to request help
The mand directly benefits the speaker; thus as- in locating the missing stool. A common feature
sessment of the people, items, and events that of these procedures is a systematic manipulation
function as reinforcers for the learner is critical of the learner’s environment to promote a func-
before teaching the mand. Researchers have devel- tional manding repertoire (Carnett et al., 2017; In-
oped systematic strategies to assess learners’ prefer- gvarsson & Hollobaugh, 2013; Landa et al., 2017;
ences for reinforcers, though a detailed description Lechago, Carr, Grow, Love, & Almason, 2010;
of these is beyond the scope of this chapter (see Shillingsburg, Bowen, Valentino, & Pierce, 2014;
Saini, Retzlaff, Roane, & Piazza, Chapter 10, this Shillingsburg et al., 2019).
volume). Extensive research has focused on teach- Three basic response forms that beginning com-
ing the mand as a component of functional-com- municators need to learn are asking for a break,
munication training (Durand & Merges, 2001; asking for help, and saying (or otherwise commu-
Greer, Fisher, Saini, Owen, & Jones, 2016; Mancil, nicating) no to an offered item or activity. These
2006; Saini, Miller, & Fisher, 2016). The primary simple skills are important, because they allow
purpose of functional-communication training the learner to exert control over his or her envi-
is to reduce challenging behavior by teaching al- ronment and bring the learner into contact with
ternative responses (i.e., mands) that produce the contingencies that will be critical for developing
same reinforcing consequences as challenging be- other skills (Hixson, 2004). Asking for a break or
havior (see Fisher, Greer, & Bouxsein, Chapter 20, assistance and communicating no are under func-
this volume). Other researchers have evaluated tional control of aversive events such as demands,
procedures to establish independent or spontane- uncompleted or difficult tasks, or unwanted items.
ous mand repertoires for learners who do not en- Communicating no allows the learner to escape or
gage in challenging behavior. Several studies have avoid an unwanted item or activity. It functions
established the efficacy of time-delay prompt- as a qualifying autoclitic in a mand function (see
ing procedures for teaching mand repertoires to Skinner, 1957, p. 322). When a learner requests an
learners with developmental disabilities (e.g., alternative item or activity in the context of an
Halle, Baer, & Spradlin, 1981; Halle, Marshall, escape or avoidance situation, the request can also
& Spradlin, 1979; Kratzer, Spooner, & Test, 1993; function as a form of rejection maintained in part
Landa, Hansen, & Shillingsburg, 2017; Shillings- by negative reinforcement. Such choice-making
burg, Marya, Bartlett, & Thompson, 2019). For responses may have the collateral effect of reduc-
instance, Halle et al. (1981) used a 5-second time- ing challenging behavior associated with escape or
delay prompting procedure to teach learners with avoidance contingencies.
intellectual developmental disorder and language Researchers have validated several strategies for
delays to mand in a naturalistic setting. The teach- teaching requesting and rejecting behaviors. Best
er arranged a cue likely to evoke a mand, such as practice integrates instruction of communicative
approaching the learner with a cup of juice. The responses into daily routines by systematically iden-
teacher waited 5 seconds for the learner to perform tifying all potential opportunities for a learner to
the mand before delivering a prompt. engage in target behaviors and embedding instruc-
Teaching Verbal Behavior 277
tion into naturally occurring events (Carnett et 1994). Fourth, the behavior analyst must teach the
al., 2017; Sigafoos, Kerr, Roberts, & Couzens, 1994; learner that escape or avoidance will not always
Ylvisaker & Feeney, 1994). Strategies for creating be possible even when the learner has requested
opportunities for verbal behavior focus on teach- it appropriately (Sigafoos, 1998). Even though we
ing communication in real-life activities. These are suggesting that we should initially teach com-
strategies include (1) delaying access to an item or municating no as a mand, more advanced learn-
activity that is present until the learner makes a re- ers will benefit from acquiring no and yes as tacts
quest (Halle et al., 1981); (2) withholding an item and intraverbals. For example, we should teach a
necessary to complete a preferred activity (Cipani, learner who can speak to say, “No,” if the teacher
1988; Lechago et al., 2010); (3) blocking a response asks, “Is this a giraffe?” when she presents a picture
or interrupting an activity to create the need for of a cow (tact), and to say, “Yes,” when the teacher
a request (Carnett, Bravo, & Waddington, 2019; asks, “Does a cow says moo?” (intraverbal; see
Shafer, 1995; Sigafoos et al., 1994); (4) provid- Shillingsburg, Kelley, Roane, Kisamore, & Brown,
ing only part of what the learner has requested to 2009).
create a new need (Duker, Kraaykamp, & Visser,
1994); (5) intentionally giving the learner the in-
correct item (Choi, O’Reilly, Sigafoos, & Lancioni, TEACHING OTHER VERBAL OPERANTS
2010; Sigafoos et al., 1994); or (6) delaying offers
of assistance until the learner makes a request for Although much research has focused on the ac-
help (Rodriguez, Levesque, Cohrs, & Niemeier, quisition of mands, research also has shown that
2017; Sigafoos & Roberts-Pennell, 1999). the tact repertoire is foundational for the devel-
The learner escapes or avoids nonpreferred opment of other verbal and nonverbal behaviors,
items or activities in a socially acceptable man- such as mands, intraverbals, stimulus categoriza-
ner if he or she can make a rejection response. tion, and analogical reasoning (e.g., Finn, Miguel,
Learners who do not communicate this need may & Ahearn, 2012; Greer & Du, 2010; Miguel et
adopt an idiosyncratic behavior that is difficult to al., 2015; Miguel & Kobari-Wright, 2013; Miguel,
interpret (Iacono, Carter, & Hook, 1998), or may Petursdottir, & Carr, 2005; Sprinkle & Miguel,
learn to escape or avoid activities through chal- 2012). Additionally, tacts may not readily emerge
lenging behavior (Carr, 1994). The learner is more after receptive-discrimination instruction (Con-
likely to acquire socially appropriate escape or treras, Cooper, & Kahng, 2020; Petursdottir &
avoidance communication in situations where he Carr, 2011), especially if a child lacks a general-
or she is motivated to escape or avoid an item or ized echoic repertoire (Horne & Lowe, 1996). For
activity. The behavior analyst can create a context this reason, a behavior analyst should prioritize
for teaching the learner to reject an offer appropri- tacts when attempting to expand a learner’s vocal
ately if the analyst can identify items or activities repertoire (e.g., Greer & Du, 2010). Tact training
the learner is motivated to escape or avoid (Che- should focus initially on preferred, familiar, and
zan, Drasgow, Martin, & Halle, 2016; Sigafoos & functional three-dimensional stimuli readily found
Roberts-Pennell, 1999). in the learner’s environment; it should then move
There are several techniques for teaching rejec- toward two-dimensional complex stimuli, includ-
tion or refusal behavior. First, the behavior analyst ing functions, features, relations, and private
can strengthen an existing appropriate behavior events (LeBlanc, Dillon, & Sautter, 2009).
to make it more effective, specific, or consistent Behavior analysts should teach echoics early in
(Warren, Yoder, Gazdag, Kim, & Jones, 1993). Sec- programming when teaching vocal tacts, because
ond, the behavior analyst can teach new commu- therapists then can use echoic prompts for other
nicative behaviors by chaining the new response verbal operants (e.g., Kodak & Clements, 2009).
to an existing, inefficient behavior (Keen, Siga- Although researchers have studied generalized
foos, & Woodyatt, 2001) or by prompting (Dras- vocal imitation (e.g., Kymissis & Poulson, 1990),
gow, Halle, Ostrosky, & Harbers, 1996). Third, the more recent investigations have focused on pro-
behavior analyst can replace socially unacceptable cedures to establish vocalizations in learners who
communicative behaviors with acceptable, func- are otherwise nonvocal. One of these procedures,
tionally equivalent behaviors. For example, the called stimulus–stimulus pairing, establishes vo-
behavior analyst could teach a learner to point calizations as conditioned reinforcers. Thus the
to the word stop on a communication board to response-produced auditory stimulus may func-
end a task instead of throwing a tantrum (Carr, tion as reinforcement for the vocalizations that
278 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
produced them (Lepper & Petursdottir, 2017; Shil- preference assessment when initiating mand train-
lingsburg, Hollander, Yosick, Bowen, & Muskat, ing (DeLeon & Iwata, 1996; Fisher et al., 1992;
2015). Other more naturalistic procedures evoke Frost & Bondy, 2002). Other factors that may be
vocalizations by capturing the learner’s interest important in selecting verbal responses to teach
while modeling vocal sounds (Charlop-Christy, relate to the learner’s behavioral development. For
LeBlanc, & Carpenter, 1999). example, learners tact items and events that are
Behavior analysts can teach simple intraver- part of the public environment before they can
bals in the form of fill-in-the-blanks at the same tact private events (see Skinner, 1957, p. 131). In
time they are teaching simple mands and tacts. By other words, learners name common items, such as
contrast, the behavior analyst should teach more toys, furniture, and important people, before they
complex intraverbals after the learner can respond learn to comment about things happening within
to the same topography or word as a listener (i.e., them, such as pain, pleasure, or other emotional
receptive discrimination) and speaker (i.e., tact). changes. Therefore, we must ensure that a learner
For example, before the analyst teaches a learner can comment about items or events in his or her
to say, “Carrot,” when the learner hears, “What environment before teaching comments about a
does a rabbit eat?”, he or she should first learn to possibly painful knee.
tact carrot and eat and respond as a listener when Strategies for teaching verbal behavior such as
hearing these words (Petursdottir, Ólafsdóttir, & incidental teaching and pivotal-response training
Aradóttir, 2008). As with other verbal operants, (Koegel & Koegel, 2005) depend on a teacher’s
the behavior analyst can teach intraverbals by awareness of the learner’s current repertoire; that
using time-delay procedures with either vocal or is, these strategies seek to expand upon current
visual prompts, although the learner’s prior experi- skill sets in small steps. Each lesson depends more
ence with prompts seems to determine which will on what the learner is doing than on what the
be more effective (Coon & Miguel, 2012; Roncati, teacher wants the learner to do. Thus, if a learner
Souza, & Miguel, 2019). has demonstrated a clear preference for a large
Behavior analysts should assess learners’ current as opposed to a small ball and can ask for a ball,
repertoires before teaching these verbal operants teaching the learner to request the large ball rep-
and use assessment results to guide programming. resents a viable target response. For learners who
Unfortunately, many language assessments focus either do not like balls or cannot request a ball in
on the form rather than function of verbal re- any modality, trying to teach large in this way is
sponses (Esch, LaLonde, & Esch, 2010), so they do not likely to be effective.
not measure the strength and breadth of a learner’s Response production issues also may influence
verbal repertoire (Carr & Miguel, 2013). Thus a which skills behavior analysts teach. Importantly,
behavior analyst might use criterion-referenced as- behavior analysts must determine how production
sessments like the Assessment of Basic Language issues may interfere with learning. For example,
and Learning Skills—Revised (ABLLS-R; Par- learners who display some vocal responses may
tington, 2006) and the Verbal Behavior Milestones have difficulty in sound production or blending;
Assessment and Placement Program (VB-MAPP; similarly, learners who sign may have difficulty
Sundberg, 2008) to evaluate a learner’s repertoire, forming or stringing signs together. Some learners
suggest goals for intervention, and track progress. may select letters on a keyboard more easily than
Although many behavior analysts often use these they can produce them with a writing implement.
assessments, research has not evaluated their psy- Research on fluency (Johnson & Layng, 1992)
chometric properties and interobserver reliability suggests that when core response production rates
(Carr & Miguel, 2013), so analysts should not rely (such as the rate of producing individual sounds,
on them solely when making clinical decisions. writing individual letters, or selecting specific pic-
tures) are very low, acquisition of more complex
skills (such as stringing together letters to spell
OTHER ISSUES words or speaking in increasingly complex sen-
tences) will be difficult.
Learner characteristics may influence how we Attributing characteristics to an individual
teach verbal skills and which ones we teach. Some learner based on a general characteristic of the
of these variables may be learner-specific. For ex- population to which he or she belongs is often
ample, a behavior analyst should use a learner’s risky. Nevertheless, a behavior analyst may con-
most potent reinforcers identified via a formal sider such characteristics on a probabilistic basis.
Teaching Verbal Behavior 279
For example, learners with autism spectrum disor- animals, identifying sounds, and describing items,
der (ASD) tend to be less sensitive than typically may be reinforcing and acquired readily. A learn-
developing learners to social reinforcers, especially er’s diagnostic or educational classification does
when they first enter treatment programs. As not set or fix educational goals, but it may suggest
such, praise, smiles, and words of encouragement consequences that do and do not function as re-
are usually not highly motivating for learners with inforcers.
ASD. Skills that produce social consequences, Behavior analysts should use many reinforcers
such as tacts, intraverbals, and most autoclitics, when teaching manding. Although manipulating
will be difficult for these learners to acquire (Fish- access to concrete rewards such as food, drink, and
er et al., 2019). If a teacher replaces social conse- toys may be relatively easy, lessons should extend
quences with more concrete but powerful reinforc- to activities and social events with many poten-
ers, such as food or toys, a learner with ASD may tial communicative partners. Specifically, we
acquire a verbal operant that may remain partially should teach learners to request items from peers
controlled by the item rather than an operant (Schwartz, Garfinkle, & Bauer, 1998) and from
controlled by generalized conditioned reinforc- adults; to talk about and play games and related
ers. Simply substituting one reinforcer for another activities with siblings (Taylor, Levin, & Jasper,
can have a significant impact on the acquisition of 1999); and to seek information (Williams, Pérez-
verbal operants. In this case, the limited availabil- González, & Vogt, 2003). Thus two general train-
ity of reinforcers may affect the behavior analyst’s ing aims are to broaden the array of reinforcers
selection of target skills. If social consequences associated with manding and to increase potential
do not function as reinforcement, then teaching mediators for various reinforcers.
mands for items and activities may be more ef- Finally, much of this chapter has focused on ac-
fective than teaching verbal operants with social quisition of verbal behavior and related interven-
reinforcers. tions in childhood; however, certain developmen-
Learners may have difficulty acquiring sophis- tal disabilities, such as ASD, produce deficits in
ticated language skills if social consequences do verbal behavior that may persist across the lifespan
not function as reinforcers at all or function only (Shattuck et al., 2007). These learners are likely
as weak reinforcers (Greer & Keohane, 2006). to need empirically supported interventions to
For example, learners with Asperger syndrome or increase and maintain functional verbal behavior
other mild forms of ASD typically develop global well into adulthood. The demonstrated continuity
language skills on par with their typically devel- in principles of behavior across the lifespan sug-
oping peers. However, some of these learners may gests that Skinner’s (1957) analysis is just as rele-
have difficulty acquiring communication skills vant to teaching verbal behavior skills to adults as
that relate to social effectiveness—from using and it is to children. Nonetheless, research on strate-
understanding puns and other word play, to more gies to teach such skills to adults with developmen-
general skills associated with successful dating and tal disabilities is far less abundant than research
romantic involvement. Thus a learner may tell a with children (e.g., Bishop-Fitzpatrick, Minshew,
joke, but may not discriminate why it is funny. & Eack, 2014). Preliminary evidence suggests that
Skinner (1957) pointed out the critical role of the conceptual approach and teaching strategies
social reinforcement in the development of lan- based in Skinner’s analysis that produce func-
guage and self-knowledge: “As we have noted, it tional verbal behavior in children also produce
is social reinforcement which leads the individual functional verbal behavior in adults with similar
to know himself. It is only through the gradual impairments (Bracken & Rohrer, 2014; Nepo, Tin-
growth of a verbal community that the individual cani, Axelrod, & Meszaros, 2015). Nevertheless,
becomes ‘conscious.’ He comes to see himself only researchers should focus on the specific nature of
as others see him, or at least only as others insist developmentally compatible and effective strate-
that he see himself” (p. 140, emphasis added). gies for improving verbal behavior in adults.
Learners who show limited responsivity to social
consequences are likely to show concomitant lim-
its in language development. SUMMARY
For other populations, such as children with
Down syndrome, social consequences may be Skinner’s (1957) work provides a platform on
particularly powerful reinforcers. In such cases, which behavior analysts can develop effective
lessons associated with tacting, such as naming targets and strategies to help children and adults
280 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
acquire or improve verbal repertoires. Considering device. International Journal of Developmental Dis-
the relevant motivational and stimulus conditions abilities, 65, 98–107.
is essential in teaching verbal responses. Behavior Carnett, A., Waddington, H., Hansen, S., Bravo, A.,
analysts can provide critical guidance regarding Sigafoos, J., & Lang, R. (2017). Teaching mands to
children with autism spectrum disorder using be-
the importance of reinforcement and how context
havior chain interruption strategies: A systematic
affects the function of new responses when design- review. Advances in Neurodevelopmental Disorders, 1,
ing lessons. Our emphasis on the function rather 203–220.
than the form of verbal behavior also suggests that Carr, E. G. (1994). Emerging themes in the functional
behavior analysts should be adept at helping learn- analysis of problem behavior. Journal of Applied Be-
ers develop verbal repertoires in several modalities, havior Analysis, 27, 393–399.
even while recognizing the societal importance of Carr, J. E., & Miguel, C. F. (2013). The analysis of ver-
speech. Indeed, the techniques for teaching verbal bal behavior and its therapeutic applications. In G.
responses described in this chapter accommodate J. Madden (Ed.), APA handbook of behavior analysis:
Vol. 2. Translating principles into practice (pp. 329–
several response modalities, including speech, sign
352). Washington, DC: American Psychological As-
language, picture-based systems, and other aug- sociation.
mentative devices. Charlop-Christy, M. H., LeBlanc L. A., & Carpenter, M.
H. (1999). Naturalistic teaching strategies (NaTS) to
teach speech to children with autism: Historical per-
REFERENCES spective, development, and current practice. Califor-
nia School Psychologist, 4, 30–46.
Alzrayer, N., Banda, D. R., & Koul, R. K. (2014). Use Chezan, L. C., Drasgow, E., Martin, C. A., & Halle,
of iPad/iPods with individuals with autism and other J. W. (2016). Negatively-reinforced mands: An ex-
developmental disabilities: A meta-analysis of com- amination of resurgence to existing mands in two
munication interventions. Review Journal of Autism children with autism and language delays. Behavior
and Developmental Disorders, 1, 179–191. Modification, 40, 922–953.
American Speech–Language–Hearing Association. Choi, H., O’Reilly, M., Sigafoos, J., & Lancioni, G.
(2016). Scope of practice in speech–language pathol- (2010). Teaching requesting and rejecting sequences
ogy. Retrieved from www.asha.org/policy/SP2016- to four children with developmental disabilities using
00343. augmentative and alternative communication. Re-
Axe, J. B. (2008). Conditional discrimination in the search in Developmental Disabilities, 31, 560–567.
intraverbal relation: A review and recommendations Chomsky, N. (1965). Aspects of a theory of syntax. Cam-
for future research. Analysis of Verbal Behavior, 24, bridge, MA: MIT Press.
159–174. Cipani, E. (1988). The missing item format. Teaching
Barbera, M. L., & Rasmussen, T. (2007). The verbal be- Exceptional Children, 21, 25–27.
havior approach: How to teach children with autism and Contreras, B. P., Cooper, A. J., & Kahng, S. (2020).
related disorders. London: Jessica Kingsley. Recent research on the relative efficiency of speak-
Bishop-Fitzpatrick, L., Minshew, N. J., & Eack, S. M. er and listener instruction for children with autism
(2014). A systematic review of psychosocial interven- spectrum disorder. Journal of Applied Behavior Analy-
tions for adults with autism spectrum disorders. In sis, 53, 584–589.
F. R. Volkmar, B. Reichow, & J. McPartland (Eds.), Coon, J. T., & Miguel, C. F. (2012). The role of increased
Adolescents and adults with autism spectrum disorders exposure to transfer of stimulus control procedures
(pp. 315–327). New York: Springer. on the acquisition of intraverbal behavior. Journal of
Bondy, A., Tincani, M., & Frost, L. (2004). Multiply Applied Behavior Analysis, 45, 657–666.
controlled verbal operants: An analysis and exten- DeLeon, I. G., & Iwata, B. A. (1996). Evaluation of a
sion to the Picture Exchange Communication Sys- multiple-stimulus presentation format for assessing
tem. Behavior Analyst, 27, 247–261. reinforcer preferences. Journal of Applied Behavior
Bracken, M., & Rohrer, N. (2014). Using an adapted Analysis, 29, 519–532.
form of the Picture Exchange Communication Sys- DeSouza, A. A., Fisher, W. W., & Rodriguez, N. M.
tem to increase independent requesting in deafblind (2019). Facilitating the emergence of convergent in-
adults with learning disabilities. Research in Develop- traverbals in children with autism. Journal of Applied
mental Disabilities, 35, 269–277. Behavior Analysis, 52, 28–49.
Brown, R. (1973). A first language. Cambridge, MA: Drasgow, E., Halle, J. W., Ostrosky, M. M., & Harbers,
Harvard University Press. H. M. (1996). Using behavioral indication and func-
Carnett, A., Bravo, A., & Waddington, H. (2019). tional communication training to establish initial
Teaching mands for actions to children with autism sign repertoire with a young child with severe dis-
spectrum disorder using systematic instruction, be- abilities. Topics in Early Childhood Special Education,
havior chain interruption, and a speech-generating 16, 500–521.
Teaching Verbal Behavior 281
Duker, P. C., Kraaykamp, M., & Visser, E. (1994). A Hall, G., & Sundberg, M. L. (1987). Teaching mands by
stimulus control procedure to increase requesting manipulating conditioned establishing operations.
with individuals who are severely/profoundly intel- Analysis of Verbal Behavior, 5, 41–53.
lectually disabled. Journal of Intellectual Disability Halle, J. W., Baer, D., & Spradlin, J. (1981). Teachers’ gen-
Research, 38, 177–186. eralized use of delay as a stimulus control procedure to
Durand, V. M., & Merges, E. (2001). Functional commu- increase language use in handicapped children. Jour-
nication training: A contemporary behavior analytic nal of Applied Behavior Analysis, 14, 389–409.
intervention for problem behavior. Focus on Autism Halle, J. W., Marshall, A. M., & Spradlin, J. E. (1979).
and Other Developmental Disabilities, 16, 110–119. Time delay: A technique to increase language use
Esch, B. E., LaLonde, K. B., & Esch, J. W. (2010). Speech and generalization in retarded children. Journal of
and language assessment: A verbal behavior analysis. Applied Behavior Analysis, 12, 431–439.
Journal of Speech–Language Pathology–Applied Behav- Hixson, M. (2004). Behavioral cusps, basic behavioral
ior Analysis, 5, 166–191. repertoires, and cumulative-hierarchical learning.
Fey, M. (1986). Language intervention with young chil- Psychological Record, 54, 387–403.
dren. Boston: College Hill Press. Horne, P. J., & Lowe, C. F. (1996). On the origins of
Finn, H. E., Miguel, C. F., & Ahearn, W. H. (2012). The naming and other symbolic behavior. Journal of the
emergence of untrained mands and tacts in children Experimental Analysis of Behavior, 65, 185–241.
with autism. Journal of Applied Behavior Analysis, Iacono, T., Carter, M., & Hook, J. (1998). Identifica-
45(2), 265–280. tion of intentional communication in students with
Fiorile, C. A., & Greer, R. D. (2007). The induction of severe and multiple disabilities. Augmentative and Al-
naming in children with no prior tact responses as ternative Communication, 14, 417–431.
a function of multiple exemplar histories of instruc- Ingvarsson, E. T., & Hollobaugh, T. (2010). Acquisi-
tion. Analysis of Verbal Behavior, 23, 71–87. tion of intraverbal behavior: Teaching children with
Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. autism to mand for answers to questions. Journal of
P., Owens, J. C., & Slevin, I. (1992). A comparison of Applied Behavior Analysis, 43, 1–17.
two approaches for identifying reinforcers for persons Johnson, K. R., & Layng, T. J. (1992). Breaking the
with severe and profound disabilities. Journal of Ap- structuralist barrier: Literacy and numeracy with flu-
plied Behavior Analysis, 25, 491–498. ency. American Psychologist, 47, 1475–1490.
Fisher, W. W., Retzlaff, B. J., Akers, J. S., DeSouza, A. Karsten, A. M., & Carr, J. E. (2009). The effects of dif-
A., Kaminski, A. J., & Machado, M. A. (2019). Es- ferential reinforcement of unprompted responding
tablishing initial auditory–visual conditional dis- on the skill acquisition of children with autism. Jour-
criminations and emergence of initial tacts in young nal of Applied Behavior Analysis, 42, 327–334.
children with autism spectrum disorder. Journal of Keen, D., Sigafoos, J., & Woodyatt, G. (2001). Replacing
Applied Behavior Analysis, 52, 1089–1106. prelinguistic behaviors with functional communica-
Frost, L., & Bondy, A. (2002). The Picture Exchange tion. Journal of Autism and Developmental Disorders,
Communication System (PECS) training manual (2nd 31, 385–398.
ed.). Newark, DE: Pyramid Products. Kelley, M. E., Shillingsburg, M. A., Castro, M. J., Addi-
Garcia, E. E., Baer, D. M., & Firestone, I. (1971). The son, L. R., & LaRue, R. H., Jr. (2007). Further evalu-
development of generalized imitation in topographi- ation of emerging speech in children with develop-
cally determined boundaries. Journal of Applied Be- mental disabilities: Training verbal behavior. Journal
havior Analysis, 4, 101–113. of Applied Behavior Analysis, 40, 431–445.
Goldstein, H., Kaczmarek, L., Pennington, R., & Shafer, Kodak, T., & Clements, A. (2009). Acquisition of mands
K. (1992). Peer-mediated intervention: Attending to, and tacts with concurrent echoic training. Journal of
commenting on, and acknowledging the behavior of Applied Behavior Analysis, 42, 839–843.
preschoolers with autism. Journal of Applied Behavior Koegel, R. L., & Koegel, L. K. (2005). Pivotal response
Analysis, 25, 289–305. treatments for autism: Communication, social, and
Greer, B. D., Fisher, W. W., Saini, V., Owen, T. M., & academic development. Baltimore: Brookes.
Jones, J. K. (2016). Functional communication train- Kratzer, D. A., Spooner, F., & Test, D. W. (1993). Ex-
ing during reinforcement schedule thinning: An tending the application of constant time delay:
analysis of 25 applications. Journal of Applied Behav- Teaching a requesting skill to students with severe
ior Analysis, 49, 105–121. multiple disabilities. Education and Treatment of Chil-
Greer, R. D., & Du, L. (2010). Generic instruction ver- dren, 16, 235–253.
sus intensive tact instruction and the emission of Kymissis, E., & Poulson, C. L. (1990). The history of
spontaneous speech. Journal of Speech–Language Pa- imitation in learning theory: The language acquisi-
thology–Applied Behavior Analysis, 5, 1–19. tion process. Journal of the Experimental Analysis of
Greer, R. D., & Keohane, D.-D. (2006). The evolution Behavior, 54, 113–127.
of verbal behavior in children. Journal of Speech and LaFrance, D. L., & Miguel, C. F. (2014). Teaching lan-
Language Pathology–Applied Behavior Analysis, 1, guage to children with autism spectrum disorder. In
111–140. J. Tarbox, D. R. Dixon, P. Sturmey, & J. L. Matson
282 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
(Eds.), Handbook of early intervention for autism spec- A review of the literature related to children with
trum disorders: Research, practice, and policy (pp. 403– autism. Education and Training in Developmental Dis-
436). New York: Springer. abilities, 41, 213–224.
LaFrance, D. L., Wilder, D. A., Normand, M. P., & Maurice, C., Green, G., & Luce, S. C. (1996). Behavioral
Squires, J. L. (2009). Extending the assessment of intervention for young children with autism: A manual
functions of vocalizations in children with limited for parents and professionals. Austin, TX: PRO-ED.
verbal repertoires. Analysis of Verbal Behavior, 25, McLay, L., Schäfer, M. C., van der Meer, L., Couper, L.,
19–32. McKenzie, E., O’Reilly, M. F., et al. (2017). Acqui-
LaMarre, J., & Holland, J. G. (1985). The functional sition, preference and follow-up comparison across
independence of mands and tacts. Journal of the Ex- three AAC modalities taught to two children with
perimental Analysis of Behavior, 43, 5–19. autism spectrum disorder. International Journal of
Landa, R. K., Hansen, B., & Shillingsburg, M. A. (2017). Disability, Development and Education, 64, 117–130.
Teaching mands for information using ‘when’ to chil- Michael, J. (1993). Establishing operations. Behavior
dren with autism. Journal of Applied Behavior Analy- Analyst, 16, 191–206.
sis, 50, 538–551. Michael, J., & Miguel, C. F. (2020). Motivating opera-
Leaf, R., & McEachin, J. (1999). A work in progress: tions. In J. O. Cooper, T. E. Heron, & W. L. Heward
Behavior management strategies and a curriculum for (Eds.), Applied behavior analysis (3rd ed., pp. 372–
intensive behavior treatment of autism. New York: DRL 394). Hoboken, NJ: Pearson.
Books. Michael, J., Palmer, D. C., & Sundberg, M. L. (2011).
LeBlanc, L. A., Dillon, C. M., & Sautter, R. A. (2009). The multiple control of verbal behavior. Analysis of
Establishing mand and tact repertoires. In R. A. Re- Verbal Behavior, 27, 3–22.
hfeldt & Y. Barnes-Holmes (Eds.), Derived relational Miguel, C. F. (2016). Common and intraverbal bidi-
responding: Applications for learners with autism and rectional naming. Analysis of Verbal Behavior, 32,
other developmental disabilities (pp. 79–108). Oakland, 125–138.
CA: New Harbinger. Miguel, C. F., Carr, J. E., & Michael, J. (2002). The ef-
Lechago, S. A., Carr, J. E., Grow, L. L., Love, J. R., & Al- fects of a stimulus–stimulus pairing procedure on the
mason, S. M. (2010). Mands for information general- vocal behavior of children diagnosed with autism.
ize across establishing operations. Journal of Applied Analysis of Verbal Behavior, 18, 3–13.
Behavior Analysis, 43, 381–395. Miguel, C. F., Frampton, S. E., Lantaya, C. A., LaFrance,
Lepper, T. L., & Petursdottir, A. I. (2017). Effects of re- D. L., Quah, K., Meyer, C. S., et al. (2015). The ef-
sponse-contingent stimulus pairing on vocalizations fects of tact training on the development of analogi-
of nonverbal children with autism. Journal of Applied cal reasoning. Journal of the Experimental Analysis of
Behavior Analysis, 50, 756–774. Behavior, 104, 96–118.
Lorah, E. R., Parnell, A., Whitby, P. S., & Hantula, Miguel, C. F., & Kobari-Wright, V. V. (2013). The ef-
D. (2015). A systematic review of tablet computers fects of tact training on the emergence of categoriza-
and portable media players as speech generating de- tion and listener behavior in children with autism.
vices for individuals with autism spectrum disorder. Journal of Applied Behavior Analysis, 46, 669–673.
Journal of Autism and Developmental Disorders, 45, Miguel, C. F., Petursdottir, A. I., & Carr, J. E. (2005).
3792–3804. The effects of multiple-tact and receptive-discrimi-
Lorah, E. R., Tincani, M., Dodge, J., Gilroy, S., Hickey, nation training on the acquisition of intraverbal be-
A., & Hantula, D. (2013). Evaluating picture ex- havior. Analysis of Verbal Behavior, 21, 27–41.
change and the iPad™ as a speech generating device Mirenda, P. (2003). Toward functional augmentative
to teach communication to young children with au- and alternative communication for students with
tism. Journal of Developmental and Physical Disabili- autism: Manual signs, graphic symbols, and voice
ties, 25, 637–649. output communication aids. Language, Speech, and
Lord, C., Cook, E. H., Leventhal, B. L., & Amaral, D. Hearing Services in Schools, 34, 203–216.
G. (2000). Autism spectrum disorders. Neuron, 28, Mundy, P. (1995). Joint attention and social–emotional
355–363. approach behavior in children with autism. Develop-
Lovaas, O. I. (2003). Teaching individuals with develop- ment and Psychopathology, 7, 63–82.
mental delays. Austin, TX: PRO-ED. Nepo, K., Tincani, M., Axelrod, S., & Meszaros, L.
Lovaas, N. W., & Lovaas, E. E. (1999). The Reading and (2015). iPod Touch® to increase functional commu-
Writing Program: An alternative form of communica- nication of adults with autism spectrum disorder and
tion. Austin, TX: PRO-ED. significant intellectual disability. Focus on Autism
Maffei, J., Singer-Dudek, S., & Keohone, D. (2014). and Other Developmental Disabilities, 32(3), 209–217.
The effects of the establishment of adult faces and/ Partington, J. W. (2006). The Assessment of Basic Lan-
or voices as conditioned reinforcers for children with guage and Learning Skills—Revised (The ABLLS-R).
ASD and related disorders. Acta de Investigacion Psi- Pleasant Hill, CA: Partington Behavior Analysts.
cologia, 4, 1621–1641. Petursdottir, A. I., & Carr, J. E. (2011). A review of rec-
Mancil, G. (2006). Functional communication training: ommendations for sequencing receptive and expres-
Teaching Verbal Behavior 283
sive language instruction. Journal of Applied Behavior nal of Developmental and Physical Disabilities, 10,
Analysis, 44, 859–876. 133–151.
Petursdottir, A. I., Carr, J. E., & Michael, J. (2005). Sigafoos, J., Kerr, M., Roberts, D., & Couzens, D. (1994).
Emergence of mands and tacts of novel objects Increasing opportunities for requesting in classrooms
among preschool children. Analysis of Verbal Behav- serving children with developmental disabilities.
ior, 21, 59–74. Journal of Autism and Developmental Disorders, 24,
Petursdottir, A. I., Ólafsdóttir, A. R., & Aradóttir, B. 631–645.
(2008). The effects of tact and listener training on Sigafoos, J., & Roberts-Pennell, D. (1999). Wrong-item
the emergence of bidirectional intraverbal relations. format: A promising intervention for teaching so-
Journal of Applied Behavior Analysis, 41, 411–415. cially appropriate forms of rejecting to children with
Piaget, J. (1951). The psychology of intelligence. London: developmental disabilities? Augmentative and Alter-
Routledge & Kegan Paul. native Communication, 15, 135–140.
Rodriguez, N. M., Levesque, M. A., Cohrs, V. L., & Nie- Skinner, B. F. (1957). Verbal behavior. Englewood Cliffs,
meier, J. J. (2017). Teaching children with autism to NJ: Prentice-Hall.
request help with difficult tasks. Journal of Applied Sprinkle, E. C., & Miguel, C. F. (2012). The effects of
Behavior Analysis, 50, 717–732. listener and speaker training on emergent relations
Roncati, A. L., Souza, A. C., & Miguel, C. F. (2019), in children with autism. Analysis of Verbal Behavior,
Exposure to a specific prompt topography predicts its 28, 111–117.
relative efficiency when teaching intraverbal behav- Sundberg, M. L. (2008). Verbal Behavior Milestones As-
ior to children with autism spectrum disorder. Journal sessment and Placement Program (VB-MAPP): A lan-
of Applied Behavior Analysis, 52, 739–745. guage and social skills assessment program for children
Saini, V., Miller, S. A., & Fisher, W. W. (2016). Multiple with autism or other developmental disabilities. Con-
schedules in practical application: Research trends cord, CA: AVB Press.
and implications for future investigation. Journal of Sundberg, M. L., & Partington, J. W. (1998). Teaching
Applied Behavior Analysis, 49, 421–444. language to children with autism or other developmental
Schwartz, I. S., Garfinkle, A. N., & Bauer, J. (1998). disabilities. Danville, CA: Behavior Analysts.
Communicative outcomes for young children with Sundberg, M. L., & Sundberg, C. A. (2011). Intraverbal
disabilities. Topics in Early Childhood Special Educa- behavior and verbal conditional discriminations in
tion, 18, 144–159. typically developing children and children with au-
Shafer, E. (1995). A review of interventions to teach tism. Analysis of Verbal Behavior, 27, 23–43.
a mand repertoire. Analysis of Verbal Behavior, 12, Taylor, B. A., Levin, L., & Jasper, S. (1999). Increas-
53–66. ing play-related statements in children with autism
Shattuck, P. T., Seltzer, M. M., Greenberg, J. S., Ors- toward their siblings: Effects of video modeling.
mond, G. I., Bolt, D., Kring, S., et al. (2007). Change Journal of Developmental and Physical Disabilities, 11,
in autism symptoms and maladaptive behaviors in 253–264.
adolescents and adults with an autism spectrum dis- Tincani, M. (2004). Comparing sign language and pic-
order. Journal of Autism and Developmental Disorders, ture exchange training for students with autism and
37(9), 1735–1747. multiple disabilities. Focus on Autism and Other De-
Shillingsburg, M. A., Bowen, C. N., Valentino, A. L., velopmental Disabilities, 19, 162–173.
& Pierce, L. E. (2014). Mands for information using Tincani, M. (2007). Beyond consumer advocacy: Au-
“who?” and “which?” in the presence of establishing tism spectrum disorders, effective instruction, and
and abolishing operations. Journal of Applied Behav- public schooling. Intervention in School and Clinic, 43,
ior Analysis, 47, 136–150. 47–51.
Shillingsburg, M. A., Hollander, D. L., Yosick, R. N., Tincani, M. J., & Boutot, E. A. (2005). Autism and
Bowen, C., & Muskat, L. R. (2015). Stimulus–stimu- technology: Current practices and future directions.
lus pairing to increase vocalizations in children with In D. L. Edyburn, K. Higgins, & R. Boone (Eds.), The
language delays: A review. Analysis of Verbal Behav- handbook of special education technology research and
ior, 31, 215–235. practice (pp. 413–421). Whitefish Bay, WI: Knowl-
Shillingsburg, M. A., Kelley, M. E., Roane, H. S., Kisa- edge by Design.
more, A., & Brown, M. R. (2009). Evaluation and Tincani, M., & Zawacki, J. (2012). Evidence based prac-
training of yes–no responding across verbal operants. tice for communication skill acquisition. In E. E.
Journal of Applied Behavior Analysis, 42, 209–223. Barton & B. A. Harn (Eds.), Educating young children
Shillingsburg, M. A., Marya, V., Bartlett, B. L., & with autism spectrum disorder (pp. 171–189). Thou-
Thompson, T. M. (2019). Teaching mands for in- sand Oaks, CA: Corwin Press.
formation using speech generating devices: A rep- Touchette, P. E., & Howard, J. S. (1984). Errorless learn-
lication and extension. Journal of Applied Behavior ing: Reinforcement contingencies and stimulus con-
Analysis, 52, 756–771. trol transfer in delayed prompting. Journal of Applied
Sigafoos, J. (1998). Assessing conditional use of graphic Behavior Analysis, 17, 175–188.
mode requesting in a young boy with autism. Jour- Twyman, J. (1996). The functional independence of im-
284 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
pure mands and tacts of abstract stimulus properties. ment in language systems of the brain. Developmental
Analysis of Verbal Behavior, 13, 1–19. Psychobiology, 46, 233–251.
Wankoff, L. S. (2005). Innovative methods in language Williams, G., Pérez-González, L. A., & Vogt, K. (2003).
intervention. Austin, TX: PRO-ED. The role of specific consequences in the mainte-
Warren, S. F., Yoder, P. J., Gazdag, G. E., Kim, K., & nance of three types of questions. Journal of Applied
Jones, H. A. (1993). Facilitating prelinguistic com- Behavior Analysis, 36, 285–296.
munication skills in young children with develop- Ylvisaker, M., & Feeney, T. J. (1994). Communication
mental delay. Journal of Speech and Hearing Research, and behavior: Collaboration between speech–lan-
36, 83–97. guage pathologists and behavioral psychologists. Top-
Werker, J. F., & Tees, R. C. (1999). Influences on infant ics in Language Disorders, 15(1), 37–54.
speech processing: Toward a new synthesis. Annual Young, J. M., Krantz, P. J., McClannahan, L. E., &
Review of Psychology, 50, 509–535. Poulson, C. L. (1994). Generalized imitation and
Werker, J. F., & Tees, R. C. (2005). Speech perception as response-class formation in children with autism.
a window for understanding plasticity and commit- Journal of Applied Behavior Analysis, 27, 685–697.
CHAP TER 17
Staff training and management represent a long- analysis and bridge that gap (Babcock, Fleming,
standing area of focus in applied behavior analysis. & Oliver, 1998; Parsons, Rollyson, & Reid, 2012).
Soon after initial demonstrations of the efficacy Staff training and management have consti-
of behavior analysis for improving the behavior of tuted a consistent but not a large area of focus
people with special needs in the 1960s, behavior in applied behavior analysis since the field’s in-
analysis directed its attention to disseminating the ception. Investigations on staff performance in
emerging technology among human-service staff. human-service settings represent a small percent-
Concern first centered on training staff in basic age of published research in behavior analysis.
behavioral procedures to use with people with de- Nonetheless, when we consider that investigators
velopmental and related disabilities (e.g., Gardner, have been addressing staff performance for over
1972). Shortly thereafter, behavior analysts recog- 40 years, behavior analysts have conducted a sub-
nized that the same principles underlying behavior stantial amount of research. Such research has
change procedures for people with developmental produced highly relevant information for improv-
disabilities were applicable with staff members’ ing staff performance. The purpose of this chapter
work performance (Hollander, Plutchik, & Horn- is to summarize the existing knowledge base from
er, 1973; Quilitch, 1975; Welsch, Ludwig, Radiker, behavior-analytic research on staff training and
& Krapfl, 1973). management. An additional purpose is to describe
Interest in training human-service staff to use gaps in the current knowledge and suggest areas
behavioral procedures and applying behavioral for future research.
strategies to manage staff performance continues
today. Such interest is due to several factors, in-
cluding recognition that many services for people FOCUS OF CHAPTER
with disabilities warrant improvement (Reid, Par-
sons, & Green, 2012, Ch. 1). A related factor is A specialty area of applied behavior analysis, orga-
the continuing gap between evidence-based pro- nizational behavior management, has conducted the
cedures for promoting desirable behavior among behavioral research on staff training and manage-
people with special needs and caregivers’ provision ment. Although the primary emphasis in organi-
of day-to-day services to those people (Lerman, zational behavior management more recently has
2009; Neef, 1995). We consider staff training and been business and industry rather than human-
management as one way to disseminate behavior service settings, behavioral research has contin-
285
286 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
ued in the latter settings. The focus of the cur- training research. First, investigators demonstrat-
rent chapter is on staff training and management ed that behavior analysts could teach professionals
specifically in human-service settings. (For fuller and paraprofessionals to use behavioral procedures
discussion of organizational behavior management to change the behavior of individuals with signifi-
in general, see Wilder & Gravina, Chapter 32, this cant disabilities. Although this research outcome
volume.) The principles of behavior change and is not surprising now, it was noteworthy at the
many of the applications are the same in the two time. The research demonstrated that behavior
settings, but significant differences in business and analysts could apply behavioral procedures in ex-
industry versus typical human-service settings isting service settings to teach important skills to
exist (e.g., variations in outcomes reflecting suc- people with severe and profound cognitive disabil-
cessful operations; personnel policies; potential ities who previously were considered unteachable
performance incentives and reinforcers). (Whitman, Hantula, & Spence, 1990). Successful
As with investigations in behavior analysis in teaching demonstrations played an important role
general, most staff research has targeted settings in major changes in residential, vocational, and
for people with intellectual and other develop- educational opportunities subsequently offered to
mental disabilities. Such settings are likewise the people with developmental disabilities across the
focus of this chapter. However, we include research United States as part of the movements toward de-
on improving staff performance in other settings, institutionalization, rights to education, and com-
particularly schools, where relevant. munity inclusion.
A second major outcome of initial studies on
training staff to use behavioral procedures per-
FORMAT OF CHAPTER tained to what constituted effective training.
Research demonstrated that typical training ap-
The chapter consists of two primary sections. The proaches relying on verbal procedures (i.e., lec-
first section summarizes the technology of staff tures and written material) were useful for train-
training and management from behavior-analytic ing knowledge about job skills, but rarely were
research to date. The second section describes cur- effective for training staff how to perform the
rent gaps in the technology and areas in need of skills (Gardner, 1972). More performance-oriented
investigation, based on the noted gaps and emerg- training procedures were necessary, such as mod-
ing trends in human-service settings. eling and practice, to teach staff the skills neces-
sary to perform job duties (for summaries, see Jahr,
1998; Reid, 2004; Reid, Parsons, & Green, 1989,
EXISTING TECHNOLOGY OF STAFF TRAINING Ch. 3).
AND MANAGEMENT Implications of findings from early behavioral
research regarding effective staff-training pro-
We can categorize behavior-analytic research on cedures have proven especially noteworthy. The
staff performance in the human services generally findings highlight a primary reason for many prob-
in three areas: (1) training staff in work perfor- lems with nonproficient staff performance in the
mance, (2) improving ongoing work performance, human services: In many cases, staffers may not
and (3) maintaining proficient work performance. have not been trained effectively to perform the
Although these three areas are closely related and skills supervisors expected. Human-service set-
often overlap, they represent a useful means of tings usually provide staff training, but the train-
organizing and describing investigations of staff ing frequently relies on verbal approaches that
work performance. research has shown are insufficient for training
performance skills.
Early training research established the foun-
Behavioral Procedures to Train
dation for the development of a highly effective
Human‑Service Staff
training technology for human-service staff to
The first investigations using behavioral proce- perform work skills. The development and evalu-
dures to train staff members typically involved ation of that technology have been described in
teaching basic behavior modification skills to hu- several reviews (Adkins, 1996; Demchak, 1987;
man-service staff for use with people with devel- Gravina et al., 2018; Jahr, 1998; Parsons et al.,
opmental disabilities (e.g., Gardner, 1972; Koegel, 2012; Reid, 2004), and we generally refer to it as
Russo, & Rincover, 1977; Watson & Uzzell, 1980). behavioral skills training. The technology represents
Two major findings resulted from the early staff- a rather straightforward approach for training
Staff Training and Management 287
proving staff performance (for reviews, see Phil- gible consequences as potential reinforcers, such
lips, 1998; Reid & Whitman, 1983). as money (Katz, Johnson, & Gelfand, 1972), trad-
Research demonstrating the inconsistent ef- ing stamps (Bricker, Morgan, & Grabowski, 1972),
fectiveness of antecedent interventions has im- and free meals (Shoemaker & Reid, 1980). Due in
portant implications. Antecedent interventions, large part to practical considerations with tangible
such as a supervisor’s providing vocal or written consequences, such as cost to an agency and lack
instructions, represent the most commonly used of supervisory control necessary to provide money
training method for improving staff performance frequently and contingently, recent investigations
in human-service settings. Consequently, supervi- have targeted more readily available consequenc-
sors often attempt to improve staff performance by es. The most frequently investigated consequence
using procedures that research has shown are not has been performance feedback.
effective consistently (cf. Sturmey, 1998). Investigators have provided feedback contin-
The continued use of antecedent procedures, gent on target staff behavior in many formats,
despite their inconsistent effectiveness, prompts including spoken (Realon, Lewallen, & Wheeler,
the question of why supervisors employ such strat- 1983), written (Kneringer & Page, 1999), graphic
egies so frequently. One likely answer is a perva- (Miles & Wilder, 2009), privately presented to in-
sive problem in the human services: lack of super- dividual staff members (Shoemaker & Reid, 1980),
visor training in evidence-based procedures for publicly presented to groups of staffers (Cotnoir-
improving staff performance (Reid, Parsons, Lat- Bichelman, Thompson, McKerchar, & Haremza,
timore, Towery, & Reade, 2005; Sturmey, 1998). 2006), and a combination of formats (Casey &
Lacking knowledge and skills of evidence-based McWilliam, 2011; Luck, Lerman, Wu, Dupuis, &
procedures, supervisors may well resort to familiar Hussein, 2018). Although each type of feedback
strategies, such as simply telling staffers what they has relative advantages and disadvantages, each
should do. Another likely reason is that intermit- generally has been effective in increasing targeted
tent changes in staff work behavior after supervi- staff behavior over time (for reviews, see Alvero et
sor instructions episodically reinforces supervisors’ al., 2001; Balcazar, Hopkins, & Suarez, 1986; Ford,
instructional behavior. 1980). The most commonly investigated type of
Clearly, some staff members respond to super- feedback is positive comments (e.g., praise) re-
visory instructions by altering their work behavior garding target staff behavior. Some debate exists
in some situations. Increased research is warranted regarding the mechanism underlying contingent
to determine conditions in which instructions and feedback as a behavior change intervention with
other antecedent interventions are likely to be ef- staff (Alvero et al., 2001). Nonetheless, ample
fective (cf. Graff & Karsten, 2012). For example, evidence demonstrates feedback can improve staff
staff behavior may be more likely to change fol- performance.
lowing an instruction if the targeted behavior in- Despite the frequently demonstrated efficacy of
volves minimal response effort or represents a one- feedback with staff performance in the research
time, discrete event, as opposed to involving more literature, systematic use of feedback has not been
considerable effort or requiring repeated activity. common practice in many human-service settings
Staffers may be more responsive to supervisory (Harchik, Sherman, Hopkins, Strouse, & Sheldon,
instructions if supervisors have a history of follow- 1989; Reid, 2004). Two likely reasons exist for this.
ing staff (non)compliance with feedback or other One reason is that supervisors may not receive ap-
consequences. If investigations could identify con- propriate training in the skills they need to pro-
ditions in which antecedents are more and less vide effective feedback to staff (Reid et al., 2005;
effective, then we could describe the efficacious Sturmey, 1998). A second reason is that providing
procedures trainers should use routinely. systematic feedback requires consistent supervisor
effort. Consequently, effective supervisor feedback
to staff is likely to require training and feedback to
Consequence Interventions for Improving
the supervisor from upper management.
Staff Performance
Little research exists on consequences to punish
Behavior-analytic research on consequence-based inadequate staff performance. Available research
training procedures has focused on reinforcing de- has shown that punishment is not an effective
sired staff performance. Investigators have directed way to change staff behavior consistently (e.g.,
relatively little attention to punishing undesired Gardner, 1970; Repp & Deitz, 1979). Nonetheless,
behavior. Early investigations often employed tan- supervisors often attempt to use punishment to
Staff Training and Management 289
change staff behavior. One early survey indicated control procedures to improve staff performance
that over 90% of supervisors in settings serving suggests that such approaches can be effective, but
people with developmental disabilities relied on the conditions under which they are effective are
punishment procedures for managing staff perfor- unclear and warrant continued research.
mance problems (Mayhew, Enyart, & Cone, 1979).
Such results are disconcerting, because research
Multicomponent Interventions for Improving
suggests that such approaches are not effective and
Staff Performance
may even be detrimental to an agency’s staff (Stur-
mey, 1998). Most interventions for improving staff perfor-
mance have been multicomponent. Typically, the
researcher conducts an initial training or instruc-
Self‑Control Interventions for Improving
tional procedure, followed by application of perfor-
Staff Performance
mance consequences and self-control procedures,
A small but relatively persistent area of research to a lesser degree. The primary purpose of research
on improving staff performance has been the eval- with multicomponent interventions has been to
uation of self-control procedures involving staffers’ demonstrate a reliable means of improving a desig-
use of goal setting, self-recording, and (to a lesser nated area of staff performance. The rationale gen-
extent) self-reinforcement. Researchers have used erally has been that combining various procedures
each procedure with one or more other self-control enhances the likelihood of success of the interven-
strategies. A primary rationale for evaluating self- tion relative to reliance on only one procedure.
control procedures from a management perspec- Although different strategies comprise multi-
tive is that the procedures have been effective component interventions, an underlying concep-
behavior change interventions in areas other than tual basis for the interventions remains the basic
staff management. Logically, we might expect A-B-C model noted earlier. Behavior analysts
them to promote behavior change among human- have attempted to streamline and provide more
service staff. Additionally, self-control procedures systemization to multicomponent interventions
can require less time and effort for supervisors, by employing a rather generic model of behavioral
because staff members are implementing the pro- supervision (Hawkins, Burgio, Langford, & Engel,
cedures to enhance their own performance in con- 1992; Reid & Shoemaker, 1984). Coinciding with
trast to requiring implementation by a supervisor a popular movement to provide consumer-cen-
(Williams, Vittorio, & Hausherr, 2002). tered services (Ivancic & Helsel, 1998), Reid and
Investigations on self-control interventions with Parsons (2002) developed behavioral outcome man-
human-service staff have reported mixed results. agement, which is an update of behavioral super-
In some cases, self-control procedures produced vision. Outcome management, which has demon-
significant improvements in staff performance strated efficacy in several human-service settings
(e.g., Burgio, Whitman, & Reid, 1983; Plavnick, (Parsons, Rollyson, & Reid, 2004; Reid, Green,
Ferreri, & Maupin, 2010), whereas the improve- & Parsons, 2003; Reid et al., 2005), identifies de-
ments were inconsistent or temporary across staff sired outcomes for agency consumers to attain and
members in others (Doerner, Miltenberger, & specifies respective staff performances necessary
Bakken, 1989; Petscher & Bailey, 2006; Richman, to assist consumers in attaining the outcomes. We
Riordan, Reiss, Pyles, & Bailey, 1988; Suda & summarize the outcome management approach to
Miltenberger, 1993). Even in the former research, working with staff below.
separating effects of self-control procedures from
effects of supervisor behavior that was occurring
Basic Steps of Behavioral Outcome Management
simultaneously was often difficult. Another con-
cern is that supervisory behavior may be neces- The basic steps of behavioral outcome manage-
sary to promote staff use of self-control procedures ment are as follows:
(Adkins, 1996), which reduces the time-efficiency
advantage of the procedures. A potential advan- 1. Specify a target consumer outcome.
tage of self-control procedures may be to maintain 2. Specify the staff performance necessary to as-
improvements in staff performance that initially sist consumers in attaining target outcome.
accompanied supervisory interventions (Brackett, 3. Train staff members in targeted behavior using
Reid, & Green, 2007; Kissel, Whitman, & Reid, performance- and competency-based training
1983). Nonetheless, the overall research on self- procedures.
290 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
4. Monitor staffers’ job performance. sonnel to initially implement and evaluate a train-
5. Provide supportive feedback for proficient staff ing or management program and continue reliable
performance. observations for extended maintenance periods.
6. Provide corrective feedback for nonproficient Furthermore, experimental manipulations must
staff performance. occur after the initial intervention, and observa-
tions must continue to evaluate variables function-
Multicomponent programs have improved ally affecting maintenance.
staff performance in many situations. However, Despite these difficulties in conducting mainte-
because of the reliance on multiple intervention nance research, the available data are encouraging
components, these approaches require certain regarding use of behavioral procedures to maintain
supervisory skills plus consistent time and effort. effects of training and management interventions.
The latter features likely represent one reason Investigators have published several demonstra-
why systematic, multicomponent behavioral ap- tions that staff behavior change can be main-
proaches to staff training and management are tained for up to several years (e.g., Harchik, Sher-
more prevalent in the research literature than in man, Sheldon, & Strouse, 1992; Parsons, Schepis,
routine practice. Reid, McCarn, & Green, 1987; Pollack, Fleming,
& Sulzer-Azaroff, 1994; Richman et al., 1988)
and even across a 30-year period (Reid, Parsons,
Behavioral Procedures to Maintain
& Jensen, 2017). These data suggest that behav-
Staff Performance
ioral management interventions are accompanied
The third major area of behavioral research on staff by sustained improvements in staff performance
performance is maintenance of behavior change if some components of the initial interventions
after training and management interventions. remain in place. However, sustaining supervisor
Behavior analysts often consider maintenance a procedures to maintain changes in staff behavior
subcategory of training or management research, can be problematic.
rather than a specific category itself. However, we One way to maintain the effects of staff perfor-
address maintenance as its own category for two mance interventions is to incorporate behavioral
reasons. First, maintaining improvements in staff procedures, like supervisory feedback, in an agen-
performance is critical for demonstrating the so- cy’s routine management system. Although some
cial significance of behavioral research on staff research has demonstrated positive effects of in-
training. Second, maintaining target staff perfor- stitutionalizing such procedures in an agency’s op-
mance has proven difficult. eration (Christian, 1983; Fielding & Blasé, 1993)
Behavior analysts have long recognized the dif- those positive findings have not been consistent
ficulty of maintaining improved staff performance (Green, Rollyson, Passante, & Reid, 2002). More
after staff training and management interventions. specifically, contingencies seem necessary to en-
Liberman (1983), for example, acknowledged that sure that supervisors carry out the maintenance
effects of behavioral interventions on staff perfor- systems. In short, we have much to learn about in-
mance often ended as soon as investigators com- corporating effective maintenance procedures in
pleted their study in one human-service setting. an agency’s routine operation.
Subsequently, investigators have recognized the
need to promote maintenance of initial improve-
ments in staff performance after behavioral inter- GAPS IN STAFF‑TRAINING AND MANAGEMENT
ventions (Carr, Wilder, Majdalany, Mathisen, & TECHNOLOGY, AND FUTURE RESEARCH AREAS
Strain, 2013; Phillips, 1998).
Although behavior analysts recognize the im- We have gained considerable knowledge from
portance of maintaining staff performance change behavior-analytic research on staff training and
and difficulties in this regard, they have conducted management. Research has identified effective
much less research in this area than on training training programs, and we have an effective tech-
and management interventions to improve staff nology for teaching performance skills to human-
performance initially (Downs, Downs, & Rau, service staff. Behavior analysts have used the
2008; van Oorsouw, Embregts, Bosman, & Jahoda, A-B-C model as a conceptual basis for develop-
2009). One likely reason for the relative lack of ing many procedures, particularly in consequence
maintenance research is the time involved in this and multicomponent interventions. Findings have
research. Investigators must work with agency per- been promising for producing long-term perfor-
Staff Training and Management 291
mance improvements with the behavioral proce- behavior-analytic research. Behavior analysts
dures researchers have investigated. have conducted much research over the last three
The knowledge base and technology for training decades on assessing the function of challenging
and managing staff performance remain incom- behavior and developing function-based interven-
plete, however. This section provides a synthesis tions. Subsequently, research has addressed how
of significant gaps in the existing technology and to train staff in functional-assessment procedures
corresponding areas warranting future research. (Chok, Shlesinger, Studer, & Bird, 2012; Moore
We can view the applied-behavior-analytic re- et al., 2002; Schnell, Sidener, DeBar, Vladescu,
search needs on staff training and management & Kahng, 2018; Wallace, Doney, Mintz-Resudek,
from two perspectives. The first perspective is re- & Tarbox, 2004). Similar developments have oc-
search on expanding the technology for training curred in other areas, such as training staff to
staff members and managing their performance. identify the preferences of people with disabilities
The second is how to disseminate and incorporate (Ausenhus & Higgins, 2019; Lavie & Sturmey,
the technology more effectively in the routine op- 2002; Roscoe, Fisher, Glover, & Volkert, 2006).
eration of human-service settings. Training and managing staff in the use of new
technologies from behavior-analytic research rep-
resent an area of continuing importance.
Expanding the Technology of Staff Training
and Management
Expanding Staff‑Training and Management Research
Like its parent discipline of applied behavior anal-
to Other Problematic Performance Areas
ysis, organizational behavior management and
related areas of behavioral staff training and man- Although behavioral research has addressed nu-
agement are continually evolving. The evolution merous types of staff behavior in the human ser-
occurs as research enhances behavior analysts’ vices, several important performance areas remain
understanding of human behavior and their abil- frequently problematic but have not been exam-
ity to promote socially valued behavior in different ined thoroughly. One notable example is staff
contexts. The evolution also occurs as behavioral turnover. High rates of staff turnover represent
researchers target new or previously unaddressed one of the most troublesome issues facing many
problem areas. Knowledge and technology derived human-service settings, yet turnover has received
from research represents the evidence-based foun- infrequent attention from behavioral researchers
dation that separates behavioral staff training and (Strouse, Carroll-Hernandez, Sherman, & Shel-
management from almost every other approach to don, 2003).
supervision in human services (Reid & Parsons, Another area of concern that has received
2002). As such, we anticipate and desire that re- relatively little attention pertains to agencies
searchers will continue to expand and refine the that provide support for adults with autism spec-
behavioral technology of staff training and man- trum disorder (ASD). Behavior-analytic research
agement. and application involving children with ASD
has grown tremendously in the last two decades,
with a corresponding increase (albeit to a lesser
Training and Managing Staff in the Use of New
degree) in training staff to work in this area (e.g.,
Behavioral Technologies
Catania, Almeida, Liu-Constant, & DiGennaro
In one way, research on staff training and manage- Reed, 2009; DiGennaro Reed, Codding, Catania,
ment in human-service settings follows research & Maguire, 2010; Graff & Karsten, 2012). How-
on changing the behavior of people with special ever, many children with ASD who have received
needs. As research in behavior analysis demon- intensive behavioral services have grown up and
strates new or better means of teaching individuals entered service systems for adults, which do not
and overcoming challenging behavior, for exam- provide the same degree of behavioral support.
ple, behavior analysts conduct staff training and Adults with disabilities such as ASD also have
management research to disseminate those means needs that go beyond children’s, and these needs
to human-service staff. require support staff with special skills (Reid,
Research on the functional assessment of 2016). To illustrate, many adults with ASD and
challenging behavior among people with devel- other severe disabilities spend much of their time
opmental disabilities provides an illustration of in center-based settings (Wehman, 2011). A long-
how research on staff performance follows other standing concern with these settings is a lack of
292 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
meaningful activity involvement among partici- ing videos and DVDs, and interactive software,
pating consumers, exemplified by frequent occur- for example, are now readily available to many
rences of consumers’ putting the same puzzle to- human-service settings (Severtson & Carr, 2012).
gether day after day, coloring in children’s coloring Correspondingly, researchers have directed signif-
books, and being involved in activities with no icant attention to evaluating and demonstrating
apparent purpose other than to keep them busy the effectiveness of these approaches to training
(Reid et al., 2017). Research on staff training and (e.g., Catania et al., 2009; Moore & Fisher, 2007;
performance identified this problem relatively Rosales, Gongola, & Homlitas, 2015; Scott, Ler-
early and reported ways of working with staffers man, & Luck, 2018; Weldy, Rapp, & Capocasa,
to promote more meaningful activity involvement 2014).
(Dyer, Schwartz, & Luce, 1984; Parsons et al., New information technologies offer many at-
1987; Reid et al., 1985). However, recent observa- tractive features for training important work
tional studies indicate that a lack of meaningful behavior to human-service staff. To illustrate,
activity involvement among adults with ASD and training DVDs and videos in other formats may
other severe disabilities continues to be prevalent represent a means of providing relevant informa-
across the United States (Reid et al., 2017). tion and procedural demonstrations that require
minimal trainer time relative to more traditional
training procedures (Macurik, O’Kane, Malanga,
Training and Managing Staff’s Use of Behavioral
& Reid, 2008). Continued research evaluating
Technologies in New Venues
these approaches to staff training seems warrant-
Just as behavior analysis is continually evolving as ed. However, caution also is warranted. Informa-
a professional discipline, the human-service field tion technologies still rely heavily on dissemina-
tends to evolve over time. If behavior analysts in- tion of verbal information, with the addition of
tend to practice in new and altered venues, such as performance-based modeling in many cases, and
in clients’ homes with caregivers, then behavioral previous research has produced inconsistent ef-
technologies must be amenable to those venues. fects for training mastery-level performance. This
Successfully applying behavior analysis in new may explain the inconsistencies regarding the ef-
venues represents another area for future research. fectiveness of this type of training technology (e.g.,
The change from institutional to community DiGennaro Reed et al., 2010; Neef, Trachtenberg,
living for people with disabilities in the United Loeb, & Sterner, 1991). Research to determine
States is an example of how we need research ways to incorporate performance-based training
on behavioral staff training and management in with new information technologies is warranted.
new venues. Much early staff-training research
occurred in institutional settings. As the commu-
Expanding Adoption of Behavioral Training
nity-living trend became more widespread, a need
and Management
arose for demonstrations of how to improve staff
performance in community settings (Harchik & We have noted the lack of widespread adoption
Campbell, 1998). For example, much early research of the behavioral technology of staff training and
on institutional staffers’ performance involved the management in human-service settings, as have
frequent presence of a supervisor for intervention others (Babcock et al., 1998; LeBlanc, Raetz, & Fe-
implementation. Community-living arrangements liciano, 2011; Reid, 2004). The social significance
do not often include frequent supervisor involve- of the evidence-based approach for staff training
ment with staff. Consequently, a need has arisen and management will remain limited until the
for research on ways to improve staff performance human services use it consistently. The following
without frequent supervisory presence. A similar sections suggest areas for research to aid wider use
situation exists in many home-based behavior- of the existing technology.
analytic early intervention programs for children
with ASD (e.g., Ausenhu & Higgins, 2019).
Developing Personnel Preparation Programs in Staff
Training and Management
Use of New Information Technology in Staff Training
One means of incorporating the staff-training and
and Management
management technology in human-service set-
Technology has revolutionized information dis- tings is to ensure that professionals entering those
semination in recent years. Internet access, train- settings are knowledgeable about the technology
Staff Training and Management 293
and skilled in its application. Traditionally, profes- agement needs to promote supervisors’ continued
sionals have received little or no training in this use of intervention components.
area (Schell, 1998). Indications exist that person- The acceptability of behavioral management
nel preparation programs are directing more atten- procedures is another research area relevant to
tion to behavioral staff training and management. maintaining supervisor performance. Behavior
For example, some university programs for train- analysts have recognized the importance of man-
ing behavior analysts are including courses with agement procedures that are acceptable to super-
content on staff training and management, partly visors for potentially promoting continued use of
because behavior-analyst certification requires supervisory skills (Reid & Whitman, 1983). Cor-
knowledge of these areas (Moore & Shook, 2001). respondingly, several investigators have attempted
Nonetheless, only a few university programs cur- to evaluate acceptable components of supervisory
rently focus on organizational behavior manage- procedures (see Parsons, 1998, for a review). How-
ment, and their emphasis is often on business and ever, as the Parsons (1998) review indicated (see
industry in contrast to human-service settings. also Parsons et al., 2012), researchers have concerns
regarding the validity of typical measures of super-
visory acceptance (i.e., questionnaire responses).
Training Supervisors in Behavioral Staff Training
We need continued research on improving the ac-
and Management
ceptability of management procedures to supervi-
We cannot expect supervisors in the human ser- sors and identifying valid measures of acceptability.
vices to be proficient in behavioral staff training Functional assessment of staff performance is
and management unless they have had training another way of potentially promoting maintenance
in these areas. Few investigations have addressed of appropriate staff behavior. Human-service staff
training supervisors in behavioral applications members have many tasks to perform and many
with staff (e.g., Clark et al., 1985; Fleming, Oliver, contingencies on their work performance. Inves-
& Bolton, 1996; Methot, Williams, Cummings, tigations on improving staffers’ performance often
& Bradshaw, 1996; Parsons & Reid, 1995; Reid impose new contingencies without apparent re-
et al., 2005). Evidence-based curricula for train- gard for existing contingencies. As a result, staff
ing supervisors are likely to aid expansion of su- members resume responding to the more common
pervisor training. The American Association on contingencies when investigators discontinue the
Intellectual and Developmental Disabilities has research contingencies. One way of avoiding this
developed one such curriculum (Reid, Parsons, & obstacle is to identify when supervisors could in-
Green, 2009). troduce new contingencies that would not com-
In research on supervisor training, attention to pete with existing contingencies (Green, Reid,
the results of staff training research as summarized Perkins, & Gardner, 1991). For example, Green et
earlier is warranted. Results of staff-training re- al. (1991) conducted a structural analysis of staffers’
search suggest that supervisor training is not likely work performance to identify periods of nonwork
to affect supervisor behavior unless upper man- behavior. The researchers then increased perfor-
agement prompts and reinforces the supervisors’ mance of selected duties during those nonwork pe-
performance of the trained skills. With few excep- riods. Targeting periods of nonwork allowed staff-
tions (e.g., Gillat & Sulzer-Azaroff, 1994; Methot ers to increase their performance of selected duties
et al., 1996), research addressing the behavior of without competing with their other duties.
senior managers or executives in human-service The Green et al. (1991) investigation also high-
settings is lacking. lights staff performance problems due to an appar-
ent lack of contingencies on certain work duties.
Several investigations have addressed reducing
Maintaining Human‑Service Staff Performance
staff time spent in nonwork activities (Brown,
The need for research on maintaining improve- Willis, & Reid, 1981; Green et al., 1991; Iwata,
ments in staff performance after training and Bailey, Brown, Foshee, & Alpern, 1976). This re-
management interventions is related closely to search assumed that staff spent time in such ac-
the need for research with supervisors and senior tivities due to a lack of supervisor contingencies on
managers. Research to date suggests that improved such behavior or a lack of contingencies on more
staff performance is likely to be maintained only if desired performance.
supervisors implement some relevant components More detailed analysis of contingencies or lack
of the initial interventions. In turn, senior man- of contingencies on staff behavior in human-ser-
294 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
vice settings may produce information relevant feedback plus approval on institutional staff perfor-
for promoting sustained improvements in staff mance. Journal of Organizational Behavior Manage-
performance. Investigators have noted the signifi- ment, 3(1), 57–68.
cance of conducting functional assessments with Burgio, L. D., Whitman, T. L., & Reid, D. H. (1983).
A participative management approach for improving
staff performance (Austin, 2000; Sturmey, 1998),
direct-care staff performance in an institutional set-
and research in this area appears promising. Par- ting. Journal of Applied Behavior Analysis, 16, 37–53.
ticularly, investigators have used the Performance Carr, J. E., Wilder, D. A., Majdalany, L., Mathisen, D., &
Diagnostic Checklist—Human Services to design Strain, L. A. (2013). An assessment-based solution to
interventions to improve problematic staff perfor- a human-service employee performance problem: An
mance, based on systematically assessed variables evaluation of the Performance Diagnostic Check-
related to the lack of desired performance (Carr et list—Human Services. Behavior Analysis in Practice,
al., 2013; Ditzian, Wilder, King, & Tanz, 2015). We 6, 16–32.
support the continuation of this line of research Casey, A. M., & McWilliam, R. A. (2011). The impact
of checklist-based training on teachers’ use of the
and offer it as an important means of furthering
zone defense schedule. Journal of Applied Behavior
the contribution of applied behavior analysis to Analysis, 44, 397–401.
promoting high-quality staff performance. Catania, C. N., Almeida, D., Liu-Constant, B., & Di-
Gennaro Reed, F. D. (2009). Video modeling to train
staff to implement discrete-trial instruction. Journal
REFERENCES of Applied Behavior Analysis, 42, 387–392.
Chok, J. T., Shlesinger, A., Studer, L., & Bird, F. L.
Adkins, V. K. (1996). Discussion: Behavioral procedures (2012). Description of a practitioner training pro-
for training direct care staff in facilities serving de- gram on functional analysis and treatment develop-
pendent populations. Behavioral Interventions, 11, ment. Behavior Analysis in Practice, 5, 25–36.
95–100. Christian, W. P. (1983). A case study in the program-
Alavosius, M. P., & Sulzer-Azaroff, B. (1990). Acquisi- ming and maintenance of institutional change. Jour-
tion and maintenance of health-care routines as a nal of Organizational Behavior Management, 5(3–4),
function of feedback density. Journal of Applied Be- 99–153.
havior Analysis, 23, 151–162. Clark, H. B., Wood, R., Kuehnel, T., Flanagan, S., Mosk,
Alvero, A. M., Bucklin, B. R., & Austin, J. (2001). An M., & Northup, J. T. (1985). Preliminary validation
objective review of the effectiveness and essential and training of supervisory interaction skills. Jour-
characteristics of performance feedback in organi- nal of Organizational Behavior Management, 7(1–2),
zational settings. Journal of Organizational Behavior 95–115.
Management, 21(1), 3–29. Clayton, M., & Headley, A. (2019). The use of behav-
Ausenhus, J. A., & Higgins, W. J. (2019). An evalua- ioral skills training to improve staff performance of
tion of real-time feedback delivered via telemedicine: discrete trial training. Behavioral Intervention, 34,
Training staff to conduct preference assessments. Be- 136–143.
havior Analysis in Practice, 12, 643–648. Cotnoir-Bichelman, N. M., Thompson, R. H., McKer-
Austin, J. (2000). Some thoughts on the field of organi- char, P. M., & Haremza, J. L. (2006). Training stu-
zational behavior management. Journal of Organiza- dent teachers to reposition infants frequently. Journal
tional Behavior Management, 20(3–4), 191–202. of Applied Behavior Analysis, 39, 489–494.
Babcock, R. A., Fleming, R. K., & Oliver, J. R. (1998). Demchak, M. A. (1987). A review of behavioral staff
OBM and quality improvement systems. Journal of training in special education settings. Education and
Organizational Behavior Management, 21(1), 33–59. Training in Mental Retardation, 22, 205–217.
Balcazar, F., Hopkins, B. L., & Suarez, Y. (1986). A criti- DiGennaro Reed, F. D., Codding, R., Catania, C. N.,
cal, objective review of performance feedback. Jour- & Maguire, H. (2010). Effects of video modeling on
nal of Organizational Behavior Management, 7(3–4), treatment integrity of behavioral interventions. Jour-
65–89. nal of Applied Behavior Analysis, 43, 291–295.
Brackett, L., Reid, D. H., & Green, C. W. (2007). Effects Ditzian, K., Wilder, D. A., King, A., & Tanz, J. (2015).
of reactivity to observations on staff performance. An evaluation of the Performance Diagnostic
Journal of Applied Behavior Analysis, 40, 191–195. Checklist—Human Services to assess an employee
Bricker, W. A., Morgan, D. G., & Grabowski, J. G. performance problem in a center-based autism treat-
(1972). Development and maintenance of a behav- ment facility. Journal of Applied Behavior Analysis, 48,
ior modification repertoire of cottage attendants 199–203.
through TV feedback. American Journal of Mental Doerner, M., Miltenberger, R. G., & Bakken, J. (1989).
Deficiency, 77, 128–136. The effects of staff self-management on positive so-
Brown, K. M., Willis, B. S., & Reid, D. H. (1981). Dif- cial interactions in a group home setting. Behavioral
ferential effects of supervisor verbal feedback and Residential Treatment, 4, 313–330.
Staff Training and Management 295
Downs, A., Downs, R. C., & Rau, K. (2008). Effects of Harchik, A. E., & Campbell, A. R. (1998). Support-
training and feedback on discrete trial teaching skills ing people with developmental disabilities in their
and student performance. Research in Developmental homes in the community: The role of organizational
Disabilities, 29, 235–246. behavior management. Journal of Organizational Be-
Dyer, K., Schwartz, I. S., & Luce, S. C. (1984). A su- havior Management, 18(2–3), 83–101.
pervision program for increasing functional activi- Harchik, A. E., Sherman, J. A., Hopkins, B. L., Strouse,
ties for severely handicapped students in a residen- M. C., & Sheldon, J. B. (1989). Use of behavioral
tial setting. Journal of Applied Behavior Analysis, 17, techniques by paraprofessional staff: A review and
249–259. proposal. Behavioral Residential Treatment, 4, 331–
Fielding, D. L., & Blasé, K. A. (1993). Creating new 357.
realities: Program development and dissemination. Harchik, A. E., Sherman, J. A., Sheldon, J. B., & Strou-
Journal of Applied Behavior Analysis, 26, 597–613. se, M. C. (1992). Ongoing consultation as a method
Fielding, L. T., Errickson, E., & Bettin, B. (1971). Modi- of improving performance of staff members in a
fication of staff behavior: A brief note. Behavior Ther- group home. Journal of Applied Behavior Analysis, 25,
apy, 2, 550–553. 599–610.
Fleming, R. K., Oliver, J. R., & Bolton, D. (1996). Hawkins, A. M., Burgio, L. D., Langford, A., & Engel,
Training supervisors to train staff: A case study in a B. T. (1992). The effects of verbal and written super-
human service organization. Journal of Organizational visory feedback on staff compliance with assigned
Behavior Management, 16(1), 3–25. prompted voiding in a nursing home. Journal of Or-
Ford, J. E. (1980). A classification system for feedback ganizational Behavior Management, 13(1), 137–150.
procedures. Journal of Organizational Behavior Man- Hogan, A., Knez, N., & Kahng, S. (2015). Evaluat-
agement, 2(3), 183–191. ing the use of behavioral skills training to improve
Gardner, J. M. (1970). Effects of reinforcement condi- school staffs’ implementation of behavior interven-
tions on lateness and absence among institutional tion plans. Journal of Behavioral Education, 24, 242–
personnel. Ohio Research Quarterly, 3, 315–316. 254.
Gardner, J. M. (1972). Teaching behavior modification Hollander, M., Plutchik, R., & Horner, V. (1973). Inter-
to nonprofessionals. Journal of Applied Behavior Anal- action of patient and attendant reinforcement pro-
ysis, 5, 517–521. grams: The “piggyback” effect. Journal of Consulting
Gillat, A., & Sulzer-Azaroff, B. (1994). Promoting prin- and Clinical Psychology, 41, 43–47.
cipals’ managerial involvement in instructional im- Ivancic, M. T., & Helsel, W. J. (1998). Organizational
provement. Journal of Applied Behavior Analysis, 27, behavior management in large residential organiza-
115–129. tions: Moving from institutional to client-centered
Gladstone, B. W., & Spencer, C. J. (1977). The effects of care. Journal of Organizational Behavior Management,
modeling on the contingent praise of mental retarda- 18(2–3), 61–82.
tion counselors. Journal of Applied Behavior Analysis, Iwata, B. A., Bailey, J. S., Brown, K. M., Foshee, T. J., &
10, 75–84. Alpern, M. (1976). A performance-based lottery to
Graff, R. B., & Karsten, A. M. (2012). Evaluation of improve residential care and training by institutional
a self-instruction package for conducting stimulus staff. Journal of Applied Behavior Analysis, 9, 417–431.
preference assessments. Journal of Applied Behavior Jahr, E. (1998). Current issues in staff training. Research
Analysis, 45, 69–82. in Developmental Disabilities, 19, 73–87.
Gravina, N., Villacorta, J., Albert, K., Clark, R., Curry, Johnson, M. D., & Fawcett, S. B. (1994). Courteous ser-
S., & Wilder, D. (2018). A literature review of orga- vice: Its assessment and modification in a human ser-
nizational behavior management interventions in vice organization. Journal of Applied Behavior Analy-
human service settings from 1990 to 2016. Journal of sis, 27, 145–152.
Organizational Behavior Management, 38, 191–224. Katz, R. C., Johnson, C. A., & Gelfand, S. (1972).
Green, C. W., Reid, D. H., Perkins, L. I., & Gardner, S. Modifying the dispensing of reinforcers: Some impli-
M. (1991). Increasing habilitative services for persons cations for behavior modification with hospitalized
with profound handicaps: An application of struc- patients. Behavior Therapy, 3, 579–588.
tural analysis to staff management. Journal of Applied Kissel, R. C., Whitman, T. L., & Reid, D. H. (1983).
Behavior Analysis, 24, 459–471. An institutional staff training and self-management
Green, C. W., Rollyson, J. H., Passante, S. C., & Reid, program for developing multiple self-care skills in se-
D. H. (2002). Maintaining proficient supervisor per- verely/profoundly retarded individuals. Journal of Ap-
formance with direct support personnel: An analysis plied Behavior Analysis, 16, 395–415.
of two management approaches. Journal of Applied Kneringer, M., & Page, T. J. (1999). Improving staff nu-
Behavior Analysis, 35, 205–208. tritional practices in community-based group homes:
Greene, B. F., Willis, B. S., Levy, R., & Bailey, J. S. Evaluation, training, and management. Journal of
(1978). Measuring client gains from staff-implement- Applied Behavior Analysis, 32, 221–224.
ed programs. Journal of Applied Behavior Analysis, 11, Koegel, R. L., Russo, D. C., & Rincover, A. (1977). As-
395–412. sessing and training teachers in the generalized use of
296 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
behavior modification with autistic children. Journal ture directions. Journal of Applied Behavior Analysis,
of Applied Behavior Analysis, 10, 197–205. 28, 297–299.
Lavie, T., & Sturmey, P. (2002). Training staff to con- Neef, N. A., Trachtenberg, S., Loeb, J., & Sterner, K.
duct a paired-stimulus preference assessment. Journal (1991). Video-based training of respite care provid-
of Applied Behavior Analysis, 35, 209–211. ers: An interactional analysis of presentation format.
LeBlanc, L. A., Raetz, P. B., & Feliciano, L. (2011). Be- Journal of Applied Behavior Analysis, 24, 473–486.
havioral gerontology. In W. W. Fisher, C. C. Piazza, Parsons, M. B. (1998). A review of procedural accept-
& H. S. Roane (Eds.), Handbook of applied behavior ability in organizational behavior management. Jour-
analysis (pp. 472–486). New York: Guilford Press. nal of Organizational Behavior Management, 18(2–3),
Lerman, D. C. (2009). An introduction to the Volume 173–190.
2, Number 2 of Behavior Analysis in Practice (BAP). Parsons, M. B., & Reid, D. H. (1995). Training residen-
Behavior Analysis in Practice, 2, 2–3. tial supervisors to provide feedback for maintaining
Liberman, R. P. (1983). Guest editor’s preface. Analysis staff teaching skills with people who have severe
and Intervention in Developmental Disabilities, 3, iii– disabilities. Journal of Applied Behavior Analysis, 28,
iv. 317–322.
Luck, K. M., Lerman, D. C., Wu, W. L., Dupuis, D. L., Parsons, M. B., Rollyson, J. H., & Reid, D. H. (2004).
& Hussein, L. A. (2018). A comparison of written, Improving day-treatment services for adults with se-
vocal, and video feedback when training teachers. vere disabilities: A norm-referenced application of
Journal of Behavioral Education, 27, 124–144. outcome management. Journal of Applied Behavior
Macurik, K. M., O’Kane, N. P., Malanga, P., & Reid, D. Analysis, 37, 365–377.
H. (2008). Video training of support staff in interven- Parsons, M. B., Rollyson, J. H., & Reid, D. H. (2012).
tion plans for challenging behavior: Comparison with Evidence-based staff training: A guide for practitio-
live training. Behavioral Interventions, 23, 143–163. ners. Behavior Analysis in Practice, 5, 2–11.
Mayhew, G. L., Enyart, P., & Cone, J. D. (1979). Ap- Parsons, M. B., Schepis, M. M., Reid, D. H., McCarn,
proaches to employee management: Policies and pref- J. E., & Green, C. W. (1987). Expanding the impact
erences. Journal of Organizational Behavior Manage- of behavioral staff management: A large-scale, long
ment, 2(2), 103–111. term application in schools serving severely handi-
Methot, L. L., Williams, W. L., Cummings, A., & Brad- capped students. Journal of Applied Behavior Analysis,
shaw, B. (1996). Measuring the effects of a manager– 20, 139–150.
supervisor training program through the generalized Petscher, E. S., & Bailey, J. S. (2006). Effects of training,
performance of managers, supervisors, front-line prompting, and self-monitoring on staff behavior in
staff, and clients in a human service setting. Journal a classroom for students with disabilities. Journal of
of Organizational Behavior Management, 16(2), 3–34. Applied Behavior Analysis, 39, 215–226.
Miles, N. I., & Wilder, D. A. (2009). The effects of be- Phillips, J. F. (1998). Applications and contributions of
havioral skills training on caregiver implementation organizational behavior management in schools and
of guided compliance. Journal of Applied Behavior day treatment settings. Journal of Organizational Be-
Analysis, 42, 405–410. havior Management, 18(2–3), 103–129.
Moore, J. W., Edwards, R. P., Sterling-Turner, H. E., Plavnick, J. B., Ferreri, S. J., & Maupin, A. N. (2010).
Riley, J., DuBard, M., & McGeorge, A. (2002). The effects of self-monitoring on the procedural in-
Teacher acquisition of functional analysis methodol- tegrity of a behavioral intervention for young chil-
ogy. Journal of Applied Behavior Analysis, 35, 73–77. dren with developmental disabilities. Journal of Ap-
Moore, J. W., & Fisher, W. W. (2007). The effects of plied Behavior Analysis, 43, 315–320.
videotape modeling on staff acquisition of functional Pollack, M. J., Fleming, R. K., & Sulzer-Azaroff, B.
analysis methodology. Journal of Applied Behavior (1994). Enhancing professional performance through
Analysis, 40, 197–202. organizational change. Behavioral Interventions, 9,
Moore, J., & Shook, G. L. (2001). Certification, accredi- 27–42.
tation, and quality control in behavior analysis. Be- Quilitch, H. R. (1975). A comparison of three staff-
havior Analyst, 24, 45–55. management procedures. Journal of Applied Behavior
Mozingo, D. B., Smith, T., Riordan, M. R., Reiss, M. L., Analysis, 8, 59–66.
& Bailey, J. S. (2006). Enhancing frequency record- Realon, R. E., Lewallen, J. D., & Wheeler, A. J. (1983).
ing by developmental disabilities treatment staff. Verbal feedback vs. verbal feedback plus praise: The
Journal of Applied Behavior Analysis, 39, 253–256. effects on direct care staff’s training behaviors. Men-
Nabeyama, B., & Sturmey, P. (2010). Using behav- tal Retardation, 21, 209–212.
ioral skills training to promote safe and correct staff Reid, D. H. (2004). Training and supervising direct sup-
guarding and ambulation distance of students with port personnel to carry out behavioral procedures. In
multiple physical disabilities. Journal of Applied Be- J. L. Matson, R. B. Laud, & M. L. Matson (Eds.), Be-
havior Analysis, 43, 341–345. havior modification for persons with developmental dis-
Neef, N. A. (1995). Research on training trainers in abilities: Treatments and supports (pp. 73–99). Kings-
program implementation: An introduction and fu- ton, NY: NADD Press.
Staff Training and Management 297
Reid, D. H. (2016). Promoting happiness among adults Reid, D. H., & Whitman, T. L. (1983). Behavioral staff
with autism and other severe disabilities: Evidence-based management in institutions: A critical review of ef-
strategies. Morganton, NC: Habilitative Management fectiveness and acceptability. Analysis and Interven-
Consultants. tion in Developmental Disabilities, 3, 131–149.
Reid, D. H. (2017). Competency-based staff training. In Repp, A. C., & Deitz, D. E. D. (1979). Improving admin-
J. K. Luiselli (Ed.), Applied behavior analysis advanced istrative-related staff behaviors at a state institution.
guidebook: A manual for professional practice (pp. 21– Mental Retardation, 17, 185–192.
40). San Diego, CA: Elsevier/Academic Press. Richman, G. S., Riordan, M. R., Reiss, M. L., Pyles, D.
Reid, D. H., Green, C. W., & Parsons, M. B. (2003). A., & Bailey, J. S. (1988). The effects of self-moni-
An outcome management program for extending toring and supervisor feedback on staff performance
advances in choice research into choice opportuni- in a residential setting. Journal of Applied Behavior
ties for supported workers with severe multiple dis- Analysis, 21, 401–409.
abilities. Journal of Applied Behavior Analysis, 36, Rosales, R., Gongola, L., & Homlitas, C. (2015). An
575–578. evaluation of video modeling with embedded instruc-
Reid, D. H., & Parsons, M. B. (2002). Working with staff tions to teach implementation of stimulus preference
to overcome challenging behavior among people who assessments. Journal of Applied Behavior Analysis, 48,
have severe disabilities: A guide for getting support plans 209–214.
carried out. Morganton, NC: Habilitative Manage- Roscoe, E. M., Fisher, W. W., Glover, A. C., & Volk-
ment Consultants. ert, V. M. (2006). Evaluating the relative effects of
Reid, D. H., Parsons, M. B., & Green, C. W. (1989). feedback and contingent money for staff training of
Staff management in human services: Behavioral re- stimulus preference assessments. Journal of Applied
search and application. Springfield, IL: Charles C Behavior Analysis, 39, 63–77.
Thomas. Sarokoff, R. A., & Sturmey, P. (2004). The effects of
Reid, D. H., Parsons, M. B., & Green, C. W. (2009). The behavioral skills training on staff implementation
supervisor training curriculum: Evidence-based ways of discrete-trial teaching. Journal of Applied Behavior
to promote work quality and enjoyment among support Analysis, 37, 535–538.
staff. Washington, DC: American Association on In- Schell, R. M. (1998). Organizational behavior manage-
tellectual and Developmental Disabilities. ment: Applications with professional staff. Journal
Reid, D. H., Parsons, M. B., & Green, C. W. (2012). The of Organizational Behavior Management, 18(2–3),
supervisor’s guidebook: Evidence-based strategies for 157–171.
promoting work quality and enjoyment among human Schnell, L. K., Sidener, T. M., DeBar, R. M., Vladescu,
service staff. Morganton, NC: Habilitative Manage- J. C., & Kahng, S. (2018). Effects of computer-based
ment Consultants. training on procedural modifications to standard
Reid, D. H., Parsons, M. B., & Jensen, J. M. (2017). functional analyses. Journal of Applied Behavior Anal-
Maintaining staff performance following a training ysis, 51, 87–98.
intervention: Suggestions from a 30-year case exam- Scott, J., Lerman, D. C., & Luck, K. (2018). Comput-
ple. Behavior Analysis in Practice, 10, 12–21. er-based training to detect antecedents and conse-
Reid, D. H., Parsons, M. B., Lattimore, L. P., Towery, quences of problem behavior. Journal of Applied Be-
D. L., & Reade, K. K. (2005). Improving staff per- havior Analysis, 51, 784–801.
formance through clinician application of outcome Severtson, J. M., & Carr, J. E. (2012). Training nov-
management. Research in Developmental Disabilities, ice instructors to implement errorless discrete-trial
26, 101–116. teaching: A sequential analysis. Behavior Analysis in
Reid, D. H., Parsons, M. B., McCarn, J. M., Green, C. Practice, 5, 13–23.
W., Phillips, J. F., & Schepis, M. M. (1985). Provid- Shoemaker, J., & Reid, D. H. (1980). Decreasing chron-
ing a more appropriate education for severely handi- ic absenteeism among institutional staff: Effects of
capped persons: Increasing and validating functional a low-cost attendance program. Journal of Organiza-
classroom tasks. Journal of Applied Behavior Analysis, tional Behavior Management, 2(4), 317–328.
18, 289–301. Smith, T., Parker, T., Taubman, M., & Lovaas, O. I.
Reid, D. H., Rotholz, D. A., Parsons, M. B., Morris, L., (1992). Transfer of staff training from workshops to
Braswell, B. A., Green, C. W., et al. (2003). Training group homes: A failure to generalize across settings.
human service supervisors in aspects of PBS: Evalu- Research in Developmental Disabilities, 13, 57–71.
ation of a statewide, performance-based program. Sneed, T. J., & Bible, G. H. (1979). An administrative
Journal of Positive Behavior Interventions, 5, 35–46. procedure for improving staff performance in an in-
Reid, D. H., & Shoemaker, J. (1984). Behavioral super- stitutional setting for retarded persons. Mental Retar-
vision: Methods of improving institutional staff per- dation, 17, 92–94.
formance. In W. P. Christian, G. T. Hannah, & T. J. Strouse, M. C., Carroll-Hernandez, T. A., Sherman, J.
Glahn (Eds.), Programming effective human services: A., & Sheldon, J. B. (2003). Turning over turnover:
Strategies for institutional change and client transition The evaluation of a staff scheduling system in a com-
(pp. 39–61). New York: Plenum Press. munity-based program for adults with developmental
298 In t e r v e n t i o n s f o r In c r e a s i n g D e s i r ab l e B e h av i o r
disabilities. Journal of Organizational Behavior Man- ing behavior modification skills to institutional staff.
agement, 23(2–3), 45–63. Applied Research in Mental Retardation, 1, 41–53.
Sturmey, P. (1998). History and contribution of organi- Wehman, P. (2011). Employment for persons with dis-
zational behavior management to services for persons abilities: Where are we now and where do we need
with developmental disabilities. Journal of Organiza- to go? Journal of Vocational Rehabilitation, 35, 145–
tional Behavior Management, 18(2–3), 7–32. 151.
Suda, K. T., & Miltenberger, R. G. (1993). Evaluation Weldy, C. R., Rapp, J. T., & Capocasa, K. (2014). Train-
of staff management strategies to increase positive ing staff to implement brief stimulus preference as-
interactions in a vocational setting. Behavioral Resi- sessments. Journal of Applied Behavior Analysis, 47,
dential Treatment, 8, 69–88. 214–218.
van Oorsouw, W. M., Embregts, P. J., Bosman, A. M., & Welsch, W. V., Ludwig, C., Radiker, J. E., & Krapfl, J.
Jahoda, A. (2009). Training staff serving clients with E. (1973). Effects of feedback on daily completion of
intellectual disabilities: A meta-analysis of aspects behavior modification projects. Mental Retardation,
determining effectiveness. Research in Developmental 11, 24–26.
Disabilities, 30, 503–511. Whitman, T. L., Hantula, D. A., & Spence, B. H.
Wallace, C. J., Davis, J. R., Liberman, R. P., & Baker, V. (1990). Current issues in behavior modification with
(1973). Modeling and staff behavior. Journal of Con- mentally retarded persons. In J. L. Matson (Ed.),
sulting and Clinical Psychology, 41, 422–425. Handbook of behavior modification with the mentally
Wallace, M. D., Doney, J. K., Mintz-Resudek, C. M., & retarded (pp. 9–50). New York: Plenum Press.
Tarbox, R. S. F. (2004). Training educators to imple- Williams, W. L., Vittorio, T. D., & Hausherr, L. (2002).
ment functional analyses. Journal of Applied Behavior A description and extension of a human services
Analysis, 37, 89–92. management model. Journal of Organizational Behav-
Watson, L. S., & Uzzell, R. (1980). A program for teach- ior Management, 22(1), 47–71.
PAR T VI
INTERVENTIONS
FOR DECREASING
PROBLEM BEHAVIOR
T he chapters and their order in Part VI remain the same in this edition of the
handbook as in the first edition. In Chapter 18, Smith introduces the reader
to the term antecedent stimuli and describes the many ways these stimuli affect
behavior, their role in functional analysis, and the mechanisms by which they
change behavior. The effects of motivating operations, which are one type of
antecedent stimuli, constitute a recurring theme in the subsequent chapters of
Part VI; thus Smith’s chapter sets the stage for new research that the authors of
subsequent chapters discuss. An exciting addition to Chapter 18 in this edition
is a review of novel research on the effects of environmental enrichment on the
brain and on human behavior. In Chapter 19, Vollmer, Athens, and Fernand
use the procedure that Iwata et al. described in their classic study on functional
analysis as the framework for their discussion of function-based extinction proce-
dures. The chapter includes reviews of classic and more recent studies that have
incorporated extinction as treatment for problem behavior. The latter half of the
chapter is a review of the response patterns associated with extinction and of why
a sound understanding of these patterns is important in clinical practice.
Fisher, Greer, and Bouxsein have updated Chapter 20 on developing
function-based positive reinforcement procedures with a focus on new research
on functional-communication training. They review studies on the effects of the
duration of exposure to the establishing operation and on the effects of multiple
schedules on reinforcement schedule thinning, transfer of treatment effects,
and relapse. Lerman and Toole note in Chapter 21 that the advent of function
analysis has improved the effectiveness of treatments for problem behavior. Thus
research on punishment has not grown substantially in recent years outside of
299
300 In t e r v e n t i o n s f o r D e c r e a s i n g D e s i r ab l e B e h av i o r
that on response interruption and redirection, which Lerman and Toole review.
As with punishment, investigators conducted the bulk of research on the token
economy in the 1960s and 1970s. Reitman, Boerke, and Vassilopoulos review
that research in Chapter 22. They also introduce the reader to studies in which
researchers have combined tokens with medication for children with attention-
deficit/hyperactivity disorder, and have evaluated children’s preference for
programs in which children earn versus lose tokens. In sum, readers will find a
plethora of useful information in Part VI that they can apply in clinical practice.
CHAP TER 18
Richard G. Smith
Behavioral interventions to treat problem behav- tivating operations for problem behavior (Fisher
ior typically involve the manipulation or manage- et al., 2018; Laraway, Snycerski, Michael, & Pol-
ment of some environmental event or condition, ing, 2003), whereas others may alter discrimina-
with the intended result to eliminate or reduce tive functions. Motivating operations temporarily
problem behavior. We use the term antecedent in- alter the effectiveness of consequences and the
tervention when we manipulate the events or con- momentary probability of behavior that has pro-
ditions that occur before the behavior. duced those consequences in the past (Laraway et
We can broadly classify antecedent interven- al., 2003). For example, NCR involves repeated,
tions into two categories. Default interventions response-independent presentation of functional
do not depend on identification of the variables reinforcers; therefore, NCR may decrease problem
that set the occasion for and maintain the prob- behavior by abolishing the reinforcing effective-
lem behavior. Default interventions can be effec- ness of those consequences via satiation or habitu-
tive for problem behaviors maintained by a range ation (Murphy, McSweeney, Smith, & McComas,
of reinforcers. Examples are antecedent exercise, 2003). Thus we would classify NCR as a motivat-
environmental enrichment, protective equipment, ing-operation-based procedure. Other antecedent
and restraint. By contrast, function-based inter- interventions may manage conditions or stimuli
ventions involve identifying the antecedents and associated with differential consequences for prob-
consequences that maintain problem behavior, lem behavior, thus altering discriminative control
and then directly manipulating at least one com- over the behavior. For example, problem behavior
ponent of that operant contingency. We use the will decrease in contexts in which it fails to pro-
term functional reinforcer for the consequence that duce functional reinforcers, and the context will
maintains problem behavior (Saini, Fisher, Ret- become an S-delta.1 Finally, some interventions
zlaff, & Keevy, 2020). For example, noncontingent
reinforcement (NCR; Vollmer, 1999) is a function-
1 Interventions based on discriminative control require ma-
based intervention in which the behavior analyst
schedules delivery of the functional reinforcer on a nipulation of both antecedent and consequent events to
establish and maintain their effectiveness; thus we do not
time-based, response-independent schedule. properly characterize them as exclusively antecedent inter-
We can further classify antecedent interven- ventions (i.e., the behavior analyst must actively control
tions according to the mechanism by which they consequences to maintain the effectiveness of the anteced-
decrease behavior. Some interventions affect mo- ent stimulus).
301
302 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
arrange the environment so that problem behavior observations during or within a few minutes after
is difficult or impossible to emit. Examples include exercise completion. Researchers have used ante-
protective equipment and mechanical restraint. cedent exercise to decrease self-injurious behavior
In the following sections, I describe anteced- (SIB; Baumeister & MacLean, 1984), aggression
ent behavioral interventions designed to reduce or (McGimsey & Favell, 1988; Powers et al., 1992),
eliminate behavior. I present a brief review of the inappropriate vocalizations (Bachman & Fuqua,
literature for each intervention and discuss proce- 1983; Powers et al., 1992), off-task behavior (Bach-
dural variations, functional properties (i.e., the be- man & Fuqua, 1983), out-of-seat behavior (Celib-
havioral principles that describe their effects), and erti et al., 1997), and stereotypy (Celiberti et al.,
strengths and limitations, starting with default 1997; Kern et al., 1984; Powers et al., 1992) exhib-
interventions and following with function-based ited by persons with developmental disabilities.
interventions. Researchers have also used antecedent exercise to
treat depression, with varying degrees of reported
success (Doyne, Chambless, & Beutler, 1983);
DEFAULT INTERVENTIONS panic disorder (Broocks et al., 1998); and pain dis-
order (Turner & Clancy, 1988) in persons without
Some antecedent procedures will decrease behav- developmental disabilities.
ior, regardless of the operant function of problem Antecedent exercise differs from other inter-
behavior. That is, identification of the functional ventions involving effortful activity such as over-
reinforcer is not necessary when using these inter- correction (Foxx & Azrin, 1972, 1973) because
ventions. As a result, we may not understand the participants engage in it independently of occur-
precise mechanisms associated with these proce- rences of the problem behavior, typically before
dures’ effects, compared to our understanding of observation sessions. By contrast, overcorrection
procedures that correspond with a reinforcement prescribes effortful activities contingent on prob-
contingency. Although prescribing intervention lem behavior. Unlike overcorrection, the process
without requiring a prior functional analysis seems of punishment cannot account for the effective-
clinically expedient, default procedures have sig- ness of antecedent exercise in decreasing problem
nificant limitations and produce side effects. behavior, because we do not present it as a conse-
We use the term default interventions for inter- quence for problem behavior.
ventions whose effects do not depend on the oper-
ant function of problem behavior. The term default
Mechanisms Underlying the Effects
does not mean that problem behavior or the in-
of Antecedent Exercise
tervention effects are not operant. Like function-
based interventions, some default interventions Although the production of a general state of fa-
may alter discriminative stimuli (Sds) or motivat- tigue is an intuitively appealing account of the
ing operations for problem behavior. However, effects of antecedent exercise, research outcomes
identifying the contingency maintaining problem seem inconsistent with this interpretation. For ex-
behavior is not a necessary component of a poten- ample, increases in on-task behavior (Powers et al.,
tially effective default intervention. 1992), increases in appropriate responding (Kern,
Koegel, Dyer, Blew, & Fenton, 1982), and the ab-
sence of overt signs of fatigue (Baumeister & Ma-
Antecedent Exercise
cLean, 1984) after bouts of exercise indicate that
Research has shown that antecedent exercise can the response-decreasing effects of this interven-
decrease problem behavior (e.g., Allison, Basile, tion are at least somewhat specific to problem be-
& MacDonald, 1991; Bachman & Fuqua, 1983; havior. In fact, researchers have reported increases
Baumeister & MacLean, 1984; Celiberti, Bobo, in several forms of appropriate behavior after ante-
Kelly, Harris, & Handleman, 1997; Kern, Koe- cedent exercise in several studies (e.g., Baumeister
gel, & Dunlap, 1984; Lochbaum & Crews, 2003, & MacLean, 1984; Celiberti et al., 1997; Kern et
McGimsey & Favell, 1988; Powers, Thibadeau, & al., 1982; Powers et al., 1992). These outcomes are
Rose, 1992). Antecedent exercise engages partici- somewhat paradoxical. This specificity of action
pants in a program of effortful activities such as appears to correlate most clearly with the social
aerobic exercise (e.g., jogging, walking, dancing, acceptability of the classes of behavior affected:
roller skating) or strength training (e.g., weight Problematic behavior decreases and socially ac-
lifting). The behavior analyst typically conducts ceptable behavior increases after exercise.
Developing Antecedent Interventions 303
Antecedent exercise may alter the reinforcing tracked problem behavior over entire school days
effectiveness of the consequences that maintain and showed that the effects of antecedent exercise
problem behavior (i.e., a motivating-operations can persist for several hours (Cannella-Malone,
effect) (Smith & Iwata, 1997). Some research- Tullis, & Kazee, 2011). These inconsistent results
ers have suggested that antecedent exercise may indicate that we need additional research to deter-
function as matched stimulation, in which exercise mine what alters the durability of the effects of an-
produces free access to stimulation like that pro- tecedent exercise on problem behavior. Another
duced by problem behavior, presumably function- potential limitation of antecedent exercise is that
ing as an abolishing operation for the maintaining it may be inconvenient or impossible to implement
reinforcer. For example, Morrison, Roscoe, and in some situations (e.g., during academic instruc-
Atwell (2011) observed decreases in automatically tion).
reinforced problem behavior both during and after
antecedent exercise for three of four participants,
Enriched Environment
suggesting that exercise devalued the automatical-
ly reinforcing consequences of problem behavior. Another way of arranging antecedent condi-
However, few studies have directly investigated the tions to decrease problem behavior is to provide
mechanisms underlying the effects of antecedent a stimulus-enriched environment. Environmen-
exercise. Behavior analysts tend to embrace an tal enrichment involves making preferred items,
abolishing-operation account due to conceptual toys, educational materials, leisure and recreation
inconsistencies associated with a stimulus control items, activities, social interaction, or a combina-
(or discrimination) account of this effect. How- tion available on a continuous, response-indepen-
ever, we should consider a motivating-operations dent schedule. Several studies have demonstrated
account tentative until more definitive evidence that environmental enrichment can be an effec-
about the behavioral mechanisms underlying tive intervention for SIB and stereotypic problem
the effectiveness of antecedent exercise becomes behavior (Berkson & Davenport, 1962; Berkson
available. & Mason, 1963, 1965; Cuvo, May, & Post, 2001;
Horner, 1980; Rapp, 2006; Ringdahl, Vollmer,
Marcus, & Roane, 1997; Saini et al., 2016). Al-
Strengths and Limitations of Antecedent Exercise
though environmental enrichment may reduce
Strengths of antecedent exercise are that it de- problem behavior maintained by social reinforce-
creases problem behavior, increases appropriate ment, much of the literature about it has focused
behavior, and improves both physical and psy- on stereotypic problem behavior, suggesting that
chological health. Exercise programs have obvi- it may be most appropriate to treat automatically
ous physiological and medical benefits, including reinforced behavior.
improved cardiovascular fitness, muscle tone, and
adaptive skills. Results of some studies suggest that
Mechanisms Underlying the Effects
antecedent exercise may decrease depression and
of Environmental Enrichment
anxiety and improve measures of general psycho-
logical health (Lochbaum & Crews, 2003). At least two reasonable accounts exist for the
A limitation of antecedent exercise is that its reductive effects of environmental enrichment.
effects appear to be temporary, limited to a brief First, environmental enrichment may involve
period immediately following the exercise. Most competition between behavior allocated toward
studies have analyzed only the short-term effects the enriching stimuli and problem behavior. That
of exercise, often during or just after bouts of ex- is, environmental enrichment may reduce problem
ercise; however, the results of more temporally ex- behavior indirectly by providing alternative, com-
tended analyses suggest that the effects of exercise peting sources of reinforcement. Results of studies
on problem behavior may be transient (Bachman indicate that environmental enrichment is more
& Fuqua, 1983; Mays, 2013). For example, two of effective when researchers use highly preferred
four participants in one study showed large and im- versus less preferred stimuli (e.g., Vollmer, Marcus,
mediate decreases in problem behavior after vigor- & LeBlanc, 1994). This finding is consistent with
ous exercise; however, these results waned across the notion that highly preferred stimuli compete
15-minute observation sessions immediately, 1 effectively with the consequences maintaining
hour, and 2 hours after exercise periods (Bach- problem behavior. One study showed that rotating
man & Fuqua, 1983). By contrast, a recent study sets of noncontingently available stimuli produced
304 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
more durable decreases in problem behavior than intervention for problem behavior may involve
did continuous availability of one set of stimuli both operant conditioning and basic neurological
(DeLeon, Anders, Rodriguez-Catter, & Neidert, mechanisms; however, direct evidence for the con-
2000). These results suggest that the participant tributions of neurological changes for these effects
engaged in problem behavior to produce automatic does not currently exist.
reinforcement when the effectiveness of alterna-
tive reinforcement sources waned, due to repeated
Strengths and Limitations
or extended contact with the single-stimulus set.
of Environmental Enrichment
Rotating alternative stimuli apparently main-
tained the relative effectiveness of those stimuli There are several clear benefits to the use of envi-
and more effectively decreased problem behavior. ronmental enrichment in practice. First, environ-
On the other hand, environmental enrichment mental enrichment is simple, straightforward, eas-
may function as an abolishing operation and may ily implemented, and cost-effective. It appears to
reduce problem behavior if its consequences and be especially effective to decrease stereotypic and
those of the environment-enriching materials are automatically reinforced behavior, as it provides
similar. In this case, we would consider environ- either a source of competing reinforcement or an
mental enrichment a function-based intervention, alternative means to access reinforcement similar
because its effectiveness would depend on a func- to that maintaining problem behavior. Providing
tional match or a relation of substitutability between a wide array of alternative forms of stimulation
the reinforcers that environmental enrichment makes sense when the operant function of problem
produces and those that maintain problem behav- behavior is unclear or difficult to assess directly.
ior. The effectiveness of the reinforcer for problem In such cases, the probability that a stimulus will
behavior may be reduced temporarily or abolished function to compete effectively with or replace the
through satiation or habituation, because environ- reinforcer that problem behavior produces can be
mental enrichment produces the same or similar increased simply by providing many alternatives.
reinforcement (Murphy et al., 2003). Research We can further increase the probability of find-
indicating that antecedent availability of stimuli ing an effective alternative stimulus by matching
such as those suspected of maintaining problem the sensory properties of those stimuli with those
behavior (matched stimuli) more effectively sup- associated with the problem behavior. Although
presses problem behavior than availability of un- environmental enrichment does not teach new
matched stimuli (e.g., Favell, McGimsey, & Schell, or alternative behavior directly, the availability of
1982; Piazza et al., 1998) is consistent with this ac- alternative activities and items appears to be as-
count. Indeed, Piazza, Adelinis, Hanley, Goh, and sociated with improvements in appropriate object-
Delia (2000) demonstrated that matched stimuli directed behavior (Horner, 1980).
were more effective in decreasing problem behav-
ior than stimuli that participants indicated were
Restraint, Protective Equipment, and Other Forms
more preferred during a preintervention prefer-
of Response Restriction
ence assessment. These results indicate that mere
competition among reinforcing options may not Researchers have evaluated the effects of restraint,
completely account for the effects of environmen- protective equipment, and other means to miti-
tal enrichment, and that environmental enrich- gate or prevent injury from occurrences of prob-
ment may qualify as a function-based intervention lem behavior (DeRosa, Roane, Wilson, Novak, &
in some cases. Silkowski, 2015; Fisher, Piazza, Bowman, Hanley,
A substantial body of research suggests that & Adelinis, 1997). These interventions physically
environmental enrichment can have direct and impede the occurrence or completion of problem
beneficial effects on the brain, such as increased behavior. Although these interventions are un-
plasticity in the cerebral cortex and increases in questionably effective in decreasing or eliminating
synaptic density (Alwis & Rajan, 2014). Moreover, problem behavior, most consider them a highly in-
research results have shown that environmen- trusive, undesirable, and inadequate approach that
tal enrichment can produce improved learning behavior analysts should only use in emergency
and memory, and that it may be useful to treat a situations, such as when problem behavior poses
range of neurological disorders, such as Alzheim- immediate and serious risk to the participant or
er’s disease and autism spectrum disorder. Thus others or will produce substantial property dam-
the effects of environmental enrichment as an age.
Developing Antecedent Interventions 305
We typically categorize procedures for physi- four-point restraints, straightjackets) are noncon-
cally restricting problem behavior by their form. structive because they do not teach or encourage
Personal restraint involves caregivers’ physically alternative, replacement behavior; in fact, they
securing and holding body parts so that prob- may actively impede acquisition of alternative be-
lem behavior cannot occur, and caregivers often havior if they restrict appropriate behavior. Sec-
use it after an episode of problem behavior has ond, restraint appears to have aversive properties
begun (i.e., as a consequence-based intervention). for some people. For example, some research has
However, they can also use personal restraint as indicated that contingent application of restraint
an antecedent intervention in situations where can suppress behavior that produces it (e.g., Ler-
problem behavior is highly likely to occur. For ex- man, Iwata, Shore, & DeLeon, 1997), which is a
ample, caregivers may physically restrain a child defining effect of aversive stimulation. Therefore,
during dental visits if that child has engaged in it is important to consider issues associated with
problem behavior during previous dental exams aversive intervention before using restraint. For
and procedures. Mechanical restraint involves se- example, attempts to apply antecedent restraint
curing limbs and body parts with devices designed may and often do occasion avoidance or escape
for this purpose, such as four-point restraints, arm behaviors such as running away or aggression. Ad-
splints, and straightjackets. Caregivers most fre- ditionally, a caregiver may threaten to use contin-
quently use mechanical restraint, such as personal gent restraint if it appears aversive (e.g., “If you do
restraint, to stop ongoing episodes of problem be- that again, I’ll put you in a straightjacket!”) or use
havior, but they can also use it as a proactive, an- it as punishment.
tecedent intervention when they anticipate severe Another limitation of restraint is that some
problem behavior. Finally, protective equipment is individuals who engage in SIB also actively seek
like restraint in that mechanical devices are used; opportunities to be placed into restraints or self-
however, protective equipment typically permits restrain (e.g., Baroff & Tate, 1969). In extreme
the wearer to engage in unrestricted motion but cases, individuals may engage in problem behavior
prevents problem behavior from producing dam- to produce access to restraint (Smith, Lerman, &
age (e.g., a padded helmet to prevent trauma from Iwata, 1996; Vollmer & Vorndran, 1998). Some
head banging). Protective equipment is usually less individuals engage in self-restraint almost con-
confining than restraints and may include the use tinuously, interfering with active engagement in
of devices such as safety goggles, helmets, and lap habilitative or other desirable behavior. Finally,
or wheelchair belts. Caregivers can use restraint restricting an individual’s ability to engage in a
devices to prevent the occurrence of a range of behavior may increase his or her motivation to
problem behaviors (e.g., aggression, property de- engage in that behavior. For example, research
struction, SIB), but protective equipment is used outcomes have shown that limiting access to lei-
primarily for SIB, as the devices generally protect sure activities can produce subsequent increases
the wearer from injury. in engagement in such activities, presumably due
to deprivation of their reinforcing aspects (Klatt,
Sherman, & Sheldon, 2000). Similarly, restricted
Strengths and Limitations of Restraint
access to problem behavior such as SIB may in-
and Protective Equipment
crease subsequent levels of the behavior for some
Restraint and protective equipment eliminate individuals (Blevins, 2003; Rapp, 2006). Thus re-
problem behavior effectively when they are used. straint may merely postpone and ultimately exac-
They are designed specifically to restrict occur- erbate the very behavior it is intended to reduce
rences of problem behavior. Thus no targeted in some cases.
problem behavior typically occurs during restraint
periods. However, restraints and protective equip-
ment have serious limitations and side effects. FUNCTION‑BASED INTERVENTIONS
First, many restraint procedures are intrusive,
in that they disrupt ongoing activities and often If we know the operant function of problem be-
prevent the occurrence of appropriate alternative havior (i.e., if we can identify the functional re-
behavior. That is, the use of restraint may limit inforcer), we can develop a function-based inter-
many activities, including participation in learn- vention. Function-based interventions correspond
ing opportunities and appropriate social interac- specifically to a maintaining contingency of rein-
tions. Thus restrictive restraint procedures (e.g., forcement. As such, intervention procedures vary
306 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
according to the contingency identified to main- searchers continue to use the term NCR and have
tain problem behavior. Function-based antecedent established its utility as shorthand for a general
interventions manipulate motivating operations, class of interventions involving response-inde-
discriminative stimuli, or both. Below, I describe pendent delivery of stimuli and events (Vollmer,
function-based antecedent interventions and their 1999). Therefore, I use the term NCR in the fol-
procedural variations. After introducing each gen- lowing discussion to describe general procedures,
eral strategy, I describe specific tactical variations but more descriptively accurate labels for proce-
associated with those strategies as they relate to dural details.
different types of maintaining contingencies for
problem behavior.
NCR with Problem Behavior
Maintained by Social Positive Reinforcement
Noncontingent Reinforcement
As with all function-based interventions, NCR
NCR involves presentation of the reinforcing con- procedures vary according to the functional prop-
sequence for problem behavior on a time-based, or erties of the problem behavior. For example, if care-
response-independent, schedule (Kettering, Fish- giver attention functions as positive reinforcement
er, Kelley, & LaRue, 2018). The behavior analyst for problem behavior, then NCR would consist of
often withholds the functional reinforcer when presenting attention on a time-based schedule and
the participant engages in problem behavior (i.e., withholding attention following problem behav-
the problem behavior is placed on extinction). Ini- ior. In an early and influential application of NCR,
tial NCR schedules are often dense, in that the Vollmer et al. (1993) showed that providing no dif-
behavior analyst provides the functional reinforc- ferential consequence following their participants’
er frequently (Hagopian, Fisher, & Legacy, 1994). attention-maintained SIB and providing attention
Subsequently, the behavior analyst may thin the on a time-based schedule produced immediate and
NCR schedule systematically, usually based on substantial decreases in SIB. Initially, the thera-
low rates of problem behavior (e.g., Vollmer, Iwata, pist provided continuous attention; however, the
Zarcone, Smith, & Mazaleski, 1993). The goal is therapist subsequently faded the reinforcement
to reach a schedule that is manageable in natural schedule to one brief presentation of attention
environments, or the effective parameters of NCR every 5 minutes. Researchers have used NCR to
become apparent as further schedule thinning treat a range of problem behaviors maintained by
produces unacceptable increases in problem be- positive reinforcement, including SIB (Vollmer et
havior. Research has shown that NCR is an effec- al., 1993), destructive behavior (Hagopian et al.,
tive intervention for SIB (Vollmer et al., 1993), ag- 1994), and bizarre speech (Mace & Lalli, 1991).
gression (Lalli, Casey, & Kates, 1997), disruption Although most NCR procedures schedule time-
(Fisher, Ninness, Piazza, & Owen-DeSchryver, based stimulus deliveries and withhold functional
1996), food refusal (Cooper et al., 1995), inappro- reinforcement for problem behavior (i.e., extinc-
priate vocalizations (Falcomata, Roane, Hovanetz, tion), several studies have shown that stimulus
Kettering, & Keeney, 2004), pica (Goh, Iwata, & presentation alone – without extinction – can be
Kahng, 1999), and pseudoseizures (DeLeon, Uy, & sufficient to produce substantial decreases in prob-
Gutshall, 2005). See Carr and LeBlanc (2006) for lem behavior (e.g., Fisher et al., 1999; Lalli et al.,
a detailed and comprehensive review of NCR. 1997).
Some behavior analysts have criticized the A few studies have investigated NCR arrange-
term noncontingent reinforcement as technically ments in which the researchers presented stimuli
inaccurate, because (1) reinforcement involves a other than the functional reinforcer for problem
contingency by definition (i.e., the process of rein- behavior (i.e., arbitrary stimuli; Fischer, Iwata, &
forcement is defined in part by a contingency be- Mazaleski, 1997; Fisher, DeLeon, Rodriguez-Cat-
tween a response and a consequence), and (2) the ter, & Keeney, 2004; Fisher, O’Connor, Kurtz, De-
target of reinforcement in NCR is unclear (Pol- Leon, & Gotjen; 2000; Hanley, Piazza, & Fisher,
ing & Normand, 1999). Other behavior analysts 1997). For example, Hanley et al. (1997) showed
have acknowledged the validity of these criticisms that time-based presentation of a stimulus identi-
(Vollmer, 1999), and many researchers now use fied as preferred via formal preference assessment
more technically correct descriptors, referring to was as effective as presentation of attention (the
time-based delivery of stimuli or events (e.g., fixed functional reinforcer for problem behavior) to de-
time [FT] 1-minute attention). However, many re- crease two participants’ problem behavior. Fischer
Developing Antecedent Interventions 307
et al. (1997) extended these findings, showing that researchers systematically decreased the schedule
continuous access to preferred stimuli decreased of breaks based on low levels of problem behavior
the positively reinforced problem behavior of two during previous sessions until one participant re-
participants, even when the behavior continued to ceived a 30-second break once every 10 minutes,
produce the functional reinforcer. and the second participant received a 20-second
Researchers have used results of preference as- break every 2.5 minutes. Although only a few
sessments to select the stimuli for arbitrary NCR studies have evaluated the use of noncontingent
arrangements. Fisher et al. (2000) showed that escape as intervention, researchers have produced
stimuli identified as highly preferred via paired- similar outcomes with disruptive behavior during
choice preference assessment (Fisher et al., 1992) speech therapy (Coleman & Holmes, 1998), prob-
more effectively decreased problem behavior than lem behavior of children with disabilities during
did less preferred stimuli. Researchers have used instruction (Kodak, Miltenberger, & Romaniuk,
competing-stimulus assessments to empirically 2003), and disruptive behavior of children without
identify preferred stimuli that occasion low lev- disabilities during dental routines (O’Callaghan,
els of problem behavior and high levels of stimu- Allen, Powell, & Salama, 2006).
lus engagement (Fisher et al., 2000; Fisher et al., Some researchers have found that provid-
2004). For example, Piazza et al. (1998) used a ing access to positive reinforcers on time-based
paired-choice preference assessment to identify a schedules can produce decreases in problem be-
pool of preferred stimuli. Subsequently, therapists havior maintained by negative reinforcement.
provided each stimulus to participants while prob- For example, several studies have investigated the
lem behavior continued to produce the identified effectiveness of NCR schedules using positive re-
maintaining reinforcer. Effective competing stim- inforcers as intervention for negatively reinforced
uli were those stimuli that occasioned low levels food refusal (e.g., Cooper et al., 1995; Piazza, Patel,
of problem behavior and high levels of stimulus Gulotta, Sevin, & Layer, 2003; Reed et al., 2004;
engagement during the assessment. Wilder, Normand, & Atwell, 2005). The outcomes
Noncontingent presentation of arbitrary stimuli of these studies have been mixed. Whereas posi-
represents a promising alternative, because it poten- tive NCR has decreased problem behavior and
tially increases the number of available stimuli and increased food acceptance for some participants
may decrease the need for extensive preinterven- (Wilder et al., 2005), other intervention compo-
tion functional assessment. However, the empiri- nents such as extinction were necessary to produce
cal support for the use of arbitrary stimuli in NCR clinically acceptable effects for other participants
is currently limited, and several questions remain (Piazza et al., 2003; Reed et al., 2004). Ingvars-
about the conditions under which arbitrary stimuli son, Hanley, and Welter (2009) compared the
will or will not decrease problem behavior, how to effectiveness of contingent reinforcement versus
best select and present stimuli, and the mechanisms NCR, using arbitrary reinforcers to treat escape-
underlying the effects of this procedure. maintained disruptive behavior. When both re-
inforcement schedules were implemented without
extinction, results showed that NCR produced a
NCR with Problem Behavior
clinically significant decrease in disruptive be-
Maintained by Social Negative Reinforcement
havior for one participant; however, extinction
NCR for behavior maintained by escape from was ultimately necessary for a second participant,
aversive stimuli or activities consists of providing and additional procedures were necessary for the
time-based breaks from those events. For example, third participant. These results illustrate the need
following functional analyses indicating that the for further research into the effectiveness of posi-
SIB of two adult participants with developmental tive NCR to treat problem behavior maintained by
disabilities was maintained by escape from aversive negative reinforcement.
training activities, Vollmer, Marcus, and Ringdahl
(1995) implemented NCR interventions in which
NCR with Problem Behavior
they presented breaks from these activities inde-
Maintained by Automatic Reinforcement
pendently of the participants’ problem behavior.
Initially, reinforcement schedules were very dense, NCR for automatically reinforced behavior typi-
with one participant receiving no training tasks, cally involves an attempt to identify the auto-
and the second receiving only 15 seconds of train- matically produced reinforcer for the behavior, fol-
ing before each 20-second break. However, the lowed by provision of that event on a time-based
308 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
schedule. In a groundbreaking study, Favell et al. stimulation like that suspected of maintaining
(1982) provided noncontingent access to stimuli problem behavior in NCR arrangements, other
that corresponded to the SIB topographies of six findings suggest that a match may not always be
participants. For example, the researchers provid- necessary. Several researchers have used assess-
ed toys and items with striking visual properties ments to identify competing stimuli for NCR ar-
(e.g., brightly colored toys, lights) to two partici- rangements with automatically reinforced problem
pants who engaged in eye poking. The researchers behavior (e.g., Piazza, Fisher, Hanley, Hilker, &
provided toys participants could mouth or small Derby, 1996; Ringdahl et al., 1997; Shore, Iwata,
food items to participants who engaged in hand DeLeon, Kahng, & Smith, 1997). Researchers
mouthing or pica. Results suggested that provid- have provided participants with access to indi-
ing noncontingent access to items that appeared vidual stimuli to identify those that are associated
to match the hypothesized functional properties with high levels of engagement and low levels of
of automatically reinforced problem behavior pro- problem behavior. Using competing-stimulus as-
duced substantial decreases in problem behavior sessments to select arbitrary stimuli for NCR ar-
for some participants. rangements seems a promising approach, given the
Subsequent research has further investigated potential difficulty of identifying the hypothesized
the importance of providing stimuli that match functional properties of automatically reinforced
the functional properties of automatically rein- problem behavior.
forced problem behavior in NCR arrangements.
For example, Piazza et al. (1998) conducted a se-
Mechanisms Underlying the Effects of NCR
ries of analyses to identify the specific functional
reinforcers for the pica of three participants, and The behavioral principles that describe the effects
to evaluate the effects of interventions that cor- of NCR have received considerable attention in
responded to the outcomes. Results indicated that the literature. One explanation is that frequent,
the participants’ pica was maintained at least in repeated contact with the functional reinforcer for
part by automatic reinforcement. An assessment problem behavior during NCR schedules acts as an
of the effects of matched (e.g., food) versus un- abolishing operation, which temporarily decreases
matched stimuli (e.g., light-up wand) showed that the effectiveness of the functional reinforcer and
matched stimuli more effectively reduced pica for decreases the occurrence of the class of behav-
two participants. Subsequently, the researchers ior maintained by that reinforcer (Laraway et al.,
conducted analyses of specific characteristics of 2003). Although most researchers refer to satiation
matched stimuli, such as taste and texture. Results to explain this decrease in reinforcer effectiveness,
indicated that food with a firmer texture more ef- some suggest that habituation may better account
fectively decreased pica than less firm food items. for these effects (Murphy et al., 2003). Habitua-
Piazza et al. (2000) extended this approach to tion is a decrease in responsiveness to stimuli after
other topographies of problem behavior. These repeated presentation of those stimuli (Thompson
investigators compared the effects of providing ac- & Spencer, 1966) and is typically associated with
cess to matched versus unmatched stimuli to treat respondent rather than operant behavior. Howev-
diverse topographies of automatically reinforced er, Murphy et al. (2003) reviewed research findings
behavior (e.g., climbing on furniture, jumping out that appear to support a habituation account for
of windows, aggression, saliva play, hand mouth- many operant phenomena, including NCR.
ing) exhibited by three children with developmen- The two accounts have different implications
tal disabilities. Researchers compared conditions for intervention. For example, habituation is fa-
in which available stimuli matched the hypoth- cilitated by fixed rather than variable schedules
esized functional properties of automatically re- of stimulus presentation (e.g., Broster & Rankin,
inforced problem behavior; stimuli did not match 1994); therefore, a habituation account predicts
the hypothesized functional properties of auto- that FT schedules should more effectively abolish
matically reinforced problem behavior; or no toys the effectiveness of the functional reinforcer than
or leisure items were available. Matched stimuli variable-time (VT) schedules. However, some
nearly eliminated problem behavior for all three basic research outcomes suggest that VT schedules
participants, although unmatched stimuli also re- more effectively suppress responding (e.g., Lattal,
duced problem behavior. 1972; Neuringer, 1973; Ono, 1987), and the results
Although these outcomes highlight the impor- of one applied investigation showed slightly more
tance of selecting items that appear to produce rapid decreases in VT than in FT conditions (Van
Developing Antecedent Interventions 309
Camp, Lerman, Kelley, Contrucci, & Vorndran, effects: Problem behavior occurs at relatively low
2000). Another characteristic of habituated be- levels following dense NCR schedules, but in-
havior is stimulus specificity, in which unpredict- creases almost immediately following leaner NCR
able changes in the presented stimulus disrupt schedules. Thus it seems possible that the effects
habituation. Therefore, ensuring a minimum of of NCR reflect a combination of extinction and
variation in type, magnitude, and mode of stimu- motivational processes.
lus presentation should enhance NCR effects. Al- Wallace, Iwata, Hanley, Thompson, and Roscoe
though outcomes in basic research are consistent (2012) directly examined the relative contribu-
with this account (e.g., Swithers & Hall, 1994; tions of abolishing operations and extinction dur-
Whitlow, 1975), no applied research has investi- ing NCR with three participants. In Study 1, lean
gated the effects of stimulus specificity by using and dense schedules of NCR with extinction were
time-based schedules. equally effective in producing large and immediate
Regardless of whether the effects of repeated reductions in problem behavior. In Study 2, the re-
stimulus exposure represent satiation or habitua- searchers implemented NCR with and without ex-
tion, both accounts are consistent with the notion tinction in a multielement design. The procedures
that NCR schedules decrease problem behavior were equally effective initially for two participants
due to an abolishing-operation effect, in which the at dense NCR schedules. Noncontingent rein-
procedure produces a decrease in the value of the forcement without extinction was ineffective for
functional reinforcer and decreases in responses the third participant. Problem behavior increased
that have produced that reinforcer in the past. Al- during NCR without extinction as the research-
ternatively, NCR may decrease behavior due to a ers thinned the NCR schedule. These outcomes,
disruption of the reinforcement contingency. Most combined with previously reported differences in
NCR procedures contain an extinction compo- the persistence of NCR’s effects (e.g., Simmons et
nent, in that the researcher does not present the al., 2003), and the effects of dense and lean NCR
functional reinforcer contingent on problem be- schedules on responding in concurrent contingent-
havior. The response-independent presentation of reinforcement schedules (e.g., Goh et al., 2000)—
the functional reinforcer further disrupts the prob- suggest that abolishing-operation and extinction
lem behavior–reinforcer contingency. Therefore, processes may combine to produce NCR effects,
decreases in problem behavior may be a function but that the dynamics that influence their relative
of extinction, at least in part. contributions (e.g., differences in NCR schedules,
Several studies have investigated the relative amounts, etc.) await further research.
contributions of extinction and abolishing op- An alternative account of the effects of NCR
erations to NCR’s reductive effects. For example, is that time-based reinforcement delivery produces
some researchers have shown that when they adventitious reinforcement of alternative behav-
combined NCR schedules with concurrent sched- iors that compete with problem behavior. Using
ules of response-contingent access to the same an inventive laboratory preparation to model
reinforcer, responding that produced contingent NCR procedures, Ecott and Critchfield (2004)
reinforcement was typically low during dense investigated the effects of NCR on classes of tar-
NCR schedules but increased as NCR schedules geted and alternative behaviors. These research-
were thinned (e.g., Goh, Iwata, & DeLeon, 2000; ers used college students as participants and de-
Marcus & Vollmer, 1996). Others have observed signed an experimental environment in which
that problem behavior occurred during NCR they could (1) measure multiple behaviors and
schedules without extinction (e.g., Kahng, Iwata, (2) constrain the range of alternative behavior
Thompson, & Hanley, 2000), or as the research- to capture changes in alternative behavior as a
ers thinned the NCR schedule (e.g., Kahng et al., function of NCR. Results indicated that behavior
2000; Simmons, Smith, & Kliethermes, 2003). previously maintained by contingent reinforce-
Abruptly discontinuing NCR should produce ment decreased and alternative behavior increased
immediate increases in problem behavior accord- during NCR, suggesting that alternative behavior
ing to an extinction account for NCR’s behavior- was adventitiously reinforced during NCR. The
reducing effects. By contrast, in NCR functions patterns of response reallocation in this experi-
as an abolishing operation, problem behavior ini- ment were consistent with an account based on
tially occurs at low levels, followed by an increase matching law (e.g., McDowell, 1989), which holds
as deprivation from the functional reinforcer oc- that relative rates of responding among response
curs. Outcomes of investigations have shown both options corresponds to relative rates of reinforce-
310 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
vention procedures cannot be conducted. For of and establish contextual control over commu-
example, Piazza, Hanley, and Fisher (1996) imple- nicative responses (e.g., Fisher, Greer, Fuhrman, &
mented a response interruption procedure for pica Querim, 2015; Hanley, Iwata, & Thompson, 2001;
in the presence of a signal card. Subsequently, pica Tiger & Hanley, 2004).
decreased when the card was present, even when A strength of stimulus control procedures is
they did not implement the response interruption that they can bring behavior under the control of
procedure. Similarly, McKenzie, Smith, Simmons, consequence-based contingencies without neces-
and Soderlund (2008) delivered reprimands con- sitating frequent or prolonged contact with those
tingent on eye poking when their participant was contingencies. Specifically, a behavior analyst can
wearing wristbands, but not when the wristbands decrease use of aversive consequences by present-
were absent. Eye poking decreased when the par- ing antecedent stimuli that he or she has correlat-
ticipant wore the wristbands, both in the inter- ed with those consequences previously. Consider
vention environment and at times when and in the effects of visible patrol cars on highway speed-
places where the researchers did not and had never ing. Drivers are more likely to observe speed limits
delivered reprimands. in the presence of patrol cars, which are correlated
Researchers have also used discrimination- with an increased probability of receiving a speed-
training procedures to signal changes in contin- ing ticket. Frequent punishment of speeding is not
gencies during reinforcement-based interventions necessary to have this effect; police only need to
for problem behavior. For example, in some situ- present a stimulus that signals a higher probability
ations it may not be possible to deliver functional of punishment. All warning stimuli operate on the
reinforcers for appropriate communication (such same principle: Unwanted behavior decreases in
as when a child with attention-maintained SIB the moment not because it is punished, but because
requests attention when his or her caregiver is a signal has been presented warning that punish-
changing an infant sibling’s diaper), or the alter- ment is likely to follow that behavior.
native response may occur excessively (such as
when a child requests attention continuously).
Antecedent Interventions Designed Specifically
The communication response may be extin-
to Treat Escape Behavior
guished and problem behavior may increase if
the caregiver does not deliver reinforcement for Researchers have developed several distinct ante-
communication responses immediately and con- cedent interventions to decrease problem behavior
sistently (e.g., Briggs, Fisher, Greer, & Kimball, maintained by negative reinforcement in the form
2018; Hagopian, Fisher, Sullivan, Acquisto, & of escape from or avoidance of aversive events.
LeBlanc, 1998). Therefore, procedures to miti- These strategies are particularly important for sev-
gate these effects would be useful. Fisher, Kuhn, eral reasons. First, the outcomes of comprehensive
and Thompson (1998) taught two participants reviews indicate that escape from task demands
to mand for functional and alternative reinforc- maintains escape behavior for approximately 32%
ers, and subsequently correlated specific stimuli of individuals exhibiting severe problem behavior;
with the functional and alternative reinforcers. a proportion that is greater than any other con-
Results showed that problem behavior decreased tingency (Beavers, Iwata, & Lerman, 2013). Sec-
when either functional or alternative reinforcers ond, the antecedents that occasion escape and
were available, and that participants manded ap- avoidance are often more obvious and available
propriately for functional and alternative reinforc- to immediate manipulation than those that occa-
ers in the presence of the relevant discriminative sion positively reinforced behavior. For example,
stimuli (i.e., the mand for the functional reinforcer escape behavior typically is motivated by the pres-
but not the mand for the alternative reinforcer oc- ence of some obvious source of aversive stimula-
curred in the presence of the stimulus correlated tion, whereas positive reinforcement is more often
with the functional reinforcer’s availability, and motivated by less-apparent operations, such as
vice versa). Thus the stimulus control procedure deprivation. Third, caregivers and therapists often
managed problem behavior when delivery of the control those sources of aversive stimulation, such
functional reinforcer for the communication re- as task demands or requests to participate in activ-
sponse was not possible. Subsequent studies have ities such as educational, dental, medical, or other
used stimulus-control training, typically with therapeutic routines. Thus antecedent approaches
condition-correlated stimuli, to manage the rates to the management of potentially aversive situa-
312 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
tions in ways that might prevent occurrences of treatment for any medical or biological condition
problematic escape behavior is, in a sense, “low that they suspect contributes to the motivation of
hanging fruit,” in that the antecedent conditions SIB. They may consider additional interventions if
that set the occasion for problem behavior are SIB persists after medical intervention.
relatively obvious and available to the caregiver.
Below, I describe several antecedent strategies
Fading in Aversive Stimuli
tailored specifically to escape behavior. See Milt-
enberger (2006) for a comprehensive discussion of One way to return aversive events to the environ-
antecedent interventions for negatively reinforced ment following their elimination is through stimu-
behavior. lus fading (e.g., Pace, Ivancic, & Jefferson, 1994;
Pace, Iwata, Cowdery, Andree, & McIntyre, 1993;
Zarcone, Iwata, Smith, Mazaleski, & Lerman,
Elimination of Aversive Stimulation
1994; Zarcone et al., 1993). Fading is the gradual
Perhaps the most straightforward antecedent in- and systematic reintroduction of stimuli that oc-
tervention for escape behavior is to remove the casion escape behavior. Zarcone et al. (1993) com-
aversive event that motivates the behavior. By pared fading plus extinction with extinction alone
definition, escape behavior occurs in the presence to reduce the escape-maintained SIB of three
of an aversive stimulus; therefore, removal of that participants. During fading, the researchers elimi-
stimulus should eliminate escape behavior, and a nated task demands that occasioned SIB. Subse-
large body of literature indicates that it does. The quently, they reintroduced task demands slowly
functional-analytic literature is replete with ex- and systematically while providing no differential
amples in which problem behavior occurs in the consequence for SIB (i.e., extinction). Extinction
presence but not the absence of task demands (e.g., alone produced more rapid decreases in SIB, but
Carr, Newsom, & Binkoff, 1980; Iwata, Dorsey, also resulted in initial extinction bursts that were
Slifer, Bauman, & Richman, 1982/1994; Iwata et not observed during fading. Attempts to use fad-
al., 1994). Although the removal of task demands ing procedures without extinction have met with
or other aversive antecedent stimuli is undeni- mixed results. Some researchers have produced
ably effective in reducing problem behavior main- encouraging outcomes (e.g., Pace et al., 1994), and
tained by escape, it is an impractical and unreal- others have failed to maintain initial decreases in
istic approach for all but the most serious cases. problem behavior without extinction (e.g., Zar-
Eliminating all potentially aversive responsibilities cone, Iwata, Smith, et al., 1994).
and requirements is simply not possible for most The behavioral processes underlying the ef-
people. Withholding as many of these events as fectiveness of fading procedures are not well un-
possible on an emergency basis may be necessary derstood. Fading with extinction may permit
in extreme cases when the presentation of task de- behavior to contact extinction contingencies
mands or other stressful events sets the occasion in the presence of a relatively weak establishing
for severe and dangerous behavior. A behavior an- operation. That is, extinction may proceed more
alyst can reintroduce the antecedents of problem smoothly if the antecedent stimuli that motivate
behavior systematically when the behavior has de- escape are altered to decrease their aversiveness
creased to acceptable levels. (Greer, Fisher, Saini, Owen, & Jones, 2016). Slow
A notable exception in which removal of aver- but repeated exposure may reduce the overall aver-
sive stimulation is the first-choice intervention siveness of stimuli when fading is effective without
is when pain or discomfort associated with ill- extinction. Indeed, desensitization, or graduated
ness, injury, or other biological factors occasions exposure to stimuli that occasion escape or avoid-
problem behavior. Several researchers have sug- ance, is a widely used procedure to treat phobias
gested that conditions such as allergies (Kennedy (e.g., Shabani & Fisher, 2006). The precise mecha-
& Meyer, 1996), constipation (Carr & Smith, nism responsible for this effect is not understood,
1995), menstrual discomfort (Carr, Smith, Giacin, although habituation offers a potentially viable ac-
Whelan, & Pancari, 2003; Taylor, Rush, Hetrick, count (Murphy et al., 2003).
& Sandman, 1993), otitis media (Carr & Smith,
1995; Cataldo & Harris, 1982; O’Reilly, 1995), and
High‑Probability Sequence/Behavioral Momentum
sleep deprivation (Kennedy & Meyer, 1996) may
be associated with escape-maintained problem Like fading, high-probability (high-p) sequences
behavior. Caregivers should arrange immediate can be used to facilitate reintroduction of aver-
Developing Antecedent Interventions 313
sive stimuli to the environment. A behavior ana- quest sequences, the mechanism(s) responsible
lyst delivers several high-p requests (ones with a for intervention effects in these studies are un-
high probability of compliance and that do not clear. Specifically, topographical features of the
occasion escape behavior) before delivering a high- and low-p responses should not be relevant
low-probability (low-p) request (one that has a low for a behavioral-momentum-based account of the
probability of compliance, occasions problematic effects of high-p sequences. Thus additional inves-
escape behavior, or both). Researchers have used tigations will be necessary to clearly identify the
high-p sequences to increase compliance with conditions under which high-p sequences are most
low-p requests (e.g., Mace et al., 1988) and to treat efficacious, and the behavioral mechanisms that
escape-maintained problem behavior (e.g., Mace produce their effects.
& Belfiore, 1990). Researchers have used the met-
aphor of behavioral momentum to describe these
Altering the Aversive Stimulus
effects. According to this account, a high density
of reinforcement for high-p requests increases mass Altering the features of the aversive stimulus or
and creates velocity in compliance, which makes event that sets the occasion for escape behavior
this class of behavior resistant to change (Nevin, to reduce its evocative effect is a possible interven-
1996). Thus compliance persists and problematic tion. For example, Cameron, Ainsleigh, and Bird
escape behavior is unlikely to occur when the be- (1992) showed that the escape SIB of one par-
havior analyst presents a low-p request. Research- ticipant was more likely to occur during washing
ers have used high-p sequences successfully to treat routines with bar rather than liquid soap. Thus re-
noncompliance to do and don’t commands (Mace searchers simply altered the task and presented liq-
et al., 1988), academic tasks (Belfiore, Lee, Vargas, uid rather than bar soap during washing routines.
& Skinner, 1997; Wehby & Hollahan, 2000), and Similarly, researchers have shown that changes
medical routines (McComas, Wacker, & Cooper, such as using a computer instead of pencil and
1998). However, results of research suggest that paper for writing (Ervin, DuPaul, Kern, & Friman,
extinction may be necessary to treat active escape 1998), or using checkers or a calculator as count-
behavior such as SIB and aggression. In some cases, ing aids during math tasks (McComas, Hoch,
using the high-p sequence without extinction may Paone, & El-Roy, 2000), can function as abolish-
exacerbate escape behavior (e.g., Zarcone, Iwata, ing operations, effectively decreasing the aversive-
Mazaleski, & Smith, 1994). ness of tasks so that they no longer evoke escape
Recent research on high-p sequences to treat behavior. Despite these encouraging results, alter-
food refusal or selectivity illustrates both the ing aversive events to make them more benign
promise and some apparent limitations of such may not always be possible (e.g., when liquid soap,
interventions. McComas, Wacker, et al. (2000) computers, checkers, or calculators are unavail-
showed that adding a high-p sequence to escape able). Thus problem behavior will likely return to
extinction for bite refusal improved the effective- preintervention levels when avoiding contact with
ness of extinction. Subsequently, Dawson et al. aversive events is not possible.
(2003) observed increases in acceptance and de-
creases in problem behavior only during high-p
Altering the Aversive Context
sequences plus escape extinction, and not during
high-p sequences alone. However, other research- Another way to reduce the aversiveness of ante-
ers have shown that a high-p sequence without cedent stimuli is to alter features of the context
extinction can be sufficient to increase bite accep- in which a behavior analyst presents those stim-
tance for some participants (e.g., Ewry & Fryling, uli. With contextual interventions, the analyst
2015; Meier, Fryling, & Wallace, 2012; Patel et al., changes the surrounding environment to reduce
2006). Interestingly, studies reporting successful the aversiveness of antecedent stimuli. Thus con-
outcomes of high-p sequences without extinction textual interventions do not involve direct ma-
used high-p requests that were topographically like nipulation of the aversive stimulus or event per se;
the behavior targeted for increase. For example, rather, the behavior analyst alters other aspects of
Meier et al. (2012) used acceptance of highly pre- the environment, which then decrease the evoca-
ferred food as the high-p response that preceded tive function (aversiveness) of the stimulus or
low-p requests to accept bites of nonpreferred food. event. Researchers have shown that many contex-
Given that the nature of high-p requests appears tual variables alter the occurrence of problematic
to have influenced the effectiveness of such re- escape behavior and have manipulated them as
314 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
interventions. For example, results of studies have lin, Giacobbe-Grieco, & Smith, 2003), and the
shown that embedding task demands in pleasant presentation of corrective feedback before (as
stories (Carr, Newsom, & Binkoff, 1976) or in pre- antecedent prompts) rather than after task trials
ferred activities (Carr & Carlson, 1993) decreases (Ebanks & Fisher, 2003), among others. The ef-
escape behavior. Similarly, Dunlap, Kern-Dunlap, fects of most contextual strategies appear to be due
Clarke, and Robbins (1991) increased on-task be- to abolishing-operation effects, in which chang-
havior and decreased problem behavior for one ing an apparently unrelated condition alters the
participant by altering session length, type of motivational properties of aversive events; that is,
motor activity, and the functional nature of pre- they decrease the aversiveness of the events that
sented tasks. motivate escape. Additional research will be nec-
Several researchers have shown that offering essary to identify the specific mechanisms associ-
a choice among putatively aversive activities can ated with the effectiveness of various contextual
decrease escape behavior (e.g., Dunlap et al., 1994; strategies.
Dyer, Dunlap, & Winterling, 1990; Vaughn &
Horner, 1997). Some outcomes suggest that this
effect will occur even if the researchers’ choices are SUMMARY
yoked to those of participants’ (Dyer et al., 1990),
or if participants choose from low-preference tasks Researchers have used an array of antecedent in-
that have evoked high levels of escape behavior terventions to decrease problem behavior. Some
previously (Vaughn & Horner, 1997). strategies decrease the motivation to engage in the
Research outcomes have shown that temporally behavior; others signal differential consequences
distant routines and events can sometimes set the for the behavior; and still others physically im-
occasion for later occurrences of problem behav- pede the occurrence of behavior in some way.
ior. Some use the term setting events, but see Smith Although each intervention has unique benefits
and Iwata (1997) for a discussion of issues with that and limitations, all share the characteristic that
term. For example, Kennedy and Itkonen (1993) they are implemented before the occurrence of
showed that one participant’s problem behavior at problem behavior. Thus a shared strength among
school was more likely to occur on days when she antecedent strategies is that they do not require
awakened late. A second participant’s problem be- the occurrence of problem behavior for their effec-
havior correlated with the number of stops on her tiveness (certain discrimination-based interven-
ride to school. Each participant’s problem behavior tions constitute a possible exception, as previously
decreased when researchers eliminated these situ- discussed). In some cases, problem behavior may
ations from their routines. When behavior ana- not occur at all after the first implementation of
lysts cannot eliminate events that set the occasion intervention.
for problem behavior, they may be able to neutral- Antecedent interventions may also comple-
ize their influence. For example, Horner, Day, and ment or accelerate the effects of consequence-
Day (1997) found that prior postponements or de- based interventions. Researchers have paired an-
lays in planned activities exacerbated the escape- tecedent strategies with extinction procedures to
maintained problem behavior of two participants decrease negative side effects associated with ex-
during subsequent instructional routines. The re- tinction (e.g., Zarcone, Iwata, Smith, et al., 1994).
searchers conducted neutralizing routines (the op- Researchers also have incorporated antecedent
portunity to draw and write for one participant, strategies into packages including differential re-
and the opportunity to reschedule the event and inforcement (e.g., Kodak et al., 2003; Marcus &
look at a yearbook for the other participant) on Vollmer, 1996; Shabani & Fisher, 2006), punish-
days when the conditions correlated with problem ment (e.g., Thompson, Iwata, Conners, & Roscoe,
behavior were present before instructional ses- 1999), and pharmacological interventions (e.g.,
sions. Problem behavior decreased substantially on Allison et al., 1991).
the days when researchers conducted neutralizing A shared limitation of antecedent strategies is
routines. that they do not build or encourage new, alterna-
Some other contextually based interventions tive forms of behavior (i.e., they are not construc-
include altering the timing of requests to avoid in- tive), because the establishment of new behavior or
terrupting preferred activities (e.g., Fritz, DeLeon, maintenance of alternative behavior requires that
& Lazarchick, 2004), presenting stimuli associated we manage reinforcing consequences. Certainly
with a positive mood (e.g., Carr, Magito McLaugh- behavior analysts incorporate antecedent ma-
Developing Antecedent Interventions 315
nipulations such as prompts and instructions into behavior. Journal of the American Academy of Child
repertoire-building interventions. But because the Psychiatry, 7, 454–470.
analysts also must manage the consequences that Baumeister, A. A., & MacLean, W, E., Jr. (1984). Decel-
are ultimately responsible for the effectiveness of eration of self-injurious and stereotypic responding
by exercise. Applied Research in Mental Retardation,
these strategies (i.e., reinforcement), these are not
5, 385–393.
considered antecedent strategies. Also, some ante- Beavers, G. A., Iwata, B. A., & Lerman, D. C. (2013).
cedent strategies, such as NCR, may interfere with Thirty years of research on the functional analysis of
acquisition of alternative behavior due to habitu- problem behavior. Journal of Applied Behavior Analy-
ation or satiation effects. When combining NCR sis, 46, 1–21.
with procedures to increase appropriate behavior, Belfiore, P. J., Lee, D. L., Vargas, A. U., & Skinner,
behavior analysts should closely monitor the out- C. H. (1997). Effects of high-preference single-digit
comes and use alternative reinforcers or varied mathematics problem completion on multiple-digit
consequences to ensure adequate progress. Thus mathematics problem performance. Journal of Ap-
plied Behavior Analysis, 30, 327–330.
antecedent strategies—alone—do not systemati-
Berkson, G., & Davenport, R. K. (1962). Stereotyped
cally establish or maintain positive alternatives to movements of mental defectives. American Journal of
problem behavior. Mental Deficiency, 66, 849–852.
Although researchers have shown that anteced- Berkson, G., & Mason, W. A. (1963). Stereotyped move-
ent procedures are effective in treating problem ments of mental defectives: III. Situation effects.
behavior, a behavior analyst should implement American Journal of Mental Deficiency, 68, 409–412.
them as components in a comprehensive interven- Blevins, T. (2003). The effects of response restriction
tion package that includes both antecedent and on non-socially maintained self-injury. Unpublished
consequence strategies. Antecedents alone can master’s thesis, University of North Texas, Denton,
produce behavior and consequences alone can af- TX.
Briggs, A. M., Fisher, W. W., Greer, B. D., & Kimball,
fect behavior that produces it; however, combining
R. T. (2018). Prevalence of resurgence of destructive
both can enhance the effects of each, improving behavior when thinning reinforcement schedules
the ability of antecedents to evoke or suppress be- during functional communication training. Journal
havior and providing more opportunities to con- of Applied Behavior Analysis, 51, 620–633.
tact more effective consequences. Thus behavior Broocks, A., Bandelow, B., Pekrun, G., George, A.,
analysts should carefully match antecedents and Meyer, T., Bartmann, U., et al. (1998). Comparison
consequences in behavior intervention programs of aerobic exercise, chlomipramine, and placebo in
to maximize the potential effectiveness of each the treatment of panic disorder. American Journal of
strategy. As shown in this chapter, a wide range Psychiatry, 155, 603–609.
of proven antecedent strategies is available for in- Broster, B. S., & Rankin, C. H. (1994). Effects of chang-
ing interstimulus interval during habituation in
tegration in comprehensive intervention packages
Caenorhabditis elegans. Behavioral Neuroscience, 108,
to treat problem behaviors and help establish ap- 1019–1029.
propriate alternative repertoires. Cameron, M. J., Ainsleigh, S. A., & Bird, F. L. (1992).
The acquisition of stimulus control of compliance
and participation during an ADL routine. Behavioral
REFERENCES Residential Treatment, 7, 327–340.
Cannella-Malone, H. I., Tullis, C. A., & Kazee, A. R.
Allison, D. B., Basile, V. C., & MacDonald, R. B. (1991). (2011). Using antecedent exercise to decrease chal-
Brief report: Comparative effects of antecedent ex- lenging behavior in boys with developmental dis-
ercise and lorazepam on aggressive behavior of an abilities and an emotional disorder. Journal of Positive
autistic man. Journal of Autism and Developmental Behavior Interventions, 13, 230–239.
Disorders, 21, 89–94. Carr, E. G., & Carlson, J. I. (1993). Reduction of severe
Alwis, D. S., & Rajan, R. (2014). Environmental enrich- behavior problems in the community using a multi-
ment and the sensory brain: The role of enrichment component treatment approach. Journal of Applied
in remediating brain injury. Frontiers in Systems Neu- Behavior Analysis, 26, 157–172.
roscience, 8, 156. Carr, E. G., Magito McLaughlin, D., Giacobbe-Grieco,
Bachman, J. E., & Fuqua, R. W. (1983). Management T., & Smith, C. E. (2003). Using mood ratings and
of inappropriate behaviors of trainable mentally im- mood induction in assessment and intervention for
paired students using antecedent exercise. Journal of severe problem behavior. American Journal on Mental
Applied Behavior Analysis, 16, 477–484. Retardation, 108, 32–55.
Baroff, G. S., & Tate, B. G. (1968). The use of aversive Carr, E. G., Newsom, C. D., & Binkoff, J. A. (1976).
stimulation in the treatment of chronic self-injurious Stimulus control of self-destructive behavior in a psy-
316 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
chotic child. Journal of Abnormal Child Psychology, rigidity on self-injury and adaptive behavior. Journal
4, 139–153. of Applied Behavior Analysis, 48, 860–864.
Carr, E. G., Newsom, C. D., & Binkoff, J. A. (1980). Doyne, E. J., Chambless, D. L., & Beutler, L. E. (1983).
Escape as a factor in the aggressive behavior of two Aerobic exercise as a treatment for depression in
retarded children. Journal of Applied Behavior Analy- women. Behavior Therapy, 14, 434–440.
sis, 13, 101–117. Dunlap, G., dePerczel, M., Clarke, S., Wilson, D.,
Carr, E. G., & Smith, C. E. (1995). Biological setting Wright, S., White, R., et al. (1994). Choice making
events for self-injury. Mental Retardation and Develop- to promote adaptive behavior for students with emo-
mental Disabilities Research Reviews, 1, 94–98. tional and behavioral challenges. Journal of Applied
Carr, E. G., Smith, C. E., Giacin, T. A., Whelan, B. Behavior Analysis, 27, 505–518.
M., & Pancari, J. (2003). Menstrual discomfort as a Dunlap, G., Kern-Dunlap, L., Clarke, S., & Robbins, F.
biological setting event for severe problem behavior: R. (1991). Functional assessment, curricular revision,
Assessment and intervention. American Journal on and severe behavior problems. Journal of Applied Be-
Mental Retardation, 108, 117–133. havior Analysis, 24, 387–397.
Carr, J. E., & LeBlanc, L. A. (2006). Noncontingent Dyer, K., Dunlap, G., & Winterling, V. (1990). Effects
reinforcement as antecedent behavior support. In of choice making on the serious problem behaviors
J. Luiselli (Ed.), Antecedent assessment and interven- of students with severe handicaps. Journal of Applied
tion: Supporting children and adults with developmental Behavior Analysis, 23, 515–524.
disabilities in community settings (pp. 147–164). Balti- Ebanks, M. E., & Fisher, W. W. (2003). Altering the
more: Brookes. timing of academic prompts to treat destructive be-
Cataldo, M. F., & Harris, J. (1982). The biological basis havior maintained by escape. Journal of Applied Be-
of self-injury in the mentally retarded. Analysis and havior Analysis, 36, 355–359.
Intervention in Developmental Disabilities, 7, 21–39. Ecott, C. L., & Critchfield, T. S. (2004). Noncontingent
Celiberti, D. A., Bobo, H. E., Kelly, K. S., Harris, S. reinforcement, alternative reinforcement, and the
L., & Handleman, J. S. (1997). The differential and matching law: A laboratory demonstration. Journal
temporal effects of antecedent exercise on the self- of Applied Behavior Analysis, 37, 249–265.
stimulatory behavior of a child with autism. Research Ervin, R. A., DuPaul, G. J., Kern, L., & Friman, P. C.
in Developmental Disabilities, 18, 139–150. (1998). Classroom-based functional and adjunctive
Coleman, C., & Holmes, P. (1998). The use of noncon- assessments: Proactive approaches to intervention
tingent escape to reduce disruptive behaviors in chil- selection for adolescents with attention deficit hyper-
dren with speech delays. Journal of Applied Behavior activity disorder. Journal of Applied Behavior Analysis,
Analysis, 31, 687–690. 31, 65–78.
Cooper, L. J., Wacker, D. P., McComas, J., Brown, K., Ewry, D. M., & Fryling, M. J. (2015). Evaluating the
Peck, S. M., Richman, D., et al. (1995). Use of com- high-probability instructional sequence to increase
ponent analysis to identify active variables in treat- the acceptance of foods with an adolescent with au-
ment packages for children with feeding disorders. tism. Behavior Analysis in Practice, 8, 1–4.
Journal of Applied Behavior Analysis, 28, 139–153. Falcomata, T. S., Roane, H. S., Hovanetz, A. N., Ketter-
Cuvo, A. J., May, M. E., & Post, T. M. (2001). Effects of ing, T. L., & Keeney, K. M. (2004). An evaluation of
living room, Snoezelen room, and outdoor activities response cost in the treatment of inappropriate vo-
on stereotypic behavior and engagement by adults calizations maintained by automatic reinforcement.
with profound mental retardation. Research in Devel- Journal of Applied Behavior Analysis, 37, 83–87.
opmental Disabilities, 22, 183–204. Favell, J. E., McGimsey, J. F., & Schell, R. M. (1982).
Dawson, J. E., Piazza, C. C., Sevin, B. M., Gulotta, C. Treatment of self-injury by providing alternate sen-
S., Lerman, D., & Kelly, M. L. (2003). Use of the sory activities. Analysis and Intervention in Develop-
high-probability instructional sequence and escape mental Disabilities, 2, 83–104.
extinction in a child with food refusal. Journal of Ap- Fischer, S. M., Iwata, B. A., & Mazaleski, J. L. (1997).
plied Behavior Analysis, 36, 105–108. Noncontingent delivery of arbitrary reinforcers as
DeLeon, I. G., Anders, B. M., Rodriguez-Catter, V., & treatment for self-injurious behavior. Journal of Ap-
Neidert, P. L. (2000). The effects of noncontingent plied Behavior Analysis, 30, 239–249.
access to single-versus multiple-stimulus sets on self- Fisher, W. W., DeLeon, I. G., Rodriguez-Catter, V., &
injurious behavior. Journal of Applied Behavior Analy- Keeney, K. M. (2004). Enhancing the effects of ex-
sis, 33, 623–626. tinction on attention-maintained behavior through
DeLeon, I. G., Uy, M., & Gutshall, K. (2005). Noncon- noncontingent delivery of attention or stimuli identi-
tingent reinforcement and competing stimuli in the fied via a competing stimulus assessment. Journal of
treatment of pseudoseizures and destructive behav- Applied Behavior Analysis, 37, 171–184.
iors. Behavioral Interventions, 20, 203–217. Fisher, W. W., Greer, B. D., Fuhrman, A. M., & Querim,
DeRosa, N. M., Roane, H. S., Wilson, J. L., Novak, M. A. C. (2015). Using multiple schedules during func-
D., & Silkowski, E. L. (2015). Effects of arm-splint tional communication training to promote rapid
Developing Antecedent Interventions 317
transfer of treatment effects. Journal of Applied Be- Hagopian, L. P., Crockett, J. L., van Stone, M., DeLeon,
havior Analysis, 48, 713–733. I. G., & Bowman, L. G. (2000). Effects of noncontin-
Fisher, W. W., Greer, B. D., Mitteer, D. R., Fuhrman, A. gent reinforcement on problem behavior and stimu-
M., Romani, P. W., & Zangrillo, A. N. (2018). Further lus engagement: The role of satiation, extinction, and
evaluation of differential exposure to establishing op- alternative reinforcement. Journal of Applied Behavior
erations during functional communication training. Analysis, 33, 433–449.
Journal of Applied Behavior Analysis, 51, 360–373. Hagopian, L. P., Fisher, W. W., & Legacy, S. M. (1994).
Fisher, W. W., Kuhn, D. E., & Thompson, R. H. (1998). Schedule effects of noncontingent reinforcement on
Establishing discriminative control of responding attention-maintained destructive behavior in identi-
using functional and alternative reinforcers during cal quadruplets. Journal of Applied Behavior Analysis,
functional communication training. Journal of Ap- 27, 317–325.
plied Behavior Analysis, 31, 543–560. Hagopian, L. P., Fisher, W. W., Sullivan, M. T., Ac-
Fisher, W. W., Ninness, H. A. C., Piazza, C. C., & Owen- quisto, J., & LeBlanc, L. A. (1998). Effectiveness of
DeSchryver, J. S. (1996). On the reinforcing effects functional communication training with and with-
of the content of verbal attention. Journal of Applied out extinction and punishment: A summary of 21
Behavior Analysis, 29, 235–238. inpatient cases. Journal of Applied Behavior Analysis,
Fisher, W. W., O’Connor, J. T., Kurtz, P. F., DeLeon, I. 31, 211–235.
G., & Gotjen, D. L. (2000). The effects of noncon- Hanley, G. P., Iwata, B. A., & Thompson, R. H. (2001).
tingent delivery of high- and low-preference stimuli Reinforcement schedule thinning following treat-
on attention-maintained destructive behavior. Jour- ment with functional communication training. Jour-
nal of Applied Behavior Analysis, 33, 79–83. nal of Applied Behavior Analysis, 34, 17–38.
Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. Hanley, G. P., Piazza, C. C., & Fisher, W. W. (1997).
P., Owens, J. C., & Slevin, I. (1992). A comparison of Noncontingent presentation of attention and alter-
two approaches for identifying reinforcers for persons native stimuli in the treatment of attention main-
with severe and profound disabilities. Journal of Ap- tained destructive behavior. Journal of Applied Be-
plied Behavior Analysis, 25, 491–498. havior Analysis, 30, 229–237.
Fisher, W. W., Piazza, C. C., Bowman, L. G., Hanley, G. Horner, R. D. (1980). The effects of an environmental
P., & Adelinis, J. D. (1997). Direct and collateral ef- enrichment program on the behavior of institution-
fects of restraints and restraint fading. Journal of Ap- alized profoundly retarded children. Journal of Ap-
plied Behavior Analysis, 30, 105–120. plied Behavior Analysis, 13, 473–491.
Fisher, W. W., Thompson, R. H., DeLeon, I. G., Piaz- Horner, R. H., Day, H. M., & Day, J. R. (1997). Using
za, C. C., Kuhn, D. E., Rodriquez-Catter, V., et al. neutralizing routines to reduce problem behaviors.
(1999). Noncontingent reinforcement: Effects of sa- Journal of Applied Behavior Analysis, 30, 601–614.
tiation versus choice responding. Research in Develop- Ingvarsson, E. T., Hanley, G. P., & Welter, K. M. (2009).
mental Disabilities, 20, 411–427. Treatment of escape-maintained behavior with posi-
Foxx, R. M., & Azrin, N. H. (1972). Restitution: A tive reinforcement: The role of reinforcement con-
method of eliminating aggressive disruptive behavior tingency and density. Education and Treatment of
of retarded and brain damaged patients. Behaviour Children, 32, 371–401.
Research and Therapy, 10, 15–27. Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., &
Foxx, R. M., & Azrin, N. H. (1973). The elimination of Richman, G. S. (1994). Toward a functional analysis
autistic self-stimulatory behavior by overcorrection. of self-injury. Journal of Applied Behavior Analysis, 27,
Journal of Applied Behavior Analysis, 6, 1–14. 197–209. (Reprinted from Analysis and Intervention in
Fritz, J., DeLeon, I. G., & Lazarchick, W. (2004). Sepa- Developmental Disabilities, 2, 3–20, 1982)
rating the influence of escape and access to preferred Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R.,
activities on problem behavior occurring in instruc- Vollmer, T. R., Smith, R. G., et al. (1994). The func-
tional contexts. Behavioral Interventions, 19, 159–171. tions of self-injurious behavior: An experimental-
Goh, H., Iwata, B. A., & DeLeon, I. G. (2000). Com- epidemiological analysis. Journal of Applied Behavior
petition between noncontingent and contingent re- Analysis, 27, 215–240.
inforcement schedules during response acquisition. Kahng, S., Iwata, B. A., Thompson, R. H., & Hanley, G.
Journal of Applied Behavior Analysis, 33, 195–205. P. (2000). A method for identifying satiation versus
Goh, H., Iwata, B. A., & Kahng, S. (1999). Multicom- extinction effects under noncontingent reinforce-
ponent assessment and treatment of cigarette pica. ment schedules. Journal of Applied Behavior Analysis,
Journal of Applied Behavior Analysis, 32, 297–316. 33, 419–432.
Greer, B. D., Fisher, W. W., Saini, V., Owen, T. M., & Kennedy, C. H., & Itkonen, T. (1993). Effects of setting
Jones, J. K. (2016). Functional communication train- events on the problem behavior of students with se-
ing during reinforcement schedule thinning: An vere disabilities. Journal of Applied Behavior Analysis,
analysis of 25 applications. Journal of Applied Behav- 26, 321–328.
ior Analysis, 49, 105–121. Kennedy, C. H., & Meyer, K. A. (1996). Sleep depri-
318 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
vation, allergy symptoms, and negatively reinforced Maglieri, K. A., DeLeon, I. G., Rodriguez-Catter, V., &
problem behavior. Journal of Applied Behavior Analy- Sevin, B. (2000). Treatment of covert food stealing
sis, 29, 133–135. in an individual with Prader–Willi syndrome. Journal
Kern, L., Koegel, R. L., & Dunlap, G. (1984). The in- of Applied Behavior Analysis, 33, 615–618.
fluence of vigorous versus mild exercise on autistic Marcus, B. A., & Vollmer, T. R. (1996). Combining
stereotyped behaviors. Journal of Autism and Devel- noncontingent reinforcement and differential rein-
opmental Disorders, 14, 57–67. forcement schedules as treatment for aberrant behav-
Kern, L., Koegel, R. L., Dyer, K., Blew, P. A., & Fenton, ior. Journal of Applied Behavior Analysis, 29, 43–51.
L. R. (1982). The effects of physical exercise on self- Mays, M. N. (2013). Using antecedent aerobic exercise to
stimulatory and appropriate responding in autistic decrease stereotypic behavior in children with autism.
children. Journal of Autism and Developmental Dis- Unpublished doctoral dissertation, Georgia State
abilities, 12, 399–419. University, Atlanta, GA.
Kettering, T. L., Fisher, W. W., Kelley, M. E., & LaRue, McComas, J., Hoch, H., Paone, D., & El-Roy, D. (2000).
R. H. (2018). Sound attenuation and preferred music Escape behavior during academic tasks: A prelimi-
in the treatment of problem behavior maintained by nary analysis of idiosyncratic establishing operations.
escape from noise. Journal of Applied Behavior Analy- Journal of Applied Behavior Analysis, 33, 479–493.
sis, 51, 687–693. McComas, J. J., Wacker, D. P., & Cooper, L. J. (1998).
Klatt, K. P., Sherman, J. A., & Sheldon, J. B. (2000). Increasing compliance with medical procedures: Ap-
Effects of deprivation on engagement in preferred plication of the high-probability request procedure
activities by persons with developmental disabilities. to a toddler. Journal of Applied Behavior Analysis, 31,
Journal of Applied Behavior Analysis, 33, 495–506. 287–290.
Kodak, T., Miltenberger, R. G., & Romaniuk, C. (2003). McComas, J. J., Wacker, D. P., Cooper, L. J., Peck, S.,
The effects of differential negative reinforcement of Golonka, Z., Millard, T., et al. (2000). Effects of the
other behavior and noncontingent escape on com- high-probability request procedure: Patterns of re-
pliance. Journal of Applied Behavior Analysis, 36, sponding to low-probability requests. Journal of De-
379–382. velopmental and Physical Disabilities, 12, 157–171.
Lalli, J. S., Casey, S. D., & Kates, K. (1997). Noncontin- McDowell, J. J. (1989). Matching theory in natural
gent reinforcement as treamtent for severe problem human environments. Behavior Analyst, 11, 95–108.
behavior: Some procedural variations. Journal of Ap- McGimsey, J. F., & Favell, J. E. (1988). The effects of
plied Behavior Analysis, 30, 127–137. increased physical exercise on disruptive behavior in
Laraway, S., Snycerski, S., Michael, J., & Poling, A. retarded persons. Journal of Autism and Developmen-
(2003). Motivating operations and terms to describe tal Disabilities, 18, 167–179.
them: Some further refinements. Journal of Applied McKenzie, S. D., Smith, R. G., Simmons, J. N., & Soder-
Behavior Analysis, 36, 407–414. lund, M. J. (2008). Suppressive effects of a stimulus
Lattal, K. A. (1972). Response–reinforcer independence correlated with reprimands for automatically main-
and conventional extinction after fixed-interval and tained eye poking. Journal of Applied Behavior Analy-
variable-interval schedules. Journal of the Experimen- sis, 41, 255–259.
tal Analysis of Behavior, 18, 133–140. Meier, A. E., Fryling, M. J., & Wallace, M. D. (2012).
Lerman, D. C., Iwata, B. A., Shore, B. A., & DeLeon, Using high-probability foods to increase the accep-
I. G. (1997). Effects of intermittent punishment on tance of low-probability foods. Journal of Applied Be-
self-injurious behavior: An evaluation of schedule havior Analysis, 45, 149–153.
thinning. Journal of Applied Behavior Analysis, 30, Michael, J. L. (2004). Concepts and principles of behavior
187–201. analysis. Kalamazoo: Western Michigan University,
Lochbaum, M. R., & Crews, D. J. (2003). Viability of Association for Behavior Analysis International.
cardiorespiratory and muscular strength programs for Miltenberger, R. G. (2006). Antecedent intervention
the adolescent with autism. Complementary Health for challenging behavior maintained by escape from
Practice Review, 8, 225–233. instructional activities. In J. Luiselli (Ed.), Anteced-
Mace, F. C., & Belfiore, P. (1990). Behavioral momen- ent assessment and intervention: Supporting children
tum in the treatment of escape-motivated stereotypy. and adults with developmental disabilities in community
Journal of Applied Behavior Analysis, 23, 507–514. settings (pp. 101–124). Baltimore: Brookes.
Mace, F. C., Hock, M. L., Lalli, J. S., West, B. J., Belfiore, Morrison, H., Roscoe, E. M., & Atwell, A. (2011). An
P., Pinter, E., et al. (1988). Behavioral momentum in evaluation of antecedent exercise on behavior main-
the treatment of noncompliance. Journal of Applied tained by automatic reinforcement using a three-
Behavior Analysis, 21, 123–141. component multiple schedule. Journal of Applied Be-
Mace, F. C., & Lalli, J. S. (1991). Linking descriptive havior Analysis, 44, 523–541.
and experimental analysis in the treatment of bi- Murphy, E. S., McSweeney, F. K., Smith, R. G., & Mc-
zarre speech. Journal of Applied Behavior Analysis, 24, Comas, J. J. (2003). Dynamic changes in reinforcer
553–562. effectiveness: Theoretical, methodological, and prac-
Developing Antecedent Interventions 319
tical implications for applied research. Journal of Ap- Poling, A., & Normand, M. (1999). Noncontingent re-
plied Behavior Analysis, 36, 421–438. inforcement: An inappropriate description of time-
Neuringer, A. (1973). Pigeons respond to produce peri- based schedules that reduce behavior. Journal of Ap-
ods in which rewards are independent of responding. plied Behavior Analysis, 32, 237–238.
Journal of the Experimental Analysis of Behavior, 19, Powers, S., Thibadeau, S., & Rose, K. (1992). Anteced-
39–54. ent exercise and its effects on self-stimulation. Behav-
Nevin, J. A. (1996). The momentum of compliance. ioral Residential Treatment, 7, 15–22.
Journal of Applied Behavior Analysis, 29, 535–547. Rapp, J. T. (2006). Toward an empirical method for
O’Callaghan, P. M., Allen, K. D., Powell, S., & Salama, identifying matched stimulation for automatically re-
F. (2006). The efficacy of noncontingent escape for inforced behavior: A preliminary investigation. Jour-
decreasing children’s disruptive behavior during re- nal of Applied Behavior Analysis, 39, 137–140.
storative dental treatment. Journal of Applied Behav- Reed, G. K., Piazza, C. C., Patel, M. R., Layer, S. A.,
ior Analysis, 39, 161–171. Bachmeyer, M. H., & Bethke, S. D. (2004). On the
O’Reilly, M. F. (1995). Functional analysis and treat- relative contributions of noncontingent reinforce-
ment of escape-maintained aggression correlated ment and escape extinction in the treatment of
with sleep deprivation. Journal of Applied Behavior food refusal. Journal of Applied Behavior Analysis, 37,
Analysis, 28, 225–226. 27–42.
Ono, K. (1987). Superstitious behavior in humans. Ringdahl, J. E., Vollmer, T. R., Marcus, B. A., & Roane,
Journal of the Experimental Analysis of Behavior, 47, H. S. (1997). An analogue evaluation of environ-
261–271. mental enrichment: The role of stimulus preference.
Pace, G. M., Ivancic, M. T., & Jefferson, G. (1994). Journal of Applied Behavior Analysis, 30, 203–216.
Stimulus fading as treatment for obscenity in a brain- Saini, V., Fisher, W. W., Retzlaff, B. J., & Keevy, M.
injured adult. Journal of Applied Behavior Analysis, 27, (2020). Efficiency in functional analysis of problem
301–305. behavior: A quantitative and qualitative review. Jour-
Pace, G. M., Iwata, B. A., Cowdery, G. E., Andree, P. J., nal of Applied Behavior Analysis, 53, 44–66.
& McIntyre, T. (1993). Stimulus (instructional) fad- Saini, V., Greer, B. D., Fisher, W. W., Lichtblau, K. R.,
ing during extinction of self-injurious escape behav- DeSouza, A. A., & Mitteer, D. R. (2016). Individual
ior. Journal of Applied Behavior Analysis, 26, 205–212. and combined effects of noncontingent reinforce-
Patel, M. R., Reed, G. K., Piazza, C. C., Bachmeyer, M. ment and response blocking on automatically rein-
H., Layer, S. A., & Pabico, R. S. (2006). An evalua- forced problem behavior. Journal of Applied Behavior
tion of a high-probability instructional sequence to Analysis, 49, 693–698.
increase acceptance of food and decrease inappro- Schaefer, H. H. (1970). Self-injurious behavior: Shaping
priate behavior in children with pediatric feeding head-banging in monkeys. Journal of Applied Behavior
disorders. Research in Developmental Disabilities, 27, Analysis, 3, 111–116.
430–442. Shabani, D. B., & Fisher, W. W. (2006). Stimulus fad-
Piazza, C. C., Adelinis, J. D., Hanley, G., P., Goh, H., ing and differential reinforcement for the treatment
& Delia, M. D. (2000). An evaluation of the effects of needle phobia in a youth with autism. Journal of
of matched stimuli on behaviors maintained by au- Applied Behavior Analysis, 39, 449–452.
tomatic reinforcement. Journal of Applied Behavior Shore, B. A., Iwata, B. A., DeLeon, I. G., Kahng, S., &
Analysis, 33, 13–27. Smith, R. G. (1997). An analysis of reinforcer substi-
Piazza, C. C., Fisher, W. W., Hanley, G. P., Hilker, K., tutability using object manipulation and self-injury
& Derby, K. M. (1996). A preliminary procedure as competing responses. Journal of Applied Behavior
for predicting the positive and negative effects of Analysis, 30, 21–40.
reinforcement-based procedures. Journal of Applied Simmons, J. N., Smith, R. G., & Kliethermes, L. (2003).
Behavior Analysis, 29, 137–152. A multiple-schedule evaluation of immediate and
Piazza, C. C., Fisher, W. W., Hanley, G. P., LeBlanc, subsequent effects of fixed-time food presentation on
L. A., Worsdell, A. S., Lindauer, S. E., et al. (1998). automatically maintained mouthing. Journal of Ap-
Treatment of pica through multiple analyses of its plied Behavior Analysis, 36, 541–544.
reinforcing functions. Journal of Applied Behavior Smith, R. G., & Iwata, B. A. (1997). Antecedent influ-
Analysis, 31, 165–189. ences of behavior disorders. Journal of Applied Behav-
Piazza, C. C., Hanley, G. P., & Fisher, W. W. (1996). ior Analysis, 30, 343–375.
Functional analysis and treatment of cigarette pica. Smith, R. G., Lerman, D. C., & Iwata, B. A. (1996). Self-
Journal of Applied Behavior Analysis, 29, 437–450. restraint as positive reinforcement for self-injurious
Piazza, C. C., Patel, M. R., Gulotta, C. S., Sevin, B. M., behavior. Journal of Applied Behavior Analysis, 29,
& Layer, S. A. (2003). On the relative contributions 99–102.
of positive reinforcement and escape extinction in Swithers, S. E., & Hall, W. G. (1994). Does oral experi-
the treatment of food refusal. Journal of Applied Be- ence terminate ingestion? Appetite, 23, 113–138.
havior Analysis, 36, 309–324. Taylor, D. V., Rush, D., Hetrick, W. P., & Sandman, C. A.
320 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
(1993). Self-injurious behavior within the menstrual Vollmer, T. R., Marcus, B. A., & Ringdahl, J. E. (1995).
cycle of developmentally delayed women. American Noncontingent escape as treatment for self-injurious
Journal on Mental Retardation, 97, 659–664. behavior maintained by negative reinforcement.
Thompson, R. F., & Spencer, W. A. (1966). Habitua- Journal of Applied Behavior Analysis, 28, 15–26.
tion: A model phenomenon for the study of neuro- Vollmer, T. R., Ringdahl, J. E., Roane, H. S., & Marcus,
nal substrates of behavior. Psychological Review, 73, B. A. (1997). Negative side effects of noncontingent
16–43. reinforcement. Journal of Applied Behavior Analysis,
Thompson, R. H., Iwata, B. A., Conners, J., & Roscoe, 30, 161–164.
E. M. (1999). Effects of reinforcement for alternative Vollmer, T. R., & Vorndran, C. M. (1998). Assessment
behavior during punishment of self-injury. Journal of of self-injurious behavior maintained by access to
Applied Behavior Analysis, 32, 317–328. self-restraint materials. Journal of Applied Behavior
Tiger, J. H., & Hanley, G. P. (2004). Developing stimulus Analysis, 31, 647–650.
control of preschooler mands: An analysis of sched- Wallace, M. D., Iwata, B. A., Hanley, G. P., Thompson,
ule-correlated and contingency-specifying stimuli. R. H., & Roscoe, E. M. (2012). Noncontingent rein-
Journal of Applied Behavior Analysis, 37, 517–521. forcement: A further examination of schedule effects
Turner, J. A., & Clancy, S. (1988). Comparison of oper- during treatment. Journal of Applied Behavior Analy-
ant behavioral and cognitive-behavioral group treat- sis, 45, 709–719.
ment for chronic low back pain. Journal of Consulting Wehby, J. H., & Hollahan, M. S. (2000). Effects of
and Clinical Psychology, 56, 261–266. high-probability requests on the latency to initiate
Van Camp, C. M., Lerman, D. C., Kelley, M. E., Con- academic tasks. Journal of Applied Behavior Analysis,
trucci, S. A., & Vorndran, C. M. (2000). Variable- 33, 259–262.
time reinforcement schedules in the treatment of Whitlow, J. W. (1975). Short-term memory in habitua-
socially maintained problem behavior. Journal of Ap- tion and dishabituation. Journal of Experimental Psy-
plied Behavior Analysis, 33, 545–557. chology: Animal Behavior Processes, 1, 196–209.
Vaughn, B. J., & Horner, R. H. (1997). Identifying in- Wilder, D. A., Normand, M., & Atwell, J. (2005).
structional tasks that occasion problem behaviors Noncontingent reinforcement as treatment for food
and assessing the effects of student versus teacher refusal and associated self-injury. Journal of Applied
choice among these tasks. Journal of Applied Behavior Behavior Analysis, 38, 549–553.
Analysis, 30, 299–312. Zarcone, J. R., Iwata, B. A., Mazaleski, J. L., & Smith,
Vollmer, T. R. (1999). Noncontingent reinforcement: R. G. (1994). Momentum and extinction effects on
Some additional comments. Journal of Applied Be- self-injurious escape behavior and noncompliance.
havior Analysis, 32, 239–240. Journal of Applied Behavior Analysis, 27, 649–658.
Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. Zarcone, J. R., Iwata, B. A., Smith, R. G., Mazaleski,
G., & Mazaleski, J. L. (1993). The role of attention in J. L., & Lerman, D. C. (1994). Reemergence and
the treatment of attention-maintained self-injurious extinction of self-injurious escape behavior during
behavior: Noncontingent reinforcement and differ- stimulus (instructional) fading. Journal of Applied Be-
ential reinforcement of other behavior. Journal of Ap- havior Analysis, 27, 307–316.
plied Behavior Analysis, 26, 9–21. Zarcone, J. R., Iwata, B. A., Vollmer, T. R., Jagtiani, S.,
Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994). Smith, R. G., & Mazaleski, J. L. (1993). Extinction of
Treatment of self-injury and hand mouthing follow- self-injurious escape behavior with and without in-
ing inconclusive functional analyses. Journal of Ap- structional fading. Journal of Applied Behavior Analy-
plied Behavior Analysis, 27, 331–344. sis, 26, 353–360.
CHAP TER 19
Extinction is the operation of discontinuing re- cause the procedures that define extinction for
sponse-contingent reinforcement, and the effect problem behavior dictate its functional properties
of extinction is the reduction in responding that (Iwata, Pace, Cowdery, & Miltenberger, 1994).
follows this operation (Catania, 2007). Identifica- Iwata, Pace, Cowdery, et al. (1994), for example,
tion of the reinforcer that maintains problem be- examined the effects of extinction based on and
havior is the critical first step in the development not based on the function of problem behavior.
of an extinction treatment. In fact, implementing We explain the results for one participant to illus-
extinction with any degree of certainty without trate their findings. Results of a functional analysis
identifying the reinforcer for problem behavior is showed that attention functioned as social positive
not possible. reinforcement for head banging. The investigators
Iwata, Dorsey, Slifer, Bauman, and Richman tested the effect of two types of putative extinc-
(1982/1994) described a functional-analytic pro- tion: attention extinction, in which the investiga-
cedure that improved the efficiency and specific- tor no longer delivered attention when the partici-
ity with which investigators could identify the pant engaged in head banging; and extinction of
reinforcer(s) for problem behavior. Iwata et al. used automatic reinforcement, in which the investiga-
functional analyses to assesses the sensitivity of tor used a helmet to attenuate the sensory con-
self-injurious behavior (SIB) of participants with sequences produced by head banging. Attention
developmental disabilities to (1) socially mediated extinction but not extinction of automatic rein-
positive reinforcement, (2) socially mediated nega- forcement decreased head banging. These findings
tive reinforcement, and (3) automatic reinforce- showed that (1) the functional analysis identified
ment. Subsequently, investigators have conducted the reinforcer that maintained head banging, (2)
functional analyses across many target behaviors head banging decreased when the investigator
(e.g., aggression, property destruction, elopement, discontinued delivery of the functional reinforcer
tantrums) and populations (e.g., individuals with for head banging, and (3) head banging did not
autism spectrum disorder, with dementia, with decrease when the investigator discontinued de-
traumatic brain injury, and without disabilities; see livery of a stimulus that was not a functional rein-
Beavers & Iwata, 2013, for a comprehensive review). forcer. Thus the functional analysis was useful for
Pretreatment functional analyses have im- prescribing the correct form and function of the
proved the efficacy of extinction treatments, be- extinction treatment.
321
322 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
In this chapter, we discuss the three functional decreased to near-zero rates. The results of Wolf
variations of extinction for problem behavior de- et al. are difficult to interpret, however, because
scribed by Iwata, Pace, Cowdery, et al. (1994): the procedure was structurally similar to time out.
extinction of problem behavior maintained by That is, Wolf et al. placed the child in social isola-
(1) socially mediated positive reinforcement, (2) tion (time out), which also resulted in discontinu-
socially mediated negative reinforcement, and (3) ing attention delivery (extinction). Thus, whether
automatic reinforcement. In addition, we discuss the discontinuation of attention, placement into
response patterns associated with extinction, fac- social isolation, or both was responsible for the ef-
tors influencing the efficacy of extinction, and fects is not clear.
other practical considerations. Lovaas and Simmons (1969) hypothesized that
social consequences reinforced the problem be-
havior of two children. Lovaas and Simmons iso-
FUNCTIONAL VARIATIONS OF EXTINCTION lated each child in a room noncontingently, which
resulted in relatively high rates of problem behav-
Extinction of Problem Behavior Maintained
ior initially, followed by a gradual decrease in rates
by Socially Mediated Positive Reinforcement
of problem behavior over time. These results sup-
Positive reinforcement is both the operation of pre- ported the presumption that social positive rein-
senting a stimulus (i.e., a positive reinforcer) con- forcement maintained problem behavior and the
tingent on the occurrence of a behavior, and the efficacy of extinction as treatment.
resulting increase in responding that produced the Mazaleski, Iwata, Vollmer, Zarcone, and Smith
reinforcer (Catania, 2007). Socially mediated rein- (1993) evaluated the effects of differential rein-
forcement is that which another individual delivers forcement of other behavior (DRO) and extinction
(Iwata, Pace, Dorsey, et al., 1994). Extinction is a on SIB emitted by three women living in a state
logical approach for decreasing problem behavior residential facility for individuals with disabilities.
maintained by social positive reinforcement, be- The pretreatment functional analysis indicated
cause the source of reinforcement is delivered by that socially mediated positive reinforcement (e.g.,
other people (hence the term social). If individu- attention) maintained each woman’s SIB. The in-
als can deliver reinforcement, then we should be vestigators demonstrated that SIB persisted when
able to teach them to withhold reinforcement. For they delivered preferred stimuli (e.g., music) for
example, if access to tangible items functions as short intervals in which the women refrained from
positive reinforcement for problem behavior (e.g., SIB (i.e., DRO). However, SIB decreased when
a caregiver provides a toy when a child hits them- the necessary extinction component was in place
self), extinction would consist of discontinuing (i.e., the investigators withheld attention). These
delivery of the tangible item when problem behav- results suggested that extinction was necessary to
ior occurs. Similarly, if access to adult attention decrease problem behavior maintained by atten-
functions as positive reinforcement for problem tion and might be the critical component of DRO
behavior (e.g., a caregiver says, “Stop,” when the schedules.
child has a tantrum), extinction would consist of
discontinuing attention when problem behavior
Extinction of Problem Behavior Maintained
occurs.
by Socially Mediated Negative Reinforcement
Before the publication of studies on functional
analysis, investigators appeared to presume that Negative reinforcement, just like positive reinforce-
positive reinforcement maintained problem be- ment, is both an operation and a process. Negative
havior in many studies. Williams (1959), for exam- reinforcement is the removal of an aversive stimulus
ple, used extinction in the form of social isolation contingent on the occurrence of a behavior and
for a child who displayed tantrums at bedtime. the resulting increase in responding that produced
The caregiver placed the child into bed at bed- the removal of that stimulus (Catania, 2007).
time, left the room, and did not reenter. Tantrums Furthermore, negative reinforcement is socially
decreased to zero rates in eight sessions, suggest- mediated when another individual delivers rein-
ing that adult attention maintained the tantrums. forcement (Iwata, Pace, Dorsey, et al., 1994). The
Wolf, Risley, and Mees (1964) used social isolation functional-analytic approach has demonstrated
for one child’s tantrums. The investigators imme- that a high proportion of problem behavior is sensi-
diately placed the child in his hospital room and tive to negative reinforcement, such as escape from
left him alone until the tantrum ceased; tantrums bite presentations (e.g., Bachmeyer, Kirkwood,
Developing Function-Based Extinction Procedures 323
Food refusal is a good example of a clinical dis- validated the results of the functional analysis
order that investigators have treated effectively by matching treatment to one (i.e., attention ex-
by using escape extinction as one component of tinction with escape and escape extinction with
an intervention. In most studies, reinforcement is attention) or both (i.e., attention and escape ex-
available for alternative behavior as part of a treat- tinction) functions of inappropriate mealtime be-
ment package, and investigators have not identi- havior. They determined that attention extinction
fied the function of the behavior (Cooper et al., alone did not decrease inappropriate mealtime be-
1995; Hoch, Babbitt, Coe, Krell, & Hackbert, havior or increase bite acceptance. In comparison,
1994; Piazza, Patel, Gulotta, Sevin, & Layer, 2003). escape extinction alone effectively decreased in-
For example, Patel, Piazza, Martinez, Volkert, and appropriate mealtime behavior and increased bite
Santana (2002) compared the effects of delivering acceptance, even though inappropriate mealtime
reinforcement for acceptance or for mouth clean, behavior produced delivery of attention. However,
which is a product measure of swallowing. The the combined attention and escape extinction
investigators found that differential reinforcement procedure decreased inappropriate mealtime be-
without escape extinction was ineffective in in- havior to near-zero levels and increased bite accep-
creasing acceptance or mouth clean or decreasing tance to 100% of trials. These pediatric feeding
inappropriate mealtime behavior. Second, they studies show that escape extinction may be a nec-
added escape extinction in the form of nonre- essary component in the treatment of food refusal;
moval of the spoon or cup to the reinforcement they also emphasize the importance of identifying
procedures, such that inappropriate mealtime be- functional relations to inform the successful treat-
havior no longer produced escape. A feeder held ment of inappropriate mealtime behavior.
the spoon or cup touching or near a child’s lips There is no definitive study on the effects of
until the child opened their mouth and allowed escape extinction, because early studies on extinc-
the feeder to deposit the bite or drink. The feeder tion of negatively reinforced behavior generally
re-presented expelled food or drink. Acceptance confounded it with other treatment components
and mouth clean increased for the three partici- (e.g., punishment, reinforcement), and more recent
pants, independently of the target of differential approaches have used extinction in combination
reinforcement, after the feeder implemented the with procedures derived from a functional analysis
escape extinction procedure. (e.g., differential negative reinforcement). A pure
In a similar study, Piazza, Patel, et al. (2003) application of escape extinction would involve a
compared the individual effects of positive rein- continuation of the aversive stimulation with-
forcement alone, escape extinction alone, and the out the introduction of some other variable (e.g.,
combined effect of positive reinforcement with es- physical guidance). Therefore, from a conceptual
cape extinction on food and liquid refusal exhibit- standpoint, future investigators should isolate the
ed by four children with total food refusal. Results effects of continuation of aversive stimulation and
indicated that positive reinforcement alone did physical guidance.
not increase acceptance or decrease inappropri-
ate mealtime behavior, whereas escape extinction
Extinction of Problem Behavior Maintained
alone increased acceptance. However, the inves-
by Automatic Reinforcement
tigators showed that the addition of positive re-
inforcement to escape extinction produced lower The defining characteristic of automatic rein-
levels of inappropriate mealtime behavior relative forcement is that reinforcement is not mediated
to escape extinction alone for some participants in socially; rather, the problem behavior produces
some phases. reinforcement (Vaughan & Michael, 1982). Note
Other investigators have conducted functional that automatic reinforcement can be positive or
analyses of inappropriate mealtime behavior and negative (Hagopian, Rooker, & Zarcone, 2015:
used the results to develop a function-based treat- Hagopian, Rooker, Zarcone, Bonner, & Arevalo,
ment (e.g., Bachmeyer et al., 2009; Berth et al., 2017). For example, head hitting can be reinforced
2019; Girolami & Scotti, 2001; Piazza, Fisher, et by endorphin release (positive reinforcement;
al., 2003). Bachmeyer et al. (2009) treated four Sandman & Hetrick, 1995), or self-scratching may
children whose inappropriate mealtime behav- attenuate an itching sensation (negative reinforce-
ior was sensitive to both social positive and so- ment). Investigators have referred to automatically
cial negative reinforcement identified by a pre- reinforced problem behavior as self-stimulation
treatment functional analysis. The investigators (e.g., Lovaas, Newsom, & Hickman, 1987) or sen-
Developing Function-Based Extinction Procedures 325
sory-reinforced behavior (e.g., Rincover, 1978). The as punishment. For example, hitting a hard helmet
term automatically reinforced problem behavior is might cause pain to an individual’s hand.
preferred, because it best describes behavior that Mazaleski, Iwata, and Rodgers (1994) con-
is not maintained socially and may be a positive- ducted a functional analysis showing that hand
or negative-reinforcement contingency. On the mouthing was automatically reinforced, and that
other hand, the terms self-stimulation and sensory- noncontingent and contingent application of oven
reinforced behavior leave out the possibility that mitts decreased hand mouthing. Mazaleski et al.
the behavior may serve a negative-reinforcement proposed that the effects of noncontingent and
function. contingent mitts may have been a function of (1)
Extinction of automatically reinforced problem the aversive properties of wearing mitts (e.g., mitts
behavior is likely to be more challenging than ex- were uncomfortable) or (2) time out from the op-
tinction of socially reinforced problem behavior. portunity to obtain automatic reinforcement from
The social environment controls socially medi- hand mouthing. Another possibility was that sen-
ated reinforcement (Vollmer, 1994); therefore, sory extinction accounted for the effect of non-
investigators can rearrange the environment to contingent mitts, and that punishment or time
discontinue reinforcement. By contrast, problem out accounted for the effect of contingent mitts
behavior rather than the social environment pro- (Mazaleski et al., 1994).
duces automatic reinforcement; thus automatically Response blocking is another procedure in-
reinforced problem behavior is more difficult to vestigators have conceptualized as extinction for
control via changes in the environment. automatically reinforced problem behavior (Saini
One approach to extinction of automatically et al., 2016). Presumably, blocking prevents the au-
reinforced behavior is called sensory extinction tomatic reinforcement produced by the response;
(Rincover, Cook, Peoples, & Packard, 1979). The however, extinction may not explain the effects
term is a misnomer, because the problem behavior of response blocking in all cases. Lerman and
producing reinforcement is extinguished, rather Iwata (1996a) varied the number of blocked SIB
than the sensory reinforcement. Nonetheless, the responses systematically to distinguish between
term has gained widespread usage, so we use it here extinction and punishment effects. They hypoth-
for consistency. During sensory extinction, the in- esized that intermittent response blocking would
vestigator terminates or blocks the putative source produce effects like an intermittent schedule of
of automatic reinforcement. For example, Rincov- reinforcement if response blocking functioned
er (1978) reduced stereotypical object spinning by as extinction. If so, intermittent response block-
blocking the auditory feedback produced by the ing would maintain or increase the frequency of
object. Rincover put carpet on the table where the SIB. Alternatively, the same intermittent sched-
participant spun the object, thereby eliminating ule of response blocking would produce effects
the sound the object produced. like an intermittent schedule of punishment if
Investigators have used sensory-extinction-like response blocking functioned as punishment. If
approaches as treatments for SIB. For example, so, intermittent response blocking would decrease
Dorsey, Iwata, Reid, and Davis (1982) used a hel- the frequency of SIB. Lerman and Iwata (1996a)
met to block the sensory products of head hitting. found that response blocking functioned as pun-
There are difficulties, however, in interpreting the ishment for an adult female engaging in chronic
results of some sensory-extinction studies. The hand mouthing, because rates of hand mouthing
source or sources of reinforcement that protective were near zero with intermittent and continuous
equipment altered are unclear, because investiga- response blocking. Alternatively, Smith, Russo,
tors implemented treatment in the absence of a and Le (1999) examined the function of response
functional analysis. For example, Rincover and blocking on the eye-poking behavior of an adult
Devany (1982) used a helmet with one partici- female. Levels of eye poking increased relative to
pant and noted decreases in head banging, pre- baseline during intermittent blocking, like the ef-
sumably as a function of extinction of automatic fect of an intermittent schedule of reinforcement.
reinforcement. An alternative explanation is that By contrast, levels of eye poking decreased during
teacher attention maintained head banging, and continuous blocking, consistent with an extinc-
the teacher ignored head banging when the child tion schedule.
wore a helmet, but not when he did not wear a Taken together, these results indicate that re-
helmet (Vollmer, 1994). Another interpretation is sponse blocking can function as extinction in
that application of protective equipment functions some cases and as punishment in others. Response
326 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
blocking would be contraindicated, then, if the that extinction bursts occurred in 36% of 113 pub-
investigator could not implement it with near- lished and unpublished cases when investigators
perfect integrity if response blocking functioned used extinction in isolation. By contrast, extinc-
as extinction. By contrast, intermittent response tion bursts occurred in only 12% of cases when
blocking still might be effective if response block- investigators combined extinction with other pro-
ing functioned as punishment. Thus a distinction cedures (e.g., differential reinforcement). Lerman,
between the extinction and punishment effects of Iwata, and Wallace (1999) found that extinction
response blocking is of clinical significance. bursts occurred in 62% of cases when investigators
used extinction in isolation, and in 39% of cases
when investigators used extinction with other
RESPONSE PATTERNS ASSOCIATED procedures. The Lerman et al. data are especially
WITH EXTINCTION important, because investigators implemented
treatment in a highly controlled environment
Skinner (1938) demonstrated in the animal labo- where extinction was only in effect during data
ratory that operant behavior decreases in frequen- collection for the study. Thus these data presum-
cy, duration, and intensity when the investigator ably measured levels of responding at the onset
terminates the delivery of reinforcement. Inves- of extinction, which is the only time extinction
tigators have observed these patterns with SIB bursts can occur. Lerman and Iwata, by contrast,
(Roscoe, Iwata, & Goh, 1998), nighttime sleep conducted their study on an inpatient unit where
disruptions (Blampied & France, 1993), and bi- clinicians might have been implementing extinc-
zarre vocalizations (Wilder, Masuda, O’Connor, tion throughout the day. Lerman and Iwata col-
& Baham, 2001), to name a few. In addition, in- lected data during short sessions; thus those data
vestigators have observed other response patterns might not have measured levels of responding at
associated with extinction—some of which are the onset of extinction. This difference in the im-
undesirable, such as the extinction or response plementation of extinction across the two studies
burst, response variation, aggression, emotional could account for the larger percentage of cases of
behavior, and spontaneous recovery. We discuss extinction bursts in Lerman et al. More generally,
these five extinction-associated response pat- investigators have found that extinction bursts are
terns because of their clinical relevance. Behavior more likely with extinction in isolation and less
analysts should recognize and anticipate these re- likely with extinction combined with other pro-
sponse patterns and prepare for them accordingly cedures.
(e.g., provide caregivers with protective equipment
in case of extinction-induced aggression).
Extinction‑Induced Response Variation
Another commonly reported response pattern
Extinction Burst
associated with extinction is response variation,
The extinction burst is the temporary increase in which is an increased tendency for novel or diverse
the frequency, intensity, or duration of a target re- behavior to occur during extinction. For example,
sponse that can occur with the onset of extinction if an adult denies a child access to reinforcement
(Cooper, Heron, & Heward, 1987). Extinction after the child asks politely for reinforcement,
bursts can be detrimental for at least two reasons. the child may engage in an alternative behavior
First, a temporary increase in the frequency, in- (e.g., whining) to access reinforcement. At times,
tensity, or duration of problem behavior may pro- extinction-induced response variation is desirable.
duce greater injury or destruction if the problem For example, response variation may produce suc-
behavior is dangerous (e.g., aggression). Second, cessive approximations of behavior that an inves-
caregivers may be less inclined to follow through tigator can differentially reinforce to shape new
with treatment if it produces a worsening of be- and desired response forms (Grow, Kelley, Roane,
havior, even if that worsening is temporary (Slo- & Shillingsburg, 2008). Response variation can be
man et al., 2005). desirable during treatment of problem behavior if
Although textbooks frequently present the ex- appropriate behavior emerges. On the other hand,
tinction burst as a common side effect, published response variation is undesirable if other forms of
data suggest that extinction bursts may not be as problem behavior emerge.
prevalent as once believed (e.g., Lerman & Iwata, To date, no studies have examined response
1995; Lerman, Iwata, & Wallace, 1999; Woods & variation explicitly during implementation of ex-
Borrero, 2019). Lerman and Iwata (1995) reported tinction for problem behavior. Nonetheless, basic
Developing Function-Based Extinction Procedures 327
research has reported response variation for de- Research on response hierarchies provides addi-
cades (e.g., Antonitis, 1951). Some applied research tional evidence for a functional-response-class in-
has reported response variation when the investi- terpretation of extinction-induced aggression. For
gators implemented extinction, although response example, Richman, Wacker, Asmus, Casey, and
variation was not the focus of the study. For ex- Andelman (1999) conducted functional analyses
ample, Lerman, Kelley, Vorndran, and Van Camp in which the reinforcement contingency for mild
(2003) blocked a participant’s head and tooth (e.g., screaming, disruption) and severe (e.g., ag-
tapping, and hand wringing increased. Similarly, gression) forms of problem behavior was the same,
Lerman, Kelley, Van Camp, and Roane (1999) and they only observed mild problem behavior.
implemented extinction for screaming, and hand Severe problem behavior emerged when Richman
clapping increased. Lerman, Kelley, et al. (1999) et al. implemented extinction for mild problem
then provided reinforcement for hand clapping as behavior but continued to deliver reinforcement
an alternative response. One explanation of these for severe problem behavior. They found that the
results is that the newly emerged responses were re- same reinforcer maintained mild and severe prob-
sults of extinction-induced response variation. An lem behavior. Two of the three participants had a
alternative explanation is that the newly emerged response pattern indicative of a response class hi-
responses were members of the same functional re- erarchy in which mild problem behavior typically
sponse class. That is, the newly emerged responses occurred before severe problem behavior. These
historically produced the same reinforcement data suggest that when severe forms of problem be-
as the response on extinction. More research is havior continue to produce reinforcement, such as
needed to determine whether extinction-induced when a caregiver cannot withhold reinforcement,
response variation occurs in applied settings. terminating the reinforcement contingency for
less severe forms is likely to produce a concomitant
increase in more severe forms of problem behavior.
Extinction‑Induced Aggression
This increased likelihood of aggression and much
Basic research has shown that aggression can more severe problem behavior when other problem
emerge as a result of withholding previously pre- behavior is placed on extinction highlights the
sented contingent or noncontingent reinforce- importance of protecting not only the client, but
ment (e.g., Azrin, Hutchinson, & Hake, 1966). therapists, caregivers, and other individuals who
The evidence for extinction-induced aggression in might be affected by implementing an extinction
humans is fairly clear (e.g., Frederiksen & Peter- procedure.
son, 1974; Kelly & Hake, 1970; Todd, Morris, &
Fenza, 1989). For example, Kelly and Hake (1970)
Extinction‑Induced Emotional Behavior
examined the effect of extinction on a punching
response in laboratory participants. Participants Extinction-induced emotional behavior is another
exhibited few punches when button pressing was pattern of responding associated with extinction
effective as an avoidance response. Punching in- (Lerman & Iwata, 1996b). Such behavior in hu-
creased for seven of nine participants when button mans can include crying, attempting to escape,
pressing was on extinction. protesting, acting upset, and other forms of agita-
An alternative explanation of extinction- tion (e.g., Baumeister & Forehand, 1971; Rovee-
induced aggression is that the emergent aggres- Collier & Capatides, 1979; Sullivan, Lewis, &
sive behavior is a member of the same functional Alessandri, 1992). For example, Sullivan et al.
response class as the behavior on extinction. If examined the emotional responding of 4-month-
so, aggression may emerge when other problem old infants during baseline, reinforcement, and
behavior no longer produces reinforcement. For extinction phases, in which the baseline and ex-
example, Vollmer, Iwata, Zarcone, Smith, and tinction phases were procedurally identical except
Mazaleski (1993) reported that noncontingent re- that baseline preceded and extinction followed
inforcement and extinction as treatment for SIB the learning history provided in the reinforcement
resulted in a burst of one participant’s aggression. phase. They concluded that positive reinforcement
One possibility is that aggression and SIB were in correlated with positive emotional expressions
the same functional response class; that is, the (e.g., facial expressions) and physiological mea-
same reinforcer (social attention in Vollmer et al.) sures (e.g., heart rate) when compared to baseline
maintained aggression and SIB. This explanation measures. By contrast, extinction correlated with
is speculative, as Vollmer et al. did not conduct a negative emotional responding and physiological
functional analysis of aggression. measures. Extinction-induced emotional behavior
328 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
in basic research may include whisker cleaning not entering the room, unless a child had a medi-
in rats, sniffing, and ambulating (Gallup & Al- cal problem. The investigators observed the ex-
tomari, 1969; Lerman & Iwata, 1996b). Whether pected extinction effect (i.e., eventual decrease of
emotional behavior in humans (e.g., crying) paral- nighttime disruptive behavior). However, brief in-
lels responding exhibited by nonhuman animals creases in nighttime disruptive behavior occurred
(e.g., whisker cleaning) is not clear. Crying and periodically across nights, which might have been
protesting in humans may be members of the same due to the time between extinction trials.
functional response class as problem behavior on There are practical reasons to be aware of
extinction, rather than induced by the extinction spontaneous recovery. First, if the recovery oc-
procedure itself. curs unexpectedly, caregivers may infer that the
Whatever the functional properties of emo- intervention does not work, and they may not
tional behavior may be, its occurrence in clinical continue implementing it. Clinicians should in-
contexts is important to note. Caregivers might form caregivers to expect spontaneous recovery,
interpret emotional behavior as a sign of discom- but also note that the recovery should be lower
fort (Blampied & France, 1993; France & Hudson, in magnitude and easier to eliminate each time
1990), and its occurrence could decrease the ac- it occurs. Second, clinicians can anticipate spon-
ceptability of extinction and negatively affect the taneous recovery and arrange the environment
integrity with which caregivers implement extinc- accordingly. For example, clinicians can imple-
tion. Furthermore, extinction-induced emotional ment safety precautions for severe aggression or
behavior could be indicative that the stimulus SIB, even though prior extinction trials have pro-
context is aversive and might increase the likeli- duced low or eventually even zero rates of prob-
hood of escape or avoidance behavior. Such be- lem behavior.
havior could have a negative impact if the context
is a learning environment or if the individual con-
sistently attempts to escape or avoid a caregiver. FACTORS THAT INFLUENCE THE EFFICACY
OF EXTINCTION
Spontaneous Recovery
Several factors influence the efficacy of extinc-
Spontaneous recovery refers to the reemergence of tion. Although the list of potential variables in-
previously extinguished behavior after a period fluencing extinction is long, some key variables
away from the context in which the investigator highlighted in this section include schedule of and
implemented extinction (Skinner, 1938). Evi- parameters of reinforcement during baseline; the
dence for spontaneous recovery is ample in basic availability of alternative sources of reinforcement
research (e.g., Hatton, 1965; Lewis, 1956; Miller & during treatment; and stimulus control, including
Stevenson, 1936; Youtz, 1938), and there are some rules.
reports of it in applied research (Lerman, Kelley,
et al., 1999). Skinner (1938) showed that exposing
Baseline Schedule of Reinforcement
a previously reinforced response to extinction de-
creased responding to pretraining levels by the end Resistance to extinction, which is persistence in re-
of the first exposure. When Skinner exposed the sponding during extinction, is one measure inves-
subject to extinction a second time, responding tigators use to evaluate the effects of extinction.
reappeared at the beginning of the session, with Catania (2007) defines resistance to extinction as
a gradual decrease to pretraining levels by the end the responses emitted, time elapsed, or number of
of the session. The subject repeated this pattern trials until responding reaches a predetermined
during subsequent exposures to extinction, with extinction criterion, such as the number of re-
gradually decreasing response curves. sponses emitted before 2 minutes elapses without
Spontaneous recovery is likely an important the occurrence of a response. Behavior maintained
phenomenon to consider in application. For ex- on an intermittent-reinforcement schedule gener-
ample, suppose a parent enters a child’s bedroom ally is more resistant to extinction than behavior
contingent on disruptive behavior at bedtime, maintained on a continuous-reinforcement sched-
and parent attention functions as reinforcement. ule (Ferster & Skinner, 1957). The more intermit-
France and Hudson (1990) taught caregivers to tent the schedule, the more resistant the behavior
implement extinction that consisted of eliminat- is to extinction (Lerman & Iwata, 1996b). The in-
ing attention for bedtime disruptive behavior by fluence of intermittent reinforcement is known as
Developing Function-Based Extinction Procedures 329
the partial-reinforcement extinction effect. Reviews that the effects of number of reinforcers reach an
discussing the partial-reinforcement extinction ef- asymptote after which resistance to extinction
fect suggest that the phenomenon is quite complex does not increase further, sometimes after as few
(e.g., Lerman & Iwata, 1996b; Lewis, 1960; Mack- as 100 reinforcers (Lerman & Iwata, 1996b). We
intosh, 1974). do not know if the asymptotic effect of reinforce-
Research on the partial-reinforcement extinc- ment number on resistance to extinction during
tion effect highlights this complexity. For ex- baseline applies to humans in complex environ-
ample, some investigators have used total num- ments. If so, resistance to extinction should reach
ber of responses that occur during extinction to its highest point quickly if problem behavior has
evaluate the partial-reinforcement extinction ef- occurred at a high rate in baseline.
fect (Lerman & Iwata, 1996b). However, Nevin
(1988) has argued that proportion of response
Delay to Reinforcement
rate during baseline is a more appropriate mea-
sure of this effect. The proportional decrease in Basic research has shown that the delay to rein-
responding when extinction follows an intermit- forcement before implementation of extinction
tent-reinforcement schedule could be greater than influences its effects. According to Lerman and
the proportional decrease in responding when Iwata (1996b), the literature suggests that prob-
extinction follows a continuous-reinforcement lem behavior is more resistant to extinction in
schedule. MacDonald, Ahearn, Parry-Cruwys, conditions where delays to reinforcement are un-
Bancroft, and Dube (2013) found that mean pro- predictable and variable than in conditions with
portional rates of problem behavior were higher no reinforcement delay. Conversely, resistance
when extinction followed a continuous-reinforce- to extinction is less pronounced (i.e., responding
ment schedule than when extinction followed an does not persist as much) if the reinforcer delay is
intermittent-reinforcement schedule. Although constant and predictable. Findings related to delay
most research shows that problem behavior de- have powerful implications for extinction-based
creases more quickly when extinction follows a treatments. The consequences for problem behav-
continuous-reinforcement schedule, findings are ior are often delayed, and the length of the delay is
sometimes mixed (Lerman, Kelley, Vorndran, often variable (e.g., Borrero, Vollmer, Borrero, &
Kuhn, & LaRue, 2002). Results of some studies Bourret, 2005)—conditions that would decrease
have shown that problem behavior decreases more the efficacy of extinction. Delivery of immediate
quickly when extinction follows an intermittent- reinforcement on a continuous schedule at a con-
reinforcement schedule (e.g., Siqueland, 1968). stant delay might reduce the deleterious effects of
Future investigators might examine the conditions reinforcement delay.
under which this occurs.
Magnitude of Reinforcement
Baseline Parameters of Reinforcement
Basic research shows that if we define reinforce-
Lerman and Iwata (1996b) outlined several other ment magnitude by amount, then smaller rein-
baseline parameters of reinforcement that appear forcement magnitudes during baseline produce
to influence responding during extinction. These more resistance to extinction. If we define rein-
parameters include but are not limited to the num- forcement magnitude by intensity, then larger re-
ber of reinforcers delivered, the delay to reinforce- inforcement magnitudes during baseline produce
ment, and the magnitude of reinforcement. more resistance to extinction. The clinical impli-
cation of these findings is that the magnitude or
intensity of baseline reinforcement, depending on
Number of Reinforcers
the function of problem behavior, should influence
Basic behavioral research has shown generally resistance to extinction. For example, problem be-
that the longer the acquisition period (i.e., his- havior maintained by food might be more resistant
tory of reinforcement) or the greater the density to extinction if reinforced during baseline with a
of reinforcer delivery (even if response indepen- small snack relative to a large meal. Conversely,
dent) before extinction is implemented, the more problem behavior maintained by attention might
resistant the behavior is to extinction (e.g., Fisher, be more resistant to extinction if reinforced dur-
Saini, et al., 2019; Nevin, Tota, Torquato, & Shull, ing baseline with high-intensity attention relative
1990). Nevertheless, basic research also has shown to low-intensity attention. Some applied research
330 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
Bachmeyer, M. H., Kirkwood, C. A., Criscito, A. B., Derby, K. M., Wacker, D. P., Sasso, G., Steege, M., Nor-
Mauzy, C. R., & Berth, D. P. (2019). A comparison of thup, J., Cigrand, K., et al. (1992). Brief functional as-
functional analysis methods of inappropriate meal- sessment techniques to evaluate aberrant behavior in
time behavior. Journal of Applied Behavior Analysis, an outpatient setting: A summary of 79 cases. Journal
52, 603–621. of Applied Behavior Analysis, 25, 713–721.
Bachmeyer, M. H., Piazza, C. C., Fredrick, L. D., Reed, Dorsey, M. F., Iwata, B. A., Reid, D. H., & Davis, P. A.
G. K., Rivas, K. D., & Kadey, H. J. (2009). Functional (1982). Protective equipment: Continuous and con-
analysis and treatment of multiply controlled inap- tingent application in the treatment of self-injurious
propriate mealtime behavior. Journal of Applied Be- behavior. Journal of Applied Behavior Analysis, 15,
havior Analysis, 42, 641–658. 217–230.
Baumeister, A. A., & Forehand, R. (1971). Effects of ex- Dowdy, A., Tincani, M., Nipe, T., & Weise, M. J. (2018).
tinction of an instrumental response on stereotyped Effects of reinforcement without extinction on in-
body rocking in severe retardates. Psychological Re- creasing compliance with nail cutting: A systematic
cord, 21, 235–240. replication. Journal of Applied Behavior Analysis, 51,
Beavers, G. A., & Iwata, B. A. (2013). Thirty years of 924–930.
research on the functional analysis of problem be- Epstein, R. (1983). Resurgence of previously reinforced
havior. Journal of Applied Behavior Analysis, 46, 1–21. behavior during extinction. Behaviour Analysis Let-
Berth, D. P., Bachmeyer, M. H., Kirkwood, C. A., ters, 3, 391–397.
Mauzy, C. R., Retzlaff, B. J., & Gibson, A. L. (2019). Ferster, C. B., & Skinner, B. F. (1957). Schedules of rein-
Noncontingent and differential reinforcement in the forcement. New York: Appleton-Century-Crofts.
treatment of pediatric feeding problems. Journal of Fisher, W. W., Felber, J. M., Phillips, L. A., Craig, A. R.,
Applied Behavior Analysis, 52, 622–641. Paden, A. R., & Niemeier, J. J. (2019). Treatment of
Blampied, N. M., & France, K. G. (1993). A behavioral resistance to change in children with autism. Journal
model of infant sleep disturbance. Journal of Applied of Applied Behavior Analysis, 52, 974–993.
Behavior Analysis, 26, 477–492. Fisher, W. W., Greer, B. D., Fuhrman, A. M., Saini, V.,
Borrero, C. S. W., Vollmer, T. R., Borrero, J. C., & Bour- & Simmons, C. A. (2018). Minimizing resurgence
ret, J. (2005). A method for evaluating parameters of of destructive behavior using behavioral momen-
reinforcement during parent–child interactions. Re- tum theory. Journal of Applied Behavior Analysis, 51,
search in Developmental Disabilities, 26, 577–592. 831–853.
Borrero, C. S. W., Vollmer, T. R., Borrero, J. C., Bourret, Fisher, W. W., Greer, B. D., Mitteer, D. R., Fuhrman, A.
J. C., Sloman, K. N., Samaha, A. L., et al. (2010). A., Romani, P. W., & Zangrillo, A. N. (2018). Further
Concurrent reinforcement schedules for problem evaluation of differential exposure to establishing op-
behavior and appropriate behavior: Experimental ap- erations during functional communication training.
plications of the matching law. Journal of the Experi- Journal of Applied Behavior Analysis, 51, 360–373.
mental Analysis of Behavior, 93, 455–469. Fisher, W. W., Ninness, H. A. C., Piazza, C. C., & Owen-
Borrero, J. C., & Vollmer, T. R. (2002). An application DeSchryver, J. S. (1996). On the reinforcing effects
of the matching law to severe problem behavior. Jour- of the content of verbal attention. Journal of Applied
nal of Applied Behavior Analysis, 35, 13–27. Behavior Analysis, 29, 235–238.
Cammilleri, A. P., Tiger, J. H., & Hanley, G. P. (2008). Fisher, W. W., Saini, V., Greer, B. D., Sullivan, W. E.,
Developing stimulus control of young children’s re- Roane, H. S., Fuhrman, A. M., et al. (2019). Baseline
quests to teachers: Classwide applications of multiple reinforcement rate and resurgence of destructive be-
schedules. Journal of Applied Behavior Analysis, 41, havior. Journal of the Experimental Analysis of Behav-
299–303. ior, 111, 75–93.
Carr, E. G., & Durand, V. M. (1985). Reducing behavior France, K. G., & Hudson, S. M. (1990). Behavior man-
problems through functional communication train- agement of infant sleep disturbance. Journal of Ap-
ing. Journal of Applied Behavior Analysis, 18, 111–126. plied Behavior Analysis, 23, 91–98.
Catania, A. C. (2007). Learning (4th ed.). Cornwall-on- Frederiksen, L. W., & Peterson, G. L. (1974). Schedule-
Hudson, NY: Sloan. induced aggression in nursery school children. Psy-
Cooper, J. O., Heron, T. E., & Heward, W. L. (1987). Ap- chological Record, 24, 343–351.
plied behavior analysis. Columbus, OH: Merrill. Fritz, J. N., Jackson, L. M., Stiefler, N. A., Wimberly,
Cooper, L. J., Wacker, D. P., McComas, J. J., Brown, K., B. S., & Richardson, A. R. (2017). Noncontingent
Peck, S. M., Richman, D., et al. (1995). Use of com- reinforcement without extinction plus differential
ponent analyses to identify active variables in treat- reinforcement of alternative behavior during treat-
ment packages for children with feeding disorders. ment of problem behavior. Journal of Applied Behavior
Journal of Applied Behavior Analysis, 28, 139–153. Analysis, 50, 590–599.
Cote, C. A., Thompson, R. H., & McKerchar, P. M. Fuhrman, A. M., Fisher, W. W., & Greer, B. D. (2016).
(2005). The effects of antecedent interventions and A preliminary investigation on improving functional
extinction on toddlers’ compliance during transi- communication training by mitigating resurgence
tions. Journal of Applied Behavior Analysis, 38, 235– of destructive behavior. Journal of Applied Behavior
238. Analysis. 49, 884–889.
334 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
Fyffe, C. E., Kahng, S., Fittro, E., & Russell, D. (2004). Hoch, T. A., Babbitt, R. L., Coe, D. A., Krell, D. M., &
Functional analysis and treatment of inappropriate Hackbert, L. (1994). Contingency contacting: Com-
sexual behavior. Journal of Applied Behavior Analysis, bining positive reinforcement and escape extinction
37, 401–404. procedures to treat persistent food refusal. Behavior
Gallup, G. G., & Altomari, T. S. (1969). Activity as a Modification, 18, 106–128.
postsituation measure of frustrative nonreward. Jour- Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., &
nal of Comparative and Physiological Psychology, 68, Richman, G. S. (1994). Toward a functional analysis
382–384. of self-injury. Journal of Applied Behavior Analysis, 27,
Girolami, P. A., & Scotti, J. R. (2001). Use of analog 197–209. (Reprinted from Analysis and Intervention in
functional analysis in assessing the function of meal- Developmental Disabilities, 2, 3–20, 1982)
time behavior problems. Education and Training in Iwata, B. A., Pace, G. M., Cowdery, G. E., & Milten-
Mental Retardation and Developmental Disabilities, berger, R. G. (1994). What makes extinction work:
36, 207–223. An analysis of procedural form and function. Journal
Grow, L. L., Kelley, M. E., Roane, H. S., & Shillings- of Applied Behavior Analysis, 27, 131–144.
burg, M. A. (2008). Utility of extinction-induced re- Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R.,
sponse variability for the selection of mands. Journal Vollmer, T. R., Smith, R. G., et al. (1994). The func-
of Applied Behavior Analysis, 41, 15–24. tions of self-injurious behavior: An experimental–
Hagopian, L. P., Rooker, G. W., & Zarcone, J. R. (2015). epidemiological analysis. Journal of Applied Behavior
Delineating subtypes of self-injurious behavior main- Analysis, 27, 215–240.
tained by automatic reinforcement. Journal of Applied Iwata, B. A., Pace, G. M., Kalsher, M. J., Cowdery, G. E.,
Behavior Analysis, 48, 523–543. & Cataldo, M. F. (1990). Experimental analysis and
Hagopian, L. P., Rooker, G. W., Zarcone, J. R., Bonner, extinction of self-injurious escape behavior. Journal
A. C., & Arevalo, A. R. (2017). Further analysis of of Applied Behavior Analysis, 23, 11–27.
subtypes of automatically reinforced SIB: A replica- Kelly, J. F., & Hake, D. F. (1970). An extinction induced
tion and quantitative analysis of published datasets. increase in an aggressive response with humans.
Journal of Applied Behavior Analysis, 50, 48–66. Journal of the Experimental Analysis of Behavior, 14,
Hagopian, L. P., Wilson, D. M., & Wilder, D. A. (2001). 153–164.
Assessment and treatment of problem behavior Lalli, J. S., Casey, S. D., & Kates, K. (1997). Noncontin-
maintained by escape from attention and access to gent reinforcement as treatment for severe problem
tangible items. Journal of Applied Behavior Analysis, behavior: Some procedural variations. Journal of Ap-
34, 229–232. plied Behavior Analysis, 30, 127–137.
Hanley, G. P., Iwata, B. A., & Thompson, R. H. (2001). Lalli, J. S., Vollmer, T. R., Progar, P. R., Wright, C.,
Reinforcement schedule thinning following treat- Borrero, J., Daniel, D., et al. (1999). Competition
ment with functional communication training. Jour- between positive and negative reinforcement in the
nal of Applied Behavior Analysis, 34, 17–38. treatment of escape behavior. Journal of Applied Be-
Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci, havior Analysis, 32, 285–296.
S. A., & Maglieri, K. A. (1997). Evaluation of client Lerman, D. C., & Iwata, B. A. (1995). Prevalence of
preference for function-based treatment packages. the extinction burst and its attenuation during treat-
Journal of Applied Behavior Analysis, 30, 459–473. ment. Journal of Applied Behavior Analysis, 28, 93–94.
Hanley, G. P., Piazza, C. C., Fisher, W. W., & Maglieri, Lerman, D. C., & Iwata, B. A. (1996a). A methodology
K. A. (2005). On the effectiveness of and preference for distinguishing between extinction and punish-
for punishment and extinction components of func- ment effects associated with response blocking. Jour-
tion-based interventions. Journal of Applied Behavior nal of Applied Behavior Analysis, 29, 231–233.
Analysis, 38, 51–65. Lerman, D. C., & Iwata, B. A. (1996b). Developing a
Hatton, G. I. (1965). Drive shifts during extinction: Ef- technology for the use of operant extinction in clini-
fects on extinction and spontaneous recovery of bar- cal settings: An examination of the basic and ap-
pressing behavior. Journal of Comparative and Physi- plied research. Journal of Applied Behavior Analysis,
ological Psychology, 59, 385–391. 29, 345–382.
Heidorn, S. D., & Jensen, C. C. (1984). Generalization Lerman, D. C., Iwata, B. A., & Wallace, M. D. (1999).
and maintenance of the reduction of self-injurious Side effects of extinction: Prevalence of bursting and
behavior maintained by two types of reinforcement. aggression during the treatment of self-injurious be-
Behaviour Research and Therapy, 22, 581–586. havior. Journal of Applied Behavior Analysis, 32, 1–8.
Herrnstein, R. J. (1974). Formal properties of the match- Lerman, D. C., Kelley, M. E., Van Camp, C. M., &
ing law. Journal of the Experimental Analysis of Behav- Roane, H. S. (1999). Effects of reinforcement mag-
ior, 21, 159–164. nitude on spontaneous recovery. Journal of Applied
Hineline, P. N. (1977). Negative reinforcement and Behavior Analysis, 32, 197–200.
avoidance. In W. K. Honig & J. E. R. Staddon (Eds.), Lerman, D. C., Kelley, M. E., Vorndran, C. M., Kuhn,
Handbook of operant behavior (pp. 364–414). Engle- S. A. C., & LaRue, R. H., Jr. (2002). Reinforcement
wood Cliffs, NJ: Prentice-Hall. magnitude and responding during treatment with
Developing Function-Based Extinction Procedures 335
differential reinforcement. Journal of Applied Behav- L. (1990). Alternative reinforcement increases resis-
ior Analysis, 35, 29–48. tance to change: Pavlovian or operant contingen-
Lerman, D. C., Kelley, M. E., Vorndran, C. M., & Van cies? Journal of the Experimental Analysis of Behavior,
Camp, C. M. (2003). Collateral effects of response 53, 359–379.
blocking during the treatment of stereotypic behav- Notterman, J. M., Schoenfeld, W. N., & Bersh, P. J.
ior. Journal of Applied Behavior Analysis, 36, 119–123. (1952). A comparison of three extinction procedures
Lewis, D. J. (1956). Acquisition, extinction, and spon- following heart rate conditioning. Journal of Abnor-
taneous recovery as a function of percentage of rein- mal and Social Psychology, 47, 674–677.
forcement and intertrial intervals. Journal of Experi- Patel, M. R., Piazza, C. C., Martinez, C. J., Volkert, V.
mental Psychology, 51, 45–53. M., & Santana, C. M. (2002). An evaluation of two
Lewis, D. J. (1960). Partial reinforcement: A selective differential reinforcement procedures with escape
review of the literature since 1950. Psychological Bul- extinction to treat food refusal. Journal of Applied Be-
letin, 57, 1–28. havior Analysis, 35, 363–374.
Lovaas, I., Newsom, C., & Hickman, C. (1987). Self- Piazza, C. C., Bowman, L. G., Contrucci, S. A., Delia,
stimulatory behavior and perceptual reinforcement. M. D., Adelinis, J. D., & Goh, H. (1999). An evalu-
Journal of Applied Behavior Analysis, 20, 45–68. ation of the properties of attention as reinforcement
Lovaas, O. I., & Simmons, J. Q. (1969). Manipulation of for destructive and appropriate behavior. Journal of
self-destruction in three retarded children. Journal of Applied Behavior Analysis, 32, 437–449.
Applied Behavior Analysis, 2, 143–157. Piazza, C. C., Fisher, W. W., Brown, K. A., Shore, B. A.,
MacDonald, J. M., Ahearn, W. H., Parry-Cruwys, D., Patel, M. R., Katz, R. M., et al. (2003). Functional
Bancroft, S., & Dube, W. V. (2013). Persistence dur- analysis of inappropriate mealtime behavior. Journal
ing extinction: Examining the effects of continuous of Applied Behavior Analysis, 36, 187–204.
and intermittent reinforcement on problem behavior. Piazza, C. C., Moes, D. R., & Fisher, W. W. (1996).
Journal of Applied Behavior Analysis, 46, 333–338. Differential reinforcement of alternative behav-
Mackintosh, N. J. (1974). The psychology of animal learn- ior and demand fading in the treatment of escape-
ing. New York: Academic Press. maintained destructive behavior. Journal of Applied
Martens, B. K., & Houk, J. L. (1989). The application Behavior Analysis, 29, 569–572.
of Herrnstein’s law of effect to disruptive and on-task Piazza, C. C., Patel, M. R., Gulotta, C. S., Sevin, B. M.,
behavior of a retarded adolescent girl. Journal of the & Layer, S. A. (2003). On the relative contributions
Experimental Analysis of Behavior, 51, 17–27. of positive reinforcement and escape extinction in
Mazaleski, J. L., Iwata, B. A., & Rodgers, T. A. (1994). the treatment of food refusal. Journal of Applied Be-
Protective equipment as treatment for stereotypic havior Analysis, 36, 309–324.
hand mouthing: Sensory extinction or punishment. Pierce, W. D., & Cheney, C. D. (2013). Behavior analysis
Journal of Applied Behavior Analysis, 27, 345–355. and learning (5th ed.). New York: Psychology Press.
Mazaleski, J. L., Iwata, B. A., Vollmer, T. R., Zarcone, Pinkston, E. M., Reese, N. M., LeBlanc, J. M., & Baer,
J. R., & Smith, R. G. (1993). Analysis of the rein- D. M. (1973). Independent control of a preschool
forcement and extinction components in DRO con- child’s aggression and peer interaction by contingent
tingencies with self-injury. Journal of Applied Behavior teacher attention. Journal of Applied Behavior Analy-
Analysis, 26, 143–156. sis, 6, 115–124.
McCord, B. E., Iwata, B. A., Galensky, T. L., Ellingson, Reed, G. K., Piazza, C. C., Patel, M. R., Layer, S. A., Bach-
S. A., & Thomson, R. J. (2001). Functional analysis meyer, M. H., Bethke, S. D., et al. (2004). On the rela-
and treatment of problem behavior evoked by noise. tive contributions of noncontingent reinforcement
Journal of Applied Behavior Analysis, 34, 447–462. and escape extinction in the treatment of food refusal.
McCord, B. E., Thomson, R. J., & Iwata, B. A. (2001). Journal of Applied Behavior Analysis, 37, 27–41.
Functional analysis and treatment of self-injury as- Rehfeldt, R. A., & Chambers, M. R. (2003). Functional
sociated with transitions. Journal of Applied Behavior analysis and treatment of verbal perseverations dis-
Analysis, 34, 195–210. played by an adult with autism. Journal of Applied
McDowell, J. J. (1988). Matching theory in natural Behavior Analysis, 36, 259–261.
human environments. Behavior Analyst, 11, 95–108. Richman, D. M., Wacker, D. P., Asmus, J. M., Casey,
Michael, J. (1993). Establishing operations. Behavior S. D., & Andelman, M. (1999). Further analysis of
Analyst, 16, 191–206. problem behavior in response class hierarchies. Jour-
Miller, N. E., & Stevenson, S. S. (1936). Agitated be- nal of Applied Behavior Analysis, 32, 269–283.
havior of rats during experimental extinction and a Rincover, A. (1978). Sensory extinction: A procedure
curve of spontaneous recovery. Journal of Compara- for eliminating self-stimulatory behavior in devel-
tive Psychology, 21, 205–231. opmentally disabled children. Journal of Abnormal
Nevin, J. A. (1988). Behavioral momentum and the Child Psychology, 6, 299–310.
practical reinforcement effect. Psychological Bulletin, Rincover, A., Cook, R., Peoples, A., & Packard, D.
103, 44–56. (1979). Sensory extinction and sensory reinforce-
Nevin, J. A., Tota, M. E., Torquato, R. D., & Shull, R. ment principles for programming multiple adaptive
336 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
behavior change. Journal of Applied Behavior Analy- Brief schedule-correlated stimuli. Journal of Applied
sis, 12, 221–233. Behavior Analysis, 41, 125–130.
Rincover, A., & Devany, J. (1982). The application of Todd, J. T., Morris, E. K., & Fenza, K. M. (1989). Tempo-
sensory extinction procedures to self-injury. Journal ral organization of extinction-induced responding in
of Applied Behavior Analysis, 2, 67–81. preschool children. Psychological Record, 39, 117–130.
Roscoe, E. M., Iwata, B. A., & Goh, H. (1998). A com- Vaughan, M. E., & Michael, J. L. (1982). Automatic
parison of noncontingent reinforcement and sensory reinforcement: An important but ignored concept.
extinction as treatments for self-injurious behavior. Behaviorism, 10, 217–227.
Journal of Applied Behavior Analysis, 31, 635–646. Vollmer, T. R. (1994). The concept of automatic rein-
Rovee-Collier, C. K., & Capatides, J. B. (1979). Positive forcement: Implications for behavioral research in
behavioral contrast in 3-month-old infants on mul- developmental disabilities. Research in Developmental
tiple conjugate reinforcement schedules. Journal of Disabilities, 15, 187–207.
the Experimental Analysis of Behavior, 32, 15–27. Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R.
Saini, V., Greer, B. D., Fisher, W. W., Lichtblau, K. R., G., & Mazaleski, J. L. (1993). The role of attention in
DeSouza, A. A., & Mitteer, D. R. (2016). Individual the treatment of attention-maintained self-injurious
and combined effects of noncontingent reinforce- behavior: Noncontingent reinforcement and differ-
ment and response blocking on automatically rein- ential reinforcement of other behavior. Journal of Ap-
forced problem behavior. Journal of Applied Behavior plied Behavior Analysis, 26, 9–21.
Analysis, 49, 693–698. Vollmer, T. R., Marcus, B. A., & Ringdahl, J. E. (1995).
Sandman, C. A., & Hetrick, W. P. (1995). Opiate mech- Noncontingent escape as treatment for self-injurious
anisms in self-injury. Mental Retardation and Develop- behavior maintained by negative reinforcement.
mental Disabilities Research Reviews, 1, 130–136. Journal of Applied Behavior Analysis, 28, 15–26.
Siqueland, E. R. (1968). Reinforcement patterns and ex- Vollmer, T. R., Peters, K. P., Kronfli, F. R., Lloveras, L.
tinction in human newborns. Journal of Experimental A., & Ibañez, V. F. (2020). On the definition of dif-
Child Psychology, 6, 431–432. ferential reinforcement of alternative behavior. Jour-
Skinner, B. F. (1938). The behavior of organisms: An ex- nal of Applied Behavior Analysis, 53, 1299–1303.
perimental analysis. Englewood Cliffs, NJ: Prentice- Vollmer, T. R., Progar, P. R., Lalli, J. S., Van Camp, C.
Hall. M., Sierp, B. J., Wright, C. S., et al. (1998). Fixed-
Slocum, S. K., & Vollmer, T. R. (2015). A comparison of time schedules attenuate extinction-induced phe-
positive and negative reinforcement for compliance nomena in the treatment of severe aberrant behavior.
to treat problem behavior maintained by escape. Journal of Applied Behavior Analysis, 31, 529–542.
Journal of Applied Behavior Analysis, 48, 563–574. Weiner, H. (1970). Instructional control of human oper-
Sloman, K. N., Vollmer, T. R., Cotnoir, N. M., Borrero, ant responding during extinction following fixed-ra-
C. S. W., Borrero, J. C., Samaha, A. L., et al. (2005). tio conditioning. Journal of the Experimental Analysis
Descriptive analyses of caregiver reprimands. Journal of Behavior, 13, 391–394.
of Applied Behavior Analysis, 38, 373–383. Wilder, D. A., Masuda, A., O’Connor, C., & Baham,
Smith, R. G., & Iwata, B. A. (1997). Antecedent influ- M. (2001). Brief functional analysis and treatment of
ences on behavior disorders. Journal of Applied Behav- bizarre vocalizations in an adult with schizophrenia.
ior Analysis, 30, 343–375. Journal of Applied Behavior Analysis, 34, 65–68.
Smith, R. G., Russo, L., & Le, D. D. (1999). Distinguish- Williams, C. D. (1959). The elimination of tantrum be-
ing between extinction and punishment effects of havior by extinction procedures. Journal of Abnormal
response blocking: A replication. Journal of Applied and Social Psychology, 59, 269.
Behavior Analysis, 32, 367–370. Wolf, M., Risley, T., & Mees, H. (1964). Application
Steege, M. W., Wacker, D. P., Cigrand, K. C., & Berg, of operant conditioning procedures to the behavior
W. K. (1990). Use of negative reinforcement in the problems of an autistic child. Behaviour Research and
treatment of self-injurious behavior. Journal of Ap- Therapy, 1, 305–312.
plied Behavior Analysis, 23, 459–467. Woods, J. N., & Borrero, C. S. W. (2019). Examining
Sullivan, M. W., Lewis, M., & Alessandri, S. M. (1992). extinction bursts in the treatment of pediatric food
Cross-age stability in emotional expressions during refusal. Behavioral Interventions, 34, 307–322.
learning and extinction. Developmental Psychology, Youtz, R. E. P. (1938). Reinforcement, extinction, and
28, 58–63. spontaneous recovery in a non-Pavlovian reaction.
Tiger, J. H., Hanley, G. P., & Heal, N. A. (2006). The ef- Journal of Experimental Psychology, 22, 305–318.
fectiveness of and preschoolers’ preference for varia- Zarcone, J. R., Iwata, B. A., Smith, R. G., Mazaleski,
tions of multiple-schedule arrangements. Journal of J. L., & Lerman, D. C. (1994). Reemergence and
Applied Behavior Analysis, 41, 475–488. extinction of self-injurious escape behavior during
Tiger, J. H., Hanley, G. P., & Larsen, K. M. (2008). A stimulus (instructional) fading. Journal of Applied Be-
practical variation of a multiple-schedule procedure: havior Analysis, 27, 307–316.
CHAP TER 20
337
338 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
The second functional component of a func- lem behavior during a functional analysis in ways
tional-analysis condition is its motivating operation that decrease its probability during treatment.
(Laraway, Snycerski, Michael, & Poling, 2003), an For example, time-based delivery of highly pre-
environmental event that has two effects. One ferred items (e.g., edibles) can reduce negatively
effect of a motivating operation is to increase or reinforced problem behavior to near-zero levels
to decrease motivation for a reinforcer (e.g., skip- by abolishing the effectiveness of escape as rein-
ping lunch increases motivation for an afternoon forcement (Lomas, Fisher, & Kelley, 2010; Mevers,
snack; eating a big lunch decreases it). The other Fisher, Kelley, & Fredrick, 2014).
effect of a motivating operation is to increase or The third functional component of a function-
decrease the probability of responses that have al-analysis condition is its reinforcing consequence.
produced that reinforcer in the past (e.g., walking A behavior analyst delivers a specific consequence
to the snack room is more probable after skipping following problem behavior in each test condition
lunch and less probable after a big lunch). Estab- of a functional analysis and does so according to a
lishing operations are motivating operations that dense schedule, usually a fixed-ratio (FR) 1 sched-
increase motivation for a given reinforcer (Peter- ule. One advantage of delivering the putative re-
son, Lerman, & Nissen, 2016), and abolishing oper- inforcer following problem behavior on a dense
ations are those that decrease motivation (Laraway schedule and not delivering that consequence in
et al., 2003). the absence of problem behavior is that the con-
Each test condition of a functional analysis has tingency should be strong and salient (Vollmer,
a specific establishing operation to increase mo- Borrero, Wright, Van Camp, & Lalli, 2001).
tivation for the reinforcer associated with that Strong and salient differential contingencies in
condition. Attention deprivation in the attention the functional-analysis conditions should lead to
condition and demands in the demand condition clearer results. A second potential advantage is
increase the effectiveness of attention and escape that response rates are generally much lower under
from demands, respectively, as reinforcement FR 1 schedules than under intermittent schedules,
and serve to evoke problem behavior. Similarly, which may reduce the risks associated with severe
no stimulation is available except for stimulation self-injurious behavior (SIB) or aggressive behav-
automatically produced by problem behavior in ior. A third potential advantage of delivering the
the alone condition. By contrast, we eliminate putative reinforcer on an FR 1 schedule is that de-
the previously mentioned establishing operations creases in problem behavior may occur more rap-
from the toy-play condition, which is the control idly if extinction is a treatment component (Ler-
condition, to decrease the probability of problem man, Iwata, Shore, & Kahng, 1996).
behavior.
Hammond, Iwata, Rooker, Fritz, and Bloom
(2013) extended this logic by examining whether TREATMENTS FOR RESPONSES REINFORCED
a specific sequence of functional-analysis condi- BY SOCIAL POSITIVE REINFORCEMENT
tions could strengthen the establishing operations
in each test condition further. Participants expe- Functional-analysis research has shown that social
rienced two functional analyses that differed only positive reinforcement maintains many problem
by condition order. Therapists conducted the func- behaviors, such as aggression, SIB, pica, and prop-
tional analysis in the fixed order of ignore, atten- erty destruction (Beavers, Iwata, & Lerman, 2013;
tion, toy play, and demand in the fixed-sequence Hanley, Iwata, & McCord, 2003). For example,
functional analysis. Therapists implemented those Iwata, Pace, Dorsey, et al. (1994) found that social
same conditions in a quasi-random order in the positive reinforcement maintained the SIB of 40
other functional analysis. Only the fixed sequence of their 152 participants (26%). Social reactions to
produced a clear function for problem behavior for problem behavior that may inadvertently function
three of the participants. as positive reinforcement in the natural environ-
Understanding how establishing operations in- ment include delivery of preferred stimuli such as
fluence the probability of problem behavior dur- attention, food, toys, music, or TV.
ing a functional analysis is important not only for Function-based treatment for problem behav-
assessing the function of problem behavior but ior maintained by social positive reinforcement
also for developing an effective treatment. That generally manipulates one or more of the three
is, behavior analysts can manipulate establishing functional components of a functional analysis
operations that increase the probability of prob- described earlier (i.e., the discriminative stimulus
Developing Function-Based Reinforcement Procedures 339
that signals reinforcement for problem behavior, as an establishing operation that evokes problem
the motivating operation that establishes the ef- behavior (Lerman & Iwata, 1995). The second
fectiveness of a consequence as reinforcement for component is important because problem behav-
problem behavior, and the consequence for prob- ior becomes less probable if the behavior analyst
lem behavior). Thus one reasonable way to begin implements extinction, so that problem behavior
the process of developing a function-based treat- no longer produces attention. Behavior analysts
ment after a functional analysis has identified often use FCT to treat problem behavior that is
the reinforcer(s) for problem behavior is to ask reinforced by social consequences (e.g., access
questions related to these three functional com- to attention or tangibles, escape from demands)
ponents: because FCT typically involves withholding the
reinforcer following problem behavior and deliv-
1. How can we arrange discriminative stimuli to ering it following an appropriate communication
signal (a) the availability of reinforcement for response. Researchers have used FCT infrequently
the alternative behavior at appropriate times, for problem behavior maintained by automatic re-
and (b) the unavailability of reinforcement for inforcement because withholding reinforcers that
problem behavior? are an automatic consequence of the behavior is
2. How can we alter the relevant motivating op- often difficult (Rapp & Lanovaz, 2014).
erations to reduce the probability of problem
behavior and increase the probability of appro-
Selecting and Teaching the FCT Response
priate behavior?
3. How can we alter the reinforcement contin- Investigators have rarely described how they se-
gency to reduce problem behavior and to in- lected the FCT response (for notable exceptions,
crease appropriate behavior? see Grow, Kelley, Roane, & Shillingsburg, 2008;
Horner & Day, 1991), although FCT is one of the
We now discuss how we can use these questions most researched operant treatments (Tiger, Han-
to guide development of two commonly prescribed ley, & Bruzek, 2008). In addition, few studies have
treatments for problem behavior reinforced by so- provided clear and replicable details on how the
cial positive reinforcement: FCT and noncontin- researchers trained the FCT response (e.g., Fisher
gent reinforcement (NCR). et al., 1993; Shirley, Iwata, Kahng, Mazaleski, &
Lerman, 1997; Wacker et al., 1990). Horner and
Day (1991) studied three components of the FCT
Functional‑Communication Training
response that have direct implications for selecting
FCT typically manipulates the consequence for an appropriate response. In the first study of the
problem behavior in two important ways: (1) A investigation, they showed that a simple and less
behavior analyst delivers the consequence the effortful response was more effective as an FCT re-
functional analysis identified as the reinforcer sponse than a more complex and effortful one was.
for problem behavior contingent on an appropri- That is, problem behavior decreased more when
ate communication response (i.e., a form of dif- the FCT response was signing a single word (i.e.,
ferential reinforcement of alternative behavior), “Break”) versus a complete sentence (i.e., “I want
and (2) the behavior analyst no longer delivers to go, please”). In the second study, Horner and
that consequence contingent on problem behav- Day showed that the response was more effective
ior (i.e., operant extinction). For example, if the when reinforced on a dense schedule (i.e., FR 1)
functional analysis indicates that contingent at- versus a leaner schedule (i.e., FR 3). Finally, in the
tention reinforces problem behavior, then the be- third study, they showed that the FCT response
havior analyst teaches the participant (1) to ob- was more effective when it produced reinforce-
tain attention via an appropriate communication ment almost immediately (after a 1-second delay)
response and (2) that problem behavior no longer than when it produced the same reinforcer after a
produces attention. The first component, deliver- longer delay (20 seconds).
ing the functional reinforcer contingent on a com- DeRosa, Fisher, and Steege (2015) extended
munication response, is important because the the research by Horner and Day (1991) by show-
participant can receive frequent reinforcement via ing that another important factor in predicting
the communication response. Providing reinforce- the effectiveness of FCT is whether the therapist
ment for the communication response minimizes can control the establishing operation for problem
periods of deprivation from attention that may act behavior. First, they compared two forms of FCT
340 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
that differed according to the type of response tar- immediately (i.e., a 0-second delay) and following
geted. Therapists modeled the vocal FCT response each FCT response (i.e., an FR 1 schedule). In
at prescribed times throughout the session in the summary, the critical components are to (1) select
vocal FCT condition. Therapists in the card FCT an FCT response that is simple for the participant,
condition physically guided a card-touch FCT re- and one we can prompt and reinforce reliably; (2)
sponse on the same schedule they used to model reinforce the response on a dense schedule; and
the vocal response. Thus, the difference between (3) deliver the reinforcer as immediately as pos-
the conditions was whether the therapist could sible following the response.
(card FCT condition) or could not (vocal FCT
condition) physically guide the FCT response.
Time‑Based Delivery of the Reinforcer
Physical guidance in the card FCT condition re-
for Problem Behavior
sulted in less exposure to the establishing opera-
tion for problem behavior for both participants. Another reinforcement-based approach for the
Decreased establishing-operation exposure in the treatment of problem behavior is to deliver the
card FCT condition corresponded to lower rates of functional reinforcer on a time-based schedule
problem behavior than in the vocal FCT condi- (e.g., Fisher, DeLeon, Rodriguez-Catter, & Kee-
tion. In a follow-up study, Fisher et al. (2018) di- ney, 2004; Vollmer, Iwata, Zarcone, Smith, &
rectly manipulated exposure to the establishing Mazaleski, 1993). Some researchers have referred
operation during initial stages of FCT and showed to this approach as noncontingent reinforcement
that establishing-operation exposures in the range (NCR), but other researchers have criticized the
of 5–20 seconds evoked extinction bursts in five of label because the intended and generally observed
six applications, whereas shorter exposures to the effect is a reduction in and weakening of problem
establishing operation produced rapid reductions behavior. By contrast, we define reinforcement as
in problem behavior without extinction bursts. an increase in responding due to the contingent
We have developed the following guidelines for presentation of a positive or negative reinforcer
selecting and training the FCT response, based (Poling & Normand, 1999). Poling and Normand
on studies by Horner and Day (1991), DeRosa et (1999) suggested the term fixed-time (FT) sched-
al. (2015), and Tiger et al. (2008). The response ules. This label, however, does not acknowledge
should be simple for the participant to emit and the prior functional relation between problem be-
easily recognized by other individuals. A response havior and the stimulus delivered on a time-based
that is already in the participant’s response reper- schedule during treatment or the fact that the re-
toire is preferable to one that is not. A short re- sults of a functional analysis are used to prescribe
quest is an appropriate FCT response (e.g., “Play treatment. We use the term NCR in this chapter
with me, please”) for participants who speak in to be consistent with the relevant applied litera-
complete sentences if the participant reliably ture, while acknowledging the inconsistency in re-
imitates a model of the vocal FCT response. We ferring to a treatment to reduce problem behavior
usually teach participants who do not speak and as reinforcement.
participants who speak but do not imitate a model We typically deliver the stimulus that previ-
of a vocal FCT response readily to touch a picture ously reinforced problem behavior on a dense
card that depicts the functional reinforcer (e.g., schedule when we initiate NCR schedules. For ex-
touching a picture card that shows the participant ample, Vollmer et al. (1993) introduced NCR as
and adult playing together). We use physical guid- a treatment for SIB reinforced by attention. The
ance to help participants who do not perform the baseline was identical to the attention condition
FCT response independently, then deliver the re- of the functional analysis, in which the therapist
inforcer. We reduce exposure to the establishing delivered 10 seconds of statements of concern or
operation by delivering reinforcement even if we disapproval contingent on SIB. The researchers
guide the participant to complete the response discontinued the contingency between SIB and
initially, which helps to decrease the probability attention (i.e., extinction), and they delivered
of evoking problem behavior while we are training praise and general conversation independently of
the FCT response. Over time, we fade the physical SIB during NCR.
prompts until the participant emits the response The delivery of attention on a dense, time-based
independently during 90% of training trials for schedule removes or lessens the establishing op-
two consecutive sessions. Finally, we deliver the eration for problem behavior, which often results
reinforcer identified during the functional analysis in immediate and large reductions in respond-
Developing Function-Based Reinforcement Procedures 341
ing. For example, Hagopian, Fisher, and Legacy and evokes problem behavior that has produced
(1994) compared dense (FT 10-second) and lean attention previously. Exposure to the establishing
(FT 5-minute) schedules of NCR and found that operation may be more likely to evoke the FCT
the dense schedule produced large and immedi- response during FCT because of its recent history
ate reductions in problem behavior, whereas the of reinforcement with attention (cf. Hoffman &
lean schedule produced smaller and less consis- Falcomata, 2014).
tent reductions. Researchers have observed simi-
lar differences with larger and smaller magnitudes
of reinforcement on time-based schedules, even TREATMENTS FOR RESPONSES REINFORCED
when they implemented NCR without extinction BY SOCIAL NEGATIVE REINFORCEMENT
(Roscoe, Iwata, & Rand, 2003). Finally, Fisher et
al. (2004) showed that NCR with extinction re- In 1987, Iwata described negative reinforcement
sulted in larger and more immediate reductions in as an emerging technology in the field of applied
problem behavior than implementation of extinc- behavior analysis. Research on the unique role of
tion alone. Behavior analysts can avoid contiguous negative reinforcement in the development, main-
pairing of problem behavior and time-based rein- tenance, and treatment of problem behavior has
forcement delivery by using a changeover delay, in grown exponentially since that time, despite criti-
which the occurrence of problem behavior at the cisms that the distinction between positive and
scheduled reinforcement interval delays reinforce- negative reinforcement is ambiguous, is without
ment delivery (Herrnstein, 1961). functional significance, and should be abandoned
(Baron & Galizio, 2005; Michael, 1975).
The field has defined negative reinforcement as
Choosing between FCT and NCR
an increase in responding due to the response-
Given that research has shown that FCT and contingent removal of a stimulus. However, Baron
NCR are effective treatments for problem behav- and Galizio (2005) and Michael (1975) argued
ior, especially when combined with extinction or that we should view reinforcement as an increase
mild punishment (Hagopian, Fisher, Sullivan, Ac- in responding due to a response-contingent envi-
quisto, & LeBlanc, 1998), how should a behavior ronmental change from one stimulus condition to
analyst decide which one to use and when? Perhaps another. Their argument is based in part on the
the first consideration should be whether problem fact that a behavior analyst may have difficulty
behavior is dangerous and likely to produce im- determining whether an individual is responding
minent harm to the participant or others. NCR to terminate one event or to gain access to the op-
may be the treatment of choice if the risk of harm posite event in some circumstances. For example,
is high, because FCT generally requires an initial does a person turn up the thermostat to escape
training period, and NCR does not. NCR also from cold air or to gain access to warm air? Other
has the advantage of requiring less monitoring behavior analysts have argued that the distinction
of the participant’s ongoing behavior than FCT. between positive and negative reinforcement is
That is, the caregiver delivers reinforcement on a useful and sufficiently engrained among behavior
time-based schedule during NCR. By contrast, the analysts that abandoning the distinction may be
caregiver must monitor and respond to the partici- unwarranted; it certainly is unlikely (Iwata, 2006;
pant’s FCT response during FCT, although most Lattal & Lattal, 2006; Sidman, 2006).
participants learn to find the caregiver and recruit Our purpose in discussing this issue is twofold.
reinforcement with the FCT response. First, applied behavior analysts should recognize
Another important consideration in choosing that there is disagreement in the field on whether
between FCT and NCR is whether establishing a the distinction between positive and negative re-
communication response in and of itself is an im- inforcement is meaningful. More importantly,
portant treatment goal. In such cases, FCT is the behavior analysts on both sides of this argument
obvious choice. FCT also may have an advantage would agree that considering, describing, and ana-
over NCR when the amount of reinforcement nec- lyzing the stimulus conditions in effect before and
essary to reduce or eliminate the establishing oper- after the target response are important, regardless
ation for problem behavior is unclear. For example, of whether the analysts use the adjectives posi-
extended exposure to attention deprivation, such tive or negative to modify the term reinforcement.
as when the caregiver is on the telephone, estab- We highlight both sides of the stimulus change
lishes the effectiveness of attention as a reinforcer by (1) describing and analyzing the relevant es-
342 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
tablishing operations and discriminative stimuli scribe procedures that increase treatment practi-
present before the target response, and (2) retain- cality. Presenting instructions and delivering dif-
ing the terms positive and negative reinforcement ferential reinforcement for compliance rather than
to describe whether the stimulus change involved for communication (Marcus & Vollmer, 1995) is
the introduction (positive reinforcement) or with- another approach behavior analysts use to reduce
drawal (negative reinforcement) of a stimulus fol- problem behavior.
lowing problem behavior. Selecting compliance as the alternative re-
Functional-analysis research has shown that sponse in differential reinforcement of alternative
termination of demands often reinforces prob- behavior has several potential advantages. First,
lem behavior. In fact, Iwata, Pace, Dorsey, et al. the individual continues to receive instructions
(1994) found that 38% of their 152 participants and is more likely to learn skills that will produce
displayed SIB maintained by social negative rein- alternative sources of reinforcement. Second, the
forcement, which was a higher percentage than for continued exposure to instructions may produce
any other behavioral function. Recent reviews of habituation and make their subsequent presenta-
the functional-analysis literature also have found tion less aversive. Third, differential reinforce-
that a high percentage of problem behavior is ment of compliance can add to the effectiveness of
negatively reinforced (Beavers et al., 2013; Hanley escape extinction in unique and important ways.
et al., 2003). Other forms of negatively reinforced For example, Cataldo and colleagues showed that
problem behavior include escape from noise (Ket- reinforcement of compliance can both increase
tering, Fisher, Kelley, & LaRue, 2018; McCord, that response and concomitantly decrease prob-
Iwata, Galensky, Ellingson, & Thomson, 2001) lem behavior, even when the consequences for
and from social interaction (Hall, DeBernardis, & problem behavior remain unchanged (Parrish,
Reiss, 2006). Cataldo, Kolko, Neef, & Egel, 1986; Russo, Catal-
do, & Cushing, 1981). Conversely, treatments that
directly target problem behavior, such as extinc-
Choosing between FCT and NCR
tion, can decrease problem behavior and simul-
FCT and NCR are both reasonable and effective taneously increase compliance, even though the
treatments for problem behavior reinforced by es- consequences for compliance remain unchanged.
cape from demands or other aversive events, and Parrish et al. (1986) and Russo et al. (1981) have
the issues and considerations discussed earlier, hypothesized that compliance and problem be-
such as selecting the FCT response, are applicable havior are inverse members of an overarching
to negatively reinforced problem behavior. We response class called instruction following, which
teach individuals whose problem behavior is rein- may include these two topographically distinct re-
forced by escape to request breaks with an FCT sponses in environmental contexts in which com-
response (e.g., saying “Break, please”; touching a pliance historically has produced reinforcement
picture card showing the child leaving the work- and problem behavior historically has produced
table; Hagopian et al., 1998) when we use FCT. punishment or extinction (Parrish et al., 1986).
We deliver escape on a dense, time-based schedule Selecting potent reinforcement is important in
(Vollmer, Marcus, & Ringdahl, 1995) when we use targeting problem behavior through differential
NCR. In most cases, we generally combine FCT reinforcement of compliance. Differential nega-
or NCR with escape extinction (discussed further tive reinforcement of compliance is one approach
by Vollmer, Athens, & Fernand, Chapter 19, this to treat problem behavior maintained by negative
volume). reinforcement. The therapist delivers escape from
One limitation common to FCT and NCR instructions and provides no differential conse-
is that a participant with negatively reinforced quence for problem behavior (i.e., escape extinc-
problem behavior frequently escapes all or most tion). For example, Vollmer et al. (1995) delivered
instructional demands during initial treatment. escape in the form of breaks from work on a dense
Thus, the participant does not benefit from in- FT schedule that produced immediate and large
struction or learn new skills (Fisher et al., 1993). reductions in SIB. They subsequently thinned the
Demand or instructional fading, in which a be- schedule of time-based breaks from work to 10
havior analyst increases the number of presented minutes for one participant and 2.5 minutes for
demands gradually, is one approach to address this the other, using an instruction-fading procedure
problem. We discuss this approach later as we de- such as the one we describe below.
Developing Function-Based Reinforcement Procedures 343
tact visual sensory–neural pathways (Hyman, Fish- stimuli associated with high levels of interaction
er, Mercugliano, & Cataldo, 1990). A reasonable and low levels of problem behavior. The behav-
hypothesis regarding the function of this unique ior analyst uses the identified stimulus or stimuli
form of SIB is that it occurs because the response by presenting them to the participant on a time-
produces a visual-like sensation that functions as based schedule when the automatically reinforced
reinforcement for the response in someone who problem behavior is most likely to occur, such as
is deprived of such stimulation. However, testing “down times” and times when alternative stimula-
whether this hypothesized automatic consequence tion is unavailable.
is the functional reinforcer for SIB is difficult if not
impossible, because it cannot be manipulated dur-
ing a functional analysis or during treatment (for a IMPROVING THE PRACTICALITY
notable exception, see Rincover, 1978). OF REINFORCEMENT‑BASED TREATMENTS
One treatment approach is to provide the indi-
vidual with alternative forms of appropriate stimu- A common criticism of behavioral treatments
lation that compete with the automatic reinforcer is that they are often labor-intensive relative to
for problem behavior (Piazza et al., 1998; Wilder, drug or milieu treatments and frequently require
Draper, Williams, & Higbee, 1997). Horner (1980) caregivers to (1) continuously monitor the be-
implemented one of the first applications of this havior and (2) accurately deliver various prompts
approach, although he did not base it on a func- and consequences. Thus, behavior analysts have
tional analysis. Horner added manipulables (such worked to develop and validate treatment compo-
as puzzles, pull and push toys, and a rocking horse) nents that make behavioral treatments easier and
to an otherwise austere institutional environment more practical to implement.
and called the treatment an enriched environment.
This simple treatment increased adaptive behavior
Alternative Reinforcement
with the manipulables and reduced SIB and ste-
reotypical behavior. One approach to increasing the practicality of
Vollmer et al. (1994) refined and extended reinforcement-based treatments is to identify al-
this treatment approach by (1) prescribing the ternative reinforcers (e.g., toys) to deliver when
enriched-environment treatment based on the a caregiver would have difficulty delivering the
results of a functional analysis in which problem consequence that previously reinforced problem
behavior persisted in the absence of social contin- behavior (Austin & Tiger, 2015; Hagopian, Con-
gencies and (2) selecting the stimuli for environ- trucci Kuhn, Long, & Rush, 2005; Hanley, Piazza,
mental enrichment based on the results of a prefer- & Fisher, 1997; Rooker, Jessel, Kurtz, & Hago-
ence assessment (Fisher et al., 1992). Piazza et al. pian, 2013). This approach is especially useful for
(1998) further refined this approach by developing problem behavior reinforced by attention because
a preference assessment, called the competing-stim- caregivers are not available to deliver attention
ulus assessment, specifically designed to identify constantly (e.g., when a caregiver is in the bath-
preferred stimuli associated with high levels of in- tub). Delivering highly preferred toys noncontin-
teraction and low levels of problem behavior. The gently can decrease the probability that attention-
competing-stimulus assessment is particularly well reinforced problem behavior will reemerge when a
suited for identifying preferred stimuli for rein- caregiver is busy.
forcement-based treatments for problem behavior Several researchers have shown that providing
maintained by automatic reinforcement and for alternative reinforcers with the functional rein-
problem behavior maintained by social positive forcer (e.g., a break with preferred toys) can pro-
reinforcement (e.g., Fisher, O’Connor, Kurtz, De- duce more robust decreases in problem behavior
Leon, & Gotjen, 2000). than delivering the functional reinforcer alone
The competing-stimulus assessment is simple (Rooker et al., 2013; Zangrillo, Fisher, Greer,
and involves a series of short sessions (e.g., 2 min- Owen, & DeSouza, 2016). For example, therapists
utes each). The behavior analyst presents a single in Zangrillo et al. (2016) delivered escape (i.e., the
competing stimulus in each session, and the par- functional reinforcer for problem behavior) follow-
ticipant can interact with the competing stimulus, ing an FCT response in one condition and escape
display the automatically reinforced problem be- plus preferred toys in the other. Rates of problem
havior, or both. Observers record stimulus interac- behavior were lower and levels of compliance were
tion and problem behavior to identify one or more higher for the two participants when the escape
Developing Function-Based Reinforcement Procedures 345
interval included preferred toys. Researchers have decrease rapidly to near-zero levels. Caregivers
reported similar findings when delivering multiple, may be unable to maintain such a dense schedule
functional reinforcers instead of a single function- in the natural environment (Fisher et al., 1993).
al reinforcer (Piazza et al., 1997; Piazza, Moes, & Thus, researchers have developed procedures to
Fisher, 1996; Rooker et al., 2013). thin the reinforcement schedule effectively and
Selecting effective reinforcement is important efficiently during FCT while (1) maintaining low
regardless of whether a behavior analyst delivers levels of problem behavior; (2) decreasing the FCT
functional reinforcement or alternative reinforce- response rate, to make treatment easier and more
ment when functional reinforcement is unavail- practical to implement; and (3) bringing the FCT
able. Fisher et al. (2000) described an adaptation response under stimulus control, so that it occurs
of the competing-stimulus assessment (Piazza et at appropriate times but does not weaken because
al., 1998) for problem behavior reinforced by so- reinforcement is periodically unavailable (Fisher,
cial contingencies. Fisher et al. presented stimuli Kuhn, & Thompson, 1998; Fyffe, Kahng, Fittro, &
individually; problem behavior produced the func- Russell, 2004; Hanley, Iwata, & Thompson, 2001;
tional reinforcer, which was attention, and they Saini, Miller, & Fisher, 2016).
measured stimulus interaction and problem behav- A practical approach to thinning the rein-
ior. Stimuli associated with high levels of interac- forcement schedule is to restrict access to the
tion and low levels of problem behavior effectively picture card immediately following reinforce-
reduced problem behavior when the researchers ment delivery when the FCT response is a pic-
delivered them on a time-based schedule, and ture exchange (Fisher, Greer, Querim, & DeRosa,
problem behavior produced attention (Fisher et 2014; Greer, Fisher, Saini, Owen, & Jones, 2016;
al., 2000). Roane, Fisher, Sgro, Falcomata, & Pabico, 2004).
Another approach to identifying effective alter- The absence of the FCT response card serves as
native stimuli involves a behavior analyst’s care- a discriminative stimulus for the unavailability
fully observing the consequences produced auto- of reinforcement and prevents the occurrence of
matically by the problem behavior. The purpose the response when it is not likely to produce re-
of these observations is to generate hypotheses inforcement. For example, a child cannot request
about the nature of the stimulation produced by a caregiver’s attention if the card is unavailable
the behavior, which the behavior analyst uses to while the caregiver is changing a sibling’s diaper.
identify stimuli that produce the same or similar The picture card should be available continu-
stimulation in the absence of problem behavior. ously during initial treatment. The caregiver initi-
Researchers have called these matched stimuli ates schedule thinning after several sessions with
because they match the type of stimulation pro- high levels of the FCT response and low levels of
duced by problem behavior (Piazza, Adelinis, problem behavior by restricting access to the card
Hanley, Goh, & Delia, 2000). Behavior analysts (e.g., by placing it out of sight) for a few seconds
can conduct a competing-stimulus assessment to after the reinforcement interval. Over time, the
identify one or more matched stimuli associated caregiver progressively increases the duration of
with high levels of interaction and low levels of the picture card’s unavailability. For example, the
problem behavior after they identify a preliminary caregiver might double the duration of the pic-
set of matched stimuli. Matched stimuli sometimes ture card’s unavailability after every two sessions
produce greater reductions in automatically rein- in which levels of problem behavior remain low
forced problem behavior than alternative stimuli until reaching a practical schedule. A reason-
do (Piazza et al., 1998, 2000), perhaps because they able schedule-thinning goal would be to present
more effectively reduce the establishing operation the FCT card once every 10–15 minutes and to
for problem behavior. provide functional reinforcement for 1–3 minutes.
Terminal reinforcement schedules are likely to
vary across participants and situations. However,
Thinning the Reinforcement Schedule
reinforcement density should remain sufficiently
FCT is an effective treatment for problem behav- high throughout and after schedule thinning,
ior that behavior analysts use frequently. Initial such that the participant does not experience long
implementation is often labor-intensive, however, periods without access to the functional reinforcer
because it involves providing reinforcement for (Roane, Falcomata, & Fisher, 2007).
the FCT response on a dense schedule, to maxi- We generally use separate discriminative stimuli
mize the likelihood that problem behavior will to signal reinforcement availability when the FCT
346 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
response is a vocal operant (e.g., “Play, please”; crease the number of instructions that we require
“Break, please”). For example, we might hang a the participant to complete before we present the
large square of green or red paper alternately on discriminative stimulus and provide escape for the
the wall to signal the availability or unavailabil- FCT response. We only increase the response re-
ity of reinforcement, respectively, according to quirement (i.e., the number of instructions) if the
the current reinforcement schedule (e.g., Fisher et rate of problem behavior remains below a criterion
al., 1998). The green paper serves as the SD (pro- (e.g., 90% below the baseline mean; Hagopian et
nounced “ess-dee”) because it signals the availabil- al., 1998).
ity of reinforcement, and the red paper serves as Similarly, several procedures are available
the SD (pronounced “ess-delta”) because it signals for increasing the practicality and ease of NCR
the unavailability of reinforcement. schedules after we achieve initial reductions in
We typically conduct training trials to expose problem behavior with a dense NCR schedule. For
the participant to the contingencies correlated example, Hagopian et al. (1994) gradually thinned
with each discriminative stimulus, such as a green the dense (FT 10-second) schedule to a lean (FT
SD to signal reinforcement and a red SD to signal 5-minute) schedule and obtained near-zero rates of
extinction. We thin the reinforcement schedule problem behavior. Importantly, treatment effects
over time by increasing the duration the SD is pres- transferred from the clinic to the classroom and
ent, decreasing the duration the SD is present, or were maintained at a 2-month follow-up.
both (Hanley et al., 2001). We continue thinning
the schedule when levels of problem behavior re-
Discriminative Stimuli
main low.
Reinforcement schedule thinning for problem Developing stimulus control over the FCT re-
behavior reinforced by escape from instructions sponse so that it occurs reliably, but only during
typically involves increasing the number of in- programmed availability of functional reinforce-
structions or trials the participant must complete ment, is important for reinforcement schedule
before escape is available (e.g., a chain schedule thinning. A behavior analyst often presents a
of reinforcement), which researchers refer to as unique discriminative stimulus (e.g., green card,
instructional or demand fading (Pace et al., 1993; green bracelet) continuously while reinforcement
Zarcone et al., 1993). Instructional fading can be is available. The behavior analyst uses a different
used alone (Pace et al., 1994) or with other treat- stimulus (e.g., red card, red bracelet) to signal that
ments, such as escape extinction (Zarcone et al., reinforcement is not available (i.e., extinction).
1993) or FCT (Hagopian et al., 1998). The thera- Incorporating discriminative stimuli during rein-
pist arranges the treatment with instructional ma- forcement schedule thinning for the FCT response
terials present when instructional fading begins, can improve the efficiency and efficacy of FCT.
but the therapist delivers no instructions initially A study by Betz, Fisher, Roane, Mintz, and
(i.e., continuous noncontingent escape). When Owen (2013) compared the effects of multiple-
problem behavior remains low in the absence of and mixed-reinforcement schedules for four par-
instruction delivery, the therapist introduces a sin- ticipants who engaged in problem behavior. The
gle instruction (e.g., “Point to red”) about halfway multiple schedule included discriminative stimuli
through the session. The therapist subsequently that signaled the availability of reinforcement,
increases the number of instructions per session but the mixed schedule did not. Therapists dur-
gradually (e.g., by adding one instruction after ing the mixed-reinforcement schedule alternated
each session in which problem behavior is at least quasi-randomly between 60-second unsignaled
90% below the baseline mean; Pace et al., 1993) periods in which functional reinforcement for the
until reaching a criterion, such as a 10-minute in- FCT response was available or unavailable. The
structional period. procedure was identical during the multiple-rein-
We generally do not reduce the number of forcement schedule, with the addition of discrimi-
instructions per session when we implement in- native stimuli such as a colored bracelet, vest, or
structional fading with FCT. Rather, we intro- card to signal the availability and unavailability
duce a chained schedule in which the participant of functional reinforcement for the FCT response.
must complete one instruction first, and then we The participants showed high rates of the FCT
present a discriminative stimulus to signal rein- response during the extinction component of the
forcement availability (i.e., escape) for the FCT mixed schedule, but relatively few responses during
response. In subsequent sessions, we gradually in- the extinction component of the multiple sched-
Developing Function-Based Reinforcement Procedures 347
ule, indicating that the stimuli facilitated discrimi- emitted considerably fewer problem behaviors in
nation between the FCT components. Next, the the condition in which the researchers signaled
researchers evaluated whether they could thin the extinction (i.e., SD present) relative to the condi-
reinforcement schedule rapidly with a multiple tion in which they did not (i.e., SD absent). Fisher,
schedule. The participants showed near-perfect Fuhrman, Greer, Mitteer, and Piazza (2020) rep-
discrimination between the components during licated and extended these findings with four ad-
rapid thinning of the schedule when discrimina- ditional participants by isolating the effects of the
tive stimuli signaled the availability and unavail- discriminative stimuli and applying those stimuli
ability of reinforcement. in a new context. These studies are important
The Betz et al. (2013) results are important be- because they show that behavior analysts can use
cause they show that behavior analysts can forgo discriminative stimuli to mitigate or even prevent
lengthy schedule thinning under some conditions treatment relapse when the functional reinforcer
and arrive more quickly at terminal FCT schedules is unavailable for extended periods.
if they establish discriminative control first. Fisher,
Greer, Fuhrman, and Querim (2015) recently ex-
tended the Betz et al. findings. They showed that ADDRESSING COMMON PROBLEMS
other individuals could implement treatment in IN REINFORCEMENT‑BASED TREATMENTS
other contexts with little loss of treatment effects
after participants readily discriminated between Hagopian et al. (1998) reviewed 21 cases in
the multiple-reinforcement-schedule components which they implemented FCT-based treatments
during FCT. That is, the FCT treatment proved and found that FCT was effective in most when
portable and easily transferred to other individuals combined with extinction. However, FCT plus
and other settings. Greer et al. (2019) replicated extinction lost its effectiveness in about one-half
and extended these findings by showing that the of the cases when they thinned the reinforce-
use of discriminative stimuli facilitated the rapid ment schedule. Similarly, Volkert, Lerman, Call,
transfer of treatment effects from behavior thera- and Trosclair-Lasserre (2009) observed resurgence
pists to the participants’ caregivers. of problem behavior in most cases when they
Greer et al. (2016) described other potential ad- placed the FCT response on extinction or rapidly
vantages of establishing stimulus control initially decreased the frequency of reinforcement for the
during FCT. The researchers summarized 25 con- response from an FR 1 to an FR 12 schedule. As
secutive applications of FCT in which they used previously mentioned, signaling extended periods
discriminative stimuli to thin the reinforcement in which reinforcement will be unavailable can
schedule. Results indicated a 96% reduction in help to mitigate resurgence (e.g., Fuhrman et al.,
levels of problem behavior from baseline, relatively 2016). In addition, providing substitute reinforcers
few FCT responses (8%) during extinction, and may also prevent occurrences of problem behavior
only one (4%) application of punishment. when the functional reinforcer is unavailable (e.g.,
Perhaps some of the most fascinating data on Fisher et al., 2000; Rooker et al., 2013). However,
discriminative stimuli during FCT show that adding a function-based punishment component
such stimuli can reduce the probability of treat- may be warranted when these reinforcement-based
ment relapse. Fuhrman, Fisher, and Greer (2016) treatments are ineffective (Hagopian et al., 1998;
exposed two participants who engaged in problem see Lerman & Toole, Chapter 21, this volume, for
behavior to two sequences of phases in which they a detailed discussion of this issue).
implemented FCT with or without discriminative
stimuli. After rates of problem behavior decreased
with each FCT type, researchers then conducted CONCLUDING COMMENTS
extinction sessions in which FCT responses never
produced reinforcement. Researchers used the Functional-reinforcement-based treatments are
same discriminative stimulus that they used dur- likely to be efficient and effective, because results
ing FCT (i.e., the SD) during the extinction ses- of the functional analysis provide specific infor-
sions that followed FCT, to determine whether mation that directs treatment development to rel-
using the discriminative stimulus to signal ex- evant antecedents and consequences. That is, the
tinction during some phases (i.e., SD present) but functional-analysis results allow a behavior ana-
not others (i.e., SD absent) would reduce the re- lyst to focus the treatment on the contexts (e.g.,
emergence of problem behavior. Both participants demand contexts when escape is the reinforcer,
348 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
low-attention contexts when attention is the re- DeLeon, I. G., Neidert, P. L., Anders, B. M., & Rodri-
inforcer), rather than implementing a treatment guez-Catter, V. (2001). Choices between positive and
across environmental contexts. Focusing on one or negative reinforcement during treatment for escape-
a few specific contexts allows for easier treatment maintained behavior. Journal of Applied Behavior
Analysis, 34, 521–525.
implementation in the natural environment. The
DeRosa, N. M., Fisher, W. W., & Steege, M. W. (2015).
functional analysis also specifies a relatively small An evaluation of time in establishing operation
number of procedures to reduce problem behavior on the effectiveness of functional communication
(e.g., eliminate the contingency between problem training. Journal of Applied Behavior Analysis, 48,
behavior and its reinforcer) and increase appropri- 115–130.
ate alternative behavior (e.g., provide access to Dunlap, G., Kern-Dunlap, L., Clarke, S., & Robbins, F.
the functional reinforcer following an appropriate R. (1991). Functional assessment, curricular revision,
communication response). This leads to more effi- and severe behavior problems. Journal of Applied Be-
cient treatment development, in that results of the havior Analysis, 24, 387–397.
Durand, V. M., Berotti, D., & Weiner, J. (1993). Func-
functional analysis prescribe a specific treatment
tional communication training: Factors affecting ef-
and avoid a trial-and-error approach to selecting fectiveness, generalization, and maintenance. In J.
treatments. Reichle & D. P. Wacker (Eds.), Communicative alter-
In conclusion, accurate functional-analysis re- natives to challenging behavior: Integrating functional
sults promote quicker development of effective assessment and treatment strategies (pp. 317–340). Bal-
treatments; easier implementation of the treat- timore: Brookes.
ment in the natural environment, by specifying Ebanks, M. E., & Fisher, W. W. (2003). Altering the
the contexts in which the treatment is relevant; timing of academic prompts to treat destructive be-
and generalization and maintenance of treatment havior maintained by escape. Journal of Applied Be-
effects, by using the functional reinforcer to main- havior Analysis, 36, 355–359.
Fisher, W. W., DeLeon, I. G., Rodriguez-Catter, V., &
tain an appropriate alternative behavior in the
Keeney, K. M. (2004). Enhancing the effects of ex-
natural environment (Durand, Berotti, & Weiner, tinction on attention-maintained behavior through
1993). noncontingent delivery of attention or stimuli identi-
fied via a competing stimulus assessment. Journal of
Applied Behavior Analysis, 37, 171–184.
REFERENCES Fisher, W. W., Fuhrman, A. M., Greer, B. D., Mitteer, D.
R., & Piazza, C. C. (2020). Mitigating resurgence of
Austin, J. E., & Tiger, J. H. (2015). Providing alterna- destructive behavior using the discriminative stimuli
tive reinforcers to facilitate tolerance to delayed re- of a multiple schedule. Journal of the Experimental
inforcement following functional communication Analysis of Behavior, 113, 263–277.
training. Journal of Applied Behavior Analysis, 48, Fisher, W. W., Greer, B. D., Fuhrman, A. M., & Querim,
663–668. A. C. (2015). Using multiple schedules during func-
Baron, A., & Galizio, M. (2005). Positive and negative tional communication training to promote rapid
reinforcement: Should the distinction be preserved? transfer of treatment effects. Journal of Applied Be-
Behavior Analyst, 28, 85–98. havior Analysis, 48, 713–733.
Beavers, G. A., Iwata, B. A., & Lerman, D. C. (2013). Fisher, W. W., Greer, B. D., Mitteer, D. R., Fuhrman, A.
Thirty years of research on the functional analysis of M., Romani, P. W., & Zangrillo, A. N. (2018). Further
problem behavior. Journal of Applied Behavior Analy- evaluation of differential exposure to establishing op-
sis, 46, 1–21. erations during functional communication training.
Betz, A. M., Fisher, W. W., Roane, H. S., Mintz, J. C., Journal of Applied Behavior Analysis, 51, 360–373.
& Owen, T. M. (2013). A component analysis of Fisher, W. W., Greer, B. D., Querim, A. C., & DeRosa,
schedule thinning during functional communica- N. (2014). Decreasing excessive functional commu-
tion training. Journal of Applied Behavior Analysis, nication responses while treating destructive behav-
46, 219–241. ior using response restriction. Research in Develop-
Carr, E. G., & Durand, V. M. (1985). Reducing behavior mental Disabilities, 35, 2614–2623.
problems through functional communication train- Fisher, W. W., Kuhn, D. E., & Thompson, R. H. (1998).
ing. Journal of Applied Behavior Analysis, 18, 111–126. Establishing discriminative control of responding
Conners, J., Iwata, B. A., Kahng, S. W., Hanley, G. P., using functional and alternative reinforcers during
Worsdell, A. S., & Thompson, R. H. (2000). Differ- functional communication training. Journal of Ap-
ential responding in the presence and absence of dis- plied Behavior Analysis, 31, 543–560.
criminative stimuli during multielement functional Fisher, W. W., O’Connor, J. T., Kurtz, P. F., DeLeon, I.
analyses. Journal of Applied Behavior Analysis, 33, G., & Gotjen, D. L. (2000). The effects of noncon-
299–308. tingent delivery of high- and low-preference stimuli
Developing Function-Based Reinforcement Procedures 349
on attention-maintained destructive behavior. Jour- al analyses. Journal of Applied Behavior Analysis, 46,
nal of Applied Behavior Analysis, 33, 79–83. 22–30.
Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003).
P., Owens, J. C., & Slevin, I. (1992). A comparison of Functional analysis of problem behavior: A review.
two approaches for identifying reinforcers for persons Journal of Applied Behavior Analysis, 36, 147–185.
with severe and profound disabilities. Journal of Ap- Hanley, G. P., Iwata, B. A., & Thompson, R. H. (2001).
plied Behavior Analysis, 25, 491–498. Reinforcement schedule thinning following treat-
Fisher, W., Piazza, C., Cataldo, M., Harrell, R., Jefferson, ment with functional communication training. Jour-
G., & Conner, R. (1993). Functional communication nal of Applied Behavior Analysis, 34, 17–38.
training with and without extinction and punish- Hanley, G. P., Piazza, C. C., & Fisher, W. W. (1997).
ment. Journal of Applied Behavior Analysis, 26, 23–36. Noncontingent presentation of attention and alter-
Fuhrman, A. M., Fisher, W. W., & Greer, B. D. (2016). native stimuli in the treatment of attention-main-
A preliminary investigation on improving functional tained destructive behavior. Journal of Applied Be-
communication training by mitigating resurgence havior Analysis, 30, 229–237.
of destructive behavior. Journal of Applied Behavior Herrnstein, R. J. (1961). Relative and absolute strength
Analysis, 49, 884–899. of response as a function of frequency of reinforce-
Fyffe, C. E., Kahng, S. W., Frittro, E., & Russell, D. ment. Journal of the Experimental Analysis of Behavior,
(2004). Functional analysis and treatment of inap- 4, 267–272.
propriate sexual behaviors. Journal of Applied Behav- Hoffman, K., & Falcomata, T. S. (2014). An evalua-
ior Analysis, 37, 401–404. tion of resurgence of appropriate communication in
Greer, B. D., Fisher, W. W., Briggs, A. M., Lichtblau, individuals with autism who exhibit severe problem
K. R., Phillips, L. A., & Mitteer, D. R. (2019). Using behavior. Journal of Applied Behavior Analysis, 47,
schedule-correlated stimuli during functional com- 651–656.
munication training to promote the rapid transfer Horner, R. D. (1980). The effects of an environmental
of treatment effects. Behavioral Development, 24, “enrichment” program on the behavior of institu-
100–119. tionalized profoundly retarded children. Journal of
Greer, B. D., Fisher, W. W., Saini, V., Owen, T. M., & Applied Behavior Analysis, 13, 473–491.
Jones, J. K. (2016). Functional communication train- Horner, R. H., & Day, H. M. (1991). The effects of re-
ing during reinforcement schedule thinning: An sponse efficiency on functionally equivalent compet-
analysis of 25 applications. Journal of Applied Behav- ing behaviors. Journal of Applied Behavior Analysis,
ior Analysis, 49, 105–121. 24, 719–732.
Grow, L. L., Kelley, M. E., Roane, H. S., & Shillings- Horner, R. H., Day, H. M., & Day, J. R. (1997). Using
burg, M. A. (2008). Utility of extinction-induced re- neutralizing routines to reduce problem behaviors.
sponse variability for the selection of mands. Journal Journal of Applied Behavior Analysis, 30, 601–614.
of Applied Behavior Analysis, 41, 15–24. Horner, R. H., Day, H. M., Sprague, J. R., O’Brien, M., &
Hagopian, L. P., Contrucci Kuhn, S. A., Long, E. S., Heathfield, L. T. (1991). Interspersed requests: A no-
& Rush, K. S. (2005). Schedule thinning following naversive procedure for reducing aggression and self-
communication training: Using competing stimuli injury during instruction. Journal of Applied Behavior
to enhance tolerance to decrements in reinforcer Analysis, 24, 265–278.
density. Journal of Applied Behavior Analysis, 38, Hyman, S. L., Fisher, W., Mercugliano, M., & Cataldo,
177–193. M. F. (1990). Children with self-injurious behavior.
Hagopian, L. P., Fisher, W. W., & Legacy, S. M. (1994). Pediatrics, 85, 437–441.
Schedule effects of noncontingent reinforcement on Iwata, B. A. (1987). Negative reinforcement in applied
attention-maintained destructive behavior in identi- behavior analysis: An emerging technology. Journal
cal quadruplets. Journal of Applied Behavior Analysis, of Applied Behavior Analysis, 20, 361–378.
27, 317–325. Iwata, B. A. (2006). On the distinction between posi-
Hagopian, L. P., Fisher, W. W., Sullivan, M. T., Ac- tive and negative reinforcement. Behavior Analyst,
quisto, J., & LeBlanc, L. A. (1998). Effectiveness of 29, 121–123.
functional communication training with and with- Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., &
out extinction and punishment: A summary of 21 Richman, G. S. (1994). Toward a functional analysis
inpatient cases. Journal of Applied Behavior Analysis, of self-injury. Journal of Applied Behavior Analysis, 27,
31, 211–235. 197–209. (Reprinted from Analysis and Intervention in
Hall, S., DeBernardis, M., & Reiss, A. (2006). Social es- Developmental Disabilities, 2, 3–20, 1982)
cape behaviors in children with fragile X syndrome. Iwata, B. A., Pace, G. M., Cowdery, G. E., & Milten-
Journal of Autism and Developmental Disorders, 36, berger, R. G. (1994). What makes extinction work:
935–947. An analysis of procedural form and function. Journal
Hammond, J. L., Iwata, B. A., Rooker, G. W., Fritz, J. N., of Applied Behavior Analysis, 27, 131–144.
& Bloom, S. E. (2013). Effects of fixed versus random Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R.,
condition sequencing during multielement function- Vollmer, T. R., Smith, R. G., et al. (1994). The func-
350 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
tions of self-injurious behavior: An experimental- with sleep deprivation. Journal of Applied Behavior
epidemiological analysis. Journal of Applied Behavior Analysis, 28, 225–226.
Analysis, 27, 215–240. Pace, G. M., Ivancic, M. T., & Jefferson, G. (1994).
Kettering, T. L., Fisher, W. W., Kelley, M. E., & LaRue, Stimulus fading as treatment for obscenity in a brain-
R. H. (2018). Sound attenuation and preferred music injured adult. Journal of Applied Behavior Analysis, 27,
in the treatment of problem behavior maintained by 301–305.
escape from noise. Journal of Applied Behavior Analy- Pace, G. M., Iwata, B. A., Cowdery, G. E., Andree, P. J.,
sis, 51, 687–693. & McIntyre, T. (1993). Stimulus (instructional) fad-
Kodak, T., Lerman, D. C., Volkert, V. M., & Trosclair, ing during extinction of self-injurious escape behav-
N. (2007). Further examination of factors that influ- ior. Journal of Applied Behavior Analysis, 26, 205–212.
ence preference for positive versus negative reinforce- Parrish, J. M., Cataldo, M. F., Kolko, D. J., Neef, N. A.,
ment. Journal of Applied Behavior Analysis, 40, 25–44. & Egel, A. L. (1986). Experimental analysis of re-
Lalli, J. S., Casey, S., & Kates, K. (1995). Reducing es- sponse covariation among compliant and inappro-
cape behavior and increasing task completion with priate behaviors. Journal of Applied Behavior Analysis,
functional communication training, extinction, and 19, 241–254.
response chaining. Journal of Applied Behavior Analy- Payne, S. W., & Dozier, C. L. (2013). Positive reinforce-
sis, 28, 261–268. ment as treatment for problem behavior maintained
Lalli, J. S., Vollmer, T. R., Progar, P. R., Wright, C., by negative reinforcement. Journal of Applied Behav-
Borrero, J., Daniel, D., et al. (1999). Competition ior Analysis, 46, 699–703.
between positive and negative reinforcement in the Peterson, C., Lerman, D. C., & Nissen, M. A. (2016). Re-
treatment of escape behavior. Journal of Applied Be- inforcer choice as an antecedent versus consequence.
havior Analysis, 32, 285–296. Journal of Applied Behavior Analysis, 49, 286–293.
Laraway, S., Snycerski, S., Michael, J., & Poling, A. Piazza, C. C., Adelinis, J. D., Hanley, G. P., Goh, H. L.,
(2003). Motivating operations and terms to describe & Delia, M. D. (2000). An evaluation of the effects
them: Some further refinements. Journal of Applied of matched stimuli on behaviors maintained by au-
Behavior Analysis, 36, 407–414. tomatic reinforcement. Journal of Applied Behavior
Lattal, K. A., & Lattal, A. D. (2006). And yet . . . : Fur- Analysis, 33, 13–27.
ther comments on distinguishing positive and nega- Piazza, C. C., Fisher, W. W., Hanley, G. P., LeBlanc,
tive reinforcement. Behavior Analyst, 29, 129–134. L. A., Worsdell, A. S., Lindauer, S. E., et al. (1998).
Lerman, D. C., & Iwata, B. A. (1995). Prevalence of Treatment of pica through multiple analyses of its
the extinction burst and its attenuation during treat- reinforcing functions. Journal of Applied Behavior
ment. Journal of Applied Behavior Analysis, 28, 93–94. Analysis, 31, 165–189.
Lerman, D. C., Iwata, B. A., Shore, B. A., & Kahng, S. Piazza, C. C., Fisher, W. W., Hanley, G. P., Remick, M.
W. (1996). Responding maintained by intermittent L., Contrucci, S. A., & Aitken, T. L. (1997). The use
reinforcement: Implications for the use of extinction of positive and negative reinforcement in the treat-
with problem behavior in clinical settings. Journal of ment of escape-maintained destructive behavior.
Applied Behavior Analysis, 29, 153–171. Journal of Applied Behavior Analysis, 30, 279–298.
Lomas, J. E., Fisher, W. W., & Kelley, M. E. (2010). The Piazza, C. C., Moes, D. R., & Fisher, W. W. (1996).
effects of variable-time delivery of food items and Differential reinforcement of alternative behav-
praise on problem behavior reinforced by escape. ior and demand fading in the treatment of escape-
Journal of Applied Behavior Analysis, 43, 425–435. maintained destructive behavior. Journal of Applied
Marcus, B. A., & Vollmer, T. R. (1995). Effects of dif- Behavior Analysis, 29, 569–572.
ferential negative reinforcement on disruption and Poling, A., & Normand, M. (1999). Noncontingent re-
compliance. Journal of Applied Behavior Analysis, 28, inforcement: An inappropriate description of time-
229–230. based schedules that reduce behavior. Journal of Ap-
McCord, B. E., Iwata, B. A., Galensky, T. L., Ellingson, plied Behavior Analysis, 32, 237–238.
S. A., & Thomson, R. J. (2001). Functional analysis Rapp, J. T., & Lanovaz, M. J. (2014). Introduction to the
and treatment of problem behavior evoked by noise. special issue on assessment and treatment of stereo-
Journal of Applied Behavior Analysis, 34, 447–462. typy. Behavior Modification, 38, 339–343.
Mevers, J. E. L., Fisher, W. W., Kelley, M. E., & Fredrick, Rincover, A. (1978). Sensory extinction: A procedure
L. D. (2014). The effects of variable-time versus con- form eliminating self-stimulatory behavior in devel-
tingent reinforcement delivery on problem behavior opmentally disabled children. Journal of Abnormal
maintained by escape. Journal of Applied Behavior Child Psychology, 6, 299–310.
Analysis, 47, 277–292. Roane, H. S., Falcomata, T. S., & Fisher, W. W. (2007).
Michael, J. (1975). Positive and negative reinforcement, Applying the behavioral economics principle of unit
a distinction that is no longer necessary: Or better price to DRO schedule thinning. Journal of Applied
ways to talk about bad things. Behaviorism, 3, 33–45. Behavior Analysis, 40, 529–534.
O’Reilly, M. F. (1995). Functional analysis and treat- Roane, H. S., Fisher, W. W., Sgro, G. M., Falcomata, T.
ment of escape-maintained aggression correlated S., & Pabico, R. R. (2004). An alternative method
Developing Function-Based Reinforcement Procedures 351
of thinning reinforcer delivery during differential during treatment with functional communication
reinforcement. Journal of Applied Behavior Analysis, training. Journal of Applied Behavior Analysis, 42,
37, 213–218. 145–160.
Romaniuk, C., Miltenberger, R., Conyers, C., Jenner, Vollmer, T. R. (1994). The concept of automatic rein-
N., Jurgens, M., & Ringenberg, C. (2002). The influ- forcement: Implications for behavioral research in
ence of activity choice on problem behaviors main- developmental disabilities. Research in Developmental
tained by escape versus attention. Journal of Applied Disabilities, 15, 187–207.
Behavior Analysis, 35, 349–362. Vollmer, T. R., Borrero, J. C., Wright, C. S., Van Camp,
Rooker, G. W., Jessel, J., Kurtz, P. F., & Hagopian, L. C., & Lalli, J. S. (2001). Identifying possible contin-
P. (2013). Functional communication training with gencies during descriptive analyses of severe behav-
and without alternative reinforcement and punish- ior disorders. Journal of Applied Behavior Analysis, 34,
ment: An analysis of 58 applications. Journal of Ap- 269–287.
plied Behavior Analysis, 46, 708–722. Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R.
Roscoe, E. M., Iwata, B. A., & Rand, M. S. (2003). Ef- G., & Mazaleski, J. L. (1993). The role of attention in
fects of reinforcer consumption and magnitude on the treatment of attention-maintained self-injurious
response rates during noncontingent reinforcement. behavior: Noncontingent reinforcement and differ-
Journal of Applied Behavior Analysis, 36, 525–539. ential reinforcement of other behavior. Journal of Ap-
Russo, D. C., Cataldo, M. F., & Cushing, P. J. (1981). plied Behavior Analysis, 26, 9–21.
Compliance training and behavioral covariation in Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994).
the treatment of multiple behavior problems. Journal Treatment of self-injury and hand mouthing follow-
of Applied Behavior Analysis, 14, 209–222. ing inconclusive functional analyses. Journal of Ap-
Saini, V., Miller, S. A., & Fisher, W. W. (2016). Multiple plied Behavior Analysis, 27, 331–344.
schedules in practical application: Research trends Vollmer, T. R., Marcus, B. A., & Ringdahl, J. E. (1995).
and implications for future investigation. Journal of Noncontingent escape as treatment for self-injurious
Applied Behavior Analysis, 49, 421–444. behavior maintained by negative reinforcement.
Shirley, M. J., Iwata, B. A., Kahng, S., Mazaleski, J. L., & Journal of Applied Behavior Analysis, 28, 15–26.
Lerman, D. C. (1997). Does functional communica- Wacker, D. P., Steege, M. W., Northup, J., Sasso, G.,
tion training compete with ongoing contingencies of Berg, W., Reimers, T., et al. (1990). A component
reinforcement?: An analysis during response acqui- analysis of functional communication training across
sition and maintenance. Journal of Applied Behavior three topographies of severe behavior problems. Jour-
Analysis, 30, 93–104. nal of Applied Behavior Analysis, 23, 417–429.
Sidman, M. (2006). The distinction between positive Weeks, M., & Gaylord-Ross, R. (1981). Task difficulty
and negative reinforcement: Some additional consid- and aberrant behavior in severely handicapped stu-
erations. Behavior Analyst, 29, 135–139. dents. Journal of Applied Behavior Analysis, 14, 449–
Skinner, B. F. (1953). Science and human behavior. New 463.
York: Free Press. Wilder, D. A., Draper, R., Williams, W. L., & Higbee,
Slocum, S. K., & Vollmer, T. R. (2015). A comparison of T. S. (1997). A comparison of noncontingent rein-
positive and negative reinforcement for compliance forcement, other competing stimulation, and liquid
to treat problem behavior maintained by escape. rescheduling for the treatment of rumination. Behav-
Journal of Applied Behavior Analysis, 48, 563–574. ioral Treatments, 12, 55–64.
Smith, R. G., Iwata, B. A., Goh, H., & Shore, B. A. Zangrillo, A. N., Fisher, W. W., Greer, B. D., Owen, T.
(1995). Analysis of establishing operations for self- M., & DeSouza, A. A. (2016). Treatment of escape-
injury maintained by escape. Journal of Applied Be- maintained challenging behavior using chained
havior Analysis, 28, 515–535. schedules: An evaluation of the effects of thinning
Tiger, J. H., Hanley, G. P., & Bruzek, J. (2008). Func- positive plus negative reinforcement during func-
tional communication training: A review and practi- tional communication training. International Journal
cal guide. Behavior Analysis in Practice, 1, 16–23. of Developmental Disabilities, 62, 147–156.
Vaughan, M. E., & Michael, J. L. (1982). Automatic Zarcone, J. R., Iwata, B. A., Vollmer, T. R., Jagtiani, S.,
reinforcement: An important but ignored concept. Smith, R. G., & Mazaleski, J. L. (1993). Extinction of
Behaviorism, 10, 217–228. self-injurious escape behavior with and without in-
Volkert, V. M., Lerman, D. C., Call, N. A., & Trosclair- structional fading. Journal of Applied Behavior Analy-
Lasserre, N. (2009). An evaluation of resurgence sis, 26, 353–360.
CHAP TER 21
Dorothea C. Lerman and Lisa M. Toole
Numerous procedures based on the process of particularly for the most severe forms of problem
punishment are effective for treating problem be- behavior—ones that place these individuals, their
havior. When the contingent presentation of a caregivers, or the environment at risk for signifi-
stimulus decreases the future likelihood of a be- cant harm.
havior, behavior analysts use the term positive pun- However, numerous authors over the last sev-
ishment. Procedural variations of positive punish- eral decades have noted that both basic and ap-
ment examined in the applied literature include plied research on punishment is declining (e.g.,
the contingent presentation of verbal reprimands, Baron, 1991; Crosbie, 1998; Kahng, Iwata, &
brief physical restraint, and demands. When the Lewin, 2002; Lydon, Healy, Moran, & Foody,
contingent removal of a stimulus decreases the 2015; Pelios, Morren, Tesch, & Axelrod, 1999).
future likelihood of a behavior, behavior analysts Although the use of punishment has been con-
use the term negative punishment. We can divide troversial for many years, authors have attributed
negative punishment further into two procedures, the decrease in applied research to advances in
response cost and time out. Response cost is the the functional analysis of problem behavior and
contingent removal of a specific amount of a posi- greater use of function-based treatment (Kahng
tive reinforcer (e.g., loss of tokens), and time out is et al., 2002; Pelios et al., 1999). Despite advances
the contingent loss of access to reinforcement for in treatment with extinction and reinforcement
a specific time (e.g., removal of reinforcement for (see Vollmer, Athens, & Fernand, Chapter 19, and
2 minutes). Fisher, Greer, & Bouxsein, Chapter 20, this vol-
Basic research findings on punishment con- ume), research findings suggest that punishment
ducted primarily with nonhumans have been in- remains an important option for caregivers of in-
strumental for developing an effective technology dividuals with severe forms of problem behavior
of punishment. The voluminous applied litera- as defined above. Punishment may be indicated
ture on punishment, now spanning more than 45 clinically when a behavior analyst cannot identify
years, has demonstrated the advantages and dis- or control the reinforcers maintaining problem be-
advantages of this treatment approach. The goal havior (e.g., Fisher et al., 1993; Hagopian, Rooker,
of this research has been to develop safe and ef- & Zarcone, 2015; Lindberg, Iwata, & Kahng, 1999;
fective punishers for the problem behavior of indi- Raulston, Hansen, Machalicek, McIntyre, & Car-
viduals with intellectual developmental disorder, nett, 2019; Saini, Greer, & Fisher, 2015), or when
352
Developing Function-Based Punishment Procedures 353
when reprimands were provided contingent on the restraint with many of the punishers described in
problem behavior of other individuals (e.g., Rich- this section (e.g., overcorrection, aversive smells,
man et al., 2001; Van Houten et al., 1982). time out; e.g., Cipani, Brendlinger, McDowell, &
Usher, 1991). Researchers have not sufficiently
explored the contribution of physical contact or
Response Blocking and Physical Restraint
restraint to the efficacy of other procedures.
Various punishment procedures involve physical
contact between the caregiver and the behaver.
Overcorrection and Other Forms
These procedures differ in terms of the amount and
of Contingent Effort
duration of the contact. Response blocking is the use
of brief physical contact to prevent a response from Several procedural variations of punishment are
occurring and is the least intrusive of these proce- similar, in that the behavior analyst requires the
dures (e.g., Giles, St. Peter, Pence, & Gibson, 2012; individual to engage in an effortful response fol-
Lalli, Livezey, & Kates, 1996; Lerman & Iwata, lowing problem behavior. This type of punisher is
1996; Reid, Parsons, Phillips, & Green, 1993). In overcorrection if the contingent response is topo-
Reid et al. (1993) and Lerman and Iwata (1996), graphically like the problem behavior or related
for example, the therapist blocked hand mouthing to the problem behavior in some other manner
by placing his or her hand approximately 2 centi- (e.g., an appropriate replacement behavior). Foxx
meters from the participant’s mouth. This behavior and Azrin (1972) developed overcorrection, and
prevented the participant from inserting his or her it consists of two procedural components that the
hand into the mouth (i.e., the back of the thera- behavior analyst implements alone or in combi-
pist’s hand contacted the individual’s hand). How- nation, depending on the problem behavior. Res-
ever, the reductive effects of response blocking may titutional overcorrection requires the individual to
be due to extinction rather than to punishment in restore the physical environment to a better state
some cases (Smith, Russo, & Le, 1999). than its original state if problem behavior pro-
Physical restraint restricts or limits an individ- duced disruption to the physical environment. For
ual’s movement, unlike response blocking, which example, a behavior analyst would require an indi-
only prevents the response. Results of several stud- vidual who turned over a garbage can in the din-
ies have shown that numerous variations of physi- ing room to pick up the garbage can and the trash,
cal restraint effectively reduce problem behavior. and then to sweep and mop the entire dining-room
Restraint procedures have included hands down, in floor. Positive-practice overcorrection requires the
which the therapist holds the individual’s hands individual to repeatedly practice an appropriate,
to the side or in the lap for a specified time (Bit- related behavior. For example, a behavior analyst
good, Crowe, Suarez, & Peters, 1980; Hagopian et would require the individual who turned over a
al., 1998; Lerman, Iwata, Shore, & DeLeon, 1997; garbage can in the dining room to repeatedly place
Thompson et al., 1999); baskethold, in which the the garbage can gently on the floor. Researchers
therapist stands behind the individual, crosses the have implemented positive practice alone in nu-
individual’s arms across the chest, and holds them merous studies with behavior that does not disrupt
above the wrists for a specified time (Fisher, Pi- the environment, such as stereotypic behavior
azza, Bowman, Kurtz, et al., 1994), and movement (e.g., Anderson & Le, 2011; Cipani et al., 1991;
suppression time out, in which the therapist uses Foxx & Azrin, 1973; Peters & Thompson, 2013).
the least amount of physical contact necessary to For example, in Peters and Thompson (2013), the
keep the individual motionless while standing in a therapist interrupted motor stereotypy and physi-
corner (Rolider & Van Houten, 1985). cally prompted the participants to engage appro-
In most studies, the therapist used brief physical priately with leisure materials for 30 seconds. The
contact, such as 30–60 seconds, and implemented direct and indirect effects of overcorrection are
physical restraint in conjunction with other po- like those associated with other contingent-effort
tential punishers (e.g., verbal reprimands, time procedures described below, despite some claims to
out; Thompson et al., 1999). In fact, time out from the contrary (see MacKenzie-Keating & McDon-
positive reinforcement was a likely component of ald, 1990, for a discussion). For example, results
any physical-restraint procedure if access to rein- of studies examining the effects of overcorrection
forcing stimuli was unavailable while the therapist on the practiced response have been inconsistent,
restrained the individual. Furthermore, therapists showing increases, decreases, and no change (e.g.,
commonly used some form of physical contact or Peters & Thompson, 2013).
Developing Function-Based Punishment Procedures 355
Similar procedures that behavior analysts do & Agras, 1974), public masturbation (Cook, Alt-
not typically classify as overcorrection include man, Shaw, & Blaylock, 1978), and pica (Paisey
contingent demands, in which researchers required & Whitney, 1989). The therapist applied aromatic
individuals to complete tasks that were unrelated ammonia (i.e., smelling salts) by breaking an am-
to the problem behavior (Fischer & Nehs, 1978; monia capsule and holding it under the individ-
Fisher et al., 1993; Watson, 1993); negative practice, ual’s nose for a specified time. Researchers have
in which the researchers required the individuals used this treatment for SIB (Altman, Haavik, &
to exhibit the problem behavior repeatedly (Azrin, Cook, 1978; Singh, Dawson, & Gregory, 1980a;
Nunn, & Frantz, 1980); and contingent exercise, Tanner & Zeiler, 1975), aggression (Doke, Wolery,
in which the researchers required the individuals & Sumberc, 1983), and breath holding and hyper-
to perform motor movements that were unrelated ventilation (Singh, 1979; Singh, Dawson, & Greg-
to the problem behavior (Kahng, Abt, & Wilder, ory, 1980b). Researchers have not reported on the
2001; Luce, Delquadri, & Hall, 1980). Response clinical use of other types of aromatics.
interruption and redirection is a well-studied inter-
vention for vocal stereotypy that resembles con-
Noise
tingent-effort procedures (Ahearn, Clark, Mac-
Donald, & Chung, 2007). In Ahearn et al. (2007), Researchers have used the contingent presenta-
for example, the therapist delivered questions or tion of noise as a punisher for finger and thumb
instructions requiring vocal responses contingent sucking (Stricker, Miltenberger, Garlinghouse,
on occurrences of vocal stereotypy; the questions Deaver, & Anderson, 2001; Stricker, Miltenberg-
or instructions continued until a participant ex- er, Garlinghouse, & Tulloch, 2003), hair pulling
hibited three correct responses without engaging (Rapp, Miltenberger, & Long, 1998), and auditory
in vocal stereotypy. Results of research suggest that hallucinations (Fonagy & Slade, 1982). In Strick-
response interruption and redirection functions er et al. (2001), for example, participants wore a
as punishment in some cases (e.g., Ahrens, Ler- device that automatically detected when a par-
man, Kodak, Worsdell, & Keegan, 2011; Shawler ticipant’s hand moved toward the mouth and pro-
& Miquel, 2015). duced a 65-decibel tone when that occurred. The
procedure reduced the thumb sucking of the two
participants to near-zero levels. Although the re-
Water Mist
searchers attributed the findings to an increase in
Researchers have used contingent water mist to the participants’ awareness of thumb sucking, the
decrease problem behavior in individuals with de- tone may have functioned as a punishing stimulus.
velopmental disabilities in a few studies (e.g., Arn- Results of a subsequent study with the same device
tzen & Werner, 1999; Dorsey et al., 1980; Friman, were consistent with a punishment interpretation,
Cook, & Finney, 1984; Singh, Watson, & Winton, because a 90-decibel tone was necessary to reduce
1986). A therapist generally implemented the pro- the finger sucking (Stricker et al., 2003).
cedure by standing in front of an individual with a
squeeze-type bottle containing room-temperature
Shock
water, holding the water bottle at a slant to avoid
spraying water directly into the individual’s eyes, Society considers contingent electric shock the most
and delivering one mist of water for each instance intrusive and controversial of the punishment pro-
of problem behavior. cedures, but research has shown that it produces
rapid and durable reductions in severe SIB (Duker
& Seys, 1996; Linscheid, Iwata, Ricketts, Williams,
Aversive Tastes and Smells
& Griffin, 1990; Salvy, Mulick, Butter, Bartlett, &
Results of studies have shown that the contin- Linscheid, 2004) and aggression (Ball, Sibbach,
gent delivery of an aversive taste, such as vinegar Jones, Steele, & Frazier, 1975; Foxx, 2003). In pre-
or lemon juice, or aversive smell, such as aromatic vious studies, researchers delivered brief, moderate
ammonia, successfully treats problem behavior. In intensity electric shock (e.g., 84 volts) to the par-
several studies, for example, a therapist squirted a ticipant’s extremity (e.g., leg) via electrodes that
small amount of unsweetened concentrated lemon a movement detection device activated (e.g., the
juice or vinegar into the mouth contingent on self- Self-Injurious Behavior Inhibiting System; Lin-
stimulatory behavior (Cipani et al., 1991; Friman scheid et al., 1990) or remotely by a caregiver or
et al., 1984), chronic rumination (Sajwaj, Libet, therapist. Research findings indicate that shock
356 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
does not increase the risk of undesirable side ef- 1987); contingent observation, in which a researcher
fects and can be a safe alternative to other punish- required a participant to remain near the reinforc-
ment procedures (e.g., Linscheid, Pejeau, Cohen, ing environment (Porterfield, Herbert-Jackson,
& Footo-Lenz, 1994). & Risely, 1976); and item removal, in which a re-
searcher terminated ongoing stimulation sources,
such as music, leisure materials, and food (Dupuis,
Negative‑Punishment Procedures Lerman, Tsami, & Shireman, 2015; Falcomata,
Time Out from Positive Reinforcement Roane, Hovanetz, Kettering, & Keeney, 2004;
Keeney, Fisher, Adelinis, & Wilder, 2000; Ritschl,
Time out, one of the most common forms of pun- Mongrella, & Presbie, 1972). In Falcomata et al.
ishment, is the contingent loss of access to positive (2004), for example, the therapist removed con-
reinforcers or withdrawal of the opportunity to tinuous access to a radio for 5 seconds contingent
earn positive reinforcers for a period. Researchers on inappropriate vocalizations. Researchers have
have removed reinforcement typically by moving called this latter form of time out response cost in
an individual to a less reinforcing environment, some studies. However, we typically classify pro-
such as a barren room, partitioned area, or cor- cedures that produce time-based reinforcement
ner (exclusionary or seclusionary time out), or by loss as time out. During response cost, by contrast,
discontinuing reinforcement in the current envi- participants earn reinforcement for appropriate
ronment (nonexclusionary time out). Toole, Bow- behavior rather than the passage of time, and cli-
man, Thomason, Hagopian, and Rush (2003), for nicians do not return lost reinforcers.
example, used exclusionary time out to treat the Researchers have often combined physical re-
severe aggression of a 15-year-old girl with intellec- straint with exclusionary or nonexclusionary time
tual developmental disorder. The participant had out. For example, a researcher might use a basketh-
access to several highly preferred items and activi- old to restrain a participant or physically guide the
ties throughout the day. A therapist guided her to participant to remain in a prescribed stance while
a padded treatment room, using the least amount sitting or standing in the corner of a room (e.g.,
of physical assistance necessary, following each Fisher, Piazza, Bowman, Kurtz, et al., 1994; Matson
instance of aggression. The therapist required & Keyes, 1990; Rolider & Van Houten, 1985). No
the participant to stay in the room for 5 minutes. studies have examined the contribution of poten-
Research has examined a wide range of time-out tially aversive aspects of physical contact to the ef-
durations (e.g., 15 seconds to 30 minutes); how- ficacy of these procedures.
ever, results of studies on the relation between this
parameter and treatment efficacy have produced
inconsistent outcomes (see Matson & DiLorenzo, Response Cost
1984, for a review). Furthermore, research compar- The contingent removal of a specific amount of
ing the effectiveness of fixed-duration time-out reinforcement, such as tokens, can function as an
lengths to those based on the absence of problem effective punisher. Typically, the participant ob-
behavior during time out (i.e., contingent release) tains reinforcement via appropriate behavior, as in
suggests that the latter does not confer any addi- differential reinforcement of alternative behavior,
tional benefits (e.g., Donaldson & Vollmer, 2011; or independently of behavior, as in noncontingent
Mace, Page, Ivancic, & O’Brien, 1986). reinforcement, and loses these reinforcers con-
Researchers have developed numerous proce- tingent on problem behavior. Researchers have
dural variations of nonexclusionary time out to re- conducted much of the research on response cost
strict reinforcement access following problem be- in the context of token-economy systems (e.g.,
havior without removing the participant from the LeBlanc, Hagopian, & Maglieri, 2000; Truchlicka,
immediate environment. These variations include McLaughlin, & Swain, 1998). However, partici-
the visual screen, in which a researcher placed a pants have lost other reinforcers via response cost,
hand, mask, or cloth over an individual’s eyes including books and audiotapes (Kahng, Tarbox,
(Mitteer, Romani, Greer, & Fisher, 2015; Singh et & Wilke, 2001), money (Epstein & Masek, 1978;
al., 1986; Rush, Crockett, & Hagopian, 2001); the Long, Miltenberger, Ellingson, & Ott, 1999), and
time-out ribbon, in which a researcher removed a participation in tournaments (e.g., Allen, 1998).
ribbon a participant wore, and reinforcement was Kahng, Tarbox, and Wilke (2001), for example,
unavailable when the participant did not have the gave a young boy who engaged in food refusal ac-
ribbon (Foxx & Shapiro, 1978; Salend & Gordon, cess to highly preferred items, such as books and
Developing Function-Based Punishment Procedures 357
audiotapes. The therapist removed the items if the reduce a child’s disruption, because the literature
boy refused to accept a bite of food or engaged in has demonstrated the efficacy of this intervention
problem behavior, and returned the items con- (e.g., Clark, Rowbury, Baer, & Baer, 1973). If this
tingent on bite acceptance. Surprisingly little re- intervention does not reduce the child’s behavior,
search has evaluated methods for determining the the clinician may increase the time-out duration.
most appropriate amount and type of reinforcers The clinician may try a more restrictive proce-
to remove in a response cost intervention. dure, such as overcorrection, if the lengthier time
out is ineffective. The clinician may continue to
evaluate increasingly restrictive procedures, such
SELECTING PUNISHMENT PROCEDURES as restraint, until he or she identifies an effective
procedure.
Although researchers have evaluated numerous This process can be time-consuming, delay the
punishers for clinical use, efficient approaches for onset of treatment, and produce prolonged ex-
prescribing specific punishment procedures have posure to multiple intrusive procedures. The as-
received less attention in the literature. Ethical sumption that a more restrictive procedure has a
guidelines mandate that behavior analysts should greater likelihood of success than a less restrictive
give priority to the least restrictive procedure that procedure has no empirical support. The hierar-
is clinically effective (e.g., Behavior Analysis Cer- chical approach emphasizes the topography of an
tification Board, 2016; Van Houten et al., 1988; intervention versus its function; it ignores the pos-
Vollmer et al., 2011). Inherent in this approach sibility that an intrusive procedure might function
to intervention selection is the need to arrange as a punisher for some people but as a reinforcer
punishment procedures hierarchically according for others, such as water mist for some participants
to their degree of restrictiveness or intrusiveness in Fisher, Piazza, Bowman, Kurtz, et al. (1994). In
(i.e., how much each procedure limits individual fact, intrusiveness level is subjective. Of concern
freedom or intrudes into an individual’s life in is the chance that exposure to progressively intru-
some manner) or aversiveness (i.e., how much sive interventions could promote habituation to
each procedure produces discomfort, pain, or putative punishers, decreasing the effectiveness of
distress). We typically consider nonexclusionary punishment. Finally, clinicians should consider a
time out and response cost as the least restrictive variety of additional factors when selecting a pun-
of the procedures, followed by exclusionary time ishment procedure, including the immediacy of
out, overcorrection, and other physical punishers. effects, relevance to behavioral function, severity
We sometimes refer to this hierarchical arrange- of the behavior, and caregivers’ willingness to use
ment as a levels system and use it to guide inter- the procedure (Iwata, Vollmer, & Zarcone, 1990;
vention evaluation and selection. Readers can find Van Houten et al., 1988; Vollmer & Iwata, 1993).
case examples and guidelines in several sources for Assessment procedures that reliably identify
using this approach to identify effective punish- effective interventions would permit clinicians
ment procedures (e.g., Alberto & Troutman, 2006; to select punishment procedures based on other
Barton, Brulle, & Repp, 1983; Cooper, Heron, & concerns, such as restrictiveness level, behavior
Heward, 1987; Couvillon, Kane, Peterson, Ryan, severity, and caregivers’ willingness to implement
& Scheuermann, 2019; Foxx, 1982; Gaylord-Ross, the procedure. Such an assessment would avoid
Weeks, & Lepner, 1980; Gross, Wright, & Drab- the trial-and-error approach clinicians commonly
man, 1981; Lovaas & Favell, 1987; Repp & Deitz, use to select interventions. As discussed in the
1978). Some states have adopted policies that following sections, researchers have evaluated sev-
explicitly categorize procedures by restrictiveness eral potential assessments. When combined with
level (Spreat & Lipinski, 1986). a functional analysis, these strategies may provide
Nonetheless, attempts to apply the least restric- clinicians with a more reliable, efficient approach
tive treatment model may themselves raise ethical for determining the least restrictive procedure that
concerns. Clinicians using a hierarchical system is effective.
typically evaluate punishment procedures on a
trial basis, starting with the least restrictive pro-
Functional Analysis
cedure that may be effective and moving to more
restrictive procedures until they identify an effec- Knowing the function of problem behavior is es-
tive intervention. To illustrate, a clinician may pecially important once a clinician has decided to
initially evaluate a 5-minute time-out procedure to include punishment in treatment. Several punish-
358 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
ment procedures are uniquely indicated or contra- punishment for problem behavior. On the other
indicated for particular behavioral functions. Pun- hand, punishment procedures with a contingent
ishment will also be more effective if the clinician decrease in attention, such as time out, may be in-
combines it with extinction and differential rein- dicated (Hagopian et al., 1998).
forcement of alternative behavior (e.g., Azrin & Figure 21.2 shows similar outcomes for the de-
Holz, 1966; Holz, Azrin, & Ayllon, 1963; Rawson mand condition. This condition directly tests con-
& Leitenberg, 1973; Thompson et al., 1999). Thus tingent removal of interaction and instructional
clinicians should withhold identified functional materials. If levels of problem behavior are like
reinforcers for problem behavior and deliver those those in the control condition (see upper panel),
reinforcers for appropriate behavior whenever pos- attention or material loss (i.e., time out) may not
sible. be effective. On the other hand, time out may be
Iwata, Dorsey, Slifer, Bauman, and Richman viable if levels of problem behavior are lower in
(1982/1994) described a comprehensive function-
al-analytic approach that is useful for identifying 8
viable options when a clinician is considering
punishment as a treatment component. As Saini,
6
Fisher, Betz, and Piazza (Chapter 13, this volume)
describe, the experimental functional analysis ef-
4
fectively identifies the function of problem behav-
ior for most participants. Furthermore, although Demand
behavior analysts use the assessment to test puta- 2
10
ous section and each behavioral function in Table
8
21.1. However, further research is needed to test
these predictions. Results of a quantitative review
6 of applied research on punishment found similar
treatment outcomes, regardless of whether the re-
4 searchers conducted a functional analysis before
treatment (Lydon et al., 2015). Whether the re-
2
searchers selected punishers based on functional-
analytic results was not clear. Furthermore, auto-
0
0 5 10 15 matic reinforcement maintained problem behavior
in most cases, particularly in studies published
Responses per Minute (Problem Behavior)
4 Stimulus Avoidance Assessment
2 Fisher and colleagues described an efficient ap-
proach for identifying potential punishers (Fisher,
0 Piazza, Bowman, Hagopian, & Langdon, 1994;
0 5 10 15
Fisher, Piazza, Bowman, Kurtz, et al., 1994). The
Sessions
researcher evaluated participants’ responses to
various punishment procedures by using a proce-
FIGURE 21.2. Three possible outcomes for the demand dure like the one Pace, Ivancic, Edwards, Iwata,
condition of the functional analysis that would provide and Page (1985) developed for identifying poten-
important information about potential punishers (see tial reinforcers. Fisher, Piazza, Bowman, Kurtz, et
text for further details).
al. (1994) included water mist and procedural vari-
ations of time out, restraint, and contingent effort
the demand condition than in the control con- in the evaluation. The researchers presented each
dition (see middle panel). Results showing that potential punisher independently of responding
escape from demands functions as reinforcement for 15–180 seconds across 10 trials, with a buzzer
for problem behavior would indicate that punish- preceding the onset of each trial to decrease the
ment procedures that remove or delay demands, likelihood of superstitious conditioning. Observ-
such as time out, physical restraint, or protective ers measured avoidance responses (e.g., dropping
equipment, may be contraindicated (Magee & to the floor), negative vocalizations (e.g., crying),
Ellis, 2001). Procedures with a contingent increase and positive vocalizations (e.g., laughing) during
in demands, such as additional work, exercise, and these exposures. The researchers predicted that
overcorrection, may be indicated for escape-main- procedures associated with the highest rates of
tained problem behavior (Hagopian et al., 1998). avoidance responses and negative vocalizations
We summarize the relation between the vari- were most likely to function as punishers. They
ous punishment procedures described in the previ- subsequently compared the clinical efficacy of
360 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
Note. Dashes indicate that the procedure is neither indicated nor contraindicated.
procedures with low, medium, and high levels of tions. Researchers evaluate one or more potential
these responses in a multielement design for each punishers during brief sessions to predict the ef-
participant. The assessment had good predictive fectiveness of the procedure(s) when implemented
validity. This procedure may be useful for assess- over lengthier periods. For example, Fisher and
ing other types of procedures (e.g., withdrawal of colleagues (Fisher, Piazza, Bowman, Hagopian, &
preferred items) and certain parameters of punish- Langdon, 1994; Fisher, Piazza, Bowman, Kurtz, et
ment (e.g., magnitude). al., 1994) compared the effects of three punishment
An advantage of this approach is that it empiri- procedures on problem behavior by implementing
cally evaluates multiple potential punishers in a each procedure during three to six 10-minute ses-
short time. Behavior analysts can combine assess- sions, alternated in a multielement design. Results
ment results with other important considerations, showed that the assessment had good predictive
such as restrictiveness and caregiver preference, validity when they evaluated the punisher asso-
to prescribe the most appropriate intervention. ciated with the lowest levels of problem behavior
Fisher, Piazza, Bowman, Hagopian, and Langdon throughout the day in an intervention package.
(1994), for example, asked caregivers to rate each Thompson et al. (1999) evaluated the effects
of nine potential punishers as acceptable or unac- of several procedures on SIB in a brief assessment
ceptable. The researchers excluded those rated as with AB designs. They conducted an extended
unacceptable from the stimulus avoidance assess- evaluation with the least restrictive procedure
ment. Furthermore, the researchers evaluated ease that correlated with a 75% or greater reduction in
of implementation for each procedure by measuring SIB with and without reinforcement. Punishment
escape responses during the avoidance assessment alone was effective in reducing SIB below baseline
(i.e., the number of times the participants success- levels for three of four participants, providing some
fully prevented implementation). These data may support for the predictive validity of the brief as-
indicate whether (1) a high level of treatment in- sessment. However, the researchers provided few
tegrity would be possible if caregivers implemented procedural details, such as the number and length
the punisher in the participant’s natural environ- of sessions or the range of punishers evaluated,
ment, and (2) caregivers would find the procedure because the brief assessment was not the focus of
acceptable for clinical use. Both factors are impor- the investigation. Similarly, Verriden and Roscoe
tant to consider in selecting interventions. (2018) found that a punisher assessment was nec-
essary to decrease the occurrence of automatically
reinforced problem behavior, following unsuccess-
Brief Punisher Assessment
ful treatment attempts with reinforcement-based
Brief assessments of punishers, conducted in con- approaches.
junction with or instead of avoidance assessments, Like stimulus avoidance assessments (Fisher,
have been useful for identifying effective interven- Piazza, Bowman, Hagopian, & Langdon, 1994;
Developing Function-Based Punishment Procedures 361
Fisher, Piazza, Bowman, Kurtz, et al., 1994), brief of SIB—a contingency that did not establish re-
punisher assessments provide information about sponse satiation. Although only the contingency
the potential efficacy of multiple punishment pro- that produced response satiation was effective, the
cedures in an efficient manner. Clinicians can also physical contact needed to ensure that partici-
obtain other measures relevant to intervention pants manipulated the items for the required time
selection, such as the immediacy of effects, ease may have functioned as punishment. Thus dif-
of implementation, and potential for side effects, ferences in the duration of physical contact con-
while conducting the assessment. founded the analysis of response satiation. Results
of other studies, however, suggest that response
satiation may be a viable method for identifying
Activity Assessment
and arranging punishment contingencies (e.g.,
A few studies have described another potential Dougher, 1983; Holburn & Dougher, 1986).
strategy for identifying effective punishers, based Although further research is needed, activity
on work by Terhune and Premack (1970, 1974) and assessments are appealing, because they broaden
Allison and Timberlake (1974). In this assessment the range of potential punishers available for clini-
approach, baseline observations determined the cal use. This approach increases the likelihood of
relative lengths of time for which participants en- successful intervention with procedures based on
gaged in various freely available activities. Results contingent effort. Caregivers may find these pro-
of some studies predicted that an activity associ- cedures more acceptable than other punishers, es-
ated with a low probability of engagement would pecially if the contingent response is appropriate
function as a punisher. The contingency arrange- and functional.
ment required the participant to engage in the
low-probability activity following instances of the
Choice Assessments
targeted (high-probability) behavior. Krivacek and
Powell (1978), for example, required three students Behavior analysts can use additional assessments
with intellectual developmental disorder to engage of caregiver or client preference to guide interven-
in low-probability activities, such as running, trac- tion selection when multiple effective punishers
ing letters, and playing with a ball, contingent on are available. Behavior analysts can assess caregiv-
problem behavior. The researchers selected activi- er preference by obtaining verbal report or ratings
ties in which participants rarely engaged sponta- of acceptability (e.g., opinions about appropriate-
neously. Although problem behavior decreased ness and willingness to implement procedures), or
during intervention, the researchers did not re- by asking caregivers to choose among the avail-
move engagement time for the low-probability able interventions. Surprisingly few studies have
activity from the total session time in the results. evaluated the acceptability of interventions after
As such, the findings are difficult to interpret, be- caregivers or staff members have implemented
cause participants had less opportunity to engage them in the natural environment (Armstrong,
in the target behavior during intervention than Ehrhardt, Cool, & Poling, 1997; see Mueller, Ed-
during baseline. wards, & Trahant, 2003, for a notable exception).
Other researchers have predicted that any ac- Research findings suggest that many factors influ-
tivity may function as a punisher if the contin- ence acceptability ratings, including knowledge of
gency requires a participant to engage in the activ- or experience with a procedure, intrusiveness of
ity at higher than desired levels (Dougher, 1983; the procedure, procedural complexity or ease of
Holburn & Dougher, 1986; Realon & Konarski, use, number of previously unsuccessful attempts
1993). This response satiation approach to inter- to treat the behavior, and client characteristics
vention selection suggests that even activities such as age and problem severity (for reviews, see
with similar or high levels of engagement relative Foxx, McHenry, & Bremer, 1996; Lennox & Milt-
to the target behavior can function as punishers. enberger, 1990; O’Brien & Karsh, 1990). Thus
Realon and Konarski (1993), for example, required having caregivers implement each procedure dur-
two participants with developmental disabilities ing brief punisher assessments before rating the
to manipulate leisure materials for 5–15 seconds acceptability of or choosing an intervention may
following each occurrence of SIB. The researchers be beneficial.
selected the punishment duration by examining In selecting an intervention, behavior analysts
baseline levels of SIB and item manipulation. Dur- can also consider the preferences of the individu-
ing a control condition, the researchers required 1 als whose behavior they are targeting for reduc-
second of item manipulation for each occurrence tion. Several studies have evaluated methods for
362 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
assessing preferences of individuals with limited ness, although few applied studies have compared
expressive-communication skills for interven- the effects of immediate and delayed punishment
tions (Giles et al., 2012; Hanley, Piazza, Fisher, directly. In a notable exception, Abramowitz and
Contrucci, & Maglieri, 1997; Hanley et al., 2005). O’Leary (1990) found that verbal reprimands were
Individuals with developmental disabilities chose more effective in decreasing off-task behavior in
among two or three different interventions by ac- schoolchildren when a teacher delivered a repri-
tivating one of three switches or touching one of mand immediately after the onset of the behav-
two pictures paired with each procedure. In Han- ior, rather than after the behavior had occurred
ley et al. (2005), for example, two children chose continuously for 2 minutes. Basic studies with hu-
between differential reinforcement alone, differen- mans and nonhumans indicate that punishment
tial reinforcement plus punishment, and punish- procedures can fail to suppress responding when
ment alone. Interestingly, both children showed a the consequence is delayed by just 10–30 seconds
clear preference for differential reinforcement plus (e.g., Banks & Vogel-Sprott, 1965; Goodall, 1984;
punishment by allocating most responding to the Trenholme & Baron, 1975). When a punisher is
choice associated with this procedure. delayed, other responses or multiple instances of
the target behavior are likely to occur before the
consequence is delivered, weakening the contin-
USING PUNISHMENT EFFECTIVELY gency between the response and its consequence.
Nonetheless, some applied studies have shown
Research findings indicate that punishment can that delayed punishment is effective (e.g., Azrin &
be highly effective for treating many behavior dis- Powers, 1975; Maglieri et al., 2000; Van Houten &
orders and is more effective than treatment with Rolider, 1988). Maglieri et al. (2000) used delayed
reinforcement or extinction (e.g., Barton, Matson, verbal reprimands to reduce the consumption of
Shapiro, & Ollendick, 1981; Grace, Kahng, & prohibited food items by a girl with Prader–Willi
Fisher, 1994; Hagopian et al., 1998; Wacker et al., syndrome. Consumption decreased to zero when
1990). However, the relative efficacy of extinction, the therapist delivered a verbal reprimand after
reinforcement, and punishment may be difficult to a 10-minute session in which the girl consumed
predict in application, because various factors re- prohibited food. The researchers did not evaluate
lated to the use of these procedures can influence the procedural components responsible for the
clinical outcomes. efficacy of the delayed punisher (e.g., therapist
In fact, results of basic research suggest that the instructions, recent history with immediate pun-
ways clinicians commonly implement punishment ishment). Results of two studies (Rolider & Van
in applied settings can undermine these interven- Houten, 1985; Van Houten & Rolider, 1988) sug-
tions’ effectiveness. Punishment may not produce gest that delayed punishment may be effective if a
immediate, substantial, or sustained reductions in consequence is paired with stimuli associated with
problem behavior if the consequence is delayed, engaging in the behavior. Contingent on earlier
intermittent, relatively mild, paired with rein- instances of the target behavior, the researchers in
forcement for problem behavior, or preceded by these studies required participants to engage in the
exposure to a less intense form or type of punisher, response (aggression, theft) or listen to audiotape
or if punishment reduces the amount of reinforce- recordings of their behavior (disruption). A thera-
ment received (see Lerman & Vorndran, 2003, for pist or caregiver then immediately delivered the
a review). Punishment may also be associated with punisher (verbal reprimands, physical restraint).
many undesirable side effects. We provide a brief Both procedures were highly effective in decreas-
overview of factors related to the use of punish- ing problem behavior. Although the researchers
ment below, together with a discussion of current did not evaluate the effect of the delayed pun-
applied findings on strategies for using punish- isher alone for most participants, this approach
ment effectively. may have prevented the adventitious punishment
of any untargeted responses that occurred before
the delivery of the consequence. Products of the
Contiguity of the Punisher
behavior also may have acquired conditioned
In most applied studies of punishment, a therapist aversive properties due to this pairing procedure.
delivered a consequence immediately after in- Researchers should evaluate other stimuli that
stances of problem behavior. Such contiguity was might bridge the interval between a response and
probably critical to the interventions’ effective- its consequence, such as instructions and condi-
Developing Function-Based Punishment Procedures 363
tioned punishers, in further applied research (Tr- ences treatment efficacy (e.g., Hobbs, Forehand,
enholme & Baron, 1975). & Murray, 1978; Richman et al., 2001; Stricker
et al., 2003; Williams, Kirkpatrick-Sanchez, &
Iwata, 1993). Richman et al. (2001), for example,
Schedule of Punishment
showed a positive relation between the loudness of
Results of basic and applied studies suggest that verbal reprimands and reductions in breath hold-
punishment is more likely to be effective if the ing exhibited by a teenager with intellectual de-
punisher is delivered contingent on each occur- velopmental disorder. In Williams et al. (1993), a
rence of the target behavior, rather than intermit- high-intensity shock (18.5 milliamperes) was more
tently (e.g., Azrin, Holz, & Hake, 1963; Calhoun effective in decreasing SIB than a low-intensity
& Matherne, 1975; Lerman et al., 1997; Thomas, shock (3.5 milliamperes). Nonetheless, results of
1968). Nonetheless, several applied studies have applied research have been inconsistent relative
demonstrated successful treatment of problem be- to basic findings on magnitude. Lengthier dura-
havior with intermittent punishment schedules tions of time out, overcorrection, and restraint
(e.g., Barton et al., 1987; Cipani et al., 1991; Clark have not produced larger and more reliable reduc-
et al., 1973; Dominguez et al., 2014; Romanczyk, tions in problem behavior than shorter durations
1977). Researchers have used a variety of sched- (e.g., Cole, Montgomery, Wilson, & Milan, 2000;
ules, including variable-ratio (VR) and fixed-inter- Singh, Dawson, & Manning, 1981).
val (FI) schedules, as well as the differential pun- Thus increasing the magnitude of an ineffec-
ishment of high response rates. For example, a VR tive punisher may have limited clinical utility and
4 schedule of time out was as effective in decreas- may even promote resistance to punishment (e.g.,
ing disruption as a continuous schedule in Clark Cohen, 1968; Terris & Barnes, 1969). Lengthier
et al. (1973). The generality of applied findings durations of punishment per se also may lead to
on intermittent punishment is unclear, however, habituation. For these reasons, clinicians should
because punishment appeared to be confounded use punishers that are effective when presented
with extinction or other potential punishers (e.g., briefly and at magnitudes that are in the range of
verbal reprimands) in most studies. those reported in applied studies.
Punishing every instance of behavior may often
be impractical, especially if the rate of respond-
Availability of Alternative Sources of Reinforcement
ing is relatively high. However, gradually thinning
the schedule of punishment may be possible after Basic and applied research also indicates that pun-
obtaining clinically significant reductions in prob- ishment will be more effective if reinforcement
lem behavior under a continuous schedule (e.g., is available readily for engaging in alternative
Barton et al., 1987; Lerman et al., 1997). In Ler- responses (e.g., Holz et al., 1963; Rawson & Leit-
man et al. (1997), for example, the therapist suc- enberg, 1973; Thompson et al., 1999) or indepen-
cessfully thinned the schedule of punishment with dently of responding (e.g., DeRosa, Roane, Bishop,
time out or restraint from fixed-ratio (FR) 1 to FI & Silkowski, 2016). Thompson et al. (1999), for ex-
300 seconds for two participants who engaged in ample, found that levels of SIB maintained by au-
SIB maintained by automatic reinforcement. Nev- tomatic reinforcement were much lower when they
ertheless, they could not increase the schedule combined punishment with reinforcement of toy
beyond FR 1 for two other participants, limiting manipulation than when they used either punish-
the generality of the findings. Further research is ment or reinforcement alone. Results of basic stud-
needed on strategies for improving the effective- ies have shown that increased deprivation for the
ness of intermittent punishment. reinforcer maintaining the punished behavior re-
duces the effectiveness of punishment (e.g., Azrin
et al., 1963). Basic findings further suggest that the
Magnitude of the Punisher
total amount of reinforcement obtained from other
Basic research conducted primarily with electric sources should meet or exceed that obtained be-
shock indicates that a larger amount or intensity fore punishment (i.e., when the individual could
of punishment will produce larger reductions in freely engage in the unpunished response; Fantino,
behavior if the magnitude is not increased gradu- 1973). As such, punishment may be most effective
ally over time (e.g., Cohen, 1968; Terris & Barnes, when the participant can obtain the maintaining
1969). Some applied studies have shown that reinforcer(s) for problem behavior or reinforcers
the intensity or duration of punishment influ- that are highly substitutable for the functional
364 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
reinforcer(s) through many sources (e.g., both con- Piazza et al. (1996) exposed a man with devel-
tingent on and independently of responding). opmental disabilities who engaged in cigarette
pica to discrimination training with two differ-
ent-colored cards. The researchers punished pica
Reinforcement of the Punished Response
with response interruption in the presence of a
Many applied studies have shown that punishment purple card, but not in the presence of a yellow
can be effective in the absence of extinction (e.g., card. No pica occurred in the presence of the
Fisher, Piazza, Bowman, Hagopian, & Langdon, purple card when the researchers subsequently
1994; Keeney et al., 2000; Lerman et al., 1997; evaluated stimulus control in the absence of the
Thompson et al., 1999). In fact, clinicians are punishment contingency. Finally, McKenzie et al.
more likely to use punishment when the reinforcer (2008) delivered reprimands for eye poking when
maintaining behavior is unknown or cannot be a participant with intellectual developmental dis-
controlled. Nonetheless, basic research indicates order wore wristbands, and withheld punishment
that punishment will be more effective if a behav- when the participant did not wear the wristbands.
ior analyst withholds or infrequently delivers rein- Subsequently, the participant did not engage in
forcement for the target behavior (Azrin & Holz, eye poking when she wore the wristbands in other
1966). The punisher may even acquire discrimina- settings, even though the researchers provided no
tive or conditioned reinforcing properties if a be- differential consequences for the behavior. Further
havior consistently produces both punishment and research is needed to evaluate the long-term effec-
reinforcement under certain conditions (e.g., Holz tiveness of antecedent-control techniques.
& Azrin, 1961). Thus behavior analysts should
combine punishment with extinction or other
Use of Conditioned Punishers
procedures to reduce the amount of reinforcement
available for the punished response whenever pos- Behavior analysts may enhance the effectiveness
sible (see Vollmer et al., Chapter 20, this volume). of intermittent, mild, or delayed punishers by es-
tablishing and using conditioned aversive stimuli
in treatment. Results of basic and applied studies
Antecedent Control of the Punished Response
indicate that stimuli that are neutral or ineffective
Results of at least three applied studies have shown as punishers may function as punishers after they
that establishing discriminative control over the are associated with punishing stimuli (e.g., Dixon,
punished response by pairing a stimulus with pun- Helsel, Rojahn, Cipolone, & Lubestsky, 1989;
ishment for engaging in the behavior enhances an Dorsey et al., 1980; Hake & Azrin, 1965; Salvy
intervention’s effectiveness (e.g., Maglieri et al., et al., 2004; Vorndran & Lerman, 2006). For ex-
2000; McKenzie, Smith, Simmons, & Soderlund, ample, a verbal reprimand alone was effective in
2008; Piazza, Hanley, & Fisher, 1996). In these suppressing problem behavior after the researchers
studies, the researcher delayed or omitted the pun- paired the word no with an effective punisher such
isher while the researchers evaluated the effects as shock or water mist contingent on responding
of the antecedent. In a study mentioned earlier, in several applied studies (e.g., Dorsey et al., 1980;
Maglieri et al. (2000) used discrimination training Lovaas & Simmons, 1969). Establishing and using
to treat the food stealing of a young girl with Prad- conditioned punishers may be especially beneficial
er–Willi syndrome. The researchers placed an or- when the caregiver withholds or delays the pri-
ange sticker on one of two containers of prohibited mary (unconditioned) punisher periodically. Fur-
food in a refrigerator. The researchers established thermore, conditioning may permit caregivers to
the sticker as a discriminative stimulus by repri- maintain intervention effects with less intrusive
manding the girl when she consumed food from procedures if they occasionally pair the condi-
the container with the sticker, and by providing tioned punisher with the original punisher (e.g.,
no differential consequences when she consumed Vorndran & Lerman, 2006). However, we need
food from the container without the sticker. As further research on the long-term maintenance of
expected, the girl consumed food from the con- conditioned punishment.
tainer without the sticker, but not from the con-
tainer with the sticker. Food stealing immediately
Maintenance of Punishment Effects
decreased to zero when the researchers placed the
sticker on both containers and on containers in a Several applied studies have reported the long-
different refrigerator. term efficacy of punishment with electric shock,
Developing Function-Based Punishment Procedures 365
overcorrection, physical restraint, verbal repri- individual’s contact with the punishing stimulus.
mands, and water mist, among others (e.g., Alt- Ensuring that caregivers continue to implement
man et al., 1978; Arntzen & Werner, 1999; Iwata, the intervention correctly and consistently, com-
Rolider, & Dozier, 2009; Kazdin, 1971; Richman bining punishment with dense schedules of alter-
et al., 2001; Rolider, Williams, Cummings, & Van native reinforcement, and removing or reducing
Houten, 1991; Salvy et al., 2004; Zegiob, Jenkins, reinforcement for problem behavior should also
Becker, & Bristow, 1976). Researchers typically increase the likelihood of successful maintenance
collected data from 6 to 12 months after the ini- (e.g., Foxx, 2003; Foxx & Livesay, 1984; Linscheid,
tial application of punishment, during which time Hartel, & Cooley, 1993).
they continued, faded, or withdrew the original A complex or labor-intensive intervention may
treatment. Researchers have reported success- be difficult for staff or caregivers to sustain over
ful clinical outcomes across longer periods, from time, possibly leading to relapse. Nevertheless, fad-
several years to 25 years after treatment (Duker ing or modifying components of the original pro-
& Seys, 1996; Foxx, Bittle, & Faw, 1989; Foxx & cedure to simplify implementation may decrease
Livesay, 1984; McGlynn & Locke, 1997). Howev- the effectiveness of the intervention. Several stud-
er, the reliability and validity of longer-term out- ies by Foxx and colleagues have demonstrated a
comes are less clear, due to a reliance on anecdotal systematic approach for successfully fading pun-
information, archival records, indirect measures of ishment (see Foxx, 2003, for a review). Foxx et al.
outcome, and/or circumscribed observations. (1989), for example, successfully treated the severe
Despite numerous examples of long-term main- aggression exhibited by a man with a dual diag-
tenance, problem behavior sometimes reemerges nosis across 52 months. The initial intervention
during and following treatment withdrawal (e.g., included punishment with contingent electric
Arntzen & Werner, 1999; Duker & Seys, 1996; shock, escape extinction, and reinforcement of
Iwata et al., 2009; Williams et al., 1993). For ex- compliance. As part of the maintenance program,
ample, Ricketts, Goza, and Matese (1993) and the researchers replaced contingent electric shock
Williams et al. (1993) reported relapses in treat- with a nonexclusionary time-out procedure by
ment with contingent electric shock 6 months first combining the two punishers and then dis-
and 31 months, respectively, after the initiation continuing the shock after 12 months. They only
of treatment. In fact, we cannot determine the withdrew shock after the participant had been
likelihood that treatment with punishment will responding regularly to positive reinforcers in his
produce long-term effects by examining the litera- home and work settings. Over the next 30 months,
ture, because researchers generally do not submit the researchers gradually decreased the time-out
and journals generally do not publish treatment duration from 3 hours to 15 minutes, and aggres-
failures. More important, few studies have directly sion remained low. The researchers noted that the
evaluated strategies that might promote successful time-out procedure had been ineffective before
maintenance. treatment with contingent electric shock. These
Nonetheless, factors that have correlated with results suggest that a gradual, highly systematic
specific cases of successful and unsuccessful main- plan for fading intervention is a key component of
tenance may suggest some potential approaches long-term maintenance.
for ensuring that punishment effects maintain
over the long run. Researchers have attributed re-
Stimulus Generalization of Punishment Effects
lapses to habituation or tolerance to the punisher,
problems with intervention consistency or integ- Unlike the research on maintenance, research on
rity, restricted opportunities to receive reinforce- stimulus generalization of punishment effects has
ment for appropriate behavior, and continued re- found that such effects rarely generalize beyond
inforcement of the punished behavior (e.g., Duker the intervention setting or context (e.g., Doke &
& Seys, 1996; Foxx & Livesay, 1984; Ricketts et Epstein, 1975; Lovaas & Simmons, 1969; Marho-
al., 1993). Habituation to the punisher may be less lin & Townsend, 1978; Rollings, Baumeister, &
likely to occur with limited or infrequent exposure Baumeister, 1977). Reductions in problem behav-
to the punishing stimulus. Using brief punishers ior typically failed to occur in contexts that were
that are highly effective in reducing problem be- not associated with punishment, despite attempts
havior, or randomly alternating among several ef- to promote such transfer (e.g., Birnbrauer, 1968;
fective punishers (e.g., Charlop, Burgio, Iwata, & Corte, Wolf, & Locke, 1971; Tate & Baroff, 1966).
Ivancic, 1988; Toole et al., 2004), would restrict an Basic studies with humans have also had difficulty
366 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
demonstrating generalization (e.g., O’Donnell & caregivers’ misuse of punishment (e.g., Kazdin,
Crosbie, 1998; O’Donnell, Crosbie, Williams, & 2001). Although basic and applied studies have
Saunders, 2000). reported these indirect effects, researchers have
However, researchers have conducted surpris- conducted few systematic analyses with clinical
ingly few studies in this area. Further research is problems. Nonetheless, research findings indicate
needed on strategies to promote generalization, be- that punishment can produce short-lived increases
cause caregivers are often unable to closely moni- or decreases in unpunished appropriate and inap-
tor behavior or implement interventions across set- propriate responses. For example, although several
tings and situations. Problem behavior is likely to studies have reported increases in aggression or cry-
emerge wherever a caregiver withholds a punisher, ing during treatment with punishment (e.g., Duker
especially if the behavior continues to produce re- & Seys, 1996; Hagopian & Adelinis, 2001), other
inforcement. Current knowledge about generaliza- studies have shown collateral decreases in unpun-
tion suggests that punishment effects may be more ished topographies of targeted problem behavior,
likely to transfer to contexts that closely resemble such as stereotypy (Cook, Rapp, Gomes, Frazer,
the intervention setting or situation (e.g., Guttman & Lindblad, 2014), as well as increases in positive
& Kalish, 1956; Hoffman & Fleshler, 1965; Honig affect or appropriate behavior, such as compliance
& Slivka, 1964). In fact, the presence of stimuli and toy play (e.g., Koegel, Firestone, Kramme, &
that have acquired tight control over responding Dunlap, 1974; Rolider et al., 1991; Toole et al.,
may help ensure that problem behavior rarely oc- 2003). Interestingly, researchers have reported that
curs in the absence of the punishment contingency some of these response forms, such as aggression,
(see Rollings & Baumeister, 1981). crying, and toy play, decrease under punishment
Results of three studies described previously (Bitgood et al., 1980; Lerman, Kelley, Vorndran, &
suggest that researchers can establish discrimina- Van Camp, 2003; Linscheid et al., 1990; Thomp-
tive control over the punished response by pairing son et al., 1999). Thus the likelihood of obtaining
a stimulus with punishment for engaging in the indirect effects and the forms that they may take
behavior (Maglieri et al., 2000; McKenzie et al., may be difficult to predict in application.
2008; Piazza et al., 1996). In these studies, prob- Basic findings suggest that aggression and emo-
lem behavior did not occur when the researchers tional responses may be more likely to occur when
introduced the discriminative stimulus (a sticker, individuals are exposed to unavoidable, intense
colored card, or wristbands) in settings that had aversive stimulation (e.g., Azrin, Hutchinson,
never been associated with punishment. Although & Hake, 1966; Hunt & Brady, 1995). Collateral
these findings are promising, we need further re- changes also may occur among unpunished re-
search on the durability of this strategy, because sponses that (1) are functionally equivalent to the
the researchers examined generalization in only a punished behavior (e.g., Baker, Woods, Tait, &
few sessions. Gardiner, 1986; St. Peter, Byrd, Pence, & Foreman,
Finally, generalization may be more likely to 2016), (2) occur in the same context as the pun-
occur if the intervention conditions are modified ished behavior (e.g., Bolles, Holtz, Dunn, & Hill,
systematically to resemble those present in the 1980), or (3) tend to immediately follow the pun-
generalization contexts. For example, a behavior ished behavior (e.g., Dunham, 1977, 1978). Both
analyst could fade the magnitude, schedule, im- basic and applied studies suggest that minimizing
mediacy, or type of punishment gradually (Foxx et exposure to the punishing stimulus with brief pun-
al., 1989) and could incorporate stimuli from the ishers that are highly effective, combining punish-
generalization setting (i.e., people, activities, ma- ment with rich reinforcement schedules for alter-
terials) into the intervention setting before testing native behavior, and withholding reinforcement
for generalization. for functionally equivalent problem behavior may
decrease the likelihood of increases in aggression
and other undesirable indirect effects.
Indirect Effects of Punishment
The most commonly described disadvantages of
punishment include the risk of elicited and oper- CONCLUSIONS
ant aggression; other emotional responses, such
as crying; decreases in appropriate behavior or Numerous procedural variations of punishment
generalized response suppression; escape from or produce durable reductions in problem behavior,
avoidance of the punishing agent or situation; and even when the contingencies maintaining the
Developing Function-Based Punishment Procedures 367
behavior are unknown. Knowledge of behavioral Arntzen, E., & Werner, S. B. (1999). Water mist punish-
function, however, should increase the likelihood ment for two classes of problem behaviour. Scandina-
of selecting an effective punisher, obtaining long- vian Journal of Behaviour Therapy, 28, 88–93.
term reductions in behavior, and successfully fad- Azrin, N. H., & Holz, W. C. (1966). Punishment. In W.
K. Honig (Ed.), Operant behavior: Areas of research
ing treatment. Furthermore, basic and applied
and application (pp. 380–447). New York: Appleton.
findings suggest that we can improve the effective- Azrin, N. H., Holz, W. C., & Hake, D. F. (1963). Fixed-
ness of punishment by (1) selecting punishers via ratio punishment. Journal of the Experimental Analy-
pretreatment avoidance, choice, or activity assess- sis of Behavior, 6, 141–148.
ments; (2) delivering the consequence immedi- Azrin, N. H., Hutchinson, R. R., & Hake, D. F. (1966).
ately after each instance of the behavior; (3) en- Extinction induced aggression. Journal of the Experi-
suring that alternative reinforcement sources are mental Analysis of Behavior, 9, 191–204.
available readily; (4) establishing discriminative Azrin, N. H., Nunn, R. G., & Frantz, S. E. (1980). Habit
control over the punished response; and (5) de- reversal vs. negative practice treatment of nailbiting.
veloping and using conditioned punishers. None- Behavoiur Research and Therapy, 18, 281–285.
Azrin, N. H., & Powers, M. A. (1975). Eliminating class-
theless, we need further research on strategies to
room disturbances of emotionally disturbed children
promote the long-term and generalized effects of by positive practice procedures. Behavior Therapy, 6,
punishment in applied settings. 525–534.
Baker, A. G., Woods, W., Tait, R., & Gardiner, K.
(1986). Punishment suppression: Some effects on al-
REFERENCES ternative behavior. Quarterly Journal of Experimental
Psychology: Comparative and Physiological Psychology,
Abramowitz, A. J., & O’Leary, S. G. (1990). Effective- 38B, 191–215.
ness of delayed punishment in an applied setting. Be- Ball, T. S., Sibbach, L., Jones, R., Steele, B., & Frazier,
havior Therapy, 21, 231–239. L. (1975). An accelerometer-activated device to con-
Ahearn, W. H., Clark, K. M., MacDonald, R. P. F., & trol assaultive and self-destructive behaviors in retar-
Chung, B. I. (2007). Assessing and treating vocal ste- dates. Journal of Behavior Therapy and Experimental
reotypy in children with autism. Journal of Applied Psychiatry, 6, 223–228.
Behavior Analysis, 40, 263–275. Banks, R. K., & Vogel-Sprott, M. (1965). Effect of de-
Ahrens, E. N., Lerman, D. C., Kodak, T., Worsdell, layed punishment on an immediately rewarded re-
A. S., & Keegan, C. (2011). Further evaluation of sponse in humans. Journal of Experimental Psychol-
response interruption and redirection as treatment ogy, 70, 357–359.
for stereotypic behavior. Journal of Applied Behavior Baron, A. (1991). Avoidance and punishment. In I. H.
Analysis, 44, 95–108. Iverson & K. A. Lattal (Eds.), Experimental analysis
Alberto, P. A., & Troutman, A. C. (2006). Applied be- of behavior (Pt. 1, pp. 173–217). Amsterdam: Elsevier
havior analysis for teachers (7th ed.). Upper Saddle Science.
River, NJ: Pearson/Merrill Prentice Hall. Barton, L. E., Brulle, A. R., & Repp, A. C. (1987). Ef-
Allen, K. D. (1998). The use of an enhanced simplified fects of differential schedule of time-out to reduce
habit reversal procedure to reduce disruptive out- maladaptive responding. Exceptional Children, 53,
bursts during athletic performance. Journal of Applied 351–356.
Behavior Analysis, 31, 489–492. Barton, R. P., Matson, J. L., Shapiro, E. S., & Ollendick,
Allison, J., & Timberlake, W. (1974). Instrumental and T. H. (1981). A comparison of punishment and DRO
contingent saccharin licking in rats: Response depri- procedures for treating stereotypic behavior of men-
vation and reinforcement. Learning and Motivation, tally retarded children. Applied Research in Mental
5, 231–247. Retardation, 2, 247–256.
Altman, K., Haavik, S., & Cook, J. W. (1978). Punish- Behavior Analyst Certification Board. (2014). BACB
ment of self-injurious behavior in natural settings professional and ethical compliance code for behav-
using contingent aromatic ammonia. Behaviour Re- ior analysts. Retrieved March 30, 2019, from www.
search and Therapy, 16, 85–96. bacb.com/wp-content/uploads/BACB-Compliance-
Anderson, J., & Le, D. D. (2011). Abatement of intrac- Code-english_190318.pdf.
table vocal stereotypy using an overcorrection proce- Birnbrauer, J. S. (1968). Generalization of punishment
dure. Behavior Interventions, 26, 134–146. effects. A case study. Journal of Applied Behavior
Armstrong, K. J., Ehrhardt, K. E., Cool, R. T., & Poling, Analysis, 1, 201–211.
A. (1997). Social validity and treatment integrity Bitgood, S. C., Crowe, M. J., Suarez, Y., & Peters, R. D.
data: Reporting in articles published in the Journal (1980). Immobilization: Effects and side effects on
of Developmental and Physical Disabilities, 1992–1995. stereotyped behavior in children. Behavior Modifica-
Journal of Developmental and Physical Disabilities, 9, tion, 4, 187–208.
359–367. Bolles, R. C., Holtz, R., Dunn, T., & Hill, W. (1980).
368 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
Comparisons of stimulus learning and response tion: Side effects and extended applications. Journal
learning in a punishment situation. Learning and of Experimental Child Psychology, 20, 496–511.
Motivation, 11, 78–96. Doke, L. A., Wolery, M., & Sumberc, C. (1983). Treat-
Calhoun, K. S., & Matherne, P. (1975). The effects of ing chronic aggression: Effects and side effects of
varying schedules of time-out on aggressive behavior response-contingent ammonia spirits. Behavior Modi-
of a retarded girl. Journal of Behavior Therapy and Ex- fication, 7, 531–556.
perimental Psychiatry, 6, 139–143. Dominguez, A., Wilder, D. A., Cheung, K., & Rey, C.
Charlop, M. H., Burgio, L. D., Iwata, B. A., & Ivancic, (2014). The use of a verbal reprimand to decrease ru-
M. T. (1988). Stimulus variation as a means of en- mination in a child with autism. Behavioral Interven-
hancing punishment effects. Journal of Applied Be- tions, 29, 339–345.
havior Analysis, 21, 89–95. Donaldson, J. M., & Vollmer, T. R. (2011). An evalu-
Cipani, E., Brendlinger, J., McDowell, L., & Usher, S. ation and comparison of time-out procedures with
(1991). Continuous vs. intermittent punishment: A and without release contingencies. Journal of Applied
case study. Journal of Developmental and Physical Dis- Behavior Analysis, 44, 693–705.
abilities, 3, 147–156. Dorsey, M. F., Iwata, B. A., Ong, P., & McSween, T. E.
Clark, H. B., Rowbury, T., Baer, A. M., & Baer, D. M. (1980). Treatment of self-injurious behavior using
(1973). Timeout as a punishing stimulus in continu- a water mist: Initial response suppression and gen-
ous and intermittent schedules. Journal of Applied eralization. Journal of Applied Behavior Analysis, 13,
Behavior Analysis, 6, 443–455. 343–353.
Cohen, P. S. (1968). Punishment: The interactive ef- Dougher, M. J. (1983). Clinical effects of response de-
fects of delay and intensity of shock. Journal of the privation and response satiation procedures. Behavior
Experimental Analysis of Behavior, 11, 789–799. Therapy, 14, 286–298.
Cole, G. A., Montgomery, R. W., Wilson, K. M., & 9Duker, P. C., & Seys, D. M. (1996). Long-term use of
Milan, M. A. (2000). Parametric analysis of overcor- electrical aversion treatment with self-injurious be-
rection duration effects: Is longer really better than havior. Research in Developmental Disabilities, 17,
shorter? Behavior Modification, 24, 359–378. 293–301.
Cook, J. L., Rapp, J. T., Gomes, L. A., Frazer, T. J., & Dunham, P. J. (1977). The nature of reinforcing stimuli.
Lindblad, T. L. (2014). Effects of verbal reprimands In W. K. Honig & J. E. R. Staddon (Eds.), Handbook
on targeted and untargeted stereotypy. Behavioral In- of operant behavior (pp. 98–124). Englewood Cliffs,
terventions, 29, 106–124. NJ: Prentice-Hall.
Cook, J. W., Altman, K., Shaw, J., & Blaylock, M. Dunham, P. J. (1978). Changes in unpunished respond-
(1978). Use of contingent lemon juice to eliminate ing during response-contingent punishment. Animal
public masturbation by a severely retarded boy. Be- Learning and Behavior, 6, 174–180.
haviour Research and Therapy, 16, 131–134. Dupuis, D. L., Lerman, D. C., Tsami, L., & Shireman, M.
Cooper, J. O., Heron, T. E., & Heward, W. L. (1987). Ap- L. (2015). Reduction of aggression evoked by sounds
plied behavior analysis. Columbus, OH: Merrill. using noncontingent reinforcement and time-out.
Corte, H. E., Wolf, M. M., & Locke, B. J. (1971). A com- Journal of Applied Behavior Analysis, 48, 669–674.
parison of procedures for eliminating self-injurious Epstein, L. H., & Masek, B. J. (1978). Behavioral control
behavior of retarded adolescents. Journal of Applied of medicine compliance. Journal of Applied Behavior
Behavior Analysis, 4, 201–213. Analysis, 11, 1–9.
Couvillon, M. A., Kane, E. J., Peterson, R. L., Ryan, J. B., Falcomata, T. S., Roane, H. S., Hovanetz, A. N., Ketter-
& Scheuermann, B. (2019). Policy and program con- ing, T. L., & Keeney, K. M. (2004). An evaluation of
siderations for choosing crisis intervention programs. response cost in the treatment of inappropriate vo-
Journal of Disability Policy Studies, 30, 35–45. calizations maintained by automatic reinforcement.
Crosbie, J. (1998). Negative reinforcement and punish- Journal of Applied Behavior Analysis, 37, 83–87.
ment. In K. A. Lattal & M. Perone (Eds.), Hand- Fantino, E. (1973). Aversive control. In J. A. Nevin &
book of research methods in human operant behavior G. S. Reynolds (Eds.), The study of behavior: Learn-
(pp. 163–189). New York: Plenum Press. ing, motivation, emotion, and instinct (pp. 239–279).
DeRosa, N. M., Roane, H. S., Bishop, J. R., & Silkowski, Glenview, IL: Scott, Foresman.
E. L. (2016). The combined effects of noncontingent Fischer, J., & Nehs, R. (1978). Use of a commonly avail-
reinforcement and punishment on the reduction of able chore to reduce a boy’s rate of swearing. Journal
rumination. Journal of Applied Behavior Analysis, 49, of Behavior Therapy and Experimental Psychiatry, 9,
680–685. 81–83.
Dixon, M. J., Helsel, W. J., Rojahn, J., Cipollone, R., & Fisher, W. W., Piazza, C. C., Bowman, L. G., Hagopian,
Lubetsky, M. J. (1989). Aversive conditioning of visu- L. P., & Langdon, N. A. (1994). Empirically derived
al screening with aromatic ammonia for treating ag- consequences: A data-based method for prescribing
gressive and disruptive behavior in a developmental- treatments for destructive behavior. Research in De-
ly disabled child. Behavior Modification, 13, 91–107. velopmental Disabilities, 15, 133–149.
Doke, L. A., & Epstein, L. H. (1975). Oral overcorrec- Fisher, W. W., Piazza, C. C., Bowman, L. G., Kurtz, P. F.,
Developing Function-Based Punishment Procedures 369
Sherer, M. R., & Lachman, S. R. (1994). A prelimi- fectiveness of an intrusive procedure: A case report.
nary evaluation of empirically derived consequences Journal of Applied Behavior Analysis, 27, 171–172.
for the treatment of pica. Journal of Applied Behavior Gross, A. M., Wright, B., & Drabman, R. S. (1980). The
Analysis, 27, 447–457. empirical selection of a punisher for a retarded child’s
Fisher, W. W., Piazza, C. C., Cataldo, M. F., Harrell, R., self-injurious behavior: A case study. Child Behavior
Jefferson, G., & Conner, R. (1993). Functional com- Therapy, 2, 59–65.
munication training with and without extinction Guttman, N., & Kalish, H. I. (1956). Discriminability
and punishment. Journal of Applied Behavior Analy- and stimulus generalization. Journal of Experimental
sis, 26, 23–36. Psychology, 51, 79–88.
Fonagy, P., & Slade, P. (1982). Punishment vs. negative Hagopian, L. P., & Adelinis, J. D. (2001). Response
reinforcement in the aversive conditioning of audi- blocking with and without redirection for the treat-
tory hallucinations. Behaviour Research and Therapy, ment of pica. Journal of Applied Behavior Analysis, 34,
20, 483–492. 527–530.
Foxx, R. M. (1982). Decreasing behaviors of retarded and Hagopian, L. P., Fisher, W. W., Thibault-Sullivan, M.,
autistic persons. Champaign, IL: Research Press. Acquisto, J., & LeBlanc, L. A. (1998). Effectiveness
Foxx, R. M. (2003). The treatment of dangerous behav- of functional communication training with and
ior. Behavioral Interventions, 18, 1–21. without extinction and punishment: A summary of
Foxx, R. M., & Azrin, N. H. (1972). Restitution: A 21 inpatient cases. Journal of Applied Behavior Analy-
method of eliminating aggressive–disruptive behav- sis, 31, 211–235.
ior of retarded and brain damaged patients. Behaviour Hagopian, L. P., Rooker, G. W., & Zarcone, J. R. (2015).
Research and Therapy, 10, 15–27. Delineating subtypes of self-injurious behavior main-
Foxx, R. M., & Azrin, N. H. (1973). The elimination of tained by automatic reinforcement. Journal of Applied
autistic self-stimulatory behavior by overcorrection. Behavior Analysis, 48, 523–543.
Journal of Applied Behavior Analysis, 6, 1–14. Hake, D. F., & Azrin, N. H. (1965). Conditioned pun-
Foxx, R. M., Bittle, R. G., & Faw, G. D. (1989). A ishment. Journal of the Experimental Analysis of Be-
maintenance strategy for discontinuing aversive pro- havior, 8, 279–293.
cedures: A 32-month follow-up of the treatment of Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci,
aggression. American Journal on Mental Retardation, S. A., & Maglieri, K. A. (1997). Evaluation of client
94, 27–36. preference for function-based treatment packages.
Foxx, R. M., & Livesay, J. (1984). Maintenance of re- Journal of Applied Behavior Analysis, 30, 459–473.
sponse suppression following overcorrection: A 10- Hanley, G. P., Piazza, C. C., Fisher, W. W., & Maglieri,
year retrospective examination of eight cases. Analy- K. A. (2005). On the effectiveness of and preference
sis and Intervention in Developmental Disabilities, 4, for punishment and extinction components of func-
65–79. tion-based interventions. Journal of Applied Behavior
Foxx, R. M., McHenry, W. C., & Bremer, B. A. (1996). Analysis, 38, 51–65.
The effects of a video vignette on increasing treat- Hobbs, S. A., Forehand, R., & Murray, R. G. (1978).
ment acceptability. Behavioral Interventions, 11, Effects of various durations of timeout on the non-
131–140. compliant behavior of children. Behavior Therapy, 9,
Foxx, R. M., & Shapiro, S. T. (1978). The timeout rib- 652–656.
bon: A nonexclusionary timeout procedure. Journal Hoffman, H. S., & Fleshler, M. (1965). Stimulus aspects
of Applied Behavior Analysis, 11, 125–136. of aversive controls: The effects of response contin-
Friman, P. C., Cook, J. W., & Finney, J. W. (1984). Ef- gent shock. Journal of the Experimental Analysis of
fects of punishment procedures on the self-stimulato- Behavior, 8, 89–96.
ry behavior of an autistic child. Analysis and Interven- Holburn, C. S., & Dougher, M. J. (1986). Effects of re-
tion in Developmental Disabilities, 4, 39–46. sponse satiation procedures in the treatment of aero-
Gaylord-Ross, R. J., Weeks, M., & Lipner, C. (1980). An phagia. American Journal of Mental Deficiency, 91,
analysis of antecedent, response, and consequence 72–77.
events in the treatment of self-injurious behavior. Holz, W. C., & Azrin, N. H. (1961). Discriminative
Education and Training of the Mentally Retarded, 15, properties of punishment. Journal of the Experimental
35–42. Analysis of Behavior, 4, 225–232.
Giles, A. F., St. Peter, C. C., Pence, S. T., & Gibson, Holz, W. C., Azrin, N. H., & Ayllon, T. (1963). Elimina-
A. B. (2012). Preference for blocking or response tion of behavior of mental patients by response-pro-
redirection during stereotypy treatment. Research in duced extinction. Journal of the Experimental Analysis
Developmental Disabilities, 33, 1691–1700. of Behavior, 6, 407–412.
Goodall, G. (1984). Learning due to response-shock Honig, W. K., & Slivka, R. M. (1964). Stimulus gener-
contingency in signaled punishment. Quarterly Jour- alization of the effects of punishment. Journal of the
nal of Experimental Psychology, 36B, 259–279. Experimental Analysis of Behavior, 7, 21–25.
Grace, N. C., Kahng, S. W., & Fisher, W. W. (1994). Hunt, H. F., & Brady, J. V. (1995). Some effects of pun-
Balancing social acceptability with treatment ef- ishment and intercurrent “anxiety” on a single oper-
370 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
ant. Journal of Comparative and Physiological Psychol- Lerman, D. C., & Iwata, B. A. (1996). A methodology
ogy, 48, 305–310. for distinguishing between extinction and punish-
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & ment effects associated with response blocking. Jour-
Richman, G. S. (1994). Toward a functional analysis nal of Applied Behavior Analysis, 29, 231–234.
of self-injury. Journal of Applied Behavior Analysis, 27, Lerman, D. C., Iwata, B. A., Shore, B. A., & DeLeon,
197–209. (Reprinted from Analysis and Intervention in I. G. (1997). Effects of intermittent punishment on
Developmental Disabilities, 2, 3–20, 1982) self-injurious behavior: An evaluation of schedule
Iwata, B. A., Rolider, N. U., & Dozier, C. L. (2009). thinning. Journal of Applied Behavior Analysis, 30,
Evaluation of timeout programs through phased 187–201.
withdrawal. Journal of Applied Research in Intellectual Lerman, D. C., Kelley, M. E., Vorndran, C. M., & Van
Disabilities, 22(2), 203–209. Camp, C. M. (2003). Collateral effects of response
Iwata, B. A., Vollmer, T. R., & Zarcone, J. R. (1990). The blocking during the treatment of stereotypic behav-
experimental (functional) analysis of behavior disor- ior. Journal of Applied Behavior Analysis, 36, 119–123.
ders: Methodology, applications, and limitations. In Lerman, D. C., & Vorndran, C. M. (2002). On the sta-
A. C. Repp & N. N. Singh (Eds.), Perspectives on the tus of knowledge for using punishment: Implications
use of nonaversive and aversive interventions for persons for treating behavior disorders. Journal of Applied Be-
with developmental disabilities (pp. 301–330). Syca- havior Analysis, 35, 431–464.
more, IL: Sycamore. Lindberg, J. S., Iwata, B. A., & Kahng, S. (1999). On the
Kahng, S., Abt, K. A., & Wilder, D. (2001). Treatment relation between object manipulation and stereotyp-
of self-injury correlated with mechanical restraints. ic self-injurious behavior. Journal of Applied Behavior
Behavioral Interventions, 16, 105–110. Analysis, 32, 51–62.
Kahng, S., Iwata, B. A., & Lewin, A. B. (2002). Behav- Linscheid, T. R., Hartel, F., & Cooley, N. (1993). Are
ioral treatment of self-injury, 1964 to 2000. American aversive procedures durable?: A five year follow-up
Journal on Mental Retardation, 107, 212–221. of three individuals treated with contingent electric
Kahng, S., Tarbox, J., & Wilke, A. E. (2001). Use of a shock. Child and Adolescent Mental Health Care, 3,
multicomponent treatment for food refusal. Journal of 67–76.
Applied Behavior Analysis, 34, 93–96. Linscheid, T. R., Iwata, B. A., Ricketts, R. W., Williams,
Kazdin, A. E. (1971). The effect of response cost in sup- D. E., & Griffin, J. C. (1990). Clinical evaluation
pressing behavior in a pre-psychotic retardate. Jour- of the Self-Injurious Behavior Inhibiting System
nal of Behavior Therapy and Experimental Psychiatry, (SIBIS). Journal of Applied Behavior Analysis, 23,
2, 137–140. 53–78.
Kazdin, A. E. (2001). Behavior modification in applied set- Linscheid, T. R., Pejeau, C., Cohen, S., & Footo-Lenz,
tings (6th ed.). Belmont, CA: Wadsworth/Thomson M. (1994). Positive side effects in the treatment of
Learning. SIB using the Self-Injurious Behavior Inhibiting Sys-
Keeney, K. M., Fisher, W. W., Adelinis, J. D., & Wilder, tem (SIBIS): Implications of operant and biochemi-
D. A. (2000). The effects of response cost in the cal explanations of SIB. Research in Developmental
treatment of aberrant behavior maintained by nega- Disabilities, 15, 81–90.
tive reinforcement. Journal of Applied Behavior Analy- Long, E. S., Miltenberger, R. G., Ellingson, S. A., & Ott,
sis, 33, 255–258. S. M. (1999). Augmenting simplified habit reversal
Koegel, R. L., Firestone, P. B., Kramme, K. W., & Dun- in the treatment of oral–digital habits exhibited by
lap, G. (1974). Increasing spontaneous play by sup- individuals with mental retardation. Journal of Ap-
pressing self-stimulation in autistic children. Journal plied Behavior Analysis, 32, 353–365.
of Applied Behavior Analysis, 7, 521–528. Lovaas, O. I., & Favell, J. E. (1987). Protection of clients
Krivacek, D., & Powell, J. (1978). Negative prefer- undergoing aversive/restrictive interventions. Educa-
ence management: Behavioral suppression using tion and Treatment of Children, 10, 311–325.
Premack’s punishment hypothesis. Education and Lovaas, O. I., & Simmons, J. Q. (1969). Manipulation of
Treatment of Children, 1, 5–13. self-destruction in three retarded children. Journal of
Lalli, J. S., Livezey, K., & Kates, K. (1996). Functional Applied Behavior Analysis, 2, 143–157.
analysis and treatment of eye poking with response Luce, S. C., Delquadri, J., & Hall, R. V. (1980). Contin-
blocking. Journal of Applied Behavior Analysis, 29, gent exercise: A mild but powerful procedure for sup-
129–132. pressing inappropriate verbal and aggressive behav-
LeBlanc, L. A., Hagopian, L. P., & Maglieri, K. A. ior. Journal of Applied Behavior Analysis, 13, 583–594.
(2000). Use of a token economy to eliminate exces- Lydon, S., Healy, O., Moran, L., & Foody, C. (2015). A
sive inappropriate social behavior in an adult with quantitative examination of punishment research.
developmental disabilities. Behavioral Interventions, Research in Developmental Disabilities, 36, 470–484.
15, 135–143. Mace, F. C., Page, T. J., Ivancic, M. T., & O’Brien, S.
Lennox, D. B., & Miltenberger, R. G. (1990). On the (1986). Effectiveness of brief time-out with and with-
conceptualization of treatment acceptability. Educa- out contingent delay: A comparative analysis. Jour-
tion and Training in Mental Retardation, 25, 211–224. nal of Applied Behavior Analysis, 19(1), 79–86.
Developing Function-Based Punishment Procedures 371
MacKenzie-Keating, S. E., & McDonald, L. (1990). ment maintenance involving life-threatening pica.
Overcorrection: Reviewed, revisited, and revised. Be- Behavioral Residential Treatment, 4, 191–211.
havior Analyst, 13, 39–48. Pelios, L., Morren, J., Tesch, D., & Axelrod, S. (1999). The
Magee, S. K., & Ellis, J. (2001). The detrimental effects impact of functional analysis methodology on treat-
of physical restraint as a consequence for inappropri- ment choice for self-injurious and aggressive behavior.
ate classroom behavior. Journal of Applied Behavior Journal of Applied Behavior Analysis, 32, 185–195.
Analysis, 34, 501–504. Peters, L. C., & Thompson, R. H. (2013). Some indirect
Maglieri, K. A., DeLeon, I. G., Rodriguez-Catter, V., & effects of positive practice overcorrection. Journal of
Sevin, B. M. (2000). Treatment of covert food steal- Applied Behavior Analysis, 46, 613–625.
ing in an individual with Prader–Willi syndrome. Piazza, C. C., Hanley, G. P., & Fisher, W. W. (1996).
Journal of Applied Behavior Analysis, 33, 615–618. Functional analysis and treatment of cigarette pica.
Marholin, D., & Townsend, N. M. (1978). An experi- Journal of Applied Behavior Analysis, 29, 437–450.
mental analysis of side effects and response mainte- Porterfield, J. K., Herbert-Jackson, E., & Risley, T. R.
nance of a modified overcorrection procedure. Be- (1976). Contingent observation: An effective and
havior Therapy, 9, 383–390. acceptable procedure for reducing disruptive behav-
Matson, J. L., & DiLorenzo, T. M. (1984). Punishment ior of young children in a group setting. Journal of
and its alternatives: New perspectives for behavior modi- Applied Behavior Analysis, 9, 55–64.
fication. New York: Springer. Rapp, J. T., Miltenberger, R. G., & Long, E. S. (1998).
Matson, J. L., & Keyes, J. B. (1990). A comparison of Augmenting simplified habit reversal with an aware-
DRO to movement suppression time-out and DRO ness enhancement device: Preliminary findings.
with two self-injurious and aggressive mentally re- Journal of Applied Behavior Analysis, 31, 665–668.
tarded adults. Research in Developmental Disabilities, Raulston, T. J., Hansen, S. G., Machalicek, W., Mc-
11, 111–120. Intyre, L. L., & Carnett, A. (2019). Interventions for
McGlynn, A. P., & Locke, B. J. (1997). A 25-year follow- repetitive behavior in young children with autism: A
up of a punishment program for severe self-injury. survey of behavioral practices. Journal of Autism and
Behavioral Interventions, 12, 203–207. Developmental Disorders, 49, 3047–3059.
McKenzie, S. D., Smith, R. G., Simmons, J. N., & Soder- Rawson, R. A., & Leitenberg, H. (1973). Reinforced
lund, M. J. (2008). Using a stimulus correlated with alternative behavior during punishment and extinc-
reprimands to suppress automatically maintained tion with rats. Journal of Comparative and Physiologi-
eye poking. Journal of Applied Behavior Analysis, 41, cal Psychology, 85, 593–600.
255–259. Realon, R. E., & Konarski, E. A. (1993). Using decel-
Mitteer, D. R., Romani, P. W., Greer, B. D., & Fisher, erative contingencies to reduce the self-injurious be-
W. W. (2015). Assessment and treatment of pica and havior of people with multiple handicaps: The effects
destruction of holiday decorations. Journal of Applied of response satiation? Research in Developmental Dis-
Behavior Analysis, 48, 912–917. abilities, 14, 341–357.
Mueller, M. M., Edwards, R. P., & Trahant, D. (2003). Reid, D. H., Parsons, M. B., Phillips, J. F., & Green, C.
Translating multiple assessment techniques into an W. (1993). Reduction of self-injurious hand mouth-
intervention selection model for classrooms. Journal ing using response blocking. Journal of Applied Behav-
of Applied Behavior Analysis, 36, 563–573. ior Analysis, 26, 139–140.
O’Brien, S., & Karsh, K. G. (1990). Treatment accept- Repp, A. C., & Deitz, D. E. D. (1978). On the selective
ability: Consumer, therapist, and society. In A. C. use of punishment: Suggested guidelines for adminis-
Repp & N. N. Singh (Eds.), Perspectives on the use of trators. Mental Retardation, 16, 250–254.
nonaversive and aversive interventions for persons with Richman, D. M., Lindauer, S. E., Crosland, K. A.,
developmental disabilities (pp. 503–516). Sycamore, IL: McKerchar, T. L., & Morse, P. S. (2001). Functional
Sycamore. analysis and treatment of breath holding maintained
O’Donnell, J., & Crosbie, J. (1998). Punishment gen- by nonsocial reinforcement. Journal of Applied Behav-
eralization gradients with humans. Psychological Re- ior Analysis, 34, 531–534.
cord, 48, 211–232. Ricketts, R. W., Goza, A. B., & Matese, M. (1993). A
O’Donnell, J., Crosbie, J., Williams, D. C., & Saunders, 4-year follow-up of treatment of self-injury. Journal
K. J. (2000). Stimulus control and generalization of of Behavior Therapy and Experimental Psychiatry, 24,
point-loss punishment with humans. Journal of the 57–62.
Experimental Analysis of Behavior, 73, 261–274. Ritschl, C., Mongrella, J., & Presbie, R. J. (1972). Group
Pace, G. M., Ivancic, M. T., Edwards, G. L., Iwata, B. A., time-out from rock and roll music and out-of-seat be-
& Page, T. J. (1985). Assessment of stimulus prefer- havior of handicapped children while riding a school
ence and reinforcer value with profoundly retarded bus. Psychological Reports, 31, 967–973.
individuals. Journal of Applied Behavior Analysis, 18, Rolider, A., & Van Houten, R. (1985). Suppressing tan-
249–255. trum behavior in public places through the use of
Paisey, T. L., & Whitney, R. B. (1989). A long-term case delayed punishment mediated by audio recordings.
study of analysis, response suppression, and treat- Behavior Therapy, 16, 181–194.
372 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
Rolider, A., Williams, L., Cummings, A., & Van Singh, N. N., Dawson, M. J., & Manning, P. J. (1981). The
Houten, R. (1991). The use of a brief movement re- effects of physical restraint on self-injurious behavior.
striction procedure to eliminate severe inappropriate Journal of Mental Deficiency Research, 25, 207–216.
behavior. Journal of Behavior Therapy and Experimen- Singh, N. N., Watson, J. E., & Winton, A. S. W. (1986).
tal Psychiatry, 22, 23–30. Treating self-injury: Water mist spray versus facial
Rollings, J. P., & Baumeister, A. A. (1981). Stimulus screening or forced arm exercise. Journal of Applied
control of stereotypic responding: Effects on target Behavior Analysis, 19, 403–410.
and collateral behavior. American Journal of Mental Smith, R. G., Russo, L., & Le, D. D. (1999). Distinguish-
Deficiency, 86, 67–77. ing between extinction and punishment effects of
Rollings, J. P., Baumeister, A. A., & Baumeister, A. A. response blocking: A replication. Journal of Applied
(1977). The use of overcorrection procedures to elimi- Behavior Analysis, 32, 367–370.
nate the stereotyped behaviors of retarded individuals: Spreat, S., & Lipinski, D. P. (1986). A survey of state
An analysis of collateral behaviors and generalization policies regarding use of restrictive/aversive behavior
of suppressive effects. Behavior Modification, 1, 29–46. modification procedures. Behavioral Residential Treat-
Romanczyk, R. G. (1977). Intermittent punishment of ment, 1, 137–152.
self-stimulation: Effectiveness during application and St. Peter, C. C., Byrd, J. D., Pence, S. T., & Foreman, A.
extinction. Journal of Consulting and Clinical Psychol- P. (2016). Effects of treatment-integrity failures on a
ogy, 45, 53–60. response-cost procedure. Journal of Applied Behavior
Rooker, G. W., Jessel, J., Kurtz, P. F., & Hagopian, L. Analysis, 49, 308–328.
P. (2013). Functional communication training with Stricker, J. M., Miltenberger, R. G., Garlinghouse, M.
and without alternative reinforcement and punish- A., Deaver, C. M., & Anderson, C. A. (2001). Evalu-
ment: An analysis of 58 applications. Journal of Ap- ation of an awareness enhancement device for the
plied Behavior Analysis, 46, 708–722. treatment of thumb sucking in children. Journal of
Rush, K. S., Crockett, J. L., & Hagopian, L. P. (2001). Applied Behavior Analysis, 34, 77–80.
An analysis of the selective effects of NCR with pun- Stricker, J. M., Miltenberger, R. G., Garlinghouse, M., &
ishment targeting problem behavior associated with Tulloch, H. E. (2003). Augmenting stimulus inten-
positive affect. Behavioral Interventions, 16, 127–135. sity with an awareness enhancement device in the
Saini, V., Greer, B. D., & Fisher, W. W. (2015). Clari- treatment of finger sucking. Education and Treatment
fying inconclusive functional analysis results: As- of Children, 26, 22–29.
sessment and treatment of automatically reinforced Tanner, B. A., & Zeiler, M. (1975). Punishment of self-
aggression. Journal of Applied Behavior Analysis, 48, injurious behavior using aromatic ammonia as the
315–330. aversive stimulus. Journal of Applied Behavior Analy-
Sajwaj, T., Libet, J., & Agras, S. (1974). Lemon-juice sis, 8, 53–57.
therapy: The control of life-threatening rumination Tate, B. G., & Baroff, G. S. (1966). Aversive control of
in a six-month-old infant. Journal of Applied Behavior self-injurious behavior in a psychotic boy. Behaviour
Analysis, 7, 557–563. Research and Therapy, 4, 281–287.
Salend, S. J., & Gordon, B. D. (1987). A group-oriented Terhune, J. G., & Premack, D. (1970). On the propor-
timeout ribbon procedure. Behavioral Disorders, 12, tionality between the probability of not-running and
131–137. the punishment effect of being forced to run. Learn-
Salvy, S., Mulick, J. A., Butter, E., Bartlett, R. K., & ing and Motivation, 1, 141–149.
Linscheid, T. R. (2004). Contingent electric shock Terhune, J. G., & Premack, D. (1974). Comparison of
(SIBIS) and a conditioned punisher eliminate severe reinforcement and punishment functions produced
head banging in a preschool child. Behavioral Inter- by the same contingent event in the same subject.
ventions, 19, 59–72. Learning and Motivation, 5, 221–230.
Shawler, L. A., & Miquel, C. F. (2015). The effects of Terris, W., & Barnes, M. (1969). Learned resistance to
motor and vocal response interruption and redirec- punishment and subsequent responsiveness to the
tion on vocal stereotypy and appropriate vocaliza- same and novel punishers. Psychonomic Science, 15,
tions. Behavioral Interventions, 30, 112–134. 49–50.
Singh, N. N. (1979). Aversive control of breath-holding. Thomas, J. R. (1968). Fixed-ratio punishment by time-
Journal of Behavior Therapy and Experimental Psychia- out of concurrent variable-interval behavior. Journal
try, 10, 147–149. of the Experimental Analysis of Behavior, 11, 609–616.
Singh, N. N., Dawson, M. J., & Gregory, P. R. (1980a). Thompson, R. H., Iwata, B. A., Conners, J., & Roscoe,
Self-injury in the profoundly retarded: Clinically sig- E. M. (1999). Effects of reinforcement for alternative
nificant versus therapeutic control. Journal of Mental behavior during punishment of self-injury. Journal of
Deficiency Research, 24, 87–97. Applied Behavior Analysis, 32, 317–328.
Singh, N. N., Dawson, M. J., & Gregory, P. R. (1980b). Toole, L. M., Bowman, L. G., Thomason, J. L., Hago-
Suppression of chronic hyperventilation using pian, L. P., & Rush, K. S. (2003). Observed increases
response-contingent aromatic ammonia. Behavior in positive affect during behavioral treatment. Be-
Therapy, 11, 561–566. havioral Interventions, 18, 35–42.
Developing Function-Based Punishment Procedures 373
Toole, L. M., DeLeon, I. G., Kahng, S., Ruffin, G. E., F., Hanley, G. P., Lennox, D., et al. (2011). The Asso-
Pletcher, C. A., & Bowman, L. G. (2004). Re-evalua- ciation for Behavior Analysis International position
tion of constant versus varied punishers using empiri- statement on restraint and seclusion. Behavior Ana-
cally derived consequences. Research in Developmen- lyst, 34, 103–110.
tal Disabilities, 25, 577–586. Vollmer, T. R., & Iwata, B. A. (1993). Implications of a
Trenholme, I. A., & Baron, A. (1975). Immediate and functional analysis technology for the use of restric-
delayed punishment of human behavior by loss of re- tive behavioral interventions. Child and Adolescent
inforcement. Learning and Motivation, 6, 62–79. Mental Health Care, 3, 95–113.
Truchlicka, M., McLaughlin, T. F., & Swain, J. C. Vorndran, C. M., & Lerman, D. C. (2006). Establishing
(1998). Effects of token reinforcement and response and maintaining treatment effects with less intrusive
cost on the accuracy of spelling performance with consequences via a pairing procedure. Journal of Ap-
middle-school special education students with be- plied Behavior Analysis, 39, 35–48.
havior disorders. Behavioral Interventions, 13, 1–10. Wacker, D. P., Steege, M. W., Northup, J., Sasso, G.,
Van Houten, R., Axelrod, S., Bailey, J. S., Favell, J. E., Berg, W., Reimers, T., et al. (1990). A component
Foxx, R. M., Iwata, B. A., et al. (1988). The right to analysis of functional communication training across
effective behavioral treatment. Journal of Applied Be- three topographies of severe behavior problems. Jour-
havior Analysis, 21, 381–384. nal of Applied Behavior Analysis, 23, 417–429.
Van Houten, R., Nau, P. A., MacKenzie-Keating, S. E., Watson, T. S. (1993). Effectiveness of arousal and arous-
Sameoto, D., & Colavecchia, B. (1982). An analysis of al plus overcorrection to reduce nocturnal bruxism.
some variables influencing the effectiveness of repri- Journal of Behavior Therapy and Experimental Psychia-
mands. Journal of Applied Behavior Analysis, 15, 65–83. try, 24, 181–185.
Van Houten, R., & Rolider, A. (1988). Recreating the Williams, D. E., Kirkpatrick-Sanchez, S., & Iwata, B.
scene: An effective way to provide delayed punish- A. (1993). A comparison of shock intensity in the
ment for inappropriate motor behavior. Journal of Ap- treatment of longstanding and severe self-injurious
plied Behavior Analysis, 21, 187–192. behavior. Research in Developmental Disabilities, 14,
Verriden, A. L., & Roscoe, E. M. (2018). An evaluation 207–219.
of a punisher assessment for decreasing automatically Zegiob, L. E., Jenkins, J., Becker, J., & Bristow, A.
reinforced problem behavior. Journal of Applied Be- (1976). Facial screening: Effects on appropriate and
havior Analysis, 52, 205–226. inappropriate behaviors. Journal of Behavior Therapy
Vollmer, T. R., Hagopian, L. P., Bailey, J. S., Dorsey, M. and Experimental Psychiatry, 7, 355–357.
CHAP TER 22
Token Economies
Researchers developed the first token-economy behaviors, and consequences that modify or influ-
programs in the early 1960s, and these represented ence behavior through the delivery of conditioned
some of the earliest applications of experimental reinforcement (Hackenberg, 2018). Antecedents
and conceptual innovations in applied behavior (i.e., discriminative stimuli or cues) are events pre-
analysis (Ferster & Skinner, 1957; Skinner, 1953). ceding a behavior that indicate that an individual
A general dissatisfaction with the quality of care should perform a behavior specified as part of a
provided to individuals in institutionalized settings, token economy. Behaviors are the actions speci-
and a desire to construct environments that would fied in the contingency relations. Consequences in
be more conducive to the development and mainte- the form of backup reinforcers should follow the
nance of adaptive behavior in these individuals (i.e., performance of the behavior reliably. Often the
a “total motivating environment”; Ayllon & Azrin, performance of the behavior signals the delivery
1968, p. 24), inspired the development of token of both immediate consequences and progress to-
economies. Ayllon and Azrin (1965) were among ward long-term, larger incentives.
the first to report positive results with a token-based A key feature of a token economy is the delivery
reward program. Allyon and Azrin initially imple- of symbolic or token reinforcers (e.g., poker chips,
mented a token economy at Anna State Hospital points) after the performance of the specified be-
in Illinois with individuals with severe impairments havior (Miltenberger, 2001). Most conditioned
such as schizophrenia. They showed that the con- reinforcers lack inherent value and influence
tingent delivery of token reinforcement increased behavior by affording individuals opportunities
work and self-care behavior, but that noncontin- to exchange them for backup reinforcers, such as
gent delivery of tokens or program discontinuation special privileges, edibles, and activities (Fiske,
resulted in drastic reductions in adaptive behavior. Isenhower, Bamond, & Lauderdale-Littin, 2020;
Ayllon and Azrin (1968) reported their extended Leon, Borrero, & DeLeon, 2016; Russell, Ingvars-
findings in The Token Economy: A Motivational Sys- son, Haggar, & Jessel, 2018). Thus the purpose of
tem for Therapy and Rehabilitation. tokens and conditioned reinforcers generally is to
bridge the delay between delivery of backup rein-
forcers and performance of the behavior (Clark,
DEFINING THE TOKEN ECONOMY Lachowicz, & Wolf, 1968; Dickerson, Tenhula, &
Green-Paden, 2005; Jones, Downing, Latkowski,
Token economies are formal descriptions of con- Ferre, & McMahon, 1992; Kazdin & Bootzin,
tingency relations, defined here as antecedents, 1972).
374
Token Economies 375
The following elements seem to be common fea- the request of a unit psychiatrist. Their analysis
tures of well-formulated contingency management revealed ambiguous target behaviors, such as fol-
or token-economy programs (Ivy, Meindi, Overley, lowing directions, being nice, and being where you
& Robson, 2017). First, a behavior analyst sets are supposed to be. The researchers also noted that
goals and specifies behaviors in observable terms. lengthy delays between the delivery and redemp-
Second, the behavior analyst identifies potential tion of tokens reduced program effectiveness. The
reinforcers and punishers, if applicable. Third, the researchers recommended the development of
behavior analyst or his or her designee monitors operational definitions for each target behavior.
target behaviors frequently and delivers conse- For example, they defined following directions as
quences consistently. Fourth, the behavior analyst “making eye contact with the speaker and initi-
develops a program that is flexible and can change ating compliance within 5–7 seconds of the re-
as the needs of the individual or program change. quest.” Staff delivered backup reinforcement based
Finally, collaboration among all parties, a written on points each patient earned the previous day
agreement, and formal monitoring processes are in the original program. Instead, the researchers
other program elements considered necessary to divided days into four blocks, and behavior dur-
maximize effectiveness (Miltenberger, 2001). The ing one block determined the privileges during
goal of a token-economy program is to strengthen the next block. These adjustments produced sub-
adaptive or desirable behavior and decrease prob- stantial reductions in time out and an increase in
lem behavior simultaneously (Miltenberger, 2001). points earned for target behavior such as following
Stokes and Baer (1977) recommended that behav- directions, but did not reduce problem behavior
ior analysts fade token-economy programs as rap- directly.
idly as possible after the programs are successful.
Interestingly, a meta-analysis of prize-based rein-
Identify Conditioned Reinforcers
forcement programs for persons with substance
abuse problems revealed that consistent monitor- Researchers have used tokens, imitation dollar
ing and ongoing praise were needed to sustain bills, buttons, stickers, and poker chips as con-
gains after program fading. Although token-econ- ditioned reinforcers (Reynolds & Kelley, 1997).
omy programs yielded abstinence in the immedi- The selection of conditioned reinforcers should be
ate postintervention follow-up, the participants sensitive to the implementation context. Hand-
did not maintain abstinence at a 6-month follow- ing a poker chip to a student who is working on
up (Benishek et al., 2014). an academic task, such as writing an essay, could
disrupt the behavior (Drabman & Tucker, 1974).
Some researchers have delivered checkmarks, smi-
Identify and Define Behavior
ley faces, or stars on a chart to reduce the poten-
The first and arguably most important step in im- tial disruption of token delivery (Anhalt, McNeil,
plementing a token-economy program is to iden- & Bahl, 1998; Higgins, Williams, & McLaughlin,
tify and define the target behavior objectively. In 2001; McGinnis, Friman, & Carlyon, 1999; Sul-
fact, Ayllon and Azrin (1968) dedicated 28 pages livan & O’Leary, 1990). For example, Higgins et
of their groundbreaking text to the topic. Am- al. (2001) developed a token economy for a third-
biguous or poorly described behavior promotes grade student with learning disabilities, in which
misunderstanding and confusion regarding pro- a teacher recorded a checkmark contingent on ap-
gram objectives and could fail to cue the consis- propriate behavior on a piece of paper taped to the
tent delivery of reinforcement for appropriate be- top left-hand corner of the student’s desk. Thus
havior. One study showed that classroom teachers the student could receive performance feedback
provided inconsistent feedback on students’ be- without significant interruption. Hupp, Reitman,
haviors and infrequent reinforcement for positive Northup, O’Callaghan, and LeBlanc (2002) modi-
classroom behaviors before they implemented a fied Pringles potato chip containers and attached
token economy. Both teacher behaviors improved them to clipboards. Researchers delivered tokens
immediately after token-economy implementation after the students demonstrated the ready position
(Kowalewicz & Coffee, 2014). during kickball games. A key feature of token de-
Moore, Tingstrom, Doggett, and Carlyon (2001) livery was that students could hear the token as
further examined problems with the specification it reached the bottom of the cylinder, but did not
of the target behavior and feedback. The research- need to divert attention from the game to receive
ers first analyzed an ineffective token economy at or retain the token.
376 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
Inclusion of response cost also may influence that “any stimulus to which the species responds
choice of conditioned reinforcers (Jowett Hirst, can be used as a reinforcer, provided only that the
Dozier, & Payne, 2016). This consideration is im- rate of the response governed by the stimulus is
portant, because many token programs include greater than that of some other response” (p. 227).
such a contingency (see Musser, Bray, Kehle, & The Premack principle became the basis for Al-
Jenson, 2001). Token programs with response lyon and Azrin’s (1968) rule about the probability
cost may include conditioned reinforcers that the of behavior, which suggests that we ought to “ob-
caregiver can remove easily, because conflict may serve what the individual does when the oppor-
ensue when the caregiver removes tokens. Erasing tunity exists. Those activities that are probable at
a point from a blackboard, flipping a card, or tak- a given time will serve as reinforcers” (Ayllon &
ing a chip from a container may be easier than re- Azrin, 1968, p. 60).
moving a token from a client’s hand or pocket. For Satiation, a reduction in reinforcement efficacy
example, Salend and Allen (1985) used 2.5-cen- after the repeated delivery of the reinforcer, can
timeter × 7.6-centimeter strips of paper taped to lead to the failure of any token-economy program.
the tops of second graders’ desks in a study evalu- One method of preventing reinforcer satiation
ating the difference between externally managed is to make conditioned reinforcers exchange-
and self-managed response cost programs. The able for a variety of backup reinforcers (Ayllon &
teacher removed the strips of paper contingent on Azrin, 1968; Bowman, Piazza, Fisher, Hagopian,
the performance of problem behavior during the & Kogan, 1997; Egel, 1980, 1981; Reese, Sher-
externally managed condition. man, & Sheldon, 1998). For example, Sran and
Most authors recommend pairing verbal praise Borrero (2010) assigned children to no-selection,
with conditioned reinforcement, although the em- single-selection, or varied-reinforcer-selection con-
pirical merits of verbal praise and its role in fading ditions in a token program. Results indicated that
conditioned reinforcement is unknown (Alberto children preferred conditions in which researchers
& Troutman, 2006; Drabman & Lahey, 1974; gave them the option to select from a variety of re-
Drabman & Tucker, 1974; Kirby & Shields, 1972; inforcers. The best outcomes were associated with
Paul & Lentz, 1977). Finally, behavior analysts maximum variety.
may base the selection of conditioned reinforce- Another method of preventing or delaying rein-
ment on safety or health issues, because individu- forcer satiation is to make the achievement of the
als may swallow small tokens, and circulating to- backup reinforcers effortful. A study by DeLeon
kens may facilitate the spread of illness. et al. (2011) revealed that stimuli associated with
high or moderate effort were more preferred and
retained their value longer than stimuli without
Identify Backup Reinforcers
an earning requirement. Nevertheless, reevalu-
Backup reinforcers are items, activities, or privileges ation of backup reinforcers may be necessary as
that the individual can exchange for conditioned often as once a week in certain contexts (Drab-
reinforcers (Kazdin, 2001). There are several con- man & Tucker, 1974).
siderations in their selection. First, backup rein-
forcers should have established reinforcing proper-
Establish a Schedule of Reinforcement
ties. One common method of identifying backup
and Exchange Rate
reinforcers is to ask the target individual about
preferred consequences. However, some have ques- There are two methods of manipulating the de-
tioned the validity of verbal assessment methods livery of backup reinforcers to maximize a token
(Northup, 2000). Furthermore, verbal assessments economy’s effectiveness. The first involves manip-
are not always feasible for the target population. ulation of the reinforcement schedule to reduce
For instance, stimulus preference assessments that the effects of reinforcer satiation. Specifically, a
measure participant approaches to stimuli present- behavior analyst generally uses continuous rein-
ed individually (Pace, Ivancic, Edwards, Iwata, & forcement, such as a fixed-ratio 1 schedule, during
Page, 1985) or in pairs (Fisher et al., 1992) may the initiation of a token economy to establish a
be more informative than verbal methods for indi- high rate of behavior. Next, the behavior analyst
viduals with intellectual developmental disorder. should introduce intermittent reinforcement based
Behavior analysts may also use behavioral obser- on either ratio or variable-time schedules after be-
vation to identify backup reinforcers (Ayllon & havior stabilizes on the continuous-reinforcement
Azrin, 1968). Specifically, Premack (1959) noted schedule (see Kazdin, 2001). We could find little
Token Economies 377
empirical support, however, for a body of research economy. Indeed, the behavior of children with
that clearly demonstrates the utility of fading lower levels of impulsivity improved even when
reinforcement from continuous to intermittent exchange delays were relatively long.
schedules in token economies. Another approach In an extension of earlier work, Hupp et al.
is to manipulate the amount of the conditioned (2002) examined the separate and combined ef-
reinforcer needed to obtain the backup reinforcer fects of medication and a token economy on the
(i.e., the exchange rate). A behavior analyst usu- symptoms of young children with attention-deficit/
ally sets the rate of reinforcement in the token hyperactivity disorder (ADHD) during a kickball
economy by measuring the natural rate of the ap- game. More importantly, the study also compared
propriate behavior during baseline, which ensures the effects of promised delayed reward to actual
that the individual will contact the reinforcement delayed reward aided by immediate token deliv-
contingency (Ayllon & Azrin, 1968). Observ- ery. Researchers supplied tokens contingent on
ing the baseline rate allows the behavior analyst “sportsman-like” behavior. Results indicated that
to gauge the number of tokens a client will earn delay of promised rewards did not increase sports-
during a given period. For example, Salend and man-like behavior during the kickball games,
Allen (1985) reported that they gave participants although delayed rewards delivered via tokens
“a set amount of tokens that represented the estab- increased sportsman-like behavior for all five chil-
lished number of problem behaviors that the sub- dren observed. Interestingly, stimulant medication
ject could engage in before losing reinforcement” had little to no positive effect on sportsman-like
(p. 61). Thus they based the number of tokens behavior in this sample of young children.
available to each participant on baseline levels of More recently, Pelham et al. (2014) assessed the
performance. The behavior analyst then assigns effects of medication (i.e., methylphenidate) and
the price of the backup reinforcers and creates a behavioral treatment in various doses and combi-
menu of rewards with a range of backup reinforc- nations. Researchers assigned children diagnosed
ers—some inexpensive (easy to earn), others more with ADHD in a summer program to either high-
costly (difficult to earn), and still others priced at intensity, moderate-intensity, or low-intensity be-
the intermediate level. havioral treatment with a token economy, or no
Establishing the value of conditioned reinforcers behavioral treatment and high-dose, moderate-
may be challenging especially for younger clients dose, or low-dose medication or placebo. Results
or clients with developmental delays. Research- indicated that the token economy and medication
ers have developed several procedures to address produced substantial improvements in child com-
this problem. Having clients observe others during pliance. More importantly, a combination of low
token exchange can sometimes establish tokens as doses of the two modalities had large beneficial
conditioned reinforcers (Ayllon & Azrin, 1968). effects on child behavior. In addition, the highest
The noncontingent delivery of tokens followed dose of medication produced only minimal added
immediately by an opportunity to exchange the benefits when combined with high-intensity be-
tokens for backup reinforcers may be effective if havioral treatment.
difficulties arise (Kazdin, 2001).
Delays between cash-out or redemption peri-
Keep Records
ods, delivery of conditioned reinforcers, the per-
formance of the behavior, or some combination of Creating a daily or weekly chart may have many
these variables may reduce the effectiveness of a benefits, including provision of a visual record of
putative reinforcer for some clients. In these cases, client improvement that may enhance compli-
reducing the delay between task performance and ance with therapy-related tasks (Reitman & Drab-
token delivery may be sufficient to improve perfor- man, 1996). Indeed, research suggests that persons
mance (Field, Nash, Handwerk, & Friman, 2004). implementing token economies and other types
Field et al. (2004) showed that reducing the num- of behavior management programs may not rec-
ber of tokens needed to earn privileges and dou- ognize the improvements of their clients without
bling exchange times by half resulted in a decrease such aids. For example, Reitman, Murphy, Hupp,
in intense behavioral episodes and an increase in and O’Callaghan (2004) found that a classroom
points earned compared to baseline. By contrast, teacher’s ratings of child behavior generally re-
Reed and Martens (2011) found that extended de- mained unchanged, although a token program sig-
lays between the disbursement of tokens and cash- nificantly reduced classroom behavior problems.
ing out did not reduce the effectiveness of a token By contrast, the use of periodic reviews of graphs
378 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
or similar behavioral data in the presence of the Over the past 50 years, most studies have sug-
clients may serve to facilitate more robust changes gested equivalence between programs that do and
in behavior, or perhaps more consistent changes in do not use response cost. Kaufman and O’Leary
the perceptions of those who interact with persons (1972) examined the differential effects of reward
in behavioral interventions. During the chart re- and response cost on academic and social behav-
view, the professionals or paraprofessionals (in the ior. Researchers randomly assigned adolescents
case of parents, supervisors, or others) can provide in a hospital to a reward condition in which the
the clients with praise and feedback concerning adolescents began each session with no tokens and
progress, compliance with recommendations, or could earn tokens throughout the period, or to a
both. Following such consultation, therapists and response cost condition in which the adolescents
clients, including persons like classroom teachers began the period with 10 tokens, which research-
and administrators, may decide jointly to alter the ers removed after rule infractions. The adolescents
treatment plan based on performance feedback could exchange tokens for backup reinforcers from
(Hawkins & Mathews, 1999). the school store. Interestingly, the two conditions
were equally effective in reducing problem behav-
ior. A more recent study by Donaldson, DeLeon,
Decide If Response Cost Is Warranted
Kahng, and Fisher (2014) evaluated the effective-
Response cost or behavior penalty (Clark, 1996) is a ness of an earn condition (i.e., token economy),
negative-punishment procedure in which the im- a loss condition (i.e., response cost), and a choice
plementer removes conditioned token reinforcers condition in which students could select one of
in a response-contingent fashion (Azrin & Holz, the conditions. In the earn condition, students
1966). Clinicians and researchers have long used began with zero tokens and could earn tokens for
response cost in token economies to reduce the remaining on task and for engaging in nondisrup-
frequency of problem behavior (Eluri, Andrade, tive behavior. During the loss condition, students
Trevino, & Mahmoud, 2016; Kazdin, 1971, 1972; began with 10 tokens and lost a token for engaging
Miltenberger, 2001; Reynolds & Kelley, 1997; Witt in off-task or disruptive behavior. As in earlier re-
& Elliot, 1982). Response cost does not restrict ac- search, both the earn and loss conditions reduced
cess to reinforcement directly, unlike many other the students’ disruptive behaviors to near-zero lev-
reductive techniques (e.g., time out; Reynolds & els. Interestingly, students seemed to prefer the loss
Kelley, 1997). Indeed, participants in a token- condition, and the authors noted significant ad-
economy program may continue to earn condi- vantages for the loss condition relative to the time
tioned reinforcers for the performance of adaptive required to maintain the token program.
behaviors, despite losing tokens when they per-
form problem behaviors.
Engage Participants and Train Staff
An important step in implementing a token-
economy program is to determine whether the Implementation of token programs typically be-
use of response cost is necessary to effect behavior gins with a formal explanation of the rules of the
change. In deciding whether to use response cost, token economy to clients. A behavior analyst
a behavior analyst should ask whether a client’s should consult clients during program develop-
problem behavior is inhibiting the performance of ment well in advance of implementation. A for-
adaptive behavior. If the answer is yes, response mal meeting ensures that clients understand how
cost is likely to be beneficial in reducing the per- to earn tokens, when and where to redeem tokens,
formance of the problem behavior. However, Milt- what tokens can be exchanged for, and how much
enberger (2001) warns that token programs using backup reinforcers cost. Finally, the behavior ana-
a behavior penalty may fail if the behavior ana- lyst must inform clients about which behaviors re-
lyst does not establish the token program before sult in token gain and loss.
the implementation of response cost. One further Although behavior analysts regard consistent
consideration is that the penalty must result in re- token delivery as a key component for behav-
moval of tokens and access to backup reinforcers ior change, researchers have yet to establish the
without resulting in the complete loss of tokens. precise level of consistency needed to achieve ef-
Complete loss of earned tokens and the loss of fectiveness. On the other hand, research on treat-
opportunities to exchange tokens for backup re- ment fidelity has revealed that some interventions
inforcers may cause the program to fail (Kazdin, implemented with less than 100% fidelity (Noell,
1972; Miltenberger, 2001). Gresham, & Gansle, 2002; Northup, Fisher,
Token Economies 379
Kahng, & Harrell, 1997) can retain effectiveness. time as patients advanced through the levels of the
Whatever the exact level of consistency needed, program, and patients at the highest level, Level 4,
improvements in integrity may require extensive could buy themselves out of the program. Specifi-
staff training (e.g., hospital personnel, teach- cally, patients on Level 1, the entrance level, en-
ers, parents). Thus ongoing staff supervision and gaged in 6 hours of scheduled classes and activities
an additional level of contingency management each day and earned tokens for adaptive behavior
may be necessary to maximize the likelihood of (e.g., self-care, bed making, and appropriate meal-
success. This may entail praise and other incen- time behavior). Patients who progressed to Level
tives for staff members who consistently deliver 2 attended 3 hours of classes and activities, and
conditioned and backup reinforcers; constructive engaged in individual assignments for the other 3
feedback for errors; and periodic staff retraining to hours. Patients at Levels 3 and 4 engaged in 4–6
eliminate drift (Miltenberger, 2001). hours of individual assignments and did not at-
Staff training looms as a unique opportunity tend scheduled classes and activities. Patients had
to improve the generality of treatment effects for the opportunity to earn more tokens, and the cri-
behavioral interventions, including response cost. terion for earning tokens shifted upward at each
However, researchers have devoted only limited level. Patients at Level 4 could purchase a “credit
attention to this important topic. Suffice it to say card” that gave them unlimited access to reinforc-
that token programs are likely to vary greatly in ers if they continued to meet Level 4 requirements.
the fidelity of implementation and effectiveness The researchers also included a fading component
(see Noell et al., 2005). to enhance the success of the program. The pro-
gram included a delay-to-reinforcement “payday”
in which patients received a lump sum of tokens as
Phase Out the Program
they graduated from Level 1 to Level 2. This com-
Teachers periodically ask us how we could rec- ponent taught the patients to plan as they would
ommend to them that they “bribe” students to outside the hospital setting. Furthermore, Paul and
behave. They also ask us, “What message would Lentz supplied staff and patients with rules about
that teach them when they enter the real world? the implementation and use of the token program.
How could we, as professionals, recommend that Several effective techniques exist for reducing
we teach students to work only when they know reliance on extrinsic or arbitrary contingencies and
we will reward them?” Notwithstanding the ob- promoting the generalization and maintenance of
servation that few if any adults work without behavior change (O’Callaghan, Reitman, North-
compensation, we believe that we are unlikely to up, Hupp, & Murphy, 2003; Stokes & Baer, 1977).
persuade persons strongly opposed to extrinsic re- For example, a behavior analyst may employ indis-
inforcement to accept it (see Reitman, 1988, for criminable contingencies to facilitate generalization
an extended discussion of this issue). Instead, we of a token program’s contingencies. The behavior
typically suggest that the best token programs are analyst must “make unclear the limits of train-
those that we phase out deliberately. ing contingencies; in particular, conceal, when
Paul and Lentz (1977) presented one of the best possible, the point at which those contingencies
examples of such phasing out. In their now-classic stop operating, if possible by delayed reinforce-
study, the researchers worked with patients insti- ment” (Stokes & Baer, 1977, p. 287). Intermittent
tutionalized for psychosis from four state hospi- reinforcement (e.g., variable-ratio or variable-time
tals in central Illinois. The researchers assigned schedules) also appears to facilitate generalization
patients to one of three treatment groups with (Ferster & Skinner, 1957). Increasing the cost of
28 patients in each group: milieu therapy, social- the backup reinforcers or delaying the opportunity
learning therapy with a token program, and a to redeem tokens are other approaches to fading
control group. The researchers followed patients token programs. We encourage readers to consult
for over 10 years. They offered a unique option Stokes and Baer (1977) for a complete list of tech-
to patients in the social-learning group to reduce niques for promoting generalization.
dependence on the token program. In a token Finally, Sullivan and O’Leary (1990) used a re-
program with a conventional levels system, indi- versal design to study the efficacy of fading proce-
viduals earn opportunities to participate in more dures for both token and response cost programs.
reinforcing environments but remain participants Results suggested that the two types of programs
in the token economy. In a unique twist, Paul and were equally effective for increasing duration of
Lentz decreased the amount of required program on-task behavior. Only half the children in the
380 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
token program showed high levels of mainte- at a target permitted access to several reinforc-
nance, however. ing activities, and the researchers made tangibles
contingent on the behavior of either randomly se-
lected children; the group; or one of the three tar-
ADDITIONAL CONSIDERATIONS get or “star” children, the individual contingency.
During baseline, the teacher received training
Treatment Acceptability
and explained the token program to her class.
Theodore, Bray, and Kehle (2004) found that al- The token program also contained a response
though teacher ratings did not reflect changes in cost in which the teacher could move a marker
behavior that were apparent through direct obser- down (which was the negative consequence),
vation, teachers reported a favorable experience contingent on disruptive behavior. Although ini-
with the token program. Reitman et al. (2004) re- tial rates of disruptive behavior were somewhat
ported similar results. McGoey and DuPaul (2000) low (M = 15% of intervals for each of the three
suggested that an all-positive-reinforcement token participants), the results indicated that the token
program and a response cost program were about economy reduced rates of classroom rule viola-
equally effective in reducing disruptive classroom tions for both individual and group contingencies.
behavior. Interestingly, teacher acceptability rat- A study by Kowalewicz and Coffee (2014) used a
ings favored the response cost procedure. When changing-criterion design to show that a group
questioned, some teachers commented that this interdependent contingency based on a mystery
procedure was far less time-consuming than motivator as part of a Tier 1 intervention in an
“catching the children being good.” We discuss elementary school improved behavior. Important-
the relative merits of response-cost-only programs ly, effects were maintained at follow-up. Seven of
at greater length toward the end of this chapter. the eight classroom teachers indicated that they
would be likely to use the group contingency in-
tervention again, as it required minimal training,
Individual versus Group Contingencies
and the data collection and time requirements
Drabman, Spitalnik, and Spitalnik (1974) con- were practical.
ducted one of the earliest studies exploring the
relative efficacy of group and individual contin-
Follow‑Up, Maintenance, and Generalization
gencies. They included four experimental condi-
tions: individual reinforcement; group reinforce- LePage et al. (2003) used a token economy on an
ment determined by the most disruptive child in acute inpatient psychiatric unit to reduce assaults
the group; group reinforcement determined by on staff and patients. Patients voluntarily enrolled
the least disruptive child in the group; and group in the program, and they earned stamps for per-
reinforcement determined by a randomly chosen forming behavior necessary for successful transi-
child in the group. Although the conditions were tion to the community. Examples of such behavior
equally effective at reducing disruptive behavior, included taking medication, timely appointment
the teacher preferred group reinforcement deter- keeping, and showering. Researchers implemented
mined by a randomly selected child in the group, response cost for major violations of safety rules
because it was the least time-consuming and easi- requiring a police presence on the unit, such as
est to use. Children, by contrast, ranked group hitting another person or destroying property. Pa-
reinforcement determined by the least disruptive tients redeemed tokens for privileges such as off-
child in the group as most preferred; however, ground passes, movies, stereo rental, or items from
the teacher disliked that procedure. Neither the the token store (e.g., snacks, drinks, phone cards).
teacher nor the students preferred the individual- The program was effective at a 2-year follow-up.
reinforcement condition. Children and teachers Patient-to-patient and employee injuries decreased
preferred an interdependent group contingency by 48% and 21%, respectively. Finally, a study by
with response cost over one with reward in a study O’Callaghan et al. (2003) used training modifica-
by Lee, Penrod, and Price (2017). tions based on the work of Stokes and Baer (1977)
Reitman et al. (2004) provided additional data to facilitate generalization of social skills to games
on the effectiveness of individual versus group without token programs.
contingencies. Children earned opportunities to Drabman and Tucker’s (1974) critique and
play the Rewards Target Game (see Anhalt et al., recommendations concerning failures of token
1998), in which chances to throw a Velcro ball programs provide an unusually comprehensive
Token Economies 381
account of shortcomings and potential fixes for status may permit. Another factor to consider is
frequently encountered implementation problems the right to effective treatment, which may dictate
in school settings. They identified three major the short-term use of behavior change strategies
classes of failures: program-based, teacher-based, to facilitate the long-term goal of greater freedom
and setting-based. Program-based failures included and independence (Reitman, 1988).
poor definition or poor monitoring of target be- The implementation of a token program may
havior, or both. Teacher- or personnel-based prob- present considerable training and resource chal-
lems included failure to reward approximations of lenges. For example, persons charged with the ad-
the behavior (shaping), altering programs without ministration of a token economy must comply with
consulting the data, and ignoring disruptive be- the terms of the agreement or contract. Teachers
havior that produced social reinforcement from or staff providing direct care must guard against
classmates. Setting-based failures included prob- reacting to noncompliance or aggression in ways
lems with discriminating changes in behavior for that are inconsistent with the rules of the token
many children simultaneously. Most importantly, program. Response cost requires the dispassionate
Drabman and Tucker offered recommendations removal of points or tokens from a potentially vola-
for choosing and distributing tokens and manag- tile person. Emotional reactions from the program
ing day-to-day changes in a classroom-based token administrator may undermine the token program
program. by supplying social or attention-based reinforcers
Although many years have passed since the for problem behavior. Thus administrators may
introduction of the token economy, questions re- become unable or unwilling to continue a token
main about its efficacy. Specifically, Maggin, Cha- program, because they may require caregivers and
fouleas, Goddard, and Johnson (2011) conducted staff to endure significant antisocial behavior and
a critical review, using criteria developed through commit material resources to support the token
the What Works Clearinghouse, to evaluate the program’s contingencies. Behavior analysts should
experimental rigor of existing research on token seek support from relatives, counselors, or other
programs. Maggin et al. concluded that token staff for a caregiver, staff person, or teacher who
economies in academic settings did not meet ev- appears overwhelmed.
idence-based standards, although they concluded Although an additional challenge is somewhat
that the data were supportive of its use. Specifi- beyond the scope of this chapter, some researchers
cally, they recommended the development of sys- have argued that a failure to account for motivating
tematic procedures for evaluating implementation operations (events that influence the effectiveness
fidelity and guidelines for staff training. of conditioned stimuli as reinforcers) may compro-
mise the effectiveness of contingency management
procedures (Laraway, Snycerski, Michael, & Pol-
CHALLENGES: PROBLEMS OF APPLICATION ing, 2003). For example, receiving $20 for cutting
a neighbor’s lawn would reduce the likelihood of a
There are many limitations of token programs, child’s completing chores at home if the child was
and most are like those of behavior modification participating in a home-based program based pri-
procedures more generally (see Reitman, Hupp, marily on monetary reinforcers. Earlier researchers
& O’Callaghan, 2005). One of the more durable have noted that failure to consider deprivation,
but less devastating is the no-cure criticism (Ken- satiation, and competing sources of reinforcement
dall, 1989)—the notion that treatment effects has doomed many a token program (Allyon &
typically do not persist after treatment withdrawal. Azrin, 1968; Drabman & Tucker, 1974).
Of course, we can apply this criticism with equal Most token programs require well-developed
measure to psychotropic medications regarded as repertoires of organizational, communication,
highly efficacious, such as stimulant medication. and negotiation skills. Thus persons charged with
Another concern relates to the ethical issues administering token programs must either teach
raised by the imposition of contingencies on vul- the prerequisite skills as needed or modify the
nerable groups, such as children and adults with token program to reduce demands on persons im-
intellectual developmental disorder. Fortunately, plementing the program. The data suggest, how-
most behavior analysts seek to minimize this risk ever, that token programs can be beneficial even
by fostering a collaborative process between a in the most difficult circumstances (see Ayllon &
treater and client in which both parties contribute Azrin, 1968; Drabman & Tucker, 1974; Field et
equally, or as equally as the client’s disability or al., 2004).
382 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
and consultation—subjects of great relevance for uli and the point(s) at which satiation for those
bringing effective treatments to the public with stimuli occurs. Many stimulus preference methods
greater regularity. Reitman and colleagues con- are available, but consensus about which methods
ducted a series of studies to evaluate whether they are best suited to which populations is lacking.
could use token programs to increase appropriate Similarly, some stimuli that function as reinforce-
social behavior in athletic settings (see Reitman et ment may be impractical (e.g., a hospital setting
al., 2005). For example, Hupp and Reitman (1999) in which patients may earn iPods as backup rein-
implemented a token economy during a basketball forcers) or require considerable staff resources to
camp for children diagnosed with ADHD. Tokens deliver consistently (e.g., rewarding a child with
consisted of “B-Ball Bucks” that the researchers 15 minutes of a staff person’s time during each
delivered contingent on sportsman-like behavior, school day). Even relatively straightforward issues
such as cheering for peers, and children could of reinforcer satiation (e.g., frequency of reinforcer
exchange them for a variety of backup reinforc- variation, satiation monitoring in applied settings)
ers. The token program improved sportsman-like appear under researched, thus leaving the behav-
behavior, whereas discussions about the value of ior analyst with little practical research-informed
good sportsmanship did not. Perhaps more im- guidance. Fortunately, recent work by applied be-
portantly, reinforcing appropriate behavior ap- havior analysts has begun to take a fresh look at
peared to decrease un-sportsman-like behavior, token programs, with an eye toward blending ele-
even though the researchers did not target un- ments of basic and applied research. These stud-
sportsman-like behavior (e.g., verbal or physical ies could lead to greater clarity concerning how
aggression) directly. Clearly, further studies are to maximize the clinical yield and practicality of
necessary to establish whether social skills learned token programs in applied settings (cf. DeLeon et
and mastered in the context of a sports setting can al., 2011; Donaldson et al., 2014; Reed & Martens,
be generalized to and maintained in nonathletic 2011; Sran & Borrero, 2010).
settings. Consult Conner-Smith and Weisz (2003), On a final note, Paul and Lentz (1977) conduct-
Noell et al. (2005), or both for excellent introduc- ed a groundbreaking longitudinal study of the im-
tions to these topics. pact of a level-based token program. Over 40 years
later, it remains one of the few studies to have re-
ported long-term, socially meaningful impact from
BASIC RESEARCH AND CLINICAL FOLKLORE a token-based instructional program. Moreover,
the researchers’ emphasis on fading to promote
Although this chapter has reviewed a substantial maintenance and generalization of the program
number of studies evaluating the efficacy of token has remained surprisingly novel. A cursory ex-
and contingency management procedures and amination of the many classrooms that employ
their application to a variety of human behaviors, token programs is likely to reveal little attention
a surprising amount of speculation and clinical to efforts to fade such programs over the course of
folklore seems to guide much of the application of the school year, or to do so systematically across
this popular procedure. For example, researchers the school years. Thus, although this chapter has
often extrapolate from basic research that condi- documented an extensive literature concerned
tioned reinforcers may become powerful reinforc- with the token economy, its many variations, and
ers in their own right (e.g., Jones et al., 1992). Yet the many settings in which researchers and clini-
few, if any, rigorous experimental studies support cians use it suggest that there is an ongoing need
this assertion. Additionally, little research on ex- for evaluation of this well-known intervention.
actly how token economies exert their influence
exists, even though researchers hypothesize that
token programs may heighten awareness of the REFERENCES
target individual’s appropriate behavior, possibly
Alberto, P. A., & Troutman, A. C. (2006). Applied be-
conferring conditioned reinforcing properties on
havior analysis for teachers (7th ed.). Upper Saddle
the behavior manager (see Risley, 2005). The fact
River, NJ: Pearson/Merrill Prentice Hall.
that many contingency management procedures Anhalt, K., McNeil, C. B., & Bahl, A. B. (1998). The
also rely extensively on verbal stimuli complicates ADHD classroom kit: A whole-classroom approach
this issue (see Hayes, 1989). for managing disruptive behavior. Psychology in the
Additional bridge studies are needed to inform Schools, 35, 67–79.
issues such as identification of reinforcing stim- Ayllon, T., & Azrin, N. H. (1965). The measurement
384 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
and reinforcement of behavior of psychotics. Journal Drabman, R. S., & Tucker, R. D. (1974). Why classroom
of the Experimental Analysis of Behavior, 8, 357–383. token economies fail. Journal of School Psychology, 12,
Ayllon, T., & Azrin, N. H. (1968). The token economy: A 178–188.
motivational system for therapy and rehabilitation. New Egel, A. L. (1980). The effects of constant vs. varied
York: Appleton-Century-Crofts. reinforcer presentation on responding by autistic
Azrin, N. H., & Holz, W. C. (1966). Punishment. In W. children. Journal of Experimental Child Psychology,
K. Honig (Ed.), Operant behavior: Areas of research 30, 455–463.
and application (pp. 790–826). New York: Appleton- Egel, A. L. (1981). Reinforcer variation: Implications for
Century-Crofts. motivating developmentally disabled children. Jour-
Benishek, L. A., Dugosh, K. L., Kirby, K. C., Matejkows- nal of Applied Behavior Analysis, 14, 345–350.
ki, J., Clements, N. T., Seymour, B. L., et al. (2014). Eluri, Z., Andrade, I., Trevino, N., & Mahmoud, E.
Prize-based contingency management for the treat- (2016). Assessment and treatment of problem behav-
ment of substance abusers: A meta-analysis. Addic- ior maintained by mand compliance. Journal of Ap-
tion, 109, 1426–1436. plied Behavior Analysis, 49, 383–387.
Bowman, L. G., Piazza, C. C., Fisher, W. W., Hagopian, Ferster, C. B., & Skinner, B. F. (1957). Schedules of rein-
L. P., & Kogan, J. S. (1997). Assessment of preference forcement. New York: Appleton-Century-Crofts.
for varied versus constant reinforcers. Journal of Ap- Field, C. E., Nash, H. M., Handwerk, M. L., & Friman, P.
plied Behavior Analysis, 30, 451–458. C. (2004). A modification of the token economy for
Budney, A. J., Higgins, S. T., Delany, D. D., Kent, L., nonresponsive youth in family-style residential care.
& Bickel, W. K. (1991). Contingent reinforcement Behavior Modification, 28, 438–457.
of abstinence with individuals abusing cocaine and Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L.
marijuana. Journal of Applied Behavior Analysis, 24, P., Owens, J. C., & Slevin, I. (1992). A comparison of
657–665. two approaches for identifying reinforcers for persons
Clark, L. (1996). SOS!: Help for parents. Bowling Green, with severe and profound disabilities. Journal of Ap-
KY: Parents Press. plied Behavior Analysis, 25, 491–498.
Clark, M., Lachowicz, J., & Wolf, M. (1968). A pilot Fiske, K. E., Isenhower, R. W., Bamon, M. J., & Lauder-
basic education program for school dropouts incor- dale-Littin, S. (2020). An analysis of token reinforce-
porating a token reinforcement system. Behaviour ment using a multiple-schedule assessment. Journal of
Research and Therapy, 6, 183–188. Applied Behavior Analysis, 53, 563–571.
Conner-Smith, J. K., & Weisz, J. R. (2003). Applying Fox, D. K., Hopkins, B. L., & Anger, W. K. (1987). The
treatment outcome research in clinical practice: long-term effects of a token economy on safety per-
Techniques for adapting interventions to the real formance in open-pit mining. Journal of Applied Be-
world. Child and Adolescent Mental Health, 8, 3–10. havior Analysis, 20, 215–224.
Conyers, C., Miltenberger, R., Maki, A., Barenz, R., Ju- Gendreau, P., Listwan, S. J., Kuhns, J. B., & Exum, M. L.
rgens, M., Sailer, A., et al. (2004). A comparison of (2014). Making prisoners accountable: Are contin-
response cost and differential reinforcement of other gency management programs the answer? Criminal
behavior to reduce disruptive behavior in a preschool Justice and Behavior, 41(9), 1079–1102.
classroom. Journal of Applied Behavior Analysis, 37, Gottlieb, H. (2000). Medication nonadherence: Find-
411–415. ing solutions to a costly medical problem. Drug Ben-
DeLeon, I. G., Gregory, M. K., Frank-Crawford, M. A., efit Trends, 12, 57–62.
Allman, M. J., Wilke, A. E., Carreau-Webster, A. B., Hackenberg, T. D. (2018). Token reinforcement: Trans-
et al. (2011). Examination of the influence of con- lational research and application. Journal of Applied
tingency on changes in reinforcer value. Journal of Behavior Analysis, 51, 393–435.
Applied Behavior Analysis, 44(3), 543–558. Hawkins, R. P., & Mathews, J. R. (1999). Frequent mon-
Dickerson, F. B., Tenhula, W. N., & Green-Paden, L. itoring of clinical outcomes: Research and account-
D. (2005). The token economy for schizophrenia: ability in clinical practice. Education and Treatment
Review of the literature and recommendations for of Children, 22, 117–135.
future research. Schizophrenia Research, 75, 405–416. Hayes, S. C. (Ed.). (1989). Cognition, contingencies and
Donaldson, J. M., DeLeon, I. G., Kahng, S., & Fisher, A. rule-governed behavior. New York: Plenum Press.
B. (2014). Effects of and preference for conditions of Higgins, J. W., Williams, R. L., & McLaughlin, T. F.
token earn versus token loss. Journal of Applied Be- (2001). The effects of a token economy employing
havior Analysis, 47(3), 537–548. instructional consequences for a third-grade student
Drabman, R. S., & Lahey, B. B. (1974). Feedback in with learning disabilities: A data-based case study.
the classroom behavior modification: Effects on the Education and Treatment of Children, 24, 99–106.
target and her classmates. Journal of Applied Behavior Hupp, S. D. A., & Reitman, D. (1999). Improving sports
Analysis, 7, 591–598. skills and sportsmanship in children diagnosed with
Drabman, R. S., Spitalnik, R., & Spitalnik, K. (1974). attention deficit/hyperactivity disorder. Child and
Sociometric and disruptive behavior as a function of Family Behavior Therapy, 21, 35–51.
four types of token reinforcement programs. Journal Hupp, S. D. A., Reitman, D., Northup, J., O’Callaghan,
of Applied Behavior Analysis, 7, 93–101. P., & LeBlanc, M. (2002). The effects of delayed re-
Token Economies 385
wards, tokens, and stimulant medication on sports- horn, L., Murphy, J., et al. (2003). Reducing assaults
manlike behavior with ADHD-diagnosed children. on an acute psychiatric unit using a token economy:
Behavior Modification, 26, 148–162. A 2-year follow-up. Behavioral Interventions, 18, 179–
Ivy, J. W., Meindl, J. N., Overley, E., & Robson, K. 190.
M. (2017). Token economy: A systematic review of Ludwig, T. D., & Geller, E. S. (1991). Improving the
procedural descriptions. Behavior Modification, 41, driving practices of pizza deliverers: Response gener-
708–737. alization and moderating effects of driving history.
Jones, R. N., Downing, R. H., Latkowski, M. E., Ferre, Journal of Applied Behavior Analysis, 24, 31–44.
R. C., & McMahon, W. M. (1992). Levels systems Maggin, D. M., Chafouleas, S. M., Goddard, K. M., &
as shaping and fading procedures: Use in a child in- Johnson, A. H. (2011). A systematic evaluation of
patient psychiatry setting. Child and Family Behavior token economies as a classroom management tool for
Therapy, 14, 15–37. students with challenging behavior. Journal of School
Jowett Hirst, E. S., Dozier, C. L., & Payne, S. W. (2016). Psychology, 49(5), 529–553.
Efficacy and preference for reinforcement and re- McGinnis, J. C., Friman, P. C., & Carlyon, W. D. (1999).
sponse cost in token economies. Journal of Applied The effect of token rewards on “intrinsic” motivation
Behavior Analysis, 49, 329–345. for doing math. Journal of Applied Behavior Analysis,
Kahng, S. W., Boscoe, J. H., & Byrne, S. (2003). The 32, 375–379.
use of an escape contingency and a token economy to McGoey, K. E., & DuPaul, G. J. (2000). Token rein-
increase food acceptance. Journal of Applied Behavior forcement and response cost procedures: Reducing
Analysis, 36, 349–353. the disruptive behavior of preschool children with
Kaufman, K. F., & O’Leary, K. D. (1972). Reward, cost, attention-deficit/hyperactivity disorder. School Psy-
and self-evaluation procedures for disruptive adoles- chology Quarterly, 15, 330–343.
cents in a psychiatric hospital school. Journal of Ap- McNamara, J. R. (1971). Teacher and students as sources
plied Behavior Analysis, 5, 293–309. for behavior modification in the classroom. Behavior
Kazdin, A. E. (1971). The effect of response cost in sup- Therapy, 2, 205–213.
pressing behavior in a pre-psychotic retardate. Jour- Mikami, A. Y., Griggs, M. S., Lerner, M. D., Emeh, C.
nal of Behavior Therapy and Experimental Psychiatry, C., Reuland, M. M., Jack, A., et al. (2013). A ran-
2, 137–140. domized trial of a classroom intervention to increase
Kazdin, A. E. (1972). Response cost: The removal of peers’ social inclusion of children with attention-def-
conditioned reinforcers for therapeutic change. Be- icit/hyperactivity disorder. Journal of Consulting and
havior Therapy, 3, 533–546. Clinical Psychology, 81(1), 100–112.
Kazdin, A. E. (2001). Behavior modification in applied set- Miltenberger, R. G. (2001). Behavior modification: Prin-
tings (6th ed.). Pacific Grove, CA: Brooks/Cole. ciples and procedures (2nd ed.). Belmont, CA: Wad-
Kazdin, A. E., & Bootzin, R. R. (1972). The token econ- sworth/Thomson Learning.
omy: An evaluative review. Journal of Applied Behav- Moore, J. W., Tingstrom, D. H., Doggett, R. A., & Car-
ior Analysis, 5, 343–372. lyon, W. D. (2001). Restructuring an existing token
Kendall, P. C. (1989). The generalization and mainte- economy in a psychiatric facility for children. Child
nance of behavior change: Comments, consider- and Family Behavior Therapy, 23, 53–59.
ations, and the “no-cure” criticism. Behavior Therapy, Musser, E. H., Bray, M. A., Kehle, T. J., & Jenson, W.
20, 357–364. R. (2001). Reducing disruptive behaviors in students
Kirby, F. D., & Shields, F. (1972). Modification of arith- with serious emotional disturbance. School Psychol-
metic response rate and attending behavior in a ogy Review, 30, 294–304.
seventh-grade student. Journal of Applied Behavior Noell, G. H., Gresham, F. M., & Gansle, K. A. (2002).
Analysis, 5, 79–84. Does treatment integrity matter?: A preliminary
Kowalewicz, E. A., & Coffee, G. (2014). Mystery mo- investigation of instructional implementation and
tivator: A Tier 1 classroom behavioral intervention. mathematics performance. Journal of Behavioral Edu-
School Psychology Quarterly, 29(2), 138–156. cation, 11, 51–67.
Laraway, S., Snycerski, S., Michael, J., & Poling, A. Noell, G. H., Witt, J. C., Slider, N. J., Connell, J. E.,
(2003). Motivating operations and terms to describe Gatti, S. L., Williams, K., et al. (2005). Treatment
them: Some further refinements. Journal of Applied implementation following behavioral consultation
Behavior Analysis, 36, 407–414. in schools: Comparison of three follow-up strategies.
Lee, K., Penrod, B., & Price, J. N. (2017). A compari- School Psychology Review, 34, 87–106.
son of cost and reward procedures with interdepen- Northup, J. (2000). Further evaluation of the accuracy
dent group contingencies. Behavior Modification, 41, of reinforcer surveys: A systematic replication. Jour-
21–44. nal of Applied Behavior Analysis, 33, 335–338.
Leon, Y., Borrero, J. C., & DeLeon, I. G. (2016). Para- Northup, J., Fisher, W. W., Kahng, S. W., & Harrell, R.
metric analysis of delayed primary and conditioned T. (1997). An assessment of the necessary strength of
reinforcers. Journal of Applied Behavior Analysis, 49, behavioral treatments for severe behavior problems.
639–655. Journal of Developmental and Physical Disabilities, 9,
LePage, J. P., Delben, K., Pollard, S., McGhee, M., Van- 1–16.
386 In t e r v e n t i o n s f o r D e c r e a s i n g P r o b l e m B e h av i o r
O’Callaghan, P. M., Reitman, D., Northup, J., Hupp, S. Reitman, D., Murphy, M. A., Hupp, S. D. A., &
D. A., & Murphy, M. A. (2003). Promoting social O’Callaghan, P. M. (2004). Behavior change and
skills generalization with ADHD-diagnosed children perceptions of change: Evaluating the effectiveness
in a sports setting. Behavior Therapy, 34, 313–330. of a token economy. Child and Family Behavior Ther-
O’Leary, K. D., & Becker, W. C. (1967). Behavior modi- apy, 26, 17–36.
fication of an adjustment class: A token reinforce- Reynolds, L. K., & Kelley, M. L. (1997). The efficacy of a
ment program. Exceptional Children, 9, 637–642. response cost-based treatment package for managing
O’Leary, K. D., Becker, W. C., Evans, M. B., & Saudar- aggressive behavior in preschoolers. Behavior Modifi-
gas, R. A. (1969). A token reinforcement program in cation, 21, 216–230.
a public school: A replication and systematic analy- Risley, T. (2005). Montrose M. Wolf (1935–2004). Jour-
sis. Journal of Applied Behavior Analysis, 2, 3–13. nal of Applied Behavior Analysis, 38, 279–287.
Pace, G. M., Ivancic, M. T., Edwards, G. L., Iwata, B. Russell, D., Ingvarsson, E. T., Haggar, J. L., & Jessel, J.
A., & Page, T. J. (1989). The use of behavioral assess- (2018). Using progressive ratio schedules to evaluate
ment to prescribe and evaluate treatments for severe- tokens as generalized conditioned reinforcers. Journal
ly handicapped children. Journal of Applied Behavior of Applied Behavior Analysis, 51, 40–52.
Analysis, 18, 173–178. Salend, S. J., & Allen, E. M. (1985). Comparative effects
Paul, G. L., & Lentz, R. J. (1977). Psychosocial treatment of externally managed and self-managed response-
of chronic mental patients: Milieu versus social-learning cost systems on inappropriate classroom behavior.
programs. Cambridge, MA: Harvard University Press. Journal of School Psychology, 23, 59–67.
Pelham, W. E., Burrows-MacLean, L., Gnagy, E. M., Fa- Sidman, M. (1989). Coercion and its fallout. Boston: Au-
biano, G. A., Coles, E. K., Wymbs, B. T., et al. (2014). thors Cooperative.
A dose-ranging study of behavioral and pharmaco- Skinner, B. F. (1953). Science and human behavior. New
logical treatment for children with ADHD. Journal of York: Macmillan.
Abnormal Child Psychology, 42(6), 1019–1031. Sran, S. K., & Borrero, J. C. (2010). Assessing the value
Premack, D. (1959). Toward empirical behavioral laws: of choice in a token system. Journal of Applied Behav-
1. Positive reinforcement. Psychological Review, 66, ior Analysis, 43(3), 553–557.
219–233. Staats, A. W., Staats, C. K., Schutz, R. E., & Wolf, M.
Reed, D. D., & Martens, B. K. (2011). Temporal dis- (1962). The conditioning of textual responses using
counting predicts student responsiveness to ex- “extrinsic” reinforcers. Journal of the Experimental
change delays in a classroom token system. Journal of Analysis of Behavior, 5, 33–40.
Applied Behavior Analysis, 44(1), 1–18. Stokes, T. F., & Baer, D. M. (1977). An implicit tech-
Reese, M. R., Sherman, J. A., & Sheldon, J. B. (1998). nology of generalization. Journal of Applied Behavior
Reducing disruptive behavior of a group home resi- Analysis, 10, 349–367.
dent with autism and mental retardation. Journal of Sullivan, M. A., & O’Leary, S. G. (1990). Maintenance
Autism and Developmental Disorders, 28, 159–165. following reward and cost token programs. Behavior
Reitman, D. (1998). Punished by misunderstanding: A Therapy, 21, 139–149.
critical evaluation of Kohn’s Punished by Rewards and Theodore, L. A., Bray, M. A., & Kehle, T. J. (2004). A
its implications for behavioral interventions with comparative study of group contingencies and ran-
children. Behavior Analyst, 21, 143–157. domized reinforcers to reduce disruptive classroom
Reitman, D., & Drabman, R. S. (1996). Read my fin- behavior. School Psychology Quarterly, 19, 253–271.
gertips: A procedure for enhancing the effectiveness Truchlicka, M., McLaughlin, T. F., & Swain, J. C.
of time-out with argumentative children. Child and (1998). Effects of token reinforcement and response
Family Behavior Therapy, 18, 35–40. cost on the accuracy of spelling performance of mid-
Reitman, D., Hupp, S. D. A., & O’Callaghan, P. M. dle school special education students with behavior
(2005). Sport skill training. In A. M. Gross & R. S. disorders. Behavioral Interventions, 13, 1–10.
Drabman (Eds.), Encyclopedia of behavior modifica- Witt, J. C., & Elliot, S. N. (1982). The response cost
tion and cognitive behavior therapy: Vol. 2. Clinical lottery: A time efficient and effective classroom
child applications (pp. 1050–1054). Thousand Oaks, intervention. Journal of School Psychology, 20, 155–
CA: SAGE. 161.
PAR T VII
SUBSPECIALTIES IN
APPLIED BEHAVIOR ANALYSIS
W e have reorganized the order of chapters in Part VII in this edition of the
Handbook of Applied Behavior Analysis. In addition, there are three new
chapters on the subspecialties of pediatric feeding disorders, teacher consultation,
and telehealth. Chapter 23 by Kodak, Grow, and Bergman opens Part VII with
a review of behavioral treatments of autism spectrum disorder (ASD). Kodak et
al. discuss the changes in the criteria for diagnosing ASD and behaviors that
are associated with a diagnosis of ASD. The chapter also provides a summary of
the most recent research on the effects of applied-behavior-analytic treatments,
which readers are sure to find instructive. Chapter 24 by Friman provides the
reader with an insightful summary of the contributions applied behavior analysts
can make to the field of pediatric medicine. He discusses three routine behavior
problems that pediatricians commonly confront—bedtime problems, enuresis,
and encopresis—and describes relevant behavior-analytic interventions to illus-
trate his points. Chapter 25 by Piazza and Kirkwood presents a behavior-analytic
conceptualization of pediatric feeding disorders. The authors also address many
practical considerations associated with assessment and treatment of feeding
disorders in children.
Martens, Daly, Begeny, and Sullivan have organized their Chapter 26
discussion of behavioral approaches to education around a model called the
learning/instructional hierarchy. This model promotes the dynamic development of
proficient skill performance. The authors have provided a historical context for
behavioral approaches to education, in addition to describing more recent stud-
ies evaluating strategies for effective teaching. Chapter 27 by DiGennaro Reed,
Hagermoser Sanetti, and Codding on teacher consultation is a new and welcome
addition to the Handbook of Applied Behavior Analysis. The authors introduce
their topic within the framework of federal legislation that has altered the
landscape of public education. The authors define behavioral consultation, discuss
different models of consultation, and provide guidance on procedures to maximize
387
388 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Behavioral Treatment
of Autism Spectrum Disorder
Autism spectrum disorder (ASD) is a neurodevel- Individuals with ASD may also have difficulties
opmental disorder that is evident in early child- with eye contact, relationships with peers, and
hood. Standard classification systems previously play skills. The second core area of impairment
included ASD in a category of related diagnoses, consists of marked abnormalities in behavior pat-
including autistic disorder, Asperger’s disorder, terns, characterized by restricted, repetitive, or
Rett’s disorder, childhood disintegrative disorder, stereotyped interests or activities. Children with
and pervasive developmental disorder not other- ASD may engage in repetitive behavior, such as
wise specified. Currently, clinicians base the diag- repeatedly placing items in lines and stereotyped
nosis of ASD on diagnostic criteria established by body movements, such as rocking and hand flap-
the Diagnostic and Statistical Manual of Mental Dis- ping. They may display restricted activities, such as
orders, fifth edition (DSM-5; American Psychiat- consuming only certain foods (e.g., chicken nug-
ric Association, 2013). DSM-5 no longer includes gets, chips); engaging in problem behavior if activ-
differential diagnoses such as Asperger’s disorder ities do not occur in a particular order; or resisting
or pervasive developmental disorder not otherwise changes in the environment, such as having a tan-
specified. Clinicians diagnose individuals with trum if a driver takes a different route to school.
ASD if they meet the revised diagnostic criteria Individuals with ASD may also display hyper- or
described below. hyporeactivity to sensory stimulation (e.g., plug-
Individuals diagnosed with ASD demonstrate ging ears in the presence of loud sounds, increased
two core areas of impairment (American Psychi- pain tolerance). Clinicians can specify ASD sever-
atric Association, 2013). First, these individuals ity at three levels (Level 1, Requiring support; Level
display persistent impairment in reciprocal social 2, Requiring substantial support; or Level 3, Requir-
communication and social interaction that may ing very substantial support), based on the degree of
cause difficulties in social and emotional reciproc- impairment in social communication and interac-
ity, reduced affect, and unresponsiveness to social tion and the severity of restricted, repetitive be-
interactions. Deficits in social communication haviors (American Psychiatric Association, 2013).
include odd and repetitive speech patterns and Over the past 30 years, the prevalence of ASD
markedly delayed speech and language skills, and has increased. Recent prevalence estimates indi-
may include a complete lack of spoken language. cate that ASD occurs in 1 in 54 children (Centers
389
390 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
for Disease Control and Prevention, 2020). This of this chapter. Refer to Kodak and Piazza (2008)
prevalence estimate is much higher than estimates for a detailed description of treatments that re-
obtained in the 1980s and early 2000s (Burd, Fish- searchers have used to treat the sleep problems of
er, & Kerbeshian, 1987; Centers for Disease Con- individuals with ASD and other developmental
trol and Prevention, 2020). Several factors may disabilities.
influence the apparent increasing trends in ASD
diagnoses, including heightened awareness of the
Feeding Disorders
disorder’s characteristics, better assessment tools,
early diagnosis, expansion of diagnostic criteria to The prevalence of feeding disorders is substantial-
include cases that were subthreshold according to ly higher in individuals diagnosed with ASD than
DSM-IV-TR criteria (American Psychiatric Asso- in typically developing children, and up to 90% of
ciation, 2000, 2013), confusion regarding measures children with ASD display at least some inappro-
of prevalence versus incidence, and issues about priate mealtime behavior (DeMeyer, 1979; Sharp
the design of studies evaluating the prevalence of et al., 2013). Food selectivity is one type of feed-
ASD (Hill, Zuckerman, & Fombonne, 2015; Volk- ing problem that is common in individuals with
mar, Lord, Bailey, Schultz, & Klin, 2004). Because ASD, perhaps due to the rigid and restricted in-
of these problems, we do not know whether the terests characteristic of this population (Schreck,
apparent growth in ASD prevalence reflects an ac- Williams, & Smith, 2004). Children who display
tual increase in the proportion of children affected food selectivity may consume enough food to meet
by the disorder. their overall caloric needs, but they may not meet
their daily nutritional needs because of limited
consumption of foods high in macro- and micro-
ASSOCIATED FEATURES OF ASD nutrients (Piazza, 2008). Treatments based on the
principles of applied behavior analysis (ABA) have
Several medical conditions and behavioral ex- been highly effective in increasing the variety of
cesses or deficits correlate with a diagnosis of foods consumed and in decreasing inappropriate
ASD (American Psychiatric Association, 2013). mealtime behavior in individuals with food se-
Although certain characteristics or conditions lectivity. For a description of effective behavioral
may be more common in individuals with ASD, interventions for children with feeding disorders,
the associated features are not part of ASD di- refer to Kodak and Piazza (2008), Volkert and Pi-
agnostic criteria. The features we review below azza (2012), and two recent randomized clinical
include sleeping and feeding disorders, delays in trials by Piazza and colleagues (Peterson, Piazza,
toilet training, severe behavior problems, and ge- Ibanez, & Fisher, 2019; Peterson, Piazza, & Volk-
netic and medical conditions (Zwaigenbaum et al., ert, 2016).
2015).
Delayed Toilet Training
Sleep Problems
Many individuals with developmental disabilities
Individuals with ASD display a higher incidence do not complete toilet training successfully before
of sleep problems, including increased daytime the age of 6. In fact, over half of parents of children
sleep (Piazza, Fisher, Kieswetter, Bowman, & with ASD report concerns regarding toilet train-
Moser, 1990) and decreased total sleep, bedtime ing and incontinence (Williams, Oliver, Allard, &
tantrums, and frequent night wakings (Kodak & Sears, 2003). Azrin and Foxx (1971) developed an
Piazza, 2008). Sleep problems can cause decreased intensive behavioral treatment package for toilet
cognitive functioning, increased levels of self-in- training for adults with developmental disabilities;
jurious behavior and other severe problem behav- it includes reinforcement for voids on the toilet,
ior, and increased levels of parental stress (Kodak fluid loading, scheduled toilet trips, and overcor-
& Piazza, 2008; McStay, Dissanayake, Scheeren, rection. Numerous studies have evaluated varia-
Koot, & Begeer, 2014). As a result, many individu- tions of this treatment package with individuals
als with ASD would benefit from treatment for with ASD and other developmental disabilities
sleep problems, given the negative impact these (e.g., LeBlanc, Carr, Crossett, Bennett, & Detwei-
problems can have on both the children and their ler, 2005).
caregivers. A description of behaviorally based Despite the effectiveness of the procedures de-
treatments for sleep problems is outside the scope veloped by Azrin and Foxx (1971) 50 years ago,
Behavioral Treatment of Autism Spectrum Disorder 391
researchers have raised concerns regarding the ap- aggressive, and disruptive behavior; noncompli-
propriateness of using punishment procedures in ance; elopement; and pica (American Psychiatric
toileting treatments (Cicero & Pfadt, 2002). As Association, 2000). Researchers have used func-
a result, researchers have developed novel treat- tional-analysis procedures to identify the environ-
ments that include antecedent manipulations mental variables that maintain the problem behav-
and reinforcement-based procedures to increase ior of individuals with developmental disabilities
the frequency of continent voids (e.g., Hagopian, (e.g., Iwata, Dorsey, Slifer, Bauman, & Richman,
Fisher, Piazza, & Wierzbicki, 1993) and stools (Pi- 1982/1994). Researchers then use the results of
azza, Fisher, Chinn, & Bowman, 1991). Hagopian these procedures to inform treatment. Although
et al. (1993) used a water-prompting procedure the assessment and treatment of severe problem
to increase urinary continence in a young male behavior is critical to intervention programs for
with intellectual developmental disorder. Results many individuals with ASD, other chapters in this
indicated that the water-prompting procedure in- book review functional analysis and treatment of
creased continent voids, while maintaining low problem behavior. Saini, Fisher, Betz, and Piazza
levels of self-injurious behavior that had occurred (Chapter 13) describe experimental functional
during other toileting treatments. analysis; Vollmer, Athens, and Fernand (Chapter
Many toilet-training treatment packages in- 19) describe function-based extinction treatments;
clude multiple components, such as scheduled Fisher, Greer, and Bouxsein (Chapter 20) describe
sits and wearing underwear. Greer, Neidert, and function-based reinforcement treatments; and Le-
Dozier (2015) conducted toilet training with 20 rman and Toole (Chapter 21) describe function-
children ages 19–39 months, including one child based punishment treatments.
diagnosed with ASD. The researchers implement-
ed a toilet-training package that included a dense
Associated Conditions and Genetic Disorders
schedule of toilet sits, wearing underwear, and dif-
ferential reinforcement for remaining dry with 6 Individuals with ASD may have some level of in-
of the 20 children, including the child with ASD. tellectual impairment, although the proportion of
Two of these children, including the child with individuals with comorbid intellectual impairment
ASD, showed improved toileting with the pack- varies across studies, with estimates ranging from
age. The remaining 14 children participated in a 18% (Levy et al., 2010) to 70% (Matson & Shoe-
component analysis and began toilet training with maker, 2009). In addition, several medical condi-
either the dense schedule of toilet sits, underwear, tions and genetic syndromes are more common
or differential reinforcement; researchers subse- in individuals with ASD. Approximately 16% of
quently added components if accidents did not de- individuals with ASD are diagnosed with epilepsy
crease. Wearing underwear improved urinating on (Levy et al., 2010), with the onset of seizures oc-
the toilet for two of the four participants. Neither curring during childhood or adolescence (Rutter,
the dense schedule of toilet sits nor differential re- 1970). Genetic disorders, including tuberous scle-
inforcement for remaining dry improved urinating rosis and fragile X syndrome, are also associated
on the toilet for any participant. Overall, results with ASD. Refer to Volkmar et al. (2004) for more
indicated that replacing diapers or pull-ups with information on the relation between these genetic
underwear improved toileting behavior, but we disorders and ASD.
need additional research to identify components
that might strengthen this effect.
Future research might focus on comparing the EARLY INTENSIVE BEHAVIORAL INTERVENTION
relative effectiveness of different approaches to
toilet training and measuring potential side effects Early intensive behavioral intervention (EIBI) is
or issues that arise with each intervention. For ex- a comprehensive treatment approach for children
ample, overcorrection or frequent scheduled sits with ASD (Roane, Fisher, & Carr, 2016). It in-
may evoke problem behavior, and a water prompt cludes a comprehensive, hierarchically arranged
may be difficult to fade. curriculum implemented for several years, with
the goal of improving a child’s overall function-
ing (Smith, 1999). Behavior analysts develop pro-
Severe Problem Behavior
cedures for EIBI programs based on principles of
Individuals with ASD are more likely to engage in operant conditioning, such as reinforcement, stim-
severe problem behavior, including self-injurious, ulus control, and generalization, to increase adap-
392 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
tive behavior and reduce problem behavior. The language skills, include greater treatment inten-
EIBI therapist provides numerous learning oppor- sity (Makrygianni & Reed, 2010; Virues-Ortega,
tunities to the child, using carefully programmed 2010), longer treatment duration (Makrygianni
instructional and reinforcement procedures. The & Reed, 2010; Virues-Ortega, 2010), inclusion of
program targets multiple areas of functioning in a caregiver training (Makrygianni & Reed, 2010),
developmental sequence to improve several broad and supervisor training with the UCLA model
skill areas. Educational targets for EIBI program- (Reichow & Wolery, 2009).
ming often include preacademic and academic Results of meta-analyses and systematic reviews
skills, verbal behavior, social skills, and inde- have produced strong consensus that EIBI is an
pendent-play skills. Typically, EIBI programs last effective treatment for many behavioral deficits
25–40 hours per week for 2–3 years and include and excesses of ASD (Reichow & Wolery, 2009).
individualized instruction in many settings, such The National Standards Project identified EIBI,
as the home, community, and school (National which it called comprehensive behavioral treatment
Autism Center, 2009, 2015). Results of surveys of for young children, as an established intervention
caregivers and service providers show that EIBI is for children with ASD (National Autism Center,
one of the most common and requested treatments 2009, 2015). An established intervention has suf-
for children with ASD (Green et al., 2006; Stah- ficient available evidence to permit investigators
mer, Collings, & Palinkas, 2005). to determine confidently that the intervention
The University of California at Los Angeles produces positive outcomes for individuals with
(UCLA) Young Autism Project is a landmark out- ASD (National Autism Center, 2015, p. 34). EIBI
come study that documented substantial improve- has the most empirical support of all compre-
ments in approximately 47% of children with ASD hensive treatment models for children with ASD
receiving EIBI (Lovaas, 1987), and those gains in (Reichow, 2012).
functioning were maintained through adolescence
(McEachin, Smith, & Lovaas, 1993). The success
ABA‑Based Models of Early Intervention
of initial studies on EIBI led to additional research
and widespread dissemination. EIBI is the most Researchers have developed several interven-
studied comprehensive treatment model for young tion models based on ABA principles to improve
children with ASD (Reichow, 2012). To date, re- outcomes for children with ASD. The models de-
search has compared (1) different intensities of scribed below have demonstrated effectiveness in
EIBI (e.g., Smith, Eikeseth, Klevstrand, & Lovaas, studies published in peer-reviewed journals, and
1997), (2) EIBI and other treatments (e.g., Eikes- their authors have published commercially avail-
eth, Smith, Jahr, & Eldevik, 2002; Howard, Spark- able treatment manuals. We review each model
man, Cohen, Green, & Stanislaw, 2005), and (3) in detail and describe key differences between the
clinic- versus caregiver-managed models (e.g., Sal- models. Finally, we provide recommendations for
lows & Graupner, 2005; Smith, Groen, & Wynn, integrating the models to maximize learning.
2000).
Researchers have established criteria to review
Natural Environment Training
the empirical support for EIBI in experimental
studies (e.g., Horner et al., 2005; Reichow, Volk- The essential features of natural environment
mar, & Cicchetti, 2008; Reichow & Wolery, 2009). training are the emphasis on child-directed in-
Although specific criteria differ somewhat, the teractions, techniques that increase motivation
criteria for experimental rigor, verification of di- to respond, and generalization of skills. Behavior
agnosis, and measures of procedural integrity have analysts typically conduct natural environment
overlapped considerably. Meta-analyses conducted training—also called natural language approach-
with various criteria have sought to identify vari- es (LeBlanc, Esch, Sidener, & Firth, 2006) and
ables that predict the outcome of behavioral inter- naturalistic-teaching strategies (National Autism
ventions with young children with ASD. Overall, Center, 2015)—in natural settings such as the
research on the relation between a child’s IQ at the home, with an emphasis on training caregivers to
start of treatment and whether the child benefits promote learning opportunities during playtime.
from EIBI has produced mixed or negligible results This model capitalizes on naturally occurring es-
(Reichow, 2012). Characteristics of EIBI programs tablishing operations to teach functional-language
that produce better outcomes, such as higher IQ skills, although the literature on the model does
scores, improved adaptive behavior, and increased not describe establishing operations as a compo-
Behavioral Treatment of Autism Spectrum Disorder 393
nent. Natural environment training incorporates interest in an item by pointing or gesturing to it.
choice-making opportunities, frequent preference The therapist physically approaches the child,
assessments, and reinforcer variation across trials engages in eye contact, and exhibits a question-
to increase the likelihood of occasioning vocaliza- ing look (Hart & Risley, 1975). If the child does
tions. Natural environment training intersperses not emit a spontaneous vocalization, the therapist
mastered tasks to maintain a child’s motivation to provides a prompt. For example, if the child does
respond. The behavior analyst programs training not say, “Toy, please,” after looking at a toy on
in free-play-like settings, to decrease the similarity the shelf, the therapist delivers the prompt “Say
between the training setting and typical academic ‘Toy, please.’ ” The therapist then fades prompts as
or work settings. Natural environment training the child responds accurately after less intrusive
typically targets requests for items in the initial prompts. Hart and Risley (1975) recommend rotat-
portion of treatment to establish a functional re- ing among a few prompts that are context-specific,
lation between the vocal response (e.g., dinosaur) such as “What is this?” or “What do you want?”
and associated reinforcers (e.g., brief access to a toy Limiting prompt variety ensures that the prompt
dinosaur). functions as a discriminative stimulus for the
Natural environment training uses recom- correct vocal response. For example, the prompt
mendations from Stokes and Baer (1977) to train “What is this?” requires the child to provide a
and promote generalization. Thus the approach label for the indicated stimulus, and the prompt
emphasizes training multiple exemplars, training “What do you want?” requires the child to provide
across settings and therapists, and use of intermit- a label for the stimulus that is of interest to the
tent contingencies to promote generalization of child (Hart & Risley, 1975). Limited prompt varia-
language (LeBlanc et al., 2006). After the child tion is one variable that distinguishes incidental
demonstrates an established verbal repertoire, the teaching from other natural environment training
behavior analyst programs intermittent contin- procedures.
gencies that more closely approximate the con- Previous research showed that incidental teach-
tingencies in natural settings (e.g., school). The ing with disadvantaged preschool children was
child is more likely to use his or her newly acquired effective in producing variability in verbal re-
language in many settings, due to natural environ- sponses, generalization to novel therapists (Hart &
ment training’s emphasis on generalization of lan- Risley, 1975), and preference for play activities that
guage skills across settings and therapists. Natural incorporated incidental teaching (Hart & Risley,
environment training includes several teaching 1968). Research on incidental teaching with chil-
strategies, such as the natural language paradigm dren with ASD indicated that training produced
and pivotal-response training. generalization of language (e.g., McGee, Krantz,
Mason, & McClannahan, 1983).
McGee, Krantz, and McClannahan (1985)
Incidental Teaching
compared incidental-teaching and traditional-
Hart and Risley (1968) developed incidental- teaching procedures (e.g., teaching at a desk in
teaching strategies from their experiences teaching a private room with minimal distractions). They
language to preschool children in natural settings. employed both strategies to train children to use
The therapist conducts incidental-teaching strat- prepositions, and results indicated that acquisition
egies in unstructured settings such as free-play of target prepositions did not vary across teach-
time in a classroom. Incidental teaching is child- ing procedures. Incidental-teaching sessions were
directed, and a therapist uses a child’s initiation as longer than traditional-teaching sessions, but inci-
a learning opportunity. The goal is to teach the dental teaching produced more correct preposition
child to respond to multiple cues in the natural use during generalization sessions than traditional
environment with spontaneous language. teaching. This study highlights the need for ad-
Incidental-teaching programs incorporate dis- ditional research comparing teaching strategies to
criminative stimuli that signal the availability of identify procedures that produce (1) rapid acquisi-
adult attention for language attempts (Hart & tion, (2) higher levels of attending, (3) generaliza-
Risley, 1975). This discrimination is important, tion of skills across settings and therapists, and (4)
because the child’s attempts to communicate are the least amount of instructional time necessary
unlikely to produce reinforcement unless an adult to produce mastery of target skills. Future research
is attending to the child’s vocalizations. The thera- also should identify child preferences for language-
pist initiates a learning trial when the child shows training strategies, such as those conducted with
394 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
sults indicated that the participants’ vocalizations the basis of its topographical characteristics (e.g.,
increased. Gillett and LeBlanc also found that vo- a child saying, “Red”), antecedent events that oc-
calizations that were prompted in the beginning of casion or evoke the response (e.g., a parent point-
the study became spontaneous after intervention. ing to a picture of a red ball), and the consequent
Similarly, Spector and Charlop (2018) trained sib- events that reinforce the response (e.g., a parent
lings of children with ASD to accurately imple- saying, “Good job, the ball is red.”).
ment the natural language paradigm, and two of Skinner’s (1957) taxonomy of verbal behavior
the three children with ASD showed improved includes seven elementary verbal operants, and
speech. Thus children may benefit from the nat- early intervention programs typically focus on
ural language paradigm, despite its limitations. five of these: mand, tact, echoic, intraverbal, and
However, we need more research on the efficacy of autoclitic. Treatment typically begins with mand
this paradigm with children who require extensive training, because this verbal operant has the most
shaping to acquire vocal verbal behavior or receive direct benefit to the speaker (Skinner, 1957). A
intervention to improve play skills. mand is a verbal operant (e.g., “Drink, please”)
that a relevant establishing operation evokes (e.g.,
walking outside on a hot day) and delivery of the
Pivotal‑Response Training
requested consequence reinforces (e.g., giving the
Another natural environment training strategy is child a drink). Behavior analysts should conduct
called pivotal-response training because it focuses mand training by using the appropriate anteced-
on pivotal areas, which are skills that produce col- ent and consequent events; that is, mand train-
lateral improvement in other skill areas. Interven- ing tends to be most effective when the relevant
tion typically occurs in natural settings such as establishing operations are present or presumed
the child’s home. Most intervention opportunities to be present, and the consequence related to the
capitalize on naturally occurring environmental child’s mand is available following an appropriate
stimuli and establishing operations. For example, response.
if a child approaches a swing set, sits on the swing, Therapists using an applied-verbal behavior ap-
and waits for an adult to push him or her, the adult proach are also concerned with stimulus control.
models a vocalization (e.g., “Swing”) and pushes Mands should occur in the presence of the rel-
the swing when the child imitates the modeled evant establishing operations and discriminative
response. Thereafter, the adult may momentarily stimuli (e.g., a listener who typically responds to
stop the swing by holding it in the air, give an the child’s vocalization). However, these therapists
expectant look, and provide an independent op- often include antecedent prompts that are com-
portunity for the child to say, “Swing.” If the child ponents of other verbal operants early in training.
emits a relevant vocalization, the adult continues For example, an applied-verbal behavior therapist
pushing the child on the swing. may include an intraverbal prompt (e.g., “What
Caregivers and teachers implement pivotal- do you want?) or an echoic prompt (e.g., “Drink”),
response training across many settings. Research to increase the likelihood that the child will emit
shows that caregivers, teachers, and other school the target response (e.g., the child saying, “Drink”)
personnel can implement pivotal-response train- and that the target response will contact the rel-
ing with high integrity after training via a video- evant reinforcer (e.g., delivery of the drink). Over
training package, ongoing feedback, and a self- time, the therapist quickly fades these prompts to
directed learning model (Nefdt, Koegel, Singer, establish a pure verbal operant.
& Gerber, 2010; Robinson, 2011; Schreibman & Once the child acquires multiple mands, the
Koegel, 2005). Although studies show that re- therapist combines antecedents that control
searchers have trained caregivers and educators to trained verbal operants with other antecedents
implement pivotal-response training successfully, to teach additional functions of language (e.g.,
we need more research to identify efficacious and combined mand–tact training). The therapist
cost-effective training strategies. programs these antecedents into learning oppor-
tunities to produce the target response, and uses
fading to transfer control of the verbal response
Applied‑Verbal Behavior
from one set of stimulus conditions to another.
The applied-verbal behavior approach to early in- Numerous studies have shown that therapists can
tervention for ASD is based on Skinner’s (1957) transfer stimulus control from one verbal operant
functional approach to verbal behavior. Skinner’s (e.g., a tact) to another (e.g., an intraverbal) dur-
functional approach defines a verbal operant on ing training (e.g., Miguel, Petursdottir, & Carr,
396 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
2005). In addition, Skinner (1957) described most interactions (Lovaas, 1987; Lovaas et al., 1973).
vocal verbal behavior as impure or under multiple Discrete-trial training often produces rapid skill
sources of control. Thus studies that evaluate ways acquisition, because it allows fast-paced instruc-
to promote the emergence of untrained operants tion to maximize the number of learning opportu-
following intervention, or that develop verbal be- nities in a brief period.
havior under multiple sources of control, should Discrete-trial training has several additional
guide treatment practices in verbal behavior (e.g., advantages relative to other early-intervention
Finn, Miguel, & Ahearn, 2012; Michael, Palmer, models. First, it involves well-defined and straight-
& Sundberg, 2011). forward training procedures, which are easy to
Skinner’s (1957) account of verbal behavior teach to inexperienced staff members. Second,
has improved communication-training procedures discrete-trial training may be well suited for teach-
by placing an emphasis on initial mand training ing certain verbal operants (e.g., tacts, echoics),
and highlighting the need for intraverbal train- because it is like the format of typical classroom
ing (Sundberg & Michael, 2001). Prior approaches instruction (e.g., sitting at a desk and answering
to communication training focused primarily on questions) than other intervention models. Fi-
tact training and largely neglected intraverbal- nally, data collection during learning trials may be
behavior training. That is, researchers taught easier with this approach than with other models
individuals with ASD to tact most items in their (Sundberg & Partington, 1999).
environment, but the children could not converse
with others. This limited training procedure did
Unique Features of Early‑Intervention Models
not teach the skills necessary for individuals with
ASD to function in typical settings or to establish Many behaviorally based early-intervention mod-
peer relationships. As such, Skinner’s conceptual- els have different language-training approaches,
ization and the inclusion of mand and intraverbal teaching methods, and terminology. Natural en-
training have advanced the quality of treatment vironment training and applied-verbal behavior
provided to children in communication-training both emphasize naturalistic-training procedures
programs (Sundberg & Michael, 2001). that focus on teaching requests (natural environ-
ment training) or mands (applied-verbal behavior)
under conditions in which a child is motivated
Discrete‑Trial Training
to obtain the requested item. The procedures
Lovaas and colleagues developed discrete-trial maximize the effects of the relevant establishing
training in the early 1970s as a behavioral inter- operations by conducting frequent preference as-
vention for children with ASD (Lovaas, Koegel, sessments, restricting access to the items outside of
Simmons, & Long, 1973). This model emphasizes teaching trials, or both. Skills taught in discrete-
a highly structured approach to teaching, in which trial training vary from those of natural environ-
the child and therapist sit at a table, and the ther- ment training and applied-verbal behavior, be-
apist trains various skills during discrete-learning cause discrete trials incorporate specific, verbal
trials with brief intervals between trials. The discriminative stimuli (e.g., “What is it?”) instead
therapist breaks the skills into components and of programming trials to bring responding under
teaches those components until the child mas- antecedent control (Skinner, 1957). Discrete-trial
ters them. The therapist subsequently teaches the training may produce correct responding on each
child to combine components into larger behavior trial, but extensive training may be necessary to
sequences. The therapist delivers highly preferred generalize these skills to other, less structured set-
items (usually small edible items) and praise to the tings. Thus proponents of natural environment
child following correct responses. Sessions, some- training and applied-verbal behavior have stated
times referred to as drills, consist of a specific num- that discrete-trial training may not promote skill
ber of trials that may involve presenting the same generalization as much as other early-intervention
discriminative stimulus in a massed-trial format models do (Sundberg & Michael, 2001).
(e.g., repeatedly presenting trials of “Touch dog” Each model uses different terminology. Natural
throughout a session). environment training and discrete-trial training
Early studies evaluating discrete-trial training make a distinction between teaching receptive
for young children with ASD reported substantial and expressive language. Teaching receptive lan-
increases in IQ scores, decreases in inappropri- guage typically refers to training auditory–visual
ate behavior (e.g., self-stimulatory and problem conditional discriminations (e.g., teaching a child
behavior), and increases in spontaneous social point to a picture following the presentation of an
Behavioral Treatment of Autism Spectrum Disorder 397
auditory stimulus). Proponents of applied-verbal assessments during trials. Instead, the therapist al-
behavior refer to this type of training as teach- ternates several highly preferred food items across
ing listener behavior. Teaching expressive language trials to decrease the likelihood of satiation.
typically refers to teaching children to tact an item
in the presence of a nonverbal stimulus (e.g., a pic-
Combining Early‑Intervention Models
ture) and a verbal stimulus (e.g., “What is it?”).
The distinction between receptive and expressive Each approach to early intervention has unique
language is consistent with a structural approach strengths that therapists can use in an eclectic
to language development (Leaf & McEachin, approach that includes several empirically vali-
1999; Maurice, Green, & Luce, 1996; see Peturs- dated, ABA-based interventions. For example, a
dottir & Carr, 2011 for a review). Alternatively, therapist can use natural environment training to
applied-verbal behavior therapists approach com- teach social skills, such as turn taking and shar-
munication training from a functional perspec- ing toys. The therapist can train the functions of
tive and consider the antecedent and consequent verbal operants using the applied-verbal behavior
events for verbal behavior during teaching (Ler- approach in the natural environment, while the
man et al., 2005). child acquires other skills (such as attending be-
The intervention setting also varies substan- havior and instruction following) during discrete-
tially, particularly early in training. Natural en- trial training. The therapist can capitalize on nat-
vironment training and applied-verbal behavior urally occurring establishing operations by only
language training occurs in natural settings and using certain highly preferred items during lan-
focuses on establishing mand repertoires. This guage training and restricting them at other times.
teaching occurs during free-play time, such as the The therapist can implement tact training during
loosely structured activities in preschool class- discrete-trial training to increase the child’s mean
rooms. In discrete-trial training, by contrast, a length of utterance in the context of the function
child is not trained in the natural environment of the vocalization (applied-verbal behavior). For
until the child masters several prerequisite skills example, training might focus on teaching tacts
during one-on-one seatwork (e.g., attending, com- and corresponding autoclitic frames while con-
pliance). That is, discrete-trial training teaches structing toys during learning trials (e.g., “That’s
verbal operants (typically tacts and echoics) dur- a tail”; Finn et al., 2012).
ing one-on-one activities at a desk for the first year The therapist can incorporate Skinner’s (1957)
or two of treatment, followed by practice in other, taxonomy of verbal behavior to teach skills with
less structured settings later in treatment (e.g., the natural environment training or discrete-trial
second year of early intervention). training. For example, a therapist may teach yes
Differences in the language-training setting and no as a mand–intraverbal operant with an ap-
substantially alter the way each approach address- plied-verbal behavior approach. The therapist can
es motivational issues. Natural environment train- ask the child whether he or she wants to watch
ing describes motivation as a child’s willingness to highly preferred videos on some trials or nonpre-
respond for preferred items. A therapist provides ferred videos on other trials. The therapist can use
frequent choices between highly preferred items discrete-trial training and program the relevant
to encourage the child to participate in treat- antecedents and consequences in the training
ment. The applied-verbal behavior approach in- context. Combining discrete-trial training with
volves establishing operations that momentarily other early-intervention models capitalizes on the
increase the effectiveness of a reinforcing stimulus strengths of several models by incorporating natu-
and increase the likelihood of behavior that has rally occurring establishing operations, and the
produced the stimulus in the past. Thus the thera- learning trials are more child-driven because the
pist programs naturally occurring establishing therapist offers the child preferred or nonpreferred
operations into mand training, such as conduct- videos before each trial.
ing training after the child has not had access to
several highly preferred items for a period (Kelley,
Shillingsburg, Castro, Addison, & LaRue, 2007). CLINICAL CONSIDERATIONS
Finally, discrete-trial training provides highly pre- AND FUTURE RESEARCH
ferred items (e.g., food) following correct responses
Establishing a Reinforcing Environment
during each trial. Discrete-trial training places less
emphasis on motivational variables, and a thera- In many cases, the identification of reinforcing
pist typically does not conduct frequent preference stimuli is a critical factor in the success of behav-
398 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
ior acquisition programs. Behavior analysts com- ment (Dunlap et al., 1994) and decreased problem
monly conduct systematic preference assessments behavior (Romaniuk et al., 2002). Choice between
to identify a hierarchy of potentially reinforcing reinforcers also can lead to more efficient skill
items or activities. Much research has focused on acquisition (Toussaint et al., 2016). Several rein-
developing effective and efficient methods for as- forcement parameters, such as reinforcement rate,
sessing preferences for stimuli among individuals quality, immediacy, and magnitude, may influence
with developmental disabilities (see Saini, Ret- an individual’s choice and alter the efficacy of the
zlaff, et al., Chapter 10, this volume). reinforcing stimuli included in early-intervention
A preference assessment can provide opportuni- learning trials. Future research should determine
ties for individuals to select the reinforcers for up- the extent to which choice variables influence the
coming teaching sessions in an early-intervention acquisition of new skills.
program. Although behavioral experts recommend
frequent, systematic preference assessments to
Assessing and Identifying Target Skills
control for potential fluctuations in preference for
reinforcers over time, early-intervention programs A behavior analyst should assess an individual’s
vary widely in how often they assess preferences. current skill level in several domains to select
A study by Love, Carr, Almason, and Petursdot- appropriate targets for intervention. Information
tir (2009) found that most early-intervention from a formal assessment of skills allows the clini-
programs (i.e., 65%) assessed preference multiple cian to (1) measure several important behavioral
times per day, whereas a few programs (i.e., 19%) repertoires, (2) identify areas that would benefit
conducted preference assessments quarterly or from intervention, and (3) track educational prog-
annually. Research indicates that frequent, brief ress over time. Several formal assessment programs
preference assessments are more effective than less are commercially available to clinicians and re-
frequent, comprehensive assessments for identi- searchers.
fying potential reinforcers (DeLeon et al., 2001).
Despite research supporting the use of frequent
Assessment of Basic Learning and Language Skills—
preference assessments, some early-intervention
Revised
programs may conduct preference assessments less
often because of time or resource constraints. Re- The Assessment of Basic Learning and Language
searchers have responded to this clinical issue by Skills—Revised (ABLLS-R) is a tool for identify-
developing brief preference assessments that are ing areas that require intervention for children
more time-efficient to implement in natural set- with autism and other developmental disabilities
tings (e.g., Carr, Nicolson, & Higbee, 2000). In ad- (Partington, 2006). The ABLLS-R includes edu-
dition, researchers have developed self-instruction cational skills grouped into 25 areas (e.g., social
packages to assist staff members who do not have skills, motor skills). Skinner’s (1957) account of
access to expert trainers to conduct preference verbal behavior forms the basis for the language
assessments with children accurately (Graff & components of the ABLLS-R. The ABLLS-R has
Karsten, 2012). guidelines clinicians can use to develop goals for
In conjunction with frequent preference assess- individualized education plans. The author of the
ments, clinicians can intersperse choice-making ABLLS-R also co-wrote a book, Teaching Language
opportunities throughout educational program- to Children with Autism or Other Developmental
ming to increase the reinforcing properties of in- Disabilities (Sundberg & Partington, 1998), as a
struction in general (Laski et al., 1988; Toussaint, companion treatment manual for many skills the
Kodak, & Vladescu, 2016). Moreover, research ABLLS-R assesses. The ABLLS-R collects infor-
suggests that individuals may prefer instructional mation (1) during interviews of caregivers and
contexts that incorporate opportunities to make other individuals who are familiar with the child
choices among reinforcers to environments that (e.g., teachers), (2) during naturalistic observa-
provide identical reinforcers selected by a therapist tions, and (3) in tests of specific skills.
(Tiger, Hanley, & Hernandez, 2006). There are
multiple opportunities to integrate choices into
Verbal Behavior Milestones Assessment
a typical teaching session during early interven-
and Placement Program
tion (e.g., task and reinforcer selection). Previous
research has demonstrated that providing choices The Verbal Behavior Milestones Assessment and
between tasks may produce increased task engage- Placement Program (VB-MAPP) assesses the cur-
Behavioral Treatment of Autism Spectrum Disorder 399
rent level of several language and related behaviors lational responses in the Transformation Module.
to identify educational target skills in individuals The behavior analyst completes PEAK assess-
with ASD and other developmental disabilities ments for each module in either an indirect (sur-
(Sundberg, 2008). The VB-MAPP includes three vey) or direct (actual client interaction) modality.
main sections: the Milestones Assessment, the The behavior analyst assesses a skill directly if the
Barriers Assessment, and the Transition Assess- caregiver does not know whether the child has
ment. Like the ABLLS-R, the VB-MAPP includes mastered the skill. The equivalence and transfor-
information gathered by (1) interviewing caregiv- mation modules also contain a preassessment of
ers and others who are familiar with the child, (2) arbitrary-stimulus relations to evaluate whether
collecting data during naturalistic observations, a child can derive novel-stimulus relations after
and (3) conducting tests of specific skills. receiving information from the behavior analyst.
The Milestones Assessment evaluates the The assessment duration can range from 1 to 4
child’s current behavioral repertoire along 16 di- hours, depending on the size and complexity of
mensions of behavior (e.g., listener responding) the child’s verbal repertoire.
arranged in a developmental sequence across Following the assessment, the behavior ana-
three levels. The three levels measure skills that lyst provides the therapists with goals and brief
are typical for children between the ages of 0 instructions on how to teach the deficit skills the
and 18 months, 18 and 30 months, and 30 and assessment has identified. Most published research
48 months, respectively. The Barriers Assess- on PEAK has focused on the first module, the Di-
ment assesses 24 language and learning barri- rect Training Module. Researchers have reported
ers (e.g., prompt dependency) that may interfere correlations between the Direct Training Module
with progress during educational programming. and IQ, expressive and receptive language, and
The behavior analyst should address barriers other behavioral assessments of language like the
identified during this assessment before focusing VB-MAPP (Dixon et al., 2014, 2015). A study of
on other objectives. The Transition Assessment PEAK implementation showed that direct-care
gathers information on specific skills needed to professionals with no history of ABA or discrete-
make educational gains in less restrictive envi- trial training implemented PEAK with 60–80%
ronments. The assessment has three main cat- accuracy after reviewing the manual. After one to
egories. Category 1 measures academic indepen- four sessions of behavioral-skills training lasting
dence (e.g., group skills) and determines the level 10–45 minutes, the same staff implemented PEAK
of support the child needs. Category 2 assesses with 100% accuracy (Belisle, Rowsey, & Dixon,
learning patterns (e.g., rate of skill acquisition) 2016).
and determines the child’s readiness for acquiring
skills outside of one-on-one instruction. Category
Autism Curriculum Encyclopedia
3 evaluates self-help skills (e.g., toileting skills),
spontaneity (e.g., adaptability to change), and Developed by professionals at the New England
self-direction (e.g., independent play skills). Center for Children, the Autism Curriculum En-
cyclopedia (ACE) is an educational software pack-
age of 2,000 skills (www.acenecc.org). The ACE
Promoting the Emergence of Advanced Knowledge
includes three assessments: the Core Skills Assess-
Relational Training System
ment, the ACE Skills Assessment, and the Pref-
The Promoting the Emergence of Advanced erence Assessment. The Core Skills Assessment
Knowledge (PEAK) Relational Training System is measures 52 basic skills such as picture–object
an assessment and curriculum tool that is based on matching, requesting assistance, and following
the principles of ABA. It includes four modules, one-step directions. The ACE Skills Assessment
each of which contains 184 skills. The modules identifies specific targets for individual learners,
are divided by the author’s conceptualization of and it includes links to protocols that behavior an-
distinct learning modalities, which include direct alysts can use to teach targeted skills. Finally, the
contingency learning in the Direct Training Mod- Preference Assessment includes 12 assessments,
ule, generalization of skills to novel stimuli and instructions and data sheets for each assessment,
learning contexts in the Generalization Module, and a tool to assist practitioners in selecting an
the formation of stimulus equivalence classes in ideal assessment to conduct. The ACE includes
the Equivalence Module, and the transformation procedures to track student progress and measure
of relational frames to more complex derived re- outcomes across years.
400 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
a novel skill. These researchers found that effec- siderations (e.g., time and effort) influence the ef-
tive prompts and prompt-fading strategies varied ficiency of various prompt-fading strategies.
across individuals, but the assessment accurately
identified ideal strategies for learners across tasks.
Maintenance and Generalization
Lerman et al. (2004) found differential efficacy of
prompts across types of skills. For example, one A primary goal of early intervention is for a child
participant required physical prompts to acquire to exhibit previously learned skills during direct
auditory–visual conditional discriminations of instruction (i.e., maintenance of the skills over
numbers and completion of puzzles, but differen- time) and engage in these skills in many set-
tial reinforcement without prompts was sufficient tings and in novel situations (i.e., generalization
to produce acquisition of matching numbers. Fu- of skills). Unfortunately, many early-intervention
ture research should examine the generality of the programs fail to program adequately for mainte-
results of assessment-based instruction across skills nance and generalization of skills (Smith, 1999).
and types of prompts. In addition, future research Furthermore, the educational gains achieved dur-
might focus on identifying specific learner skills or ing intervention often do not generalize to other
deficits predictive of the relative efficacy of various response topographies (e.g., learning to add the
stimulus and response prompts. -ing ending to a novel word) or to response varia-
tions required in other settings (Charlop, Schreib-
man, & Thibodeau, 1985). As such, future re-
Identifying Prompt‑Fading Strategies
search should evaluate variables that promote and
Behavior analysts typically incorporate response interfere with maintenance and generalization of
and stimulus prompts into instructional proce- skills acquired through early intervention.
dures during initial teaching sessions and fade the
prompts as quickly as possible (i.e., prompt fad- Maintenance. Maurice et al. (1996) recom-
ing). The purpose of prompt fading is to transfer mend assessing maintenance once per week for
stimulus control from therapist-delivered prompts 3–6 weeks to measure whether a child contin-
to stimuli in the natural environment that should ues to exhibit the recently mastered skill. When
evoke appropriate responses (Walker, 2008). data demonstrate that the child has maintained a
Researchers have developed and evaluated mul- target skill (e.g., correct responses during weekly
tiple prompt-fading procedures (e.g., least-to-most maintenance probes), the child may continue to
prompting; Horner & Keilitz, 1975). Numerous practice the skill at relevant times, but without ad-
studies have compared the relative effectiveness ditional data collection on the mastered skill.
and efficiency of prompt-fading techniques (e.g., Many early-intervention manuals recommend
Seaver & Bourret, 2014). In general, research in- modifications to ongoing teaching procedures to
dicates that most prompt-fading procedures are increase the likelihood that children with ASD
effective for teaching many new skills (Ault, Wol- will show skill maintenance after training. For
ery, Doyle, & Gast, 1989). However, the efficien- example, Lovaas (2003) recommended modifying
cy of the procedures varies; that is, the number of both the schedule and the type of reinforcement
sessions required to teach new skills varies con- provided for correct responses—procedures that
siderably across different prompting procedures. can also promote generalization. In addition, when
The specific skills of the learner or the task may the behavior analyst uses arbitrary reinforcers (e.g.,
affect the efficiency of the prompting procedures food) to train a new task, he or she should program
(Wolery & Gast, 1984). For example, within- naturally occurring reinforcers (e.g., praise, high
stimulus fading may be ideal for learners who at- fives) into treatment and maintenance, to ensure
tend to irrelevant features of stimuli (i.e., faulty that behavior is not extinguished in the natural
stimulus control), whereas an identity-matching environment once the delivery arbitrary reinforc-
prompt may be well suited for learners who have ers ceases.
poor attending skills (Fisher et al., 2007). An-
other consideration is that of the training and ef- Generalization. Generalization occurs when a
fort required to implement various prompt-fading child displays skills acquired in one setting or with
techniques. For example, within-stimulus fading one person in other settings or with other people
requires extensive preparation and materials. Fu- without direct training. Stokes and Baer (1977)
ture research might evaluate how the learner’s described seven techniques to evaluate and pro-
skill level, features of the task, and practical con- mote generalization. The specific techniques most
402 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
relevant to early intervention include training suf- increased attention in recent years (Fryling, Wal-
ficient exemplars and using indiscriminable con- lace, & Yassine, 2012). Carroll, Kodak, and Fisher
tingencies. Training sufficient exemplars involves (2013) conducted an initial study to identify the
(1) teaching multiple responses to the same stim- prevalence of treatment integrity failures when
ulus, such as teaching multiple ways to deliver a classroom teachers of children with ASD imple-
compliment; and (2) teaching the child to engage mented discrete-trial training. The results showed
in the correct response across many different, rele- that teachers and paraprofessionals made frequent
vant stimuli, such as teaching the child to tact dog errors during discrete-trial training: (1) only cor-
in the presence of different examples of dogs that recting approximately 40% of errors following in-
share some features. Sometimes generalization can correct responses; (2) not providing clear instruc-
occur when the behavior analyst includes as few as tions during approximately 50% of teaching trials;
two response exemplars, therapists, or settings dur- and (3) not delivering reinforcement consistently
ing training (e.g., Stokes, Baer, & Jackson, 1974), (teachers delivered reinforcement for only about
but more exemplars are required at other times 20% of correct responses). Although the frequent
(e.g., Wunderlich, Vollmer, Donaldson, & Phil- occurrence of instructional errors during class-
lips, 2014). Future research should focus on deter- room-based discrete-trial training is a cause for
mining the optimal conditions to achieve stimulus concern, whether these errors negatively affected
generalization. learning was unclear. Therefore, Carroll et al.
Another behavioral technique that is highly rele- (2013) conducted a second experiment to evalu-
vant to training for generalization is the use of indis- ate the impact of these common instructional er-
criminable contingencies. When using indiscrim- rors on children’s learning. The authors compared
inable contingencies, the behavior analyst typically high-integrity instruction (in which the therapist
provides reinforcement for appropriate behavior on made no errors) to low-integrity instruction (in
an intermittent schedule (e.g., variable-interval 30 which the therapist made errors during 67% of tri-
seconds). For example, Freeland and Noell (2002) als) and to a no-instruction control condition. The
evaluated whether completion of math problems researchers based the low-integrity instruction
was maintained when correct responding no longer percentage on the mean number of errors teach-
produced reinforcement following a phase in which ers and paraprofessionals made using discrete-trial
(1) each correct response produced reinforcement training with children with ASD during classroom
or (2) correct responding produced intermittent re- observations. The results showed that participants
inforcement. Results indicated that responding was rapidly acquired skills in the high-integrity instruc-
maintained only following a phase of intermittent tion condition, but did not acquire similar skills
reinforcement. Although most behavior analysts in the low-integrity instruction condition. In fact,
typically thin the schedule of reinforcement after a most participants did not learn targeted skills dur-
child masters skills in early-intervention programs, ing low-integrity instruction, despite three times
few studies have evaluated the reinforcement sched- the number of instructional sessions. Participants
ule necessary to maintain skills in settings other acquired the skills when the researchers used high-
than the training setting. integrity instruction with those skills initially as-
signed to the low-integrity condition.
Similar studies evaluating decrements in treat-
Treatment Integrity
ment integrity found that errors occurring in 33–
Many studies demonstrating the efficacy of early- 100% of trials affected learning (e.g., DiGennaro
intervention models involved rigorous experimen- Reed, Reed, Baez, & Maguire, 2011; Jenkins, Hirst,
tal protocols in which highly trained clinicians & DiGennaro Reed, 2015; Noell, Gresham, &
served as therapists. However, most direct-service Gansle, 2002). However, whether these controlled
providers in practice have less education and ex- studies that manipulated specific components of
perience than the therapists in the published re- instruction represent the types and frequency of
search. Thus therapists in practice are unlikely errors occurring during the implementation of
to implement ABA interventions with the same ABA interventions in homes, schools, and clinics
integrity levels as those reported in the literature is unclear. Additional research should evaluate the
(e.g., Reed, Osborne, & Corness, 2007; Smith, Fla- minimum amount of treatment integrity necessary
nagan, Garon, & Bryson, 2015). to produce positive treatment outcomes. Investiga-
The impact of treatment integrity on the effec- tions like these can assist in developing guidelines
tiveness of behavioral interventions has received for treatment integrity in practice.
Behavioral Treatment of Autism Spectrum Disorder 403
Daar, J. H., & Szelkey, S. (2015). Toward a behavior J. J. (1993). A water-prompting procedure for the
analysis of complex language: Evaluating the rela- treatment of urinary incontinence. Journal of Applied
tionship between PEAK and the VB-MAPP. Journal Behavior Analysis, 26, 473–474.
of Developmental and Physical Disabilities, 27, 223–233. Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci,
Dixon, M. R., Carman, J., Tyler, P. A., Whiting, S. W., S. A., & Maglieri, K. A. (1997). Evaluation of client
Enoch, M., & Daar, J. H. (2014). Peak relational preference for function-based treatment packages.
training system for children with autism and develop- Journal of Applied Behavior Analysis, 30, 459–473.
mental disabilities: Correlation with Peabody Picture Hanley, G. P., Piazza, C. C., Fisher, W. W., & Maglieri,
Vocabulary Test and assessment reliability. Journal of K. A. (2005). On the effectiveness of and preference
Developmental and Physical Disabilities, 26, 603–614. for punishment and extinction components of func-
Dunlap, G., dePerczel, M., Clarke, S., Wilson, D., White, tion-based interventions. Journal of Applied Behavior
R., & Gomez, A. (1994). Choice making to promote Analysis, 38, 51–65.
adaptive behavior for students with emotional and Hart, B. M., & Risley, T. R. (1968). Establishing use of
behavioral challenges. Journal of Applied Behavior descriptive adjectives in the spontaneous speech of
Analysis, 27, 505–518. disadvantaged preschool children. Journal of Applied
Eikeseth, S., Smith, T., Jahr, E., & Eldevik, S. (2002). Behavior Analysis, 1, 109–120.
Intensive behavioral treatment at school for 4- to Hart, B., & Risley, T. R. (1975). Incidental teaching of
7-year-old children with autism: A 1-year comparison language in the preschool. Journal of Applied Behavior
controlled study. Behavior Modification, 26, 49–68. Analysis, 8, 411–420.
Finn, H. E., Miguel, C. F., & Ahearn, W. H. (2012). The Hill, A. P., Zuckerman, K., & Fombonne, E. (2015).
emergence of untrained mands and tacts in children Epidemiology of autism spectrum disorders. In M. D.
with autism. Journal of Applied Behavior Analysis, 45, Robinson-Agramonte (Ed.), Translational approaches
265–280. to autism spectrum disorder (pp. 13–38). Cham, Swit-
Fisher, W. W., Kodak, T., & Moore, J. W. (2007). Em- zerland: Springer International.
bedding an identity-matching task within a prompt- Horner, R. D., & Keilitz, I. (1975). Training mentally
ing hierarchy to facilitate acquisition of conditional retarded adolescents to brush their teeth. Journal of
discriminations in children with autism. Journal of Applied Behavior Analysis, 25, 491–498.
Applied Behavior Analysis, 40, 489–499. Horner, R. H., Carr, E. G., Halle, J., McGee, G., Odom,
Freeland, J. T., & Noell, G. H. (2002). Programming for S., & Wolery, M. (2005). The use of single-subject re-
maintenance: An investigation of delayed intermit- search to identify evidence-based practice in special
tent reinforcement and common stimuli to create education. Exceptional Children, 71, 165–179.
indiscriminable contingencies. Journal of Behavioral Howard, J. S., Sparkman, C. R., Cohen, H. G., Green,
Education, 11, 5–18. G., & Stanislaw, H. (2005). A comparison of inten-
Fryling, M. J., Wallace, M. D., & Yassine, J. N. (2012). sive behavior analytic and eclectic treatments for
Impact of treatment integrity on intervention ef- young children with autism. Research in Developmen-
fectiveness. Journal of Applied Behavior Analysis, 45, tal Disabilities, 26, 359–383.
449–453. Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., &
Gianoumis, S., Seiverling, L., & Sturmey, P. (2012). The Richman, G. S. (1994). Toward a functional analysis
effects of behavior skills training on correct teacher of self-injury. Journal of Applied Behavior Analysis, 27,
implementation of natural language paradigm teach- 197–209. (Reprinted from Analysis and Intervention in
ing skills and child behavior. Behavioral Interven- Developmental Disabilities, 2, 3–20, 1982)
tions, 27, 57–74. Jenkins, S. R., Hirst, J. M., & DiGennaro Reed, F. D.
Gillett, J. N., & LeBlanc, L. A. (2007). Parent-imple- (2015). The effects of discrete-trial training commis-
mented natural language paradigm to increase lan- sion errors on learner outcomes: An extension. Jour-
guage and play in children with autism. Research in nal of Behavioral Education, 24, 196–209.
Autism Spectrum Disorders, 1(3), 247–255. Kelley, M. E., Shillingsburg, M. A., Castro, M. J., Addi-
Graff, R. B., & Karsten, A. M. (2012). Evaluation of son, L. R., & LaRue, R. H., Jr. (2007). Further evalu-
a self-instruction package for conducting stimulus ation of emerging speech in children with develop-
preference assessments. Journal of Applied Behavior mental disabilities: Training verbal behavior. Journal
Analysis, 45, 69–82. of Applied Behavior Analysis, 40, 431–445.
Green, V. A., Pituch, K. A., Itchon, J., Choi, A., Kodak, T., Clements, A., & LeBlanc, B. (2013). A rapid
O’Reilly, M., & Sigafoos, J. (2006). Internet survey of assessment of instructional strategies to teach audi-
treatments used by parents of children with autism. tory-visual conditional discriminations to children
Research in Developmental Disabilities, 27, 70–84. with autism. Research in Autism Spectrum Disorders,
Greer, B. D., Neidert, P. L., & Dozier, C. L. (2016). A 7, 801–807.
component analysis of toilet-training procedures Kodak, T., Clements, A., Paden, A. R., LeBlanc, B.,
recommended for young children. Journal of Applied Mintz, J., & Toussaint, K. A. (2015). Examination
Behavior Analysis, 49, 69–84. of the relation between an assessment of skills and
Hagopian, L. P., Fisher, W., Piazza, C. C., & Wierzbicki, performance on auditory–visual conditional discrim-
Behavioral Treatment of Autism Spectrum Disorder 405
inations for children with autism spectrum disorder. (1973). Some generalization and follow-up measures
Journal of Applied Behavior Analysis, 48, 52–70. on autistic children in behavior therapy. Journal of
Kodak, T., Fisher, W. W., Clements, A., Paden, A. R., & Applied Behavior Analysis, 6, 131–166.
Dickes, N. (2011). Functional assessment of instruc- Love, J. R., Carr, J. E., Almason, S. M., & Petursdot-
tional variables: Linking assessment and treatment. tir, A. I. (2009). Early and intensive behavioral in-
Research in Autism Spectrum Disorders, 5, 1059–1077. tervention for autism: A survey of clinical practices.
Kodak, T., & Piazza, C. C. (2008). Assessment and be- Research in Autism Spectrum Disorders, 3, 421–428.
havioral treatment of feeding and sleeping disorders Makrygianni, M. K., & Reed, P. (2010). A meta-analytic
in children with autism spectrum disorders. Child review of the effectiveness of behavioural early inter-
and Adolescent Psychiatric Clinics of North America, vention programs for children with autism spectrum
17, 887–905. disorders. Research in Autism Spectrum Disorders, 4,
Koegel, R. L., O’Dell, M. C., & Koegel, L. K. (1987). A 577–593.
natural language paradigm for teaching non-verbal Matson, J. L., & Shoemaker, M. (2009). Intellectual
autistic children. Journal of Autism and Developmen- disability and its relationship to autism spectrum
tal Disorders, 17, 187–199. disorders. Research in Developmental Disabilities, 30,
Laski, K. E., Charlop, M. H., & Schreibman, L. (1988). 1107–1114.
Training parents to use the natural language para- Maurice, C., Green, G., & Luce, S. C. (1996). Behavioral
digm to increase their autistic children’s speech. Jour- intervention for young children with autism: A manual
nal of Applied Behavior Analysis, 21, 391–400. for parents and professionals. Austin, TX: PRO-ED.
Leaf, R., & McEachin, J. (1999). A work in progress: McComas, J. J., Wacker, D. P., Cooper, L. J., Asmus, J.
Behavior management strategies and a curriculum for M., Richman, D., & Stoner, B. (1996). Brief experi-
intensive behavior treatment of autism. New York: DRL mental analysis of stimulus prompts for accurate re-
Books. sponding on academic tasks in an out-patient clinic.
LeBlanc, L. A., Carr, J. E., Crossett, S. E., Bennett, C. Journal of Applied Behavior Analysis, 29, 397–401.
M., & Detweiler, D. D. (2005). Intensive outpatient McEachin, J. J., Smith, T., & Lovaas, O. I. (1993). Long-
behavioral treatment of primary urinary inconti- term outcome for children with autism who received
nence of children with autism. Focus on Autism and early intensive behavioral treatment. American Jour-
Other Developmental Disabilities, 20, 98–105. nal on Mental Retardation, 97, 359–372.
LeBlanc, L. A., Esch, J., Sidener, T. M., & Firth, A. M. McGee, G. G., Krantz, P. J., Mason, D., & McClanna-
(2006). Behavioral language interventions for chil- han, L. E. (1983). A modified incidental-teaching
dren with autism: Comparing applied verbal behav- procedure for autistic youth: Acquisition and gener-
ior and naturalistic teaching approaches. Analysis of alization of receptive object labels. Journal of Applied
Verbal Behavior, 22, 49–60. Behavior Analysis, 16, 329–338.
Lerman, D. C., Parten, M., Addison, L. R., Vorndran, C. McGee, G. G., Krantz, P. J., & McClannahan, L. E.
M., Volkert, V. M., & Kodak, T. (2005). A method- (1985). The facilitative effects of incidental teaching
ology for assessing the functions of emerging speech on preposition use by autistic children. Journal of Ap-
in children with autism. Journal of Applied Behavior plied Behavior Analysis, 18, 17–31.
Analysis, 38, 303–316. McGhan, A. C., & Lerman, D. C. (2013). An assess-
Lerman, D. C., Vorndran, C., Addison, L., & Kuhn, S. ment of error-correction procedures for learners with
A. (2004). A rapid assessment of skills in young chil- autism. Journal of Applied Behavior Analysis, 46,
dren with autism. Journal of Applied Behavior Analy- 626–639.
sis, 37, 11–26. McStay, R. L., Dissanayake, C., Scheeren, A., Koot, H.
Levy, S., Giarrelli, E., Li-Ching, L., Schieve, L., Kirby, M., & Begeer, S. (2014). Parenting stress and autism:
R., Cunniff, C., et al. (2010). Autism spectrum dis- The role of age, autism severity, quality of life and
order and co-occurring developmental, psychiatric, problem behaviour of children and adolescents with
and medical conditions among children in multiple autism. Autism, 18, 502–510.
populations of the United States. Journal of Develop- Michael, J., Palmer, D. C., & Sundberg, M. L. (2011).
mental and Behavioral Pediatrics, 31, 267–275. The multiple control of verbal behavior. Analysis of
Lovaas, O. I. (1981). Teaching developmentally disabled Verbal Behavior, 27, 3–22.
children: The ME book. Baltimore: University Park Miguel, C. F., Petursdottir, A. I., & Carr, J. E. (2005).
Press. The effects of multiple-tact and receptive-discrimi-
Lovaas, O. I. (1987). Behavioral treatment and normal nation training on the acquisition of intraverbal be-
educational and intellectual functioning in young havior. Analysis of Verbal Behavior, 21, 27–41.
autistic children. Journal of Consulting and Clinical National Autism Center. (2009). National Standards
Psychology, 55, 3–9. Project. Randolph, MA: Author.
Lovaas, O. I. (2003). Teaching individuals with develop- National Autism Center. (2015). Findings and conclu-
mental delays: Basic intervention techniques. Austin, sions: National Standards Project, Phase 2. Randolph,
TX: PRO-ED. MA: Author.
Lovaas, O. I., Koegel, R., Simmons, J. Q., & Long, J. Nefdt, N., Koegel, R., Singer, G., & Gerber, M. (2010).
406 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
The use of a self-directed learning program to pro- students with autism to implement pivotal response
vide introductory training in pivotal response treat- treatment in inclusive school settings, using a brief
ment to parents of children with autism. Journal of video feedback treatment package. Focus on Autism
Positive Behavioral Interventions, 12, 23–32. and Other Developmental Disabilities, 26, 105–118.
Noell, G. H., Gresham, F. M., & Gansle, K. A. (2002). Romaniuk, C., Miltenberger, R., Conyers, C., Jenner,
Does treatment integrity matter?: A preliminary N., Jurgens, M., & Ringenberg, C. (2002). The influ-
investigation of instructional implementation and ence of activity choice on problem behaviors main-
mathematics performance. Journal of Behavioral Edu- tained by escape versus attention. Journal of Applied
cation, 11, 51–67. Behavior Analysis, 35, 349–362.
Partington, J. W. (2006). Assessment of Basic Language Rutter, M. (1970). Psychological development: Predic-
and Learning Skills—Revised. Pleasant Hill, CA: Be- tions from infancy. Journal of Child Psychology and
havior Analysts. Psychiatry, 11, 49–62.
Peterson, K. M., Piazza, C. C., Ibañez, V. F., & Fisher, W. Sallows, G. O., & Graupner, T. D. (2005). Intensive
W. (2019). Randomized controlled trial of an applied behavioral treatment for children with autism: Four-
behavior analysis intervention for food selectivity in year outcome and predictors. American Journal on
children with autism spectrum disorder. Journal of Mental Retardation, 110, 417–438.
Applied Behavior Analysis, 52, 895–917. Schmidt, A. C., Hanley, G. P., & Layer, S. A. (2009).
Peterson, K. M., Piazza, C. C., & Volkert, V. M. (2016). A further analysis of the value of choice: Control-
A comparison of a modified sequential oral sensory ling for illusory discriminative stimuli and evaluating
approach to an applied behavior-analytic approach the effects of less preferred items. Journal of Applied
in the treatment of food selectivity in children with Behavior Analysis, 42, 711–716.
autism spectrum disorder. Journal of Applied Behavior Schreck, K. A., Williams, K., & Smith, A. F. (2004).
Analysis, 49, 485–511. A comparison of eating behaviors between children
Petursdottir, A. I., & Carr, J. E. (2011). A review of rec- with and without autism. Journal of Autism and De-
ommendations for sequencing receptive and expres- velopmental Disorders, 34, 433–438.
sive language instruction. Journal of Applied Behavior Schreibman, L., & Koegel, R. L. (2005). Training for
Analysis, 44, 859–876. parents of children with autism: Pivotal responses,
Piazza, C. C. (2008). Feeding disorders and behavior: generalization, and individualization of intervention.
What have we learned? Developmental Disabilities In E. D. Hibbs & P. S. Jensen (Eds.), Psychosocial
Research Reviews, 14, 174–181. treatments for child and adolescent disorders: Empiri-
Piazza, C. C., Fisher, W., Chinn, S., & Bowman, L. cally based strategies for clinical practice (pp. 605–631).
(1991). Reinforcement of incontinent stools in the Washington, DC: American Psychological Associa-
treatment of encopresis. Clinical Pediatrics, 30, 28–32. tion.
Piazza, C. C., Fisher, W., Kiesewetter, K., Bowman, L., Seaver, J. L., & Bourret, J. C. (2014). An evaluation of
& Moser, H. (1990). Aberrant sleep patterns in chil- response prompts for teaching behavior chains. Jour-
dren with Rett syndrome. Brain and Development, 12, nal of Applied Behavior Analysis, 47, 777–792.
488–493. Sharp, W. G., Berry, R. C., McCracken, C., Nuhu, N.
Reed, P., Osborne, L. A., & Corness, M. (2007). The N., Marvel, E., et al. (2013). Feeding problems and
real-world effectiveness of early teaching interven- nutrient intake in children with autism spectrum dis-
tions for children with autism spectrum disorder. Ex- orders: A meta-analysis and comprehensive review of
ceptional Children, 73, 417–433. the literature. Journal of Autism and Developmental
Reichow, B. (2012). Overview of meta-analyses on early Disorders, 43, 2159–2173.
intensive behavioral intervention for young children Skinner, B. F. (1957). Verbal behavior. New York: Apple-
with autism spectrum disorders. Journal of Autism and ton-Century-Crofts.
Developmental Disorders, 42, 512–530. Smith, I. M., Flanagan, H. E., Garon, N., & Bryson, S.
Reichow, B., Volkmar, F. R., & Cicchetti, D. V. (2008). E. (2015). Effectiveness of community-based early
Development of the evaluative method for evaluating intervention based on pivotal response treatment.
and determining evidence-based practices in autism. Journal of Autism and Developmental Disorders, 45,
Journal of Autism and Developmental Disorders, 38, 1858–1872.
1311–1319. Smith, T. (1999). Outcome of early intervention for
Reichow, B., & Wolery, M. (2009). Comprehensive children with autism. Clinical Psychology: Science and
synthesis of early intensive behavioral interventions Practice, 6, 33–49.
for young children with autism based on the UCLA Smith, T., Eikeseth, S., Klevstrand, M., & Lovaas, O. I.
Young Autism Project. Journal of Autism and Devel- (1997). Intensive behavioral treatment for preschool-
opmental Disorders, 39, 23–41. ers with severe mental retardation and pervasive
Roane, H. S., Fisher, W. W., & Carr, J. E. (2016). Ap- developmental disorder. American Journal on Mental
plied behavior analysis as treatment for autism spec- Retardation, 102, 238–249.
trum disorder. Journal of Pediatrics, 175, 27–32. Smith, T., Groen, A., & Wynn, J. (2000). Randomized
Robinson, S. E. (2011). Teaching paraprofessionals of trial of intensive early intervention for children with
Behavioral Treatment of Autism Spectrum Disorder 407
pervasive developmental disorder. American Journal Toussaint, K. A., Kodak, T., & Vladescu, J. C. (2016).
on Mental Retardation, 105, 269–285. An evaluation of choice on instructional efficacy and
Spector, V., & Charlop, M. H. (2018). A sibling-me- individual preferences among children with autism.
diated intervention for children with autism spec- Journal of Applied Behavior Analysis, 49(1), 170–175.
trum disorder: Using the natural language paradigm Virues-Ortega, J. (2010). Applied behavior analytic in-
(NLP). Journal of Autism and Developmental Disor- tervention for autism in early childhood: Meta-anal-
ders, 48(5), 1508–1522. ysis, meta-regression and dose-response meta-analysis
Stahmer, A. C., Collings, N. M., & Palinkas, L. A. of multiple outcomes. Clinical Psychology Review, 30,
(2005). Early intervention practices for children with 387–399.
autism: Descriptions from community providers. Volkert, V. M., & Piazza, C. C. (2012). Pediatric feeding
Focus on Autism and Other Developmental Disabilities, disorders. In P. Sturmey & M. Hersen (Eds.), Hand-
20, 66–79. book of evidence-based practice in clinical psychology
Stokes, T. F., & Baer, D. M. (1977). An implicit tech- (pp. 323–338). Hoboken, NJ: Wiley.
nology of generalization. Journal of Applied Behavior Volkmar, F. R., Lord, C., Bailey, A., Schultz, R. T., &
Analysis, 10, 349–367. Klin, A. (2004). Autism and pervasive developmen-
Stokes, T. F., Baer, D. M., & Jackson, R. L. (1974). Pro- tal disorders. Journal of Child Psychology and Psychia-
gramming the generalization of a greeting response try, 45, 135–170.
in four retarded children. Journal of Applied Behavior Walker, G. (2008). Constant and progressive time delay
Analysis, 7, 599–610. procedures for teaching children with autism: A lit-
Sundberg, M. L. (2008). Verbal Behavior Milestones As- erature review. Journal of Autism and Developmental
sessment and Placement Program. Concord, CA: AVB Disorders, 38, 261–275.
Press. Williams, G., Oliver, J. M., Allard, A. M., & Sears, L.
Sundberg, M. L., & Michael, J. (2001). The benefits (2003). Autism and associated medical and familial
of Skinner’s analysis of verbal behavior for children factors: A case control study. Journal of Developmen-
with autism. Behavior Modification, 25, 698–724. tal and Physical Disabilities, 15, 335–349.
Sundberg, M. L., & Partington, J. W. (1998). Teaching Wolery, M., & Gast, D. L. (1984). Effective and efficient
language to children with autism or other developmental procedures for the transfer of stimulus control. Topics
disabilities. Pleasant Hill, CA: Behavior Analysts. in Early Childhood Special Education, 4, 52–77.
Sundberg, M. L., & Partington, J. W. (1999). The need Wunderlich, K. L., Vollmer, T. R., Donaldson, J. M., &
for both discrete trial and natural environment lan- Phillips, C. L. (2014). Effects of serial and concurrent
guage training for children with autism. In P. M. training on acquisition and generalization. Journal of
Ghezzi, W. L. Williams, & J. E. Carr (Eds.), Autism: Applied Behavior Analysis, 47, 1–15.
Behavior analytic perspectives (pp. 139–156). Reno, Zwaigenbaum, L., Bauman, M. L., Choueiri, R., Kasari,
NV: Context Press. C., Carter, A., Granpeesheh, D., et al. (2015). Early
Tiger, J. H., Hanley, G. P., & Hernandez, E. (2006). An intervention for children with autism spectrum disor-
evaluation of the value of choice with preschool chil- der under 3 years of age: Recommendations for prac-
dren. Journal of Applied Behavior Analysis, 39, 1–16. tice and research. Pediatrics, 136(Suppl. 1), S60–S81.
CHAP TER 24
Behavioral Pediatrics
Integrating Applied Behavior Analysis
with Pediatric Medicine
Patrick C. Friman
Few children in the United States have a men- of pediatrics that focuses on the relation between
tal health care provider, but all have a medical behavior and pediatric health care (Blum & Fri-
health care provider. The medical care provided man, 2000; Christophersen, 1982; Friman, 2005a,
falls under the general rubric of pediatric primary 2008; Friman & Blum, 2003). The central mes-
care, and a pediatric primary care provider is the sage of this chapter is that integrating ABA and
first professional a caregiver concerned about his behavioral pediatrics extends the scope of ABA,
or her child’s behavior problems contacts. About expands effective practice, and improves pediatric
one-fourth of the children seen in primary care health care for children.
have symptoms that meet criteria for a behav- Behavioral pediatrics is a broad field that in-
ioral or emotional disorder (Costello, Edelbrock, cludes four primary domains of research and prac-
et al., 1988; Horwitz, Leaf, Leventhal, Forsyth, tice: (1) the evaluation and treatment of high-
& Speechley, 1992), and another 40% or more frequency, low-intensity (routine) child behavior
may exhibit subclinical behaviors or emotions problems presenting in primary health care set-
that cause their caregivers concern (Costello & tings; (2) the influence of physiological variables
Shugart, 1992). These proportions are found in on child behavior problems; (3) the influence of
urban (e.g., Costello, Edelbrock, et al., 1988) and behavioral variables on child medical problems;
rural (Polaha, Dalton, & Allen, 2011) settings. As and (4) contextual variables that are central to the
a result, researchers have referred to pediatricians first three domains. The high prevalence of be-
as “gatekeepers” for child mental health services havioral problems presenting in pediatric settings,
(Dulcan et al., 1990; Costello, Burns, et al., 1988), and the increasing recognition of the reciprocal
and to the locus of their practices as “de facto relationship between medical concerns and child
mental health settings” (Green et al., 2017; Jens- behavior problems, have led to dramatic growth in
sen, Buttenheim, & Fiks, 2019; Regier, Goldberg, behavioral pediatrics over the past 40 years. ABA
& Taube, 1978; cf. Polaha et al., 2011). Applied has made significant contributions over that time,
behavior analysis (ABA) is a powerful science that and these, coupled with the continuing growth of
specializes in behavior, its analysis, and the devel- behavioral pediatrics, provide multiple opportuni-
opment of methods for influencing it in socially ties for applied behavior analysts to work in child
adaptive ways. Behavioral pediatrics is the branch health care settings.
408
Behavioral Pediatrics 409
RELATIONSHIP BETWEEN PEDIATRICS AND ABA child’s problem, stigma results from not contacting
a pediatric care provider about a child’s problem.
As indicated above, primary care providers (sever- Therefore, the pediatrician is almost always the
al subspecialties provide primary care for children first professional to learn of child behavior prob-
but the most common by far is pediatrics; hereafter, lems. Most of these are low-intensity, high-fre-
I use the term pediatrician to refer to the providers quency problems that are not necessarily represen-
and pediatrics to the settings) are the profession- tative of true pathology. But by caregivers’ reports,
als who are most likely to provide initial inter- they are often disruptive to families and difficult to
ventions for children’s behavioral or emotional solve (Christophersen, 1982; Earls, 1980; Friman,
problems (Christophersen, 1982; Costello, Burns, 2005a, 2008). Additionally, a significant percent-
et al., 1988; Costello, Edelbrock, 1988; Dulcan et age of them will deteriorate into more serious con-
al., 1990). They are most likely to use supportive ditions if left unsolved. A classic example involves
counseling, prescriptive behavioral treatment, or bedtime problems.
referral (Blum & Friman, 2000; Friman & Blum,
2003; Friman, 2005a, 2008). ABA has contributed
Bedtime Problems
substantially to the development, implementa-
tion, and evaluation of pediatric interventions, Teaching children to go to bed, go to sleep, and
especially prescriptive behavioral treatment, for stay asleep throughout the night is difficult for
many of the behavior problems children present many families in mainstream North American
initially in primary care (Christophersen, 1982; culture, in that at least 30% of families contend
Friman, 2005a, 2008; Friman & Blum, 2003). with sleep problems three or more nights a week
The cardinal principle informing these interven- (Friman & Schnoes, 2020; Lozoff, Wolf, & Davis,
tions—that behavior is influenced by its current 1985). The difficulties caregivers report include
and historical circumstances—is familiar to and bedtime struggles like resistance to going to bed;
accepted by most pediatricians. Thus primary care fussing and crying while in bed; and night waking
pediatricians have incorporated many of these in- with fussing, crying, and unauthorized departures
terventions into practice, including interventions from the bedroom. Pediatricians often address
for child discipline, incontinence, sleep disorders, these problems by prescribing soporific drugs, but
habit disorders, and symptoms of attention-deficit/ these medications produce side effects, and treat-
hyperactivity disorder (e.g., Blum & Friman, 2000; ment gains are often lost when the medication is
Blum, Williams, Friman, & Christophersen, 1995; withdrawn (Christophersen & Mortweet, 2013;
Christophersen & Friman, 2010; Christophersen Edwards & Christophersen, 1994).
& Mortweet, 2013; Friman, 2002; Friman &
Schmitt, 1989).
Behavioral Treatment for Bedtime Problems
In this chapter, I discuss four problems that be-
havior analysts treat, study, or treat and study, one The primary component of the most effective
representing each of the four primary domains of behavioral interventions for bedtime problems
behavioral pediatrics: (1) bedtime struggles, repre- involves one of the first documented and most fre-
senting routine behavior problems; (2) constipa- quently used ABA-informed procedures: extinc-
tion and retentive encopresis, representing physi- tion. As children develop sleep habits, they often
ological influences on behavior; (3) nocturnal learn to associate specific environmental factors
enuresis, representing behavioral influences on with self-quieting and the induction of sleep. Mis-
medical problems, and (4) adherence to medical informed caregiver efforts to help children sleep
regimens, representing contextual variables cen- (e.g., soothe, cuddle, or lie down with the child
tral to problems in the first three domains. until sleep onset occurs) often result in problem-
atic sleep associations that mitigate the process of
falling asleep. Unfortunately, when such a care-
ROUTINE BEHAVIOR PROBLEMS giver is absent at bedtime, the child is left with-
out the stimulus that is most powerfully associated
The pediatrician is one of the most trusted profes- with sleep. The child’s response to the caregiver’s
sionals in the United States. Almost all families absence typically involves prolonged and intensive
have pediatric care providers for their children— crying that resembles an extinction burst (Blamp-
and, contrary to the stigma that results from ied & Bootzin, 2013; Edwards & Christophersen,
contacting a mental health care provider about a 1994; Ferber, 2006; Friman, 2005b; Friman &
410 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Schnoes, 2020). This response usually motivates and a large amount of clinical experience that
the caregiver to intervene either by further sooth- Ferber (2006) has described suggest that very few
ing or by disciplining the child—both of which, do. Although the mechanism responsible for the
unfortunately, usually worsen the problem. Sooth- effectiveness of graduated extinction is unknown,
ing responses to crying can reinforce it, disciplin- one possible explanation is that increasing the re-
ary responses to crying often provoke more crying, sponse requirement to 45 minutes of crying may
and both caregiver responses interfere with the lean the schedule so much that the reinforcing ef-
child’s learning to self-quiet (Blampied & Bootzin, fects of sleep supersede the reinforcing effects of
2013; Friman, 2005b; Friman & Schnoes, 2020; caregiver visitation.
Lozoff et al., 1985; Schnoes & Reimers, 2009). Not
surprisingly, after failed attempts to solve the prob- Positive Routines. The positive-routines proce-
lem themselves, caregivers whose children exhibit dure involves a hybrid of extinction and a reinforc-
bedtime problems often ask their pediatricians for ing bedtime ritual. In this procedure, a caregiver
advice. I have described the four procedures pro- decides on a preferred bedtime for the child and
fessionals are most likely to prescribe below, which establishes the time at which the child typically
were all ABA-derived. falls asleep. Beginning shortly before the time the
child typically falls asleep, the caregiver engages
Extinction. The extinction approach to bed- the child in several quiet activities lasting no lon-
time problems involves no visits by the caregiver ger than 20 minutes total. During the activities,
to the child’s bedroom after the child has gone to the caregiver issues easily followed instructions
bed. In effect, the child is left to cry it out. Gen- and richly supplies reinforcement for compliance.
erally, extinction works more rapidly than other These are followed by the terminal instruction
approaches, but it presents problems that mitigate “Now get in bed and go to sleep,” or something
its overall effectiveness: (1) Crying can be highly equivalent—a procedure consistent with the high-
aversive to caregivers, especially during the first probability instructional sequence in research on
nights of treatment; (2) crying and screaming can behavioral momentum (Mace et al., 1988). If the
draw attention from the neighbors, with predict- child leaves the bed at any time after the comple-
ably problematic consequences; and (3) extended tion of the routines and the terminal instruction,
crying and screaming differentially affects care- the caregiver places him or her back in bed, telling
givers, which can cause marital/couple discord the child that the routine is over and it is time
(Adams & Rickert, 1989; Edwards & Christo- for bed. The caregiver ignores crying or verbal-
phersen, 1994; Rickert & Johnson, 1988). Thus izations. At specified intervals (e.g., 1 week), the
extinction is a straightforward behavioral ap- caregiver moves the positive routine back in time
proach to child bedtime problems that has limited 5–10 minutes. This backward movement continues
social validity. To improve social validity, sleep until the caregiver arrives at the preferred bedtime
researchers have developed other multicomponent for the child, which can take 6–8 weeks or more.
methods that employ extinction but include other Experimental comparison of the positive-routines
procedures to decrease its intensity and aversive- procedure with scheduled extinction showed that
ness for caregivers. both improved bedtime behavior for children, but
the caregivers using positive routines reported sig-
Graduated Extinction. Graduated extinction nificantly improved marital relations, suggesting a
involves advising caregivers to ignore bedtime more socially valid procedure (Adams & Rickert,
problem behavior for specific time intervals that 1989).
gradually increase, usually beginning with a
5-minute interval at the first episode, 10 minutes The Bedtime Pass. The bedtime pass program
at the second, and 15 minutes for subsequent epi- involves (1) requiring the child to get into bed;
sodes on Night 1 (e.g., Adams & Rickert, 1989). (3) providing the child with a small object (e.g., a
These intervals increase over the course of a week, laminated note card) exchangeable for one “free”
ending with 35 minutes for the first episode on trip out of the bedroom or one visit by the care-
Night 7, 40 minutes for the second, and 45 minutes giver after being put to bed to satisfy an accept-
for all subsequent episodes and nights. Although able request (e.g., for a drink, hug, or visit to the
children can have tantrums for longer than 45 bathroom); (3) having the child surrender the pass
minutes at night, research (Adams & Rickert, after using it; and (4) using extinction thereafter
1989; see also Edwards & Christophersen, 1994) (Schnoes, 2011). In the initial study, the program
Behavioral Pediatrics 411
eliminated the high rates of crying out, calling out, logical variable that can cause toileting problems,
and coming out of the bedroom after bedtime that ranging from resistance to a bona fide diagnostic
two children (ages 3 and 10 years) exhibited. Ad- category known as retentive encopresis.
ditionally, caregivers achieved these successful re-
sults without an accompanying “extinction burst”
during initial intervention periods, and a large
Retentive Encopresis
group of sample caregivers rated the intervention Definition
as more acceptable than extinction alone (Friman
Functional encopresis, a common presenting
et al., 1999). Investigators replicated results of the
complaint in pediatrics (representing 3–5% of
bedtime pass program in a single-subject analysis
of four 3-year-old children (Freeman, 2006) and a referrals), is a disorder in which children either
randomized trial involving 19 children ages 3–6 voluntarily or involuntarily pass feces into or onto
years (Moore, Fruzetti, & Friman, 2007). The bed- an inappropriate location, usually their clothing
time pass may achieve its effectiveness through (Christophersen & Friman, 2010; Friman, 2017,
differential reinforcement of alternative behavior 2019). Encopresis is not diagnosed if the problem
(Vollmer & Iwata, 1992), in which the request and is exclusively due to an anatomical or neurologi-
surrender of the pass allow the child to access po- cal abnormality that prevents continence. The
tent bedtime reinforcers, and the caregiver places current criteria from the Diagnostic and Statistical
bedtime problems on extinction. Manual of Mental Disorders, fifth edition (DSM-
5; American Psychiatric Association, 2013) are as
This brief discussion of child bedtime problems follows: (1) inappropriate passage of feces at least
and their treatment is by no means complete. It once a month for at least 3 months; (2) chrono-
merely involves four ABA-derived interventions logical or developmentally equivalent age of at
that professionals use most frequently and have least 4 years; and (3) not due exclusively to the
the most empirical support (for other interven- direct physiological effects of a substance (e.g., a
tions, see Burke, Kuhn, & Peterson, 2004; Fri- laxative) or a general medical condition, except
man, 2005b; Friman & Schnoes, 2020; Honaker through a mechanism involving constipation.
& Meltzer, 2014). DSM-5 distinguishes two subtypes of encopresis:
one with constipation and overflow incontinence,
and one without these symptoms. I focus on en-
INFLUENCE OF PHYSIOLOGICAL VARIABLES copresis with constipation and overflow, because
ON CHILD BEHAVIOR PROBLEMS this subtype has a physiological cause (i.e., consti-
pation). The etiology of encopresis without con-
The fundamental assumption of ABA is that be- stipation remains unknown (Beaudry-Bellefeuille,
havior occurs as a function of environmental cir- Booth, & Lane, 2017).
cumstances, but this position does not exclude the
influence of physiological variables. Rather, many Relevant Physiology
physiological variables may be causal but are the
results of environmental contingencies that oc- The large intestine or colon is the distal end of
curred in a phylogenetic context (Skinner, 1966). the alimentary tract, which is composed sequen-
Physiological variables often play an initiating tially of the esophagus, stomach, small intestine,
role in behavior problems that present in pediat- and colon. I provide a rudimentary description of
rics. For example, child stomach pain caused by the system here, because behavior analysts should
physiological variables (e.g., a flu virus) can lead to understand the physiology that supplies the logic
missing school. While home from school, reinforc- of effective treatment (for more thorough reviews,
ers such as avoidance of schoolwork and contact see Weinstock & Clouse, 1987; Whitehead &
with sympathetic responses from caregivers influ- Schuster, 1985). The colon is a tube-shaped organ
ence the child’s behavior. These influences can with a muscular wall. It connects the small intes-
in turn result in complaints of stomach pain that tine to the rectum and anus. It has three primary
do not involve physiological variables, a condition functions: fluid absorption, storage, and evacua-
sometimes referred to as recurrent abdominal pain tion. Extended storage and planned evacuation are
(Finney, Lemanek, Cataldo, Katz, & Fuqua, 1989). the defining features of fecal continence. Muscu-
There are many other examples, but the one I lar contractions of the colon walls, called peristal-
want to discuss here is constipation—a physio- sis, produce a wave-like motion that moves waste
412 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
through the colon; various external events (e.g., a tion, or a combination of these factors (Meunier,
meal, moving about) stimulate these movements. Marechal, & De Beaujeu, 1979; Wald, Chandra,
The colon absorbs moisture from the waste that Chiponis, & Gabel, 1986). The combined effects
moves through it, creating semisolid feces. of these factors are a lowered probability of volun-
The rectum, a hollow receptacle at the distal tary stool passage and a heightened probability of
end of the colon, usually contains little or no feces fecal retention.
until muscular contractions in the colonic wall Chronic fecal retention causes fecal impaction,
propel feces into it, which produces distension. which enlarges the colon, produces decreased mo-
Distension stimulates sensory receptors in the rec- tility of the bowel system, and occasionally results
tal mucosa and in the muscles of the pelvic floor, in involuntary passage of large stools and soiling
resulting in relaxation of the internal sphincter, due to seepage of soft fecal matter. Physicians
which facilitates defecation. This process is invol- often refer to the seepage as paradoxical diarrhea,
untary, but a child can constrict the anal canal because the children retain large masses of stool
and inhibit defecation by contracting the external and are functionally constipated, but their colon
anal sphincter and the functionally related pu- allows passage of soft stool around the mass, which
borectalis muscle. When the child suppresses the results in diarrhea (Levine, 1982). The relation of
urge to defecate, the rectum accommodates the fecal impaction to encopresis is well established,
retained stool through the adaptive pliance of its and 80% of patients show fecal impaction accom-
structure and terminates the reflex relaxation of panying fecal incontinence at the first clinic visit
the internal sphincter. The urge gradually decays, via clinical exam and 90% on X-ray of the lower
and some of the fecal matter in the rectum returns abdomen (Davidson, 1958; Levine, 1982).
to the descending colon by retroperistalsis.
Evaluation
Etiology
Either before or directly after the initial visit, the
Physicians can trace between 80 and 95% of en- behavior analyst should refer a child with encopre-
copresis cases to a primary causal variable, con- sis to the pediatrician for a medical examination;
stipation (Hatch, 1988; Levine, 1982). Although this usually includes a routine check of history,
definitions for constipation vary, children who abdominal palpation, rectal examination, and
frequently go 2 or more days without a bowel sometimes an X-ray of the abdomen to determine
movement are probably prone to constipation. the extent of fecal impaction. A barium enema
Caregivers of children with encopresis common- is rarely necessary unless features of the exam
ly complain that the children deliberately soil suggest Hirschsprung’s disease. Rare anatomical
their clothing, but this attribution is usually false and neurological problems can cause fecal reten-
(Levine, 1982). The primary cause of soiling is tion and soiling; neurological problems include
fecal retention (constipation), which is generally Hirschsprung’s disease, and anatomical defects in-
not caused by characterological or psychopatho- clude a variety of malformations and locations of
logical problems (Friman, 2002; Friman, Mathews, the anus that a physical exam can detect and that
Finney, & Christophersen, 1988; Gabel, Hegedus, require medical management (Hatch, 1988).
Wald, Chandra, & Chaponis, 1986). Retention In addition to routine behavior assessments, the
is usually the result of a constellation of factors, behavioral interview for encopresis should include
many of which are beyond a child’s immediate questions related to constipation. These include
control (Levine, 1982). These factors include a asking whether (1) there is ever a long period be-
constitutional predisposition (i.e., slow gastroin- tween bowel movements; (2) bowel movements are
testinal transit time); diet; insufficient leverage for atypically large (e.g., stop up the toilet); (3) fecal
passage of hard stools; and occasional or frequent matter ever has an unusually foul odor; (4) fecal
painful passage of hard stools, resulting in negative matter is ever hard, difficult, or painful to pass; (5)
reinforcement for holding stools. In rare cases, re- the child ever complains of not being able to feel
tention may be related to sexual abuse. Some chil- the movement or make it to the toilet on time; and
dren—especially those with extreme constipation, (6) the child ever hides soiled underwear. An af-
treatment failure, or both—have an increased firmative answer to one or more of these questions
threshold of awareness of rectal distension, a possi- is highly suggestive of retentive encopresis, and
bly weak internal sphincter, a tendency to contract hiding underwear suggests a history that includes
the external sphincter during the act of defeca- some form of punishment.
Behavioral Pediatrics 413
Encopresis is not well understood outside the trials (e.g., Stark et al., 1997). For example, in a
medical community, and characterological and study of 58 children with encopresis, 60% were
psychopathological interpretations prevalent in completely continent after 5 months, and those
Western culture are likely to influence a caregiver’s who did not achieve full continence averaged a
interpretation of the condition, which may influ- 90% decrease in accidents (Lowery, Srour, White-
ence how the child views the problem. The behav- head, & Schuster, 1985). However, not all children
ior analyst can begin encopresis treatment during succeed with the conventional approach, and re-
the evaluation by providing accurate information searchers have developed augmentative methods
that “demystifies” the problem. Lastly, the evalua- for these children. In a manner typical of ABA,
tion should include questions about diet and tim- developing these augmentative methods began
ing of meals. Low-fiber diets and irregular meals with the study of behaviors associated with treat-
can contribute to encopresis (Koppen et al., 2016). ment failure (Stark, Spirito, Lewis, & Hart, 1990).
Incorporating behavior management methods rel-
evant to the behaviors, teaching caregivers to use
Treatment
them, and delivering treatment in a group format
During the past 40 years, several descriptive and produced an 83% decrease in accidents in 18 treat-
controlled experimental studies have supported a ment-resistant children with encopresis, and the
multicomponent approach to treatment of chronic children maintained decreased accidents or even
retentive encopresis, partly derived from the pio- improved at a 6-month follow-up (Stark, Owens-
neering work of Davidson (1958), Christophersen Stively, Spirito, Lewis, & Guevremont, 1990).
and Rainey (1976), Levine (1982), and Wright
(1975). The behavior analyst can address the first The general premise of this section—that phys-
component during the evaluation by demystifying iological variables can influence or cause behavior
the entire elimination process and its disordered problems—is not controversial even within ABA,
manifestations. Generally, this means providing a science dedicated to environmental variables.
information about bowel dynamics and the rela- Constipation is one such variable, and there are
tion of the problem to constipation. Second, the many others (e.g., anorexia due to gastroesophage-
caregiver should remove fecal impaction with al reflux, restricted activities due to pain). Because
enemas, laxatives, or both under the direction of of the physiological component of these problems,
the pediatrician. Third, the child should sit on serious health consequences of unsuccessful treat-
the toilet for about 5 minutes once or twice a day. ment often compound the behavioral components
Fourth, the caregivers should promote proper toi- that contribute to the problem. For example, ex-
leting with encouragement and not with coercion. treme fecal retention can be life-threatening, and
Additionally, they should not reserve their praise even routine cases can seriously decrease social
and affection for proper elimination; caregivers standing and increase social isolation in affected
should provide praise for just sitting on the toi- children. Because of the behavioral components
let. Fifth, caregivers should give the child a stool of these problems, a solely medical intervention is
softener like mineral oil or MiraLAXTM under the insufficient for effective treatment. A method that
pediatrician’s direction, to ease the passage of hard delivers or aids the delivery of the medical compo-
stools. Sixth, the caregiver should increase the nents of treatment while addressing the behavioral
child’s dietary fiber. Seventh, the caregiver should components is needed, and thus behavioral pedi-
arrange for and encourage the child to increase his atrics is an ideal context. Additionally, although
or her activity level and fluid intake, to increase many types of behavior problems stem from physi-
and maintain colon motility. Eighth, the child’s ological influences, the most frequently occurring
feet should be on a flat surface during toileting problem is some form of noncompliance with a
episodes. Foot placement is crucial to the Valsalva treatment regimen for the physiological dimen-
maneuver, which is the grunting push necessary to sions. As an example, cooperation with prescribed
produce a bowel movement. Ninth, the caregiver treatment for encopresis is so necessary for success
should reward the child for bowel movements in that instructional control training is frequently a
the toilet. See Christophersen and Friman (2010) component of treatment (Christophersen & Fri-
and Friman (2019) for reviews discussing these man, 2010; Friman, 2017, 2019). Furthermore, al-
recommendations in greater depth. though researchers have made progress in improv-
The literature on this approach or variations ing pediatric compliance, it remains one of the
thereof has progressed sufficiently to lead to group most chronic problems in pediatric medicine (see
414 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
the discussion of adherence below). Therefore, this feedback that is used readily in primary care set-
domain of behavioral pediatrics provides many op- tings (Christophersen & Friman, 2010; Friman,
portunities for applied behavior analysts interested 2017, 2019). Below I briefly review nocturnal en-
in working in pediatric health care settings. uresis, its relevant physiology, and its alarm-based
treatment.
a skill that would be difficult to teach without the (Culbert et al., 1996). Additionally, evidence that
sensory feedback the alarm provides. biofeedback can generate operant responses that
The enuresis alarm produces cures slowly, and lead to control over physiological processes long
during the first few weeks of alarm use, waking thought to be outside volition, like skin tempera-
occurs only after a complete voiding if it occurs ture and blood pressure, is mounting. For example,
at all. One study using the size of the urine stain researchers have used verbally based awareness en-
on the soiled sheets as the dependent measure hancement methods to alter the level of mediators
showed that before accident-free nights, the stain of the immune system in saliva (Olness, Culbert,
grew increasingly smaller on successive nights, & Uden, 1989) and to decrease the recurrence of
suggesting a gradual process of continence attain- chronic mouth ulcers (Andrews & Hall, 1990).
ment (Ruckstuhl & Friman, 2003). In other words, Collectively, the large body of research document-
the feedback properties of the alarm gradually but ing the effectiveness of the urine alarm, and the
inexorably strengthened the skills necessary to even larger literature on the effectiveness of bio-
avoid it. The core skill involves contraction of the feedback treatment for a broad range of medical
pelvic-floor muscles, causing sustained elevation of conditions, underscore the research and clinical
the bladder neck, which stops or prevents urina- potential represented by the influence of behav-
tion. Increased sensory awareness of urinary need ioral variables on physiology. This potential, in
and waking to urinate are possible outcomes, but turn, represents an excellent opportunity for ap-
are less likely and actually inferior to sustained, plied behavior analysts interested in working in
accident-free sleep throughout the night. pediatric settings.
Increasing sensory awareness of urinary need
before daytime accidents, however, is a key compo-
nent in the most empirically supported treatment CONTEXTUAL VARIABLES
for diurnal enuresis. Only two studies are avail-
able, as researchers have studied diurnal enuresis A broad range of contextual variables influence
minimally, and the first used a much simpler con- the effectiveness of health care and its delivery.
ceptualization (Halliday, Meadow, & Berg, 1987). These variables are not specific to any particular
Specifically, this early study merely suggested that problem that presents in a health care delivery sys-
the alarm served as a reminder for urination. A tem, but they have the capacity to influence all of
colleague and I (Friman & Vollmer, 1995) con- them. For example, doctor communication is cen-
ducted a subsequent study using the biofeedback tral, but not specific, to care for problems treated
conceptualization with a young girl who was ini- in health care settings, although its role can vary
tially unresponsive to urinary urge and onset, but depending on the problem at hand. For example,
who rapidly became responsive with use of the effective communication may be less critical for
alarm. The decreasing latency between alarm treatments like “the tincture of time” (e.g., “Let’s
onset and appropriate response was characteristic just watch this for a while and see if the child grows
of learning curves during alarm-based treatment out of it”) than it would be for a complex medical
for nocturnal enuresis and biofeedback treatments intervention. A directly related contextual vari-
in general. able is patient communication. For example, if
patients incompletely report symptoms or do not
Most biofeedback treatments are much more ask critical questions, the care they receive may be
technically complex than the urine alarm, and inadequate (e.g., Finney et al., 1990). I have chosen
clinicians use them for a broad range of physi- treatment adherence for this chapter as the exem-
ologically based behavioral concerns that often plar of a contextual variable that affects care.
initially present in pediatric settings. Among the
physiological processes that we can monitor are
muscle tension, skin temperature, respiratory rate, Treatment Adherence
blood pressure, and skin moisture (perspiration)
Definition
(Friman, 2009). Researchers and clinicians have
used biofeedback devices sensitive to these pro- Treatment adherence is so central to pediatric
cesses in treating a variety of disorders, including medicine that this chapter probably should have
headaches, other varieties of chronic pain, asth- begun with it. True, routine behavior problems
ma, bruxism, anxiety disorders, sleep disorders, are very frequently concerns in pediatric medi-
and dysfunction of the autonomic nervous system cine, but treatment adherence is always a concern
Behavioral Pediatrics 417
because it begins with adhering to the pediatric ents in pediatrics, and there are empirically sup-
appointment itself. Treatment effectiveness is a ported treatment protocols for its most commonly
moot issue if families do not keep appointments presenting forms (enuresis and encopresis). These
or do not follow prescribed treatments. Estimated treatments include information on relevant physi-
rates of nonadherence hover around 50% for both ology, diet, toileting schedules, behavioral contin-
psychological (Kazdin, 1996) and medical (Rapoff, gencies, activity levels, and caregiver involvement.
2010) services. Although treatment adherence is To maximize integrity, a clinician would verbally
problematic across clinical populations, it is par- deliver information covering these details and sup-
ticularly challenging with children and adoles- plement the delivery with supportive instructional
cents, because there are usually at least two sourc- aids. Other examples include regimens for chronic
es of nonadherence: the children/adolescents and diseases and empirically supported treatment pro-
their caregivers (Watson, Foster, & Friman, 2006). tocols for behavior disorders.
Thus establishing acceptable levels of adherence
requires examining child and caregiver variables
Assessment of Adherence
that either facilitate or impede it.
The relevant literature highlights three salient Adherence is an observable and measurable be-
terms: adherence, compliance, and integrity. Com- havior, and thus well suited to behavioral as-
pliance and adherence are basically synonymous, sessment. The primordial adherence behavior
but adherence has gained favor in recent years (see is appointment keeping, and its assessment is
Rapoff, 2010), and I use it here. Adherence and straightforward: Did the patient keep the appoint-
integrity, however, are not synonymous, because ment? The other adherence behaviors to be as-
treatment integrity refers to the fidelity with which sessed are determined by the nature of the problem
a clinician delivers treatment. Adherence refers to to be treated. For acute problems like otitis media
the extent to which patients keep appointments or strep throat, the behaviors of primary concern
and accurately and consistently follow the steps are relatively simple; they typically involve rest
of prescribed treatments. For example, a routine and taking prescribed medication for an abbrevi-
prescription for functional retentive encopresis in- ated time. For chronic diseases, the behaviors of
cludes increases in fluid, fiber, scheduled toilet sits, concern are usually more complex. They typically
stool softeners, and incentives, as well as decreases involve multiple timelines and classes of behavior.
in dairy products, processed foods, coercion, pun- The vast literature on behavioral assessment of-
ishment, and irregular or extended toilet sits. As- fers some guidance on selection of target behav-
sessing treatment adherence involves determining iors that can readily inform how and what to as-
the number of prescribed steps actually followed sess when adherence to a complex regimen for a
(accuracy) and the regularity of applications (con- chronic condition is the focus (e.g., Friman, 2009).
sistency). A complete review of assessment is far beyond
Although patient nonadherence has always the scope of this chapter. I merely provide a cur-
been the primary focus in treatment adherence re- sory description. There are at least four options
search, clinicians and researchers alike now realize for selecting target regimen behaviors. The first
that treatment integrity among clinicians should is the most straightforward, but probably the most
also be a target. Clinician communication about difficult, and this is to measure all behaviors that
treatment has a significant effect on the probabil- are relevant to regimen adherence. The second
ity of adherence, and accuracy and consistency involves targeting only those behaviors that are
are dimensions of the communication. Additional most problematic or aversive to others. For exam-
variables include clinician training and experi- ple, children with enuresis are often less bothered
ence, interpersonal skill, capacity to manifest au- by soiled bedding than their caregivers are. The
thority without instigating resistance, and sense third involves targeting behaviors that are the
of timing. Other not so self-evident and possibly most immediately crucial to a patient’s health and
more manipulable variables include the specificity well-being. For example, compliance with daily
of treatment recommendations; the standardiza- insulin injections is more immediately critical to
tion of treatment protocols; and the provision of health maintenance in diabetes than foot care is.
supportive instructional aids like handouts, audio The fourth involves targeting those behaviors that
and video recordings, and e-mail or web-based are easiest to change; the rationale here is to build
communications. For example, as described above, up momentum for larger changes (cf. Mace et al.,
incontinence is a problem that frequently pres- 1988).
418 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
The next issue involves who does the assessing herence. Most measures are indirect, like assays
and who is assessed—questions particularly im- (see above) or pill counts (see below). But the lim-
portant in pediatrics, because although children ited use of direct observation is not so much an
are the targets of treatment, their caregivers are obstacle as it is an opportunity. The indirect mea-
usually the recipients of the regimen and are al- sures are likely to dominate the field, but direct
ways partly and often fully responsible for carry- observation can supplement them when this is fea-
ing it out. Thus both a child and a caregiver can sible. However, expanding assessment to include
often be the focus of assessment. Additionally, as direct observation will usually require observers
is typical in most branches of medicine, there are other than those on the research team, because
multiple sources of information, including physi- although treatments are prescribed in clinical set-
cians, nurses, therapists, and support staff. For ex- tings, they are usually carried out elsewhere (e.g.,
ample, in our early appointment-keeping research, home, school). The most likely candidates are
the staff at the outpatient check-in desk supplied family members or school personnel, but there are
critical assessment data (Friman, Finney, Rapoff, other possibilities. For example, a study on dietary
& Christophersen, 1985). compliance used camp counselors as observers
The final issue involves how to assess adher- (Lorenz, Christensen, & Pitchert, 1985). Validity
ence, and the literature describes multiple strate- is the primary advantage of direct observations.
gies. Here, I briefly discuss only the most common. They are direct measures of adherence, and they
However, this dimension of adherence assessment accrue all the scientific advantages that behavior
would seem to be limited only by the resources, analysts have touted for decades (e.g., Johnston &
ingenuity, and creativity of investigators. In short, Pennypacker, 2009; Rapoff, 2010). Their disadvan-
how best to assess adherence presents a growth tages primarily involve access. As indicated above,
potential for behavior analysts. The types I list little of the typical treatment regimen is carried
here include drug assays, direct observation, elec- out in the clinic setting. Regimen requirements
tronic monitoring, pill counts, provider estimates, are distributed temporally across the patients’ days
and patient and caregiver reports (Rapoff, 2010). and nights, and situationally across the settings
There are advantages and disadvantages to each. of their lives. The cost of resources necessary to
I then discuss functional assessment of adherence carry out representative direct observations of ad-
and some general concerns about procedures. herence behaviors necessary for most regimens is
prohibitive. Nonetheless, when accessibility is not
Drug Assays. Drug assays range in sophistica- a significant concern, behavior analysts should in-
tion from observing a bioavailable marker that is clude direct observation in adherence assessment.
either part of or added to a drug to various kinds
of testing, ranging from determining simple blood Electronic Monitors. The revolution in elec-
sugar levels (something the patient or patient’s tronic technology that has occurred over the past
caregivers do) to testing for various metabolites few decades has significantly expanded assessment
(something that only a lab technician can do). options for adherence researchers. For example,
However, proper interpretation of a drug assay researchers can now monitor whether a patient
virtually always involves some basic knowledge of retrieves pills or liquids from medication bottles or
clinical pharmacology, especially as it pertains to an inhalant from an inhaler, and researchers can
absorption, distribution, and elimination of medi- store the obtained data electronically and retrieve
cations in (or not in) the body. The advantages it later for analysis. The research on electronic
of assays include objective quantification, clinical monitoring is so extensive that a pediatric psy-
utility, and information on dose–response rela- chology task force has identified it as a well-estab-
tions. Their disadvantages include abbreviated lished measure (Quittner, Modi, Lemanek, Levers-
time horizons, expense, invasiveness, need for ex- Landis, & Rapoff, 2008). The primary advantage
pert readings, and variability of readings due to of electronic monitoring is that it can provide ob-
various causes (e.g., enteric coatings, contents of jective measures of a broad spectrum of adherence
the stomach, presence of other drugs, age, gender, behaviors continuously (if necessary) in real time.
habits). No other method has this advantage. The primary
disadvantage is that data from electronic monitor-
Observation. Although direct observation is ing provide only an indirect measure of adherence
considered a sine qua non for behavior-analytic behavior. A patient may not take pills retrieved
research, it is actually limited in research on ad- from a bottle, may not swallow liquid from a dis-
Behavioral Pediatrics 419
penser, and may dispense mist from an inhaler some well-respected behavior analysts have ac-
into the air rather than inhale it into the lungs. knowledged the potential utility of self-reports and
Thus, independent of other measures indicating have suggested methods for increasing their rigor
that the patient ingested the medicine, the behav- (e.g., reducing bias, heightening validity; Critch-
ior analyst is likely to consider electronic monitor- field, Tucker, & Vuchinich, 1998), behavior-ana-
ing only as a supplementary measure. lytic researchers remain skeptical of their use as a
source of data. Non-behavior-analytic researchers
Pill Counts. The long-standing tradition of studying adherence are much less skeptical, and
using pill counts as a measure of adherence began they have developed multiple self-report methods,
to fade with the advent of electronic monitoring. including global rating scales, structured inter-
Their use is simple: The researcher merely counts views, and daily diaries. As just one example, sleep
the pills remaining in the medication container diaries are a staple of clinical and scientific ap-
and compares it against the prescribed regimen. proaches to pediatric sleep problems (e.g., Ferber &
Obviously, this is a primitive method, compared to Kryger, 1995). The behavior analyst contemplat-
the real-time data available from electronic moni- ing research that involves assessment of adherence
toring. Thus, although pill counts are a measure should consider coupling self-report measures with
of adherence behavior and any measure is better the more objective measures favored in behavior
than no measure, they share none of the advan- analysis. Doing so could expand the scope of the
tages of electronic monitoring and all their disad- research and the possibility of having behavior-
vantages. analytically oriented adherence research accepted
in mainstream medical journals. Although the re-
Provider Estimates. Obtaining provider esti- search (or at least a portion of it) could strain the
mates typically involves asking medical provid- credulity of behavior analysts, the advantage of
ers to complete scales that assess adherence. The exporting the behavior-analytic dimension of the
most common involve Likert-type scales whose research into mainstream medicine could—and
items pertain to judgments about the likelihood would, in my opinion—be worth the cost.
adherence has occurred and its extent. In a more
primitive form, providers merely answer yes or no Functional Assessment. There is little mention
when asked whether they believe a patient has or of function in the adherence literature. Once
will follow a regimen. The primary advantage of again, behavior analysts should see this as an op-
provider estimates is feasibility. They require little portunity rather than as a reason to avoid this
effort or expense. A small amount of evidence important area of investigation. Although what
suggests that they are more accurate than global follows is elementary to behavior analysts, it bears
ratings from patients and their caregivers. The mentioning nonetheless: Determining the func-
disadvantages will be obvious to behavior ana- tion of a behavior, whether it involves self-injury
lysts. Provider estimates merely involve asking for or nonadherence, is a well-established method
providers’ opinions, and opinions, even from well- for designing treatments that are informed by the
established experts, are subject to bias and other identified functions.
well-known threats to validity (Rapoff, 2010).
Moreover, outside of some forms of social valid- General Concerns about Procedures. There are
ity, opinions are wholly unsatisfactory as a primary multiple issues to address in the pursuit of adher-
measure for behavior analysts. ence research that I only briefly mention here, but
the brevity of my remarks is inversely related to
Patient and Caregiver Reports. Patient and care- their importance. Beyond limited space, the rea-
giver reports are merely variations on self-reports— son for brief mention of these concerns is that
a method fundamental to research in mainstream they are central to all methods deployed to mea-
psychology, psychiatry, some related social sci- sure human behavior. The first is reactivity. All
ences, and clinical medical research. Yet, outside measurement systems are reactive, and the portion
of social validity, behavior analysts virtually never of data resulting from the reactivity is not reflec-
use them as a primary source of assessment data. tive of the behavior being measured. Second, the
The reasons are immediately evident to behavior measures employed must be representative of that
analysts: Behavior is the subject under consider- behavior (in this instance, adherence). For exam-
ation, and reports of behavior are only considered ple, pills missing from a pill container are not nec-
as behavior, not as measures of behavior. Although essarily an accurate representation of medication
420 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
consumed. Direct measures, when possible and The point for behavior analysts is that health
feasible, are always preferred over indirect mea- education is a primordial component of pediatric
sures, because they more completely and accurate- medicine, and that one of its central purposes is to
ly capture the essence of adherence. For example, promote adherence. The extent to which it does
directly observing a patient walking into a clinic so in its currently employed forms, the extent to
is superior to asking a caregiver whether his or her which those forms could be improved, and the de-
child kept his or her appointment. Standards of velopment of new behaviorally informed methods
measurement, especially reliability and validity, are all rich opportunities for behavior-analytic re-
are essential considerations in all measurement search.
systems, and assessment of adherence is no excep-
tion. Interpretation is important, because the data Monitoring. Monitoring, whether its target is
do not speak for themselves. Learning that a pa- subatomic dynamics (Heisenberg, 1958/1999) or
tient took a pill means little without information human behavior (Friman, 2009), affects the be-
about the importance, meaning, and relevance (to havior of the object monitored. The direction of
the regimen) of that pill. Adherence to a medical the change in human behavior is determined by
regimen is a clinical activity, and its assessment its social valence: Socially acceptable behaviors
data should have demonstrable clinical utility. As tend to increase in frequency, and socially unac-
a counterexample, much of the information solic- ceptable behaviors tend to decrease in frequency
ited in the packets that patients receive in medical (Nelson, 1977). As a thought experiment, imag-
settings has little or no bearing on their medical ine the amount and direction of behavior change
condition or its treatment. that would result from pointing a video camera at
a group. In pediatric medicine, the primary moni-
tor is almost always a caregiver. Researchers have
Improving Adherence
used monitoring to enhance adherence across a
Health Education. There are two forms of in- broad range of programs, from functional encopre-
tervention in behavioral pediatrics, and health sis (O’Brien, Ross, & Christophersen, 1986) to diet
education is one of them (prescriptive treatment and exercise (Rapoff, 2010).
is the other). Some amount of health education
accompanies all treatment prescribed in pediatrics Prompts and Reminders. Although prompts and
(e.g., see the discussions of bedtime problems, en- reminders can play a role across virtually all medi-
copresis, and enuresis above), and a major portion cal regimens, professionals use them more for ap-
of it pertains to the importance of following the pointment keeping than for any other dimension
treatment regimen. As one example, health educa- of pediatric medicine. Reminder phone calls are
tion messages about the importance of completing now a basic part of virtually all medical subspecial-
the regimen are now a standard part of prescribed ties that involve prescriptive treatment, although
antibiotic regimens. This was not always the case. this was not always the case. For example, Outpa-
Patients often discontinued the regimen as soon as tient Pediatrics at the University of Kansas School
their symptoms receded. Premature discontinua- of Medicine routinely reported high no-show rates
tion, however, can lead to symptom recurrence and for pediatric visits. The department had not used
treatment resistance. In pediatric studies evaluat- prompts or reminders. Our research group imple-
ing the use of health education to increase adher- mented an intervention involving mailed and
ence, among the most typical targets are antibiotic telephoned reminders, and it resulted in approxi-
regimens for otitis media; the results are mixed, mately a 20% increase in appointments kept and
with only about half of available studies showing a 20% decrease in appointments not kept (cancel-
a significant benefit (Rapoff, 2010). However, this lations were not included in the database). These
50% failure rate may be more a reflection of edu- improvements in appointment keeping were ac-
cational methods than of health education itself. companied by substantial cost savings (Friman et
For example, a test of three educational methods al., 1985). We replicated these results in a training
to promote adherence to a regimen for reducing clinic staffed by pediatric residents (Friman, Glass-
dangerous infant behavior showed that an educa- cock, Finney, & Christophersen, 1987). Studies
tional video with modeling was as effective as an like these have led to the virtually universal use of
in-home demonstration, and more effective than reminder systems in modern medicine and stand
educational materials supplied by the American as evidence that behaviorally based research on
Academy of Pediatrics (Linnerooth et al., 2002). prompting can exert a significant influence on
Behavioral Pediatrics 421
how professionals address nonadherence in medi- tion is one example. The literature showing that
cal settings. That influence extends far beyond behaviorally based disciplinary interventions (e.g.,
appointment keeping. Some form of prompting time out, time in) can improve compliance is so
could potentially improve adherence to any medi- well established that it need not be cited here.
cal regimen to be followed outside a clinic setting. Nonadherence to a medical regimen is a form of
As just one example of potential research in this noncompliance, and child resistance is frequently
area, research on smartphone-based prompting a factor. The extent to which disciplinary strate-
applications to improve adherence is limited. But gies can improve adherence is a subject worthy of
the ubiquity of smartphones coupled with their al- behavior-analytic investigation.
most limitless capacity for applications presents an Rule governance is another aspect of behavior
opportunity for enterprising behavior analysts to that almost certainly plays a significant role in
assess their potential for increasing adherence to adherence. Skinner (1969) described rules as con-
appointments and medical regimens. tingency-specifying stimuli, and Hayes (1989) has
characterized rule-governed behavior as behavior
Incentives. Presumably the most powerful in- under the control of verbal stimuli. When cogni-
centive for adhering to a medical regimen involves tively oriented investigators refer to self-manage-
negative reinforcement in the form of rapid relief ment for promoting adherence (cf. Rapoff, 2010),
from symptoms. However, not all appointments re- they are really referring to rule-governed behavior.
sult in treatment, and not all treatment results in However, the cognitive account places emphasis
rapid relief. As just one example of the latter, the on the self as agent and deemphasizes the role of
regimen for juvenile diabetes requires multiple be- the environment (i.e., verbal community). This
haviors that are effortful (e.g., dietary restrictions, may explain why there is so little research show-
foot care), painful (e.g., insulin shots, blood sugar ing that self-management can improve adherence.
tests), or both, but that do not provide immedi- Why there is no behavior-analytic research exam-
ate symptom relief. Thus the addition of positive ining the role rule governance plays in adherence
reinforcement in the form of tangible incentives awaits an explanation.
can be helpful. One example involves token sys- A final option involves various forms of psycho-
tems to promote adherence to regimens in juve- therapy. From a cognitive perspective, mental ill-
nile rheumatoid arthritis. The basic form involves ness involves some combination of cognitive dis-
four primary components: (1) task-analyzing the tortion and neurotransmitter levels. Despite being
regimen into readily executable steps, (2) supply- highly theoretical at this stage of its development,
ing tokens contingent upon completion of steps, this perspective appears to operate at the level of
(3) withdrawing tokens contingent upon failure to dogma in mainstream psychology and psychiatry.
complete steps, and (4) arranging for regular ex- From a behavioral perspective, mental illness in-
change of tokens for agreed-upon rewards. One of volves maladaptive verbal behavior and environ-
the most difficult regimens to follow involves ab- mental contingencies. Regardless of perspective,
staining from addictive substances, and incentive the so-called illness can interfere with adherence
programs have been more effective than virtually when present. Conventional approaches (e.g., cog-
any other method for accomplishing this outcome nitive) employ some combination of medication
(e.g., Higgins, Heil, & Sigmon, 2013). Despite this and direct psychotherapy to resolve the illness and
extraordinary range of documented effective ap- increase compliance. Behavioral approaches em-
plications, ranging from performing simple steps ploy health education to correct the problematic
in a juvenile rheumatoid arthritis regimen to com- rule governance (e.g., vaccines do not cause autism
plying with prescribed abstinence from the most spectrum disorder) and contingency management
addictive substances known to humankind, incen- to improve behaviors related to adherence. Testing
tive programs for improving adherence to medical these approaches to improving adherence against
regimens has received only modest attention from each other could be interesting.
researchers in general and even less from behavior
analysts in particular. This section on adherence is cursory. Nonad-
herence to medical regimens is the single greatest
Miscellaneous Approaches. There are various threat to the health and well-being of children in
other approaches to promoting adherence in pe- the United States. This problem is so well recog-
diatric populations, although research directly nized that it typically receives book-length treat-
evaluating them is scant. Disciplinary interven- ment (e.g., O’Donohue & Levensky, 2006; Rapoff,
422 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
2010; Stuart, 1982). Although the section ends but a subsidiary one in this chapter. Nonetheless, I
this chapter on behavioral pediatrics, it should suggest a few tactics that could help. For example,
probably have been the introductory section (as applied behavior analysts interested in collabo-
noted earlier), because the treatments discussed in rating with pediatricians at a local level should
the first three dimensions of behavioral pediatrics attend and present at pediatric conferences and
are moot if patients do not keep appointments or lectures. They should also attend case manage-
follow regimens. One of the major concerns of be- ment discussions in pediatric settings and offer to
havior analysts in the 21st century is how to bring help physicians implement behavioral assessments
our field more into the mainstream of everyday and treatments. If these contacts result in a refer-
life. Almost all U.S. children have a primary care ral, promptly sending data-based feedback on the
physician. Thus any research or intervention that effects of the ABA interventions to the referring
targets pediatrics in general, as adherence does, is pediatrician for inclusion in the patient’s medical
by definition, mainstream. Enterprising behavior chart not only mirrors standard practice between
analysts might consider taking that route. physicians in all domains of medicine; it is also
likely to lead to more referrals. Additionally, the
increasing demands by third-party payers for docu-
CONCLUSIONS mentation of treatment and its effects can make
data-based feedback necessary for continued reim-
Behavioral pediatrics is a diverse field that in- bursement for costs accruing from needed ongoing
cludes research and treatment of common child treatment. This fact, coupled with the value ABA
behavior problems, research on the interactions places on ongoing data collection, makes applied
between physiology and behavior that affect child behavior analysts increasingly attractive as col-
health, and treatments derived from the findings laborators in pediatric medicine. Applied behavior
of that research. ABA is a science that conducts analysts could also become more involved in medi-
research on interactions between environmental cal professional organizations and thereby provide
and behavioral variables and evaluates interven- a community resource for questions pertaining to
tions for socially significant problems that have behavior. More systemically, incorporating the
been derived from that research. The integration clinical phenomena referred to by hypothetical
of behavioral pediatrics and ABA not only benefits constructs like depression, anxiety, or tempera-
both fields, but actually extends the effectiveness ment into behavior-analytic theories, and making
of pediatricians—the primary guardians of child those phenomena the focus of ABA assessments
health in this country—and thereby contributes and interventions, could advance the field and in-
to the health of children. I have confined my dis- crease its acceptance (e.g., Friman, Hayes, & Wil-
cussion to four problem areas, each representing son, 1998; see also Friman, 2010). Lastly, and con-
one domain of behavioral pediatrics. This repre- sistent with the primary purpose of this chapter,
sents a much-abbreviated review of behavioral pe- designing and testing treatments for problems that
diatrics and the potential role of ABA, and there frequently present in pediatric settings contributes
are many aspects of behavioral pediatrics that I to the recognition of ABA as a resource and ac-
have not addressed (e.g., infant colic, oppositional ceptance of it by pediatricians as a valuable sci-
behaviors, habit disorders, anxiety and depressive ence (Riley & Freeman, 2019). Perhaps more than
disorders, chronic illnesses, and pain). For a broad- any other subspecialty in medicine, pediatrics is
er sample, please refer to various source documents a pragmatic specialty, and effective and efficient
(e.g., Allen, Barone, & Kuhn, 1993; Blum & Fri- treatments readily trump ideological differences.
man, 2000; Christophersen, 1982; Friman, 2005a, In conclusion, the large and mounting body of
2008; Friman & Blum, 2003). evidence documenting the many ways ABA has
Despite its many contributions to behavioral been found to influence behavior, coupled with
pediatrics, ABA is still not widely available and the mutually determinative role played by behav-
accepted in the pediatric medical community. ior and physiology, suggests that a partnership be-
Remedying the problem of limited availability is a tween applied behavior analysts and pediatricians
primary purpose of this chapter; I hope to increase would benefit the health of children in the United
the number of applied behavior analysts interested States. I have argued in this chapter that behav-
in behavioral pediatrics. Remedying the problem ioral pediatrics is the ideal locus for this partner-
of limited acceptance is an important concern, ship.
Behavioral Pediatrics 423
and therapy (Vol. 2, pp. 731–739). Thousand Oaks, Friman, P. C., Mathews, J. R., Finney, J. W., & Christo-
CA: SAGE. phersen, E. R. (1988). Do children with encopresis
Friman, P. C. (2005b). Good night, we love you, we will have clinically significant behavior problems? Pedi-
miss you, now go to bed and go to sleep: Managing bed- atrics, 82, 407–409.
time problems in young children. Boys Town, NE: Girls Friman, P. C., & Schmitt, B. D. (1989). Thumb suck-
and Boys Town Press. ing: Guidelines for pediatricians. Clinical Pediatrics,
Friman, P. C. (2008). Primary care behavioral pediat- 28, 438–440.
rics. In M. Hersen & A. Gross (Eds.), Handbook of Friman, P. C., & Schnoes, C. (2020). Pediatric preven-
clinical psychology (Vol. 2, pp. 728–758). Hoboken, tion: Sleep dysfunction. Pediatric Clinics of North
NJ: Wiley. America, 67, 559–571.
Friman, P. C. (2009). Behavioral assessment. In M. Friman, P. C., & Vollmer, D. (1995). Successful use of
Hersen, D. Barlow, & M. Nock (Eds.), Single case the nocturnal urine alarm for diurnal enuresis. Jour-
experimental designs (3rd ed., pp. 91–143). Boston: nal of Applied Behavior Analysis, 28, 89–90.
Allyn & Bacon. Gabel, S., Hegedus, A. M., Wald, A., Chandra, R., &
Friman, P. C. (2010). Come on in, the water is fine: Chaponis, D. (1986). Prevalence of behavior prob-
Achieving mainstream relevance through integra- lems and mental health utilization among encopretic
tion with primary medical care. Behavior Analyst, 33, children. Journal of Developmental and Behavioral Pe-
19–36. diatrics, 7, 293–297.
Friman, P. C. (2017). Incontinence. In A. Wetzel (Ed.), Green, C., Storfer-Isser, A., Stein, R. E. K., Garner, A.
The SAGE encyclopedia of abnormal and clinical psy- S., Kerker, B. D., Szilagyi, M., et al. (2017). Which
chology (pp. 1296–1298). Thousand Oaks, CA: SAGE. pediatricians comanage mental health conditions?
Friman, P. C. (2019). Incontinence in the child: A Academic Pediatrics, 17, 479–486.
biobehavioral perspective. In T. H. Ollendick, S. W. Halliday, S., Meadow, S. R., & Berg, I. (1987). Successful
White, & B. A. White (Eds.), The Oxford handbook of management of daytime enuresis using alarm proce-
clinical child and adolescent psychology (pp. 367–381). dures: A randomly controlled trial. Archives of Dis-
New York: Oxford University Press. eases in Children, 62, 132–137.
Friman, P. C., & Blum, N. J. (2003). Primary care be- Hatch, T. F. (1988). Encopresis and constipation in chil-
havioral pediatrics. In M. Hersen & W. Sledge (Eds.), dren. Pediatric Clinics of North America, 35, 257–281.
Encyclopedia of psychotherapy (pp. 379–399). New Hatti, M. (2007). 2 deaths spur bed-wetting drug
York: Academic Press. warning: FDA warns that some patients taking
Friman, P. C., & Christophersen, E. R. (1986). Biobe- desmpopressin may be at risk of seizure and death.
havioral prevention in primary care. In N. Krasnegor, Retrieved July 11, 2016, from www.webmd.com/chil-
J. D. Arasteh, & M. F. Cataldo (Eds.), Child health dren/news/20071204/2-deaths-spur-bedwetting-drug-
behavior: A behavioral pediatrics perspective (pp. 254– warning.
280). New York: Wiley. Hayes, S. C. (1989). Rule-governed behavior: Cognition,
Friman, P. C., Finney, J. W., Rapoff, M. A., & Christo- contingencies, and instructional control. New York:
phersen, E. R. (1985). Improving pediatric appoint- Plenum Press.
ment keeping: Cost effectiveness and social valida- Heisenberg, W. (1999). Physics and philosophy. Amherst,
tion of reminders and reduced response requirement. NY: Prometheus. (Original work published 1958)
Journal of Applied Behavior Analysis, 18, 315–323. Higgins, S. T., Heil, S. H., & Sigmon, S. C. (2013).
Friman, P. C., Glasscock, S. G., Finney, J. W., & Chris- Voucher based contingency management in the
tophersen, E. R. (1987). Reducing effort with remind- treatment of substance abuse disorders. In G.
ers and a parking pass to improve appointment keep- Madden (Ed.), APA handbook of behavior analysis
ing for patients of pediatric residents. Medical Care, (pp. 481–500). Washington, DC: American Psychi-
25, 83–86. atric Association.
Friman, P. C., Handwerk, M. L., Swearer, S. M., McGin- Honaker, S. M., & Meltzer, L. J. (2014). Bedtime prob-
nis, C., & Warzak, W. J. (1998). Do children with pri- lems and night wakings in young children: An up-
mary nocturnal enuresis have clinically significant date of the evidence. Pediatric Respiratory Reviews,
behavior problems? Archives of Pediatrics and Adoles- 15, 333–339.
cent Medicine, 152, 537–539. Horwitz, S. M., Leaf, P. J., Leventhal, J. M., Forsyth, B.,
Friman, P. C., Hayes, S. C., & Wilson, K. (1998). Why & Speechley, K. N. (1992). Identification and man-
behavior analysts should study emotion: The exam- agement of psychosocial and developmental prob-
ple of anxiety. Journal of Applied Behavior Analysis, lems in community-based, primary care pediatric
31, 137–156. practices. Pediatrics, 89, 480–485.
Friman, P. C., Hoff, K. E., Schnoes, C., Freeman, K. Houts, A. C. (1991). Nocturnal enuresis as a biobehav-
A., Woods, D. W., & Blum, N. (1999). The bedtime ioral problem. Behavior Therapy, 22, 133–151.
pass: An approach to bedtime crying and leaving the Irwin, C. E., Cataldo, M. F., Matheny, A. P., & Peterson,
room. Archives of Pediatric and Adolescent Medicine, L. (1992). Health consequences of behaviors: Injury
153, 1027–1029. as a model. Pediatrics, 90, 798–807.
Behavioral Pediatrics 425
Jenssen, B. P., Buttenheim, A. M., & Fiks, A. G. (2019). Evaluating the Bedtime Pass Program for child re-
Using behavioral economics to encourage parent sistance to bedtime: A randomized controlled trial.
behavior change: Opportunities to improve clinical Journal of Pediatric Psychology, 32, 283–287.
effectiveness. Academic Pediatrics, 19, 4–10. Nelson, R. (1977). Assessment and therapeutic func-
Johnston, J. M., & Pennypacker, H. S. (2009). Strategies tions of self-monitoring. In M. Hersen, R. M. Eisler &
and tactics of behavioral research (3rd ed.). New York: P. M. Miller (Eds.), Progress in behavior modification
Routledge. (Vol. 5, pp. 263–308). New York: Academic Press.
Kazdin, A. E. (1996). Dropping out of child therapy: O’Brien, S., Ross, L. V., & Christophersen, E. R. (1986).
Issues for research and implications for clinical Primary encopresis: Evaluation and treatment. Jour-
practice. Clinical Child Psychology and Psychiatry, 1, nal of Applied Behavior Analysis, 19, 137–145.
133–156. O’Donohue, W. T., & Levensky, E. R. (2006). Promoting
Koff, S. A. (1995). Why is desmopressin sometimes in- treatment adherence. London: SAGE.
effective at curing bed wetting. Scandinavian Jour- Olness, K., Culbert, T., & Uden, D. (1989). Self-regula-
nal of Urology and Nephrology, Supplementum, 173, tion of salivary immunoglobulin A by children. Pedi-
103–108. atrics, 83, 66–71.
Koppen, I. J., von Gontard, A., Chase, J., Cooper, C. S., Polaha, J., Dalton, W. T., & Allen, S. (2011). The preva-
Rittig, C. S., Bauer, S. B., et al. (2016). Management lence of emotional and behavior problems in pedi-
of functional nonretentive fecal incontinence in atric primary care serving rural children. Journal of
children: Recommendations from the International Pediatric Psychology, 36, 652–660.
Children’s Continence Society. Journal of Pediatrics, Quittner, A. L., Modi, A. C., Lemanek, K. L., Levers-
12, 56–64. Landis, C. E., & Rapoff, M. A. (2008). Evidence-
Lackgren, G., Neveus, T., & Stenberg, A. (1997). Di- based assessment of adherence to medical treatments
urnal plasma vasopressin and urinary output in ado- in pediatric psychology. Journal of Pediatric Psychol-
lescents with monosymptomatic nocturnal enuresis. ogy, 33, 916–936.
Acta Paediatrica, 86, 385–390. Rapoff, M. A. (2010). Adherence to pediatric medical regi-
Levine, M. D. (1982). Encopresis: Its potentiation, mens (2nd ed.). New York: Springer.
evaluation, and alleviation. Pediatric Clinics of North Regier, D. A., Goldberg, I. D., & Taube, C. A. (1978).
America, 29, 315–330. The de facto US mental health system. Archives of
Linnerooth, P. J. N., Walsh, M., Spear, S., Niccols, R., General Psychiatry, 35, 685–693.
Maestretti, J., Pulido, L., et al. (2002). Reducing in- Rickert, V. I., & Johnson, M. (1988). Reducing noc-
fant potentially dangerous behavior: A randomized con- turnal awaking and crying episodes in infants and
trolled trial of three interventions for preventing uninten- young children: A comparison between scheduled
tional child injury. Paper presented at the 36th annual awakenings and systematic ignoring. Pediatrics, 81,
convention of the Association for Advancement of 203–212.
Behavior Therapy, Reno, NV. Riley, A. R., & Freeman, K. A. (2019). Impacting pe-
Lorenz, R. A., Christensen, N. K., & Pitchert, J. W. diatric primary care: Opportunities and challenges
(1985). Diet related knowledge, skill, and adherence for behavioral research in a shifting healthcare land-
among children with insulin dependent diabetes scape. Behavior Analysis, 19, 23–38.
mellitus. Pediatrics, 75, 872–876. Rittig, S., Knudsen, U. B., Norgaard, J. P., Pedersen,
Lowery, S., Srour, J., Whitehead, W. E., & Schuster, M. E. B., & Djurhuus, J. C. (1989). Abnormal diurnal
M. (1985). Habit training as treatment of encopresis rhythm of plasma vasopressin and urinary output in
secondary to chronic constipation. Journal of Pediat- patients with enuresis. American Journal of Physiol-
ric Gastroenterology and Nutrition, 4, 397–401. ogy, 256, F644–F671.
Lozoff, B., Wolf, A. W., & Davis, N. S. (1985). Bedtime Ruckstuhl, L. E., & Friman, P. C. (2003). Evaluating
problems seen in pediatric practice. Pediatrics, 75, the effectiveness of the vibrating urine alarm: A study
477–483. of effectiveness and social validity. Paper presented at
Mace, F., Hock, M. L., Lalli, J. S., West, B. J., Belfiore, the 29th annual convention of the Association for
P., Pinter, E., et al. (1988). Behavioral momentum in Behavior Analysis, San Francisco, CA.
the treatment of noncompliance. Journal of Applied Schnoes, C. J. (2011). The bedtime pass. In M. Perlis,
Behavioral Analysis, 21, 123–142. M. Aloia, & B. Kuhn (Eds.), Behavioral treatments for
Meunier, P., Marechal, J. M., & De Beaujeu, M. J. sleep disorders: A comprehensive primer of behavioral
(1979). Rectoanal pressures and rectal sensitivity in sleep medicine interventions (pp. 293–298). Boston:
chronic childhood constipation. Gastroenterology, Elsevier.
77, 330–336. Schnoes, C. J., & Reimers, T. M. (2009). Assessment
Moffatt, M. E. (1997). Nocturnal enuresis: A review of and treatment of child and adolescent sleep disor-
the efficacy of treatments and practical advice for ders. In D. McKay & E. A. Storch (Eds.), Cognitive-
clinicians. Journal of Developmental and Behavioral behavior therapy for children: Treating complex and
Pediatrics, 18, 49–56. refractory cases (pp. 293–324). New York: Springer.
Moore, B. A., Fruzetti, A. E., & Friman, P. C. (2007). Shepard, J. A., Poler, J. E., & Grabman, J. H. (2017).
426 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Evidence-based psychosocial treatments for pediatric Stuart, R. B. (1982). Adherence, compliance, and general-
elimination disorders. Journal of Clinical Child and ization in medicine. New York: Brunner/Mazel.
Adolescent Psychology, 46, 767–797. Vincent, S. A. (1974). Mechanical, electrical and other
Skidmore-Roth, L. (2010). Mosby’s nursing drug reference aspects of enuresis. In J. H. Johnston & W. Good-
(23rd ed.). St. Louis, MO: Elsevier. win (Eds.), Reviews in pediatric urology (pp. 280–313).
Skinner, B. F. (1966). The phylogeny and ontogeny of New York: Elsevier.
behavior. Science, 153, 1205–1213. Vollmer, T. R., & Iwata, B. A. (1992). Differential rein-
Skinner, B. F. (1969). Contingencies of reinforcement. forcement as treatment for behavior disorders: Proce-
New York: Appleton-Century-Crofts. dural and functional variations. Research in Develop-
Stark, L. J., Knapp, L. G., Bowen, A. M., Powers, S. mental Disabilities, 13, 393–417.
W., Jelalian, E., Evans, S., et al. (1993). Increasing Wald, A., Chandra, R., Chiponis, D., & Gabel, S.
caloric consumption in children with cystic fibrosis: (1986). Anorectal function and continence mecha-
Replication with 2-year follow-up. Journal of Applied nisms in childhood encopresis. Journal of Pediatric
Behavior Analysis, 26, 435–450. Gastroenterology and Nutrition, 5, 346–351.
Stark, L. J., Opipari, L. C., Donaldson, D. L., Danovsky, Watson, T. S., Foster, N., & Friman, P. C. (2006). Treat-
D. A., Rasile, D. A., & DelSanto, A. F. (1997). Evalu- ment adherence in children and adolescents. In W.
ation of a standard protocol for retentive encopresis: T. Odonohue & E. R. Levensky (Eds.), Promoting
A replication. Journal of Pediatric Psychology, 22, treatment adherence (pp. 343–351). London: SAGE.
619–633. Weinstock, L. B., & Clouse, R. E. (1987). A focused
Stark, L. J., Owens-Stively, J. A., Spirito, A., Lewis, A. overview of gastrointestinal physiology. Annals of Be-
V., & Guevremont, D. (1990). Group behavioral havioral Medicine, 9, 3–6.
treatment of retentive encopresis. Journal of Pediatric Whitehead, W. E., & Schuster, M. M. (1985). Gastroin-
Psychology, 15, 659–671. testinal disorders: Behavioral and physiological basis for
Stark, L. J., Spirito, A., Lewis, A. V., & Hart, K. J. treatment. New York: Academic Press.
(1990). Encopresis: Behavioral parameters associated Wright, L. (1975). Outcome of a standardized program
with children who fail medical management. Child for treating psychogenic encopresis. Professional Psy-
Psychiatry and Human Development, 20, 169–179. chology, 6, 453–456.
CHAP TER 25
A Behavior-Analytic Approach
to Pediatric Feeding Disorders
Cathleen C. Piazza and Caitlin A. Kirkwood
Feeding is unique relative to other behaviors dis- Not surprisingly, nature has even provided some
cussed in this book, because food1 functions as organisms with the ability to adapt to unpredict-
an appetitive stimulus, and feeding is ubiquitous able changes in food availability. For example,
across species. Organisms require relatively con- some bird species that typically exhibit a narrow
stant caloric intake, and metabolic output cannot range of foraging behavior develop strategies to
exceed caloric intake for long. Many organisms identify alternative food sources when the food
have adapted unique mechanisms for the habitat in their typical habitat becomes scarce (Diquelou,
in which they feed that promote consistent acqui- Griffin, & Sol, 2016).
sition of calories and nutrients. For example, the Even when changes occur that disrupt food
hyacinth macaw, a native parrot of South Ameri- availability, motivation to feed is typically not
ca, has a beak that can exert hundreds of pounds disrupted (Diquelou et al., 2016). In fact, profes-
of pressure per square inch; this beak is ideal for sionals in the United States are often so confident
cracking the hard shells of the palm nuts that are about the reinforcing properties of food that their
a major component of the bird’s diet (Borsari & response to a caregiver of a poorly growing child is
Ottoni, 2005). Some organisms have adaptations “The child will eat when he [or she] gets hungry.”
that allow them to reduce metabolic needs, store But are there exceptions to this time-worn adage?
nutrients, or both to respond to predictable chang- And if so, why does this happen and what are the
es in their habitat that affect food availability. For stimulus conditions under which it occurs?
example, black bears suppress their basal meta-
bolic rate by 25% during the 5–7 months per year
that they hibernate (Toien et al., 2011). Despite an DEFINITION
organism’s adaptations to feed in a specific habitat
or during predictable environmental changes, un- We use the term feeding disorder for children who
predictable changes in environmental conditions, do not consume sufficient calories, hydration, or
such as drought, may threaten the organism’s abil- nutrition to gain weight and grow, to maintain
ity to access calories and nutrients consistently. hydration, or to meet their nutritional needs for
macro- and micronutrients. Feeding disorders are
1 Weuse the term food here to mean an organism’s source of heterogeneous and may include refusal to eat;
energy (kilocalories) and nutrients. refusal to eat certain types or textures of food;
427
428 S u b s p e c i a lt ie s in A p p l ie d B e h av i o r A n a ly s i s
dependence on a limited or developmentally in- thrust, pocketing food, poor lip closure, refusal,
appropriate source of nutrition, such as bottle de- spitting out food, and vomiting (Arvedson &
pendence in a 3-year-old child; and skill deficits, Brodsky, 2002).
such as inability to self-feed or transition to age- or
developmental-stage-appropriate textures. The Di-
agnostic and Statistical Manual of Mental Disorders, DEPENDENT VARIABLES
fifth edition (DSM-5; American Psychiatric As-
sociation, 2013) uses the term avoidant/restrictive Feeding consists of a complex chain of behav-
food intake disorder to refer to the following: (1) iors that includes placing bites or drinks into the
A child exhibits a feeding or eating disturbance mouth; lateralizing the food to the molars and
characterized by persistent failure to meet appro- chewing, if necessary; forming a bolus (a mass of
priate nutritional needs, energy needs, or both, food or drink) on the tongue; elevating the tongue;
with significant weight loss, significant nutritional and propelling the bolus to the pharynx (Derkay
deficiency, need for enteral feeding or oral nutri- & Schechter, 1998). A child with a feeding disor-
tional supplements, or obvious interference with der may have difficulties with one or more of the
psychosocial functioning; (2) the disturbance is behaviors in the chain. Thus intervention may
not better explained by a lack of available food or target one or more problematic feeding behaviors,
by an associated culturally permissible practice; (3) and intervention for one behavior may affect the
the eating disturbance does not occur only during occurrence of other behaviors. We provide the
the course of anorexia nervosa or bulimia nervosa, reader with operational definitions and a discus-
and there is no evidence of a disturbance in the sion of dependent variables that researchers have
child’s experience of body weight or shape; and (4) used; we realize that the list is not exhaustive and
the eating disturbance is not attributed to a con- that the field does not have standard operational
current medical condition or better explained by definitions for these behaviors.
another mental disorder, and if the eating distur- Acceptance, the occurrence of the bite or drink
bance occurs in the context of another condition entering the mouth, is often the first in the chain
or disorder, its severity exceeds what is routinely of feeding behaviors that researchers target for in-
associated with that condition or disorder and tervention. In LaRue et al. (2011), observers scored
warrants additional clinical attention. the occurrence of acceptance if a child leaned to-
A diagnostic nosology such as DSM-5 is help- ward the spoon or cup and opened his or her mouth
ful in describing the characteristics of a feeding in the absence of negative vocalizations and inap-
disorder, but it tells us little about why feeding propriate behavior, so that the feeder could deposit
disorders occur or how to develop effective inter- the entire bolus of food (except for an amount
ventions. Although multiple children may have smaller than the size of a pea) or any amount of liq-
the same topographical expression of a feeding uid in the mouth within 5 seconds of presentation.
disorder, such as poor weight gain, the etiology of Notice that LaRue et al.’s definition included qual-
the feeding disorder may be different across chil- ifiers such as “leaned toward the spoon or cup in
dren. For example, one poorly growing child may the absence of negative vocalizations” and “within
have delayed gastric emptying, complain of being 5 seconds of presentation.” LaRue et al. included
full, and refuse to eat after consuming a small these qualifiers to ensure that observers were
amount of food (Stein, Everhart, & Lacy, 2015). measuring children’s rather than feeders’ behav-
Another child, by contrast, may have oral–motor ior. Observers scored acceptance if a child leaned
skill deficits. Although the child eats frequently forward soon after the bite or drink presentation
and appears motivated to eat (e.g., asks for food), and opened his or her mouth so that the feeder
meals are lengthy, and the child cannot consume could put the bite or drink in the child’s mouth.
sufficient calories to gain weight and grow. A third By contrast, observers did not score acceptance
child may vomit frequently, and poor weight gain if the child’s mouth was open because he or she
is an indirect result of gastroesophageal reflux dis- was crying, which gave the feeder the opportunity
ease. Within those broad topographical presenta- to put the bite or drink in the mouth. Ideally, in-
tions of feeding disorders, individual children may tervention increases feeding compliance, and dis-
exhibit different behaviors that contribute to their tinguishing between child compliance and feeder
feeding disorders, such as excessive saliva produc- behavior is important for determining progress.
tion, inability to chew, inability to lateralize the Observers in LaRue et al. scored the occurrence of
tongue, an open-mouth posture, persistent tongue acceptance for each bite or drink presentation and
A Behavior-Analytic Approach to Feeding Disorders 429
converted occurrences to a percentage after divid- tion the utensil was in arm’s reach of the child.
ing the occurrences of acceptance by the number Controlling for the presence of the utensil may be
of presented bites or drinks. important in cases where the duration of utensil
Children in Peterson, Piazza, and Volkert presence may differ markedly from phase to phase
(2016) self-fed, and “observers scored the occur- or from condition to condition.
rence of acceptance when the child used the uten- Expulsion is food or liquid exiting the mouth.
sil or his fingers to put the entire bite of food in his For example, Wilkins, Piazza, Groff, and Vaz
mouth within 8 s of presentation, not including (2011) defined expulsion for liquids as each time
placement of the bite in the mouth during re-pre- any amount of liquid (pea size or larger) that the
sentation” (p. 5). Peterson et al. converted occur- child had not swallowed was visible outside the lips
rences of acceptance to a percentage after dividing after any amount of liquid had passed the plane of
the number of accepted bites by the number of the lips; and expulsion for solids as each time any
presented bites. Peterson et al. described the bite amount of food (pea size or larger) that the child
presentation as one piece of food measuring 0.6 had not swallowed was visible outside the lips after
× 0.6 × 0.6 centimeters, and the feeder restricted the entire bolus of food had passed the plane of
the operant by presenting only one bite at a time. the lips. Observers in Wilkins et al. scored the fre-
Observers scored acceptance only if the child put quency of expulsion and presented the data as ex-
the entire bite in his mouth within 8 seconds of pulsions per bite. As with many feeding behaviors,
the feeder placing the bite in front of the child. the opportunity to engage in expulsion will affect
Acceptance, in this case, conveyed information rates of expulsion, and the clinician should con-
about how much food entered the child’s mouth, sider which method of data presentation is most
and it accounted for the number of bites available appropriate to address the clinical problem or re-
to the child to accept from session to session. search question. For example, the clinician might
Inappropriate mealtime behavior may refer to calculate expulsions per bite if opportunity to
numerous behaviors that disrupt feeding and that expel is fixed (e.g., based on number of presented
differ along many dimensions. Some researchers bites), but expulsions per opportunity if opportu-
have addressed this dilemma by conceptualizing nity to expel varies, such as during re-presentation.
inappropriate mealtime behavior as behavior a Mouth clean is a product measure of swallowing
child emits with his or her external body parts, that we often use, because swallowing is difficult to
such as the head, hands, arms, and legs, that measure reliably in our clinical experience. For ex-
prevents solids, liquids, or both from entering or ample, observers in Kadey, Piazza, Rivas, and Zeleny
remaining in the mouth. These researchers have (2013) scored “mouth clean if no food larger than a
separated inappropriate mealtime behavior from pea was in [the child’s] mouth, unless the absence
behaviors the child emits with his or her mouth, of food was the result of expulsion” (p. 540). The
such as expulsion and packing, and behaviors feeder in Kadey et al. checked the child’s mouth 30
that have a putative physiological basis, such as seconds after the bite entered the mouth by saying,
coughing, gagging, and vomiting. For example, “Show me.” The mouth check gave observers the
observers in Ibañez, Piazza, and Peterson (2019) opportunity to score mouth clean or pack, which is
scored inappropriate mealtime behavior “when the converse of mouth clean. For example, observ-
the utensil was in arm’s reach of the child and the ers in Wilkins et al. (2014) scored “pack if the entire
child turned his head 45° or greater away from the bite (with the exception of food smaller than a pea)
utensil during a bite or drink presentation; used entered the child’s mouth, and food larger than a
his hand to contact the utensil, food or drink, or pea was in the child’s mouth at the 30-s check”
the feeder’s hand or arm anywhere from the elbow (p. 4). Kadey et al. and Wilkins et al. converted
down while the feeder was presenting the bite or mouth clean or pack, respectively, to a percentage
drink; threw food, liquids, or utensils; or blocked after dividing the number of occurrences of mouth
his mouth with his hand, bib, or toys” (p. 1009). clean or pack, respectively, by the number of bites
Notice that Ibañez et al.’s operational definition that entered the child’s mouth. Note that the de-
qualifies that the utensil must be in arm’s reach of nominator in Kadey et al. and Wilkins et al. was
the child, because the defined behavior can occur number of bites that entered the mouth, meaning
only when a utensil is present. Observers scored that no opportunities for mouth clean or pack oc-
the frequency of inappropriate mealtime behavior curred if no bites entered the child’s mouth; this is
and converted it to a rate by dividing the number important for readers to remember when interpret-
of inappropriate mealtime behaviors by the dura- ing data from studies that use this measure. Note
430 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
also that Kadey et al. and Wilkins et al. included example, Rommel et al. (2003) characterized the
a caveat that the child could have a mouth clean feeding disorders of 700 children referred for as-
if a small amount of food remained in the mouth, sessment and treatment of severe feeding difficul-
because even typical feeders may have residue in ties as medical (86%), oral–motor (61%), and/or
the mouth after swallowing. The researchers based behavioral (18%). Combined causes, such as medi-
the size of the residue (pea size in these studies) cal, behavioral, and/or oral–motor, for the feeding
on the amount of presented food. In both studies, disorder occurred in over 60% of children. Other
bolus sizes were a level small maroon spoon for six researchers have found that feeding disorders have
children, a half-level small maroon spoon for four neurological (62%), structural (53%), behavioral
children, and a level baby spoon for two children in (43%), cardiorespiratory (34%), and metabolic
Wilkins et al., and a level small maroon spoon for (12%) causes, with most children having causes
the child in Kadey et al. The size of the acceptable in two or more categories simultaneously (Burk-
residue might change if the presented bolus was low, McGrath, Valerius, & Rudolph, 2002; Davis,
larger or smaller. For example, the acceptable resi- Bruce, Cojin, Mousa, & Hyman, 2010). Feeding
due in Volkert, Peterson, Zeleny, and Piazza (2014) disorders are also prevalent among specific diag-
was the size of a grain of rice, because the presented nostic groups, such as children with autism spec-
bolus was a piece of food measuring 0.6 × 0.6 × 0.6 trum disorder, cerebral palsy, and Down syndrome
centimeters. (Bandini et al., 2010).
Although chewing is an important feeding skill, The high prevalence of medical conditions and
there is a relative dearth of applied-behavior-ana- oral–motor dysfunction in children with feeding
lytic research on teaching children with feeding disorders suggests that biological factors play an
disorders to chew. In a notable exception, Volkert important role in the etiology and maintenance of
et al. (2014) scored a chew “each time the child’s feeding disorders. Children with chronic medical
teeth and/or jaw completed one up-and-down mo- problems that affect the digestive system directly,
tion with the teeth parted at least 1.3 cm following such as congenital defects of the gastrointestinal
a verbal or model prompt while food was visible tract, delayed gastric emptying, food allergies,
anywhere in the mouth except the center of the gastroesophageal reflux disease, malabsorption,
tongue or between the front teeth” (p. 708). One or metabolic disorders, may associate eating with
limitation of this definition is that it did not dif- fatigue, nausea, pain, or a combination. For exam-
ferentiate rotary chews from an immature chewing ple, children with gastroesophageal reflux disease
pattern such as munching (Volkert, Piazza, Vaz, & may associate eating with the pain that occurs
Frese, 2013). One unique and important feature when excess acid erupts into the esophagus. Nau-
of Volkert et al. (2014), however, was that the re- sea plays an important role in the development
searchers included a measure of mastication. They of food aversions (Schafe & Bernstein, 1996),
defined mastication as food with pieces no larger and when nausea is paired with eating, aversions
than 0.2 × 0.2 centimeters in a liquid medium at to tastes may develop after only one or a few tri-
the mastication check. The mastication check als, may generalize to different foods, and may be
was like the mouth check described by Kadey et highly treatment-resistant.
al. (2013), but observers determined whether the Researchers estimate that feeding disorders
child had masticated rather than swallowed the occur in 40–70% of children with chronic medical
bite. One important future direction for research conditions (Davis et al., 2010; Douglas & Byron,
on pediatric feeding disorders is to develop more 1996; Lukens & Silverman, 2014; Thommessen,
sophisticated measures of chewing, perhaps using Heiberg, Kase, Larsen, & Riis, 1991), suggesting
automated methods (Hadley, Krival, Ridgel, that the presence of other chronic medical prob-
Hahn, & Tyler, 2015), which will allow investiga- lems, such as bronchopulmonary dysplasia, may
tors to determine whether a child’s chewing skills contribute to the etiology of feeding disorders. In-
are appropriate for the child’s age, developmental fants with complex medical histories are subject to
level, or a combination. numerous invasive diagnostic tests and procedures
that involve manipulation of the face and mouth,
such as laryngoscopy. Such a child may come to
ETIOLOGY associate the presentation of items to the mouth
with discomfort, pain, or both. From the child’s
Research has suggested that the etiology of feeding perspective, a spoon may be indistinguishable from
disorders is complex and multifactorial (Rommel, a laryngoscope or other devices that professionals
Meyer, Feenstra, & Veereman-Wauters, 2003). For use during invasive medical procedures and tests.
A Behavior-Analytic Approach to Feeding Disorders 431
Caregivers of chronically hospitalized and medi- that the cause of feeding disorders is multifacto-
cally fragile children often report oral aversions rial (Rommel et al., 2003). The goal of evaluation
that affect feeding and other activities associated should be to determine whether anatomical, medi-
with the face and mouth, such as face washing and cal, or oral–motor skill deficits contribute to the
tooth brushing. child’s feeding disorder and whether the child is
Oral–motor dysfunction may include difficul- a safe oral feeder. Members of such a team might
ties sucking, difficulties with bolus propulsion, include a behavior analyst, a dietitian, a gastro-
inability to lateralize food from one side to an- enterologist or the child’s primary physician, and
other, difficulties swallowing, and tongue thrust, an occupational or speech therapist. A behavior
and these problems may affect a child’s ability and analyst should not underestimate the negative
motivation to eat (Darrow & Harley, 1998). The consequences of feeding a child before appropri-
child’s refusal to eat may cause or exacerbate pre- ate evaluation. At a minimum, the behavior ana-
existing oral–motor dysfunction and further con- lyst should consult the child’s primary physician,
tribute to the child’s failure to develop appropriate describe the proposed course of assessment and
oral–motor skills. That is, a child who refuses to treatment, and obtain medical clearance to start
eat does not have the opportunity to practice the therapy. Failure to identify a medical condition or
skill of eating and does not develop the oral–motor oral–motor skill deficits before beginning therapy
skills to become a competent eater. could result in worsening of the feeding disorder or
When eating is paired with an aversive experi- even death. For example, introducing food variety
ence, the child may develop refusal behavior such to promote proper nutrient intake is a reasonable
as batting at the spoon, head turning, or covering goal for a child with a feeding disorder, but a re-
the mouth to avoid eating. These behaviors may action to an unidentified food allergy could cause
increase in frequency as a function of caregiver anaphylaxis, which is a severe, potentially life-
responses to child behavior during meals. Bor- threatening event (Sicherer & Sampson, 2010).
rero, Woods, Borrero, Masler, and Lesser (2010) In addition, some children with oral–motor dys-
conducted observations of 25 children with feed- function do not manage specific consistencies of
ing disorders and their caregivers to describe and solids, liquids, or both, and other children are not
quantify the caregivers’ responses to the children’s safe oral feeders with solids or liquids of any consis-
inappropriate mealtime behavior. Researchers tency. Oral–motor dysfunction may be associated
compared conditional probabilities of a caregiver with aspiration due to solids or liquids entering the
providing attention, escape, or a tangible item fol- airway, which can cause medical problems such as
lowing refusal or acceptance to the unconditioned pneumonia. An evaluation by a swallow special-
probabilities of each event. Observations indicated ist, usually an occupational or speech therapist,
that caregivers coaxed, removed the spoon, threat- can determine whether oral–motor dysfunction
ened to take away preferred items, presented pre- may be causing or contributing to a child’s feed-
ferred foods, or engaged in a combination of these ing disorder and determine whether the child is
events following refusal behavior. Escape in the a safe oral feeder (Schwarz, Corredor, Fisher-Me-
form of spoon removal or meal termination and dina, Cohen, & Rabinowitz, 2001). A dietitian
attention in the form of coaxing, reprimands, and calculates the child’s caloric, hydrational, and nu-
statements of concern most frequently followed tritional needs and determines whether the child
refusal behavior. Similarly, Piazza, Fisher, et al. requires diet modifications. For example, a child
(2003) observed caregivers and children with feed- with a glycogen storage disorder requires careful
ing disorders during meals. Caregivers responded monitoring of blood sugar levels and a diet that
to inappropriate mealtime behavior with one or restricts simple sugars. Drops in blood sugar levels
more of the following consequences: (1) providing may cause seizures, coma, and even death (Goto,
escape from bites of food or the meal, (2) coaxing Arah, Goto, Terauchi, & Noda, 2013).
or reprimanding (e.g., “Eat your peas, they are good Results of the interdisciplinary evaluation may
for you”), or (3) providing a toy or preferred food. indicate that the child needs medical treatment,
consistency or texture manipulations, a special
diet, or a combination, and these interventions
EVALUATION may resolve the child’s feeding disorder. Some chil-
dren, however, may not start feeding or may not
A behavior analyst should consider referring a feed well even after interdisciplinary intervention,
child with a feeding disorder for evaluation by particularly if they engage in refusal behavior. In
an interdisciplinary team, given the evidence these cases, a qualified behavior analyst can ma-
432 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
nipulate mealtime antecedents and consequences pared variations of extinction matched to the
to determine whether they affect the child’s refusal results of each functional analysis. Both analyses
behavior. identified escape and attention as reinforcement
Piazza, Fisher, et al. (2003) used analogue func- for inappropriate mealtime behavior, and es-
tional analyses to assess the effects of caregiver cape and attention extinction were necessary to
consequences on child behavior. Inappropriate achieve a clinically acceptable outcome for one
mealtime behavior, such as batting at the spoon participant. For the other two participants, the
and head turning, produced attention, such as Piazza, Fisher, et al. procedure identified multiple
coaxing and brief verbal reprimands, during the reinforcers for inappropriate mealtime behavior,
attention condition; a break from the bite or drink but the Najdowski et al. procedure identified only
presentation during the escape condition; access to escape as the reinforcer for inappropriate mealtime
a tangible item, such as a preferred food, during behavior. For those two participants, the interven-
the tangible condition; and no differential conse- tion matched to the reinforcers identified by the
quence during the control condition. Escape from Piazza, Fisher, et al. procedure produced a clinically
bite or drink presentations functioned as negative acceptable outcome, but the intervention matched
reinforcement for the inappropriate mealtime be- to the reinforcer identified by the Najdowski et al.
havior of 9 of the 10 children who showed differ- procedure did not.
ential responding during the functional analyses. The differential responding demonstrated by
Access to adult attention or tangible items func- participants across test and control conditions
tioned as positive reinforcement for the inappro- in Bachmeyer et al. (2019), Girolami and Scotti
priate mealtime behavior of 8 of the 10 children (2001), Najdowski et al. (2008), and Piazza, Fisher,
who showed differential responding during the et al. (2003) suggest that even if the etiology of a
functional analyses. Girolami and Scotti (2001) pediatric feeding disorder is multiple and complex,
found that escape from food presentation and environmental events may reinforce inappropriate
mealtime demands for two children, and contin- mealtime behavior. This finding is important, be-
gent access to toys and attention for one child, cause (1) we may not be able to identify the cause
functioned as reinforcement for mealtime behav- of the child’s feeding disorder; (2) even if we iden-
ior problems such as aggression and spitting out tify the cause, that cause may be immutable, such
food. Najdowski et al. (2008) trained caregivers to as a history of prematurity; and (3) the underlying
conduct functional analyses in which a caregiver cause may not be related to the condition(s) that
placed a plate of nonpreferred food on the table maintains the behavior (Iwata et al., 1982/1994).
in front of a child in the attention, escape, and We can, however, change how we respond to the
tangible conditions and preferred food in the con- child’s inappropriate mealtime behavior, and such
trol condition. The caregiver washed dishes in the changes may be effective as treatment.
attention and control conditions and provided at-
tention following inappropriate mealtime behav-
ior in the attention condition. The caregiver sat TREATMENT
next to the child; provided continuous prompts to
“Take a bite”; followed the child with the plate if Evaluations of treatments based on theories of op-
the child left the chair; used three-step prompting; erant conditioning have formed the bulk of the in-
and removed the bite if the child engaged in inap- tervention research on pediatric feeding disorders.
propriate mealtime behavior in the demand con- Kerwin (1999) surveyed peer-reviewed medical
dition. Unlike in the Piazza, Fisher, et al. (2003) and psychological journals to identify studies that
and Girolami and Scotti (2001) studies, escape reported on psychosocial or behavioral interven-
was the only reinforcer identified for inappropriate tions for pediatric feeding disorders. She used the
mealtime behavior. Najdowski et al. noted that the modified criteria of the Task Force on Promotion
experimental preparation they used might have and Dissemination of Psychological Procedures
accounted for the difference in findings. (1995) to identify methodologically rigorous stud-
Bachmeyer, Kirkwood, Criscito, Mauzy, and ies that met the criteria and to classify interven-
Berth (2019) conducted two functional analyses tions for pediatric feeding disorders that were well
with three children with feeding disorders: one established, probably efficacious, or promising.
with the procedure Piazza, Fisher, et al. (2003) de- Analysis of the 29 studies that met the criteria
scribed and one with the procedure Najdowski et indicated that the only well-established interven-
al. (2008) described. Bachmeyer et al. then com- tions were behavioral interventions that included
A Behavior-Analytic Approach to Feeding Disorders 433
(1) positive reinforcement of appropriate feeding pediatric feeding disorders. Researchers have used
behavior and ignoring inappropriate mealtime be- differential reinforcement in several ways that
havior, and (2) positive reinforcement of appropri- include, but are not limited to, immediate (Patel,
ate feeding behavior and physical guidance of the Piazza, Martinez, Volkert, & Santana, 2002) or
appropriate feeding behavior (e.g., Ahearn, Ker- delayed (Kern & Marder, 1996; Riordan et al.,
win, Eicher, Shantz, & Swearingin, 1996; Kerwin, 1984) reinforcement with stimuli that research-
Ahearn, Eicher, & Burd, 1995; Linscheid, Oliver, ers selected arbitrarily (Kern & Marder, 1996;
Blyler, & Palmer, 1978; Piazza, Anderson, & Fish- Casey, Cooper-Brown, Wacker, & Rankin, 2006)
er, 1993; Riordan, Iwata, Finney, Wohl, & Stan- or with systematic preference assessments (Buck-
ley, 1984; Riordan, Iwata, Wohl, & Finney, 1980; ley, Strunck, & Newchok, 2005), or with tokens
Stark, Powers, Jelalian, Rape, & Miller, 1994). the child could exchange for meal discontinuation
Volkert and Piazza (2012) extended Kerwin (Kahng, Boscoe, & Byrne, 2003). For example, Pe-
(1999), using the same criteria to identify stud- terson, Volkert, and Zeleny (2015) used differen-
ies of interventions for pediatric feeding disorders tial reinforcement for two children with a feeding
and to categorize the level of empirical support for disorder to increase self-drinking from a cup. Re-
those interventions. Volkert and Piazza identified searchers conducted a multiple-stimulus-without-
74 studies that met the inclusion criteria. Analysis replacement assessment before each session and
of those studies showed that differential reinforce- used the three most preferred stimuli as reinforce-
ment of alternative behavior, escape extinction ment for self-drinking. The researchers increased
and putative escape extinction, and physical guid- the amount of liquid in the cup after the child’s
ance for self-feeding were well-established inter- self-drinking with the smaller amount increased
ventions; non-nutritive sucking (Field et al., 1982; with differential reinforcement.
Sehgal, Prakash, Gupta, Mohan, & Anand, 1990) Stark et al. (1996) randomly assigned nine chil-
and oral stimulation (Fucile, Gisel, & Lau, 2002; dren with cystic fibrosis to a behavioral interven-
Rocha, Moreira, Pimenta, Ramos, & Lucena, tion or a wait-list control group and used calories
2007) were probably efficacious interventions; and consumed and weight gain as the dependent vari-
oral support (Boiron, Nobrega, Roux, Henrot, & ables. The multicomponent behavioral interven-
Saliba, 2007; Einarsson-Backes, Deitz, Price, Glass, tion included caregiver praise, a star chart, and
& Hays, 1994), stimulus fading (e.g., Shore, Bab- access to privileges for appropriate feeding be-
bit, Williams, Coe, & Snyder, 1998), and simul- havior. Caloric intake increased to 1,032 calories
taneous presentation without escape extinction per day and mean weight gain was 1.7 kilograms
(e.g., Ahearn, 2003; Buckley & Newchok, 2005; for the group receiving the behavioral interven-
Piazza et al., 2002) were promising interventions. tion, compared to 244 calories and 0 kilograms,
Like Kerwin, Volkert and Piazza found that most respectively, for the control group. Participants
empirically supported interventions were behav- maintained higher levels of caloric intake relative
ior-analytic. Unlike Kerwin, however, they found to baseline at 3- and 6-month follow-ups. Other
that three interventions (non-nutritive sucking, researchers have used differential reinforcement
oral stimulation, and oral support) did not incor- alone and in combination with other procedures,
porate behavior-analytic principles or procedures. such as response cost (Kahng, Tarbox, & Wilke,
These studies appeared in journals from the fields 2001), to increase acceptance of solids (Werle,
of developmental medicine, occupational therapy, Murphy, & Budd, 1993), liquids (Kelley, Piazza,
otorhinolaryngology, and pediatrics, and focused & Fisher, 2003), or both (Roth, Williams, & Paul,
on increasing oral intake in premature infants. For 2010). One study demonstrated increased accep-
the purposes of this chapter, we review interven- tance and decreased self-injurious behavior during
tions aimed at increasing acceptance, decreasing noncontingent reinforcement (Wilder, Normand,
inappropriate mealtime behavior, decreasing ex- & Atwell, 2005).
pulsion, increasing mouth clean and decreasing Manipulating antecedents is another method
packing, and teaching chewing skills. that researchers have used to increase the food
or liquid acceptance of children with feeding dis-
orders. For example, Meier, Fryling, and Wallace
Acceptance
(2012) and Patel et al. (2007) used high-probability
Results of the analysis by Volkert and Piazza (2012) instructions, such as “Put an empty spoon in your
showed that differential reinforcement of alternative mouth,” that were like those of the target behavior
behavior was a well-established intervention for (e.g., “Take a bite”) to increase acceptance. Simul-
434 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
taneous presentation is an antecedent procedure in package. Piazza, Patel, Gulotta, Sevin, and Layer
which the feeder presents preferred food and non- (2003) attempted to clarify the relative contribu-
preferred food together, like a nonpreferred pea tions of positive reinforcement and putative escape
on a preferred chip. For example, Ahearn (2003) extinction. Differential positive reinforcement of
added condiments to increase vegetable consump- mouth clean, in which the feeder provided atten-
tion, and Tiger and Hanley (2006) added choco- tion and tangible items for mouth clean and inap-
late to milk to increase milk consumption. Piazza propriate mealtime behavior produced escape, was
et al. (2002) compared simultaneous and sequential not effective for increasing acceptance or decreas-
presentation in an extension of Kern and Marder ing inappropriate mealtime behavior. Acceptance
(1996). The feeder presented a bite of nonpreferred increased and inappropriate mealtime behavior
food on a bite of preferred food in the simultaneous decreased only when the feeder implemented pu-
condition or presented the bite of preferred food if tative escape extinction. Inappropriate mealtime
the child ate the bite of nonpreferred food in the behavior and negative vocalizations were lower for
sequential condition. Acceptance increased for some children during differential reinforcement
two of three participants in the simultaneous but and putative escape extinction relative to putative
not the sequential condition. Acceptance for the escape extinction alone, but the differences were
third participant increased when the feeder imple- often small or not replicated in subsequent phases.
mented simultaneous presentation and physical Piazza, Patel, et al. concluded that putative escape
guidance, but not when the feeder implemented extinction was necessary to increase acceptance
sequential presentation and physical guidance. and decrease inappropriate mealtime behavior, but
that differential positive reinforcement for mouth
clean may have contributed to lower levels of in-
Acceptance and Inappropriate Mealtime Behavior
appropriate mealtime behavior, negative vocaliza-
Before researchers began conducting functional tions, or both for some children when combined
analyses of inappropriate mealtime behavior, they with putative escape extinction. A study by Reed
developed procedures for putative escape extinc- et al. (2004) produced similar results for noncon-
tion, based on the assumption that escape from tingent reinforcement and putative escape extinc-
bites or drinks functioned as negative reinforce- tion, except that the effects of noncontingent re-
ment for inappropriate mealtime behavior. The inforcement on inappropriate mealtime behavior
procedures researchers have studied most often and negative vocalizations were less robust than
are nonremoval of the spoon (Hoch, Babbitt, Coe, those for differential positive reinforcement in Pi-
Krell, & Hackbert, 1994) and physical guidance azza, Patel, et al. (2003).
(Ahearn et al., 1996), and both involve discon- Peterson, Piazza, Ibañez, and Fisher (2019) con-
tinuing the hypothesized response–reinforcer rela- ducted a randomized controlled trial to evaluate
tion. Ahearn et al. (1996) showed that nonremov- the effects of a behavior-analytic intervention
al of the spoon and physical guidance increased (noncontingent reinforcement and nonremoval of
acceptance. During nonremoval of the spoon, the the spoon) relative to a wait-list control to deter-
feeder held the bite near the child’s lips until the mine whether the food selectivity of young chil-
feeder could deposit the bite into the mouth. Dur- dren with autism spectrum disorder would resolve
ing physical guidance, the feeder applied gentle over time without intervention. The researchers
pressure to the mandibular junction of the jaw to randomly assigned three children to the behavior-
open the mouth and deposited the bite if the child analytic intervention and three children to the
did not accept it. Ahearn et al. assessed caregiver wait-list control group. Consumption increased for
acceptability for the interventions by asking each the three children in the behavior-analytic group,
caregiver which treatment he or she preferred. All but not for the children in the wait-list control
caregivers chose physical guidance, which was as- group.
sociated with fewer corollary behaviors (such as Researchers based the behavior-analytic inter-
disruptions) for all children and with shorter meal ventions described above on the assumption that
durations for two of the three children. escape from bites or drinks would function as re-
Studies by Hoch et al. (1994) and Ahearn et inforcement for inappropriate mealtime behavior
al. (1996) are representative of other studies on in the absence of a formal functional analysis
putative escape extinction, in that researchers (Ahearn et al., 1996; Cooper et al., 1995; Hoch
included differential or noncontingent reinforce- et al., 1994). By contrast, Bachmeyer et al. (2009)
ment (Cooper et al., 1995) in an intervention used a functional analysis to determine that es-
A Behavior-Analytic Approach to Feeding Disorders 435
cape from bites or drinks and adult attention re- an increase in acceptance and mouth clean and
inforced the inappropriate mealtime behavior of decreased inappropriate mealtime behavior. The
four children with feeding disorders. Bachmeyer researchers gradually increased the ratio of non-
et al. then evaluated the effects of variations of preferred to preferred food until the children were
extinction that matched one or both functional consuming the nonpreferred food alone.
reinforcers: (1) escape extinction and attention Mueller et al. (2004) used a fading component
following inappropriate mealtime behavior, (2) to achieve the targeted outcome; other researchers
attention extinction and escape following inap- have used various fading procedures with nonre-
propriate mealtime behavior, and (3) escape ex- moval of the spoon or physical guidance. Groff,
tinction and attention extinction. Results showed Piazza, Volkert, and Jostad (2014) used a syringe
that variations of extinction that discontinued to deposit solids and liquids into the mouth of a
delivery of the reinforcers for inappropriate meal- child who clenched his teeth during presentation,
time behavior identified by the functional analysis gradually increased the volume of solids and liq-
(escape and attention), were necessary to reduce uids in the syringe, and then faded from syringe to
inappropriate mealtime behavior to clinically ac- spoon for solids and syringe to cup for liquids. Dur-
ceptable rates and to increase acceptance to high, ing fading, the researchers taped the syringe to a
stable levels. LaRue et al. (2011) tested the effects spoon or a cup, so the tip of the syringe protruded
of a negative-reinforcement-based intervention from the tip of the spoon or lip of the cup by 5 cen-
(differential negative reinforcement and nonre- timeters, and moved the tip of the syringe and the
moval of the spoon with re-presentation) with 11 tip of the spoon or lip of the cup closer together.
children whose inappropriate mealtime behavior The child began eating from a spoon and drinking
was maintained by escape. Mouth clean produced from the cup without the syringe during probe ses-
a 30-second break from bite or drink presentations sions the feeder conducted between fading steps.
during differential negative reinforcement. Ac- Other dimensions on which researchers have con-
ceptance increased and inappropriate mealtime ducted fading include bite number (Najdowski,
behavior decreased when the feeder implemented Wallace, Doney, & Ghezzi, 2003), bottle to spoon
nonremoval of the spoon and re-presentation; dif- (Johnson & Babbitt, 1993), high- to low-probabil-
ferential negative reinforcement for mouth clean ity demands (Penrod, Gardella, & Fernand, 2012),
had no effect on behavior, either alone or in com- liquid to baby food (Bachmeyer, Gulotta, & Piazza,
bination with nonremoval of the spoon and re- 2013), preferred to nonpreferred liquid type (Lu-
presentation. iselli, Ricciardi, & Gilligan, 2005), liquid volume
The studies on escape extinction reviewed (Hagopian, Farrell, & Amari, 1996), portion size
above included consequence manipulations such (Freeman & Piazza, 1998), spoon distance (Rivas,
as differential reinforcement for acceptance and Piazza, Patel, & Bachmeyer, 2010), food variety
nonremoval of the spoon (e.g., Piazza, Patel, et al., (Valdimarsdottir, Halldorsdottir, & Sigurådóttir,
2003; Reed et al., 2004). Researchers also have 2010), spoon to cup (Babbitt, Shore, Smith, Wil-
tested antecedent manipulations with escape ex- liams, & Coe, 2001), and texture (Luiselli & Glea-
tinction or putative escape extinction to increase son, 1987; Shore et al., 1998).
acceptance. Patel et al. (2006) combined high-
probability instructions that were like those for
Alternative Interventions
the low-probability instruction with nonremoval
of the spoon to increase acceptance. Dawson et al. Recent research has compared the effects of be-
(2003), by contrast, showed that high-probability havior-analytic interventions for pediatric feed-
instructions did not differentially affect levels of ing disorders to ones that are popular among
acceptance when combined with nonremoval of non-behavior-analytic professionals. For example,
the spoon. Addison et al. (2012) compared the effects of a
Mueller, Piazza, Patel, Kelley, and Pruett (2004) behavior-analytic intervention to a sensory-inte-
used blending, a variation of simultaneous presen- gration intervention. A speech therapist and two
tation, with differential or noncontingent posi- occupational therapists developed individual-
tive reinforcement and nonremoval of the spoon ized sensory-integration interventions for the two
to increase consumption for two children with a participants. Acceptance and amount consumed
feeding disorder. Blending or mixing preferred and increased and inappropriate mealtime behavior
nonpreferred foods, such as preferred yogurt mixed decreased during the behavior-analytic but not
with nonpreferred green-bean puree, produced the sensory-integration intervention. Peterson et
436 S u b s p e c i a lt ie s in A p p l ie d B e h av i o r A n a ly s i s
al. (2016) conducted a randomized controlled trial seating, in which the feeder reclined the highchair
of a behavior-analytic versus a modified sequential- from its upright position. Expulsion decreased and
oral-sensory approach to treat the food selectivity of was equivalent for the two interventions. Patel, Pi-
six young children with autism spectrum disorder. azza, Santana, and Volkert (2002) evaluated the
Peterson et al. randomly assigned three children effects of type and texture of food (Munk & Repp,
to the behavior-analytic intervention and three 1994) on expulsion. Rates of expulsion were high-
children to the modified sequential-oral-sensory er when the feeder presented meat relative to other
intervention, compared the effects across novel, foods, and expulsion decreased when the research-
healthy target foods, and tested for generalization ers lowered the texture of meat.
of intervention effects. Consumption of novel, Utensil manipulation is another strategy research-
healthy target foods increased for the children in ers have used to decrease expulsion (Dempsey, Pi-
the behavior-analytic intervention group, but not azza, Groff, & Kozisek, 2011; Gulotta, Piazza, Patel,
for the children in the modified sequential-oral- & Layer, 2005; Hoch, Babbitt, Coe, Duncan, &
sensory group. Peterson et al. then implemented Trusty, 1995; Volkert, Vaz, Piazza, Frese, & Bar-
the behavior-analytic intervention with the chil- nett, 2011). For example, Girolami, Boscoe, and
dren previously assigned to the modified sequen- Roscoe (2007) showed that presenting and re-pre-
tial-oral-sensory intervention group. Consumption senting bites on a Nuk, which is a bristled utensil
of novel, healthy target foods increased during the caregivers use to initiate toothbrushing with in-
behavior-analytic intervention, and Peterson et al. fants, reduced expels relative to presentation and
observed a potential generalization effect for foods re-presentation of bites on a spoon or on a spoon
that had been exposed to the modified sequential- and a Nuk, respectively. Wilkins et al. (2014) com-
oral-sensory intervention. pared presentation of bites on a spoon or a Nuk
with 12 children during initial intervention. Ac-
ceptance increased and inappropriate mealtime
Expulsion
behavior decreased for 8 of 12 children. Five of the
Results of functional analysis studies show that es- 8 had lower levels of expulsion, and 4 of the 8 had
cape from bites or drinks functions as negative re- higher levels of mouth clean, when the feeder pre-
inforcement for inappropriate mealtime behavior, sented bites on the Nuk.
but inappropriate mealtime behavior is probably
not the only behavior that produces escape. For
Mouth Clean and Pack
example, Coe et al. (1997) used nonremoval of the
spoon to increase the acceptance of two children Hoch et al. (1994) proposed that the feeder should
with a feeding disorder and observed simultane- provide reinforcement for a behavior that occurs
ous increases in expulsion. Coe et al. hypothesized early in the chain of feeding behaviors, such as
that negative reinforcement in the form of escape acceptance, and then shift reinforcement to a be-
from swallowing food reinforced expulsion. Re-pre- havior that occurs later in the chain, such as swal-
sentation, or scooping up expelled food and placing lowing. To that end, Patel, Piazza, Martinez, et al.
it back into the mouth or getting a new bite of the (2002) compared the effects of differential positive
same food, resulted in near-zero levels of expulsion. reinforcement for acceptance versus mouth clean.
Although Coe et al. (1997) and Sevin, Gulotta, When differential reinforcement did not increase
Sierp, Rosica, and Miller (2002) demonstrated acceptance or mouth clean, the feeder added pu-
the effectiveness of re-presentation, others have tative escape extinction. Acceptance and mouth
found that re-presentation was not effective con- clean increased and inappropriate mealtime be-
sistently. For example, Wilkins et al. (2011) added havior decreased. Patel et al. concluded that the
a chin prompt when nonremoval of the spoon plus point in the chain in which the feeder provided
re-presentation did not decrease expulsion. During differential reinforcement was not as important as
the chin prompt, the feeder placed gentle upward putative escape extinction for increasing accep-
pressure on the child’s chin as the feeder deposited tance and mouth clean and decreasing inappropri-
the bite or drink during re-presentation, which ate mealtime behavior.
reduced expulsion. Shalev, Milnes, Piazza, and Two consequence-based interventions re-
Kozisek (2018) compared a modified chin prompt, searchers have used to increase mouth clean and
in which the feeder waited for the child’s jaw to decrease packing are redistribution (Girolami et
relax and then placed gentle upward pressure on al., 2007; Gulotta et al., 2005; Levin, Volkert,
the chin while depositing the drink, with reclined & Piazza, 2014; Sevin et al., 2002; Stubbs, Volk-
A Behavior-Analytic Approach to Feeding Disorders 437
ert, Rubio, & Ottinger, 2017) and a chaser (Vaz, had mouth clean with some pureed foods but not
Piazza, Stewart, Volkert, & Groff, 2012). For ex- others. The researchers pureed food in a smoothie
ample, Sevin et al. (2002) used a Nuk brush dur- blender during a second texture assessment, and
ing redistribution to remove packed food from the levels of mouth clean were higher than with the
participant’s mouth and place the food on the other textures.
tongue. Redistribution increased mouth clean and Finally, researchers have used fading to increase
decreased packing. Volkert et al. (2011) obtained mouth clean and decrease pack. For example, the
similar results by using a flipped spoon during redis- goal for the child in Patel, Piazza, Kelly, Ochsner,
tribution. The feeder removed packed food with a and Santana (2001) was to increase his intake of
spoon, inserted the spoon with the bite into the a calorically dense beverage, Carnation Instant
participant’s mouth, turned the spoon 180°, and Breakfast with whole milk. The child refused the
dragged the bowl of the spoon along the tongue breakfast drink, but he did drink water. Therefore,
toward the lips to deposit the previously packed the researchers added and gradually increased the
bite. The feeder in Vaz et al. (2012) gave the child amount of the drink powder in water and subse-
a chaser (a liquid or solid the child consistently quently replaced the water with milk. Other di-
accepted and swallowed) to reduce packing. The mensions along which researchers have faded to
feeder presented the chaser either immediately increase mouth clean and decrease pack are liquid
after he or she deposited the target bite into the to baby food (Bachmeyer et al., 2013), spoon to cup
child’s mouth for two children, or 15 seconds after (Groff, Piazza, Zeleny, & Dempsey, 2011), syringe
he or she deposited the target bite for a third child. to spoon, and syringe to cup (Groff et al., 2014).
Researchers in the studies described above used
a Nuk or a flipped spoon to redistribute packed
Chewing
food. Researchers have also evaluated the effects
of utensil manipulation as an antecedent inter- Chewing is a skill that emerges in typically eating
vention (Dempsey et al., 2011; Sharp, Odom, & children as the caregiver increases the texture of
Jaquess, 2012; Stubbs et al., 2017). For example, presented food, which is usually around 12 months
Sharp, Harker, and Jaquess (2010) compared the of age. In our experience, many children with feed-
effects of presentation on an upright spoon, a ing disorders do not begin chewing at or after 12
flipped spoon, or a Nuk. Levels of mouth clean in- months of age when the caregiver increases food
creased for the flipped spoon and Nuk but were not texture. Nevertheless, caregivers often base the
clinically acceptable. Other studies have shown texture of presented food on a child’s age rather
clinically acceptable increases in mouth clean and than the child’s chewing skills. A mismatch be-
decreases in packing with flipped-spoon presen- tween the texture of presented food and the child’s
tation (Rivas, Piazza, Kadey, Volkert, & Stewart, chewing skills increases the risk of aspiration, par-
2011; Sharp et al., 2012; Stubbs et al., 2017), Nuk ticularly if the child swallows the food without
presentation (Gulotta et al., 2005; Sevin et al., masticating it sufficiently (Patel et al., 2005). Chil-
2002), or a combination of flipped spoon and chin dren who lack appropriate chewing skills may de-
prompt (Dempsey et al., 2011). velop inappropriate compensatory behavior, such
Texture or food consistency is another anteced- as using the tongue to push food against the roof
ent variable that affects levels of mouth clean and of the mouth. We often see this behavior emerge
pack (Bachmeyer et al., 2013; Kadey et al., 2013; when the caregiver presents meltable solids, such
Patel, Piazza, Layer, Coleman, & Swartzwelder, as cookies, crackers, and chips. The child learns
2005; Sharp & Jaquess, 2009). For example, Kadey that he or she can use the tongue to moisten and
et al. (2013) assessed food texture and food type break apart the meltable solid, and this behavior
to identify potential causes of one young girl’s does not change when the caregiver presents foods
packing. First, the researchers compared levels of that do not melt or break apart with saliva, such as
mouth clean with chopped food (table food cut meats. These children reach an impasse in which
into small pieces), wet ground food (small chunks they consume meltable solids and small amounts
of food in a wet medium), and pureed food (table of more difficult foods, such as pizza, but they can-
food blended until smooth). The results showed not advance any further. They often have exces-
that levels of mouth clean were highest with pu- sive meal lengths and do not consume sufficient
reed food, but even those levels were not accept- calories for weight gain and growth, because their
able clinically. When the researchers presented chewing skills are not efficient and effective. We
foods individually, they determined that the child have found that teaching a child to chew is the
438 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
strategy that is most effective for advancing tex- implement a feeding intervention with above 90%
ture. integrity; the researchers used verbal and written
For example, our group evaluated a multicom- instructions, modeling, video review, and perfor-
ponent intervention to increase chews per bite, as- mance feedback during and after in-home feeding
sess mastication, and eliminate early swallowing, services. Mueller et al. (2003) evaluated four dif-
which observers scored at the mastication check ferent multicomponent training packages to in-
if no food was visible in the mouth and the food crease intervention integrity for caregivers imple-
was not absent because of expulsion (Volkert et al., menting pediatric feeding interventions. In Study
2014). Caregivers served as feeders and used grad- 1, written protocols, verbal instructions, therapist
uated verbal, model, and physical prompting to modeling, and rehearsal training increased care-
teach the children in sequential steps to chew (1) givers’ intervention integrity to high levels. Muel-
on an empty 7.6-centimeter piece of airline tubing ler et al. then examined the effects of the train-
that was 0.6 centimeters in diameter, (2) on a bite ing package’s components in Study 2. Mueller et
of food measuring 0.6 × 0.6 × 0.6 centimeters in al. assigned six caregivers to one of three train-
the tube, (3) on a strip of food measuring 0.6 × 0.6 ing conditions with two caregivers per condition:
× 5.1 centimeters on half of a tube, and (4) on a written protocols and verbal instructions; written
strip of food measuring 0.6 × 0.6 × 5.1 centimeters. protocols, verbal instructions, and modeling; and
Final steps included presenting a bite of food mea- written instructions, verbal instructions, and re-
suring 0.6 × 0.6 × 0.6 centimeters and increasing hearsal. Each training package produced high lev-
bite size for one child. els of intervention integrity and maintenance over
Volkert et al. (2013) noted that one limitation a 3-month period. Other researchers have taught
of the chewing literature is that studies have not caregivers to use general and specific prompts
measured mastication. Ensuring that a child has (Pangborn, Borrero, & Borrero, 2013; Werle et al.,
masticated accepted food is important for mini- 1993), functional-analysis procedures (Najdowski
mizing aspiration risk. Volkert et al. used a vocal et al., 2003, 2008), intervention (e.g., demand fad-
prompt, “Chew X times,” for one child and provid- ing, differential reinforcement, escape extinction;
ed praise if the child met the chew criterion. The Anderson & McMillan, 2001; Najdowski et al.,
feeder checked 30 seconds after the bite entered 2003, 2010; Pangborn et al., 2013; Seiverling, Wil-
the child’s mouth to determine if the child masti- liams, Sturmey, & Hart, 2012), and data collection
cated the bite, which Volkert et al. defined as food (Najdowski et al., 2003).
with pieces no larger than 0.2 × 0.2 centimeters in
a liquid medium after chewing. Chews per bite and
mastication increased during the intervention. CONCLUSION
and depression (Franklin & Rodger, 2003; Singer, we know little about what causes these behaviors.
Sing, Hill, & Jaffe, 1990) and are financially costly Are they part of a chain of escape and avoidance
to the families and to society (Nebraska Legisla- behaviors, and as we extinguish one behavior,
ture, 2009; Williams, Riegel, Gibbons, & Field, does another behavior emerge to take its place?
2007). Are they the result of an oral–motor skill deficit,
The etiology of feeding disorders is multiply in which the child lacks the skills to manage sol-
controlled and complex (Rommel et al., 2003), and ids or liquids effectively? Are they the result of a
the behaviors that constitute a feeding disorder combined etiology? The systematic, data-based ap-
are heterogeneous. Current diagnostic nosologies proach that behavior analysts use is ideal for an-
describe the characteristics of a feeding disorder swering these questions, but we have yet to apply
but are not prescriptive. Historically, researchers them to pediatric feeding disorders.
have hypothesized that escape from feeding func- Finally, intervention for a feeding disorder re-
tions as negative reinforcement for inappropriate quires knowledge that extends far beyond applied
mealtime behavior, based on the results of studies behavior analysis. A behavior analyst should rec-
in which putative escape extinction was effective ognize when to consult with another professional,
for increasing acceptance and decreasing inap- such as an allergist, pediatric gastroenterologist,
propriate mealtime behavior (Cooper et al., 1995; or speech and language pathologist. Inadequate
Hoch et al., 1994; Kerwin et al., 1995; Patel, Piazza, training can lead to mistakes in therapy that can
Martinez, et al., 2002; Piazza, Patel, et al., 2003; have serious consequences, such as anaphylaxis
Reed et al., 2004). Functional analysis studies have due to cross-contamination, aspiration when a
confirmed that escape functioned as negative rein- child is not a safe oral feeder for the presented food
forcement for the inappropriate mealtime behav- or liquid, or choking because the presented tex-
ior of most children in those studies (Allison et al., ture is inappropriate for the child’s chewing skills.
2012; Bachmeyer et al., 2009; Girolami & Scot- “Knowing what you don’t know” is an essential
ti, 2001; Kirkwood, Piazza, & Peterson, in press; skill for behavior analysts working with children
LaRue et al., 2011; Najdowski et al., 2008; Piazza, with feeding disorders.
Fisher, et al., 2003), but some studies have found
that adult attention and tangible items functioned
as reinforcement too (Bachmeyer et al., 2009; Gi- REFERENCES
rolami & Scotti, 2001; Piazza, Fisher, et al., 2003).
There are only a few studies of function-based in- Addison, L. R., Piazza, C. C., Patel, M. E., Bachmeyer,
terventions (Allison et al., 2012; Bachmeyer et al., M. H., Rivas, K. M., Milnes, S. M., et al. (2012). A
2009; Bachmeyer et al., 2019; Kirkwood et al., in comparison of sensory integrative and behavioral
press; LaRue et al., 2011; Najdowski et al., 2003). therapies as treatment for pediatric feeding disorders.
Results of these studies suggest that we can use the Journal of Applied Behavior Analysis, 45, 455–471.
Ahearn, W. H. (2003). Using simultaneous presentation
results of a functional analysis of inappropriate
to increase vegetable consumption in a mildly selec-
mealtime behavior to prescribe effective interven-
tive child with autism. Journal of Applied Behavior
tions to increase acceptance and decrease inap- Analysis, 36, 361–365.
propriate mealtime behavior. Ahearn, W. H., Kerwin, M. E., Eicher, P. S., Shantz,
The complexity of feeding disorders, however, J., & Swearingin, W. (1996). An alternating treat-
necessitates that we refine our functional-analysis ments comparison of two intensive interventions for
procedures to identify the characteristics of the food refusal. Journal of Applied Behavior Analysis, 29,
feeding environment that establish escape or other 321–332.
stimuli as reinforcement. Results of studies on fad- Allison, J., Wilder, D. A., Chong, I., Lugo, A., Pike, J., &
ing suggest that we can alter the properties of Rudy, N. (2012). A comparison of differential rein-
stimuli associated with high levels of acceptance forcement and noncontingent reinforcement to treat
and low levels of inappropriate mealtime behavior food selectivity in a child with autism. Journal of Ap-
plied Behavior Analysis, 45, 613–617.
to produce appropriate feeding. Can we identify
American Psychiatric Association. (2013). Diagnostic
those antecedent variables a priori? Most children and statistical manual of mental disorders (5th ed.).
emit some form of appropriate behavior. How can Arlington, VA: Author.
we identify that behavior and use it to prescribe an Anderson, C. M., & McMillan, K. (2001). Parental use
effective feeding intervention? of escape extinction and differential reinforcement
Similarly, although behavior analysts have eval- to treat food selectivity. Journal of Applied Behavior
uated procedures to reduce expulsion and packing, Analysis, 34, 511–515.
440 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Arvedson, J. C., & Brodsky, L. (2002). Pediatric swal- Peck, S. M., Richman, D., et al. (1995). Use of com-
lowing and feeding: Assessment and management (2nd ponent analyses to identify active variables in treat-
ed.). Albany, NY: Singular. ment packages for children with feeding disorders.
Babbitt, R. L., Shore, B. A., Smith, M., Williams, K. E., Journal of Applied Behavior Analysis, 28, 139–153.
& Coe, D. A. (2001). Stimulus fading in the treat- Darrow, D. H., & Harley, C. M. (1998). Evaluation of
ment of adipsia. Behavioral Interventions, 16, 197–207. swallowing disorders in children. Otolaryngologic
Bachmeyer, M. H., Gulotta, C. S., & Piazza, C. C. Clinics of North America, 31, 405–418.
(2013). Liquid to baby food fading in the treatment Davis, A. M., Bruce, A., Cocjin, J., Mousa, H., &
of food refusal. Behavioral Interventions, 28, 281–298. Hyman, P. (2010). Empirically supported treatments
Bachmeyer, M. H., Kirkwood, C. A., Criscito, A. B., for feeding difficulties in young children. Current
Mauzy, C. R., & Berth, D. P. (2019). A comparison of Gastroenterology Reports, 12, 189–194.
functional analysis methods of inappropriate meal- Dawson, J. E., Piazza, C. C., Sevin, B. M., Gulotta, C.
time behavior. Journal of Applied Behavior Analysis, S., Lerman, D., & Kelley, M. L. (2003). Use of the
52, 603–621. high-probability instructional sequence and escape
Bachmeyer, M. H., Piazza, C. C., Frederick, L. D., Reed, extinction in a child with food refusal. Journal of Ap-
G. K., Rivas, K. D., & Kadey, H. J. (2009). Functional plied Behavior Analysis, 36, 105–108.
analysis and treatment of multiply controlled inap- Dempsey, J., Piazza, C. C., Groff, R. A., & Kozisek, J. M.
propriate mealtime behavior. Journal of Applied Be- (2011). A flipped spoon and chin prompt to increase
havior Analysis, 42, 641–658. mouth clean. Journal of Applied Behavior Analysis, 44,
Bandini, L. G., Anderson, S. E., Curtin, C., Cermak, S., 961–965.
Evans, E. W., Scampini, R., et al. (2010). Food selec- Derkay, C. S., & Schechter, G. L. (1998). Anatomy and
tivity in children with autism spectrum disorders and physiology of pediatric swallowing disorders. Otolar-
typically developing children. Journal of Pediatrics, yngologic Clinics of North America, 31, 397–404.
157, 259–264. Diquelou, M. C., Griffin, A. S., & Sol, G. D. (2016).
Boiron, M., Nobrega, L. D., Roux, S., Henrot, A., & The role of motor diversity in foraging innovations:
Saliba, E. (2007). Effects of oral stimulation and oral A cross-species comparison in urban birds. Behavioral
support on non-nutritive sucking and feeding perfor- Ecology, 27, 584–591.
mance in preterm infants. Developmental Medicine Douglas, J. E., & Byron, M. (1996). Interview data on
and Child Neurology, 49, 439–444. severe behavioural eating difficulties in young chil-
Borrero, C. S. W., Woods, J. N., Borrero, J. C., Masler, dren. Archives of Disease in Childhood, 75, 304–308.
E. A., & Lesser, A. D. (2010). Descriptive analysis of Einarsson-Backes, L. M., Dietz, J., Price, R., Glass, R., &
pediatric food refusal and acceptance. Journal of Ap- Hays, R. (1994). The effect of oral support on suck-
plied Behavior Analysis, 43, 71–88. ing efficiency in preterm infants. American Journal of
Borsari, A., & Ottoni, E. B. (2005). Preliminary obser- Occupational Therapy, 48, 490–498.
vations of tool use in captive hyacinth macaws (Ano- Field, T., Ignatoff, E., Stinger, S., Brennan, J., Greenberg,
dorhynchus hyacinthinus). Animal Cognition, 8, 48–53. R., Widmayer, S., et al. (1982). Nonnutritive sucking
Buckley, S. D., & Newchok, D. K. (2005). An evalua- during tube feedings: Effects on preterm neonates in
tion of simultaneous presentation and differential an intensive care unit. Pediatrics, 70, 381–384.
reinforcement with response cost to reduce packing. Franklin, L., & Rodger, S. (2003). Parents’ perspectives
Journal of Applied Behavior Analysis, 38, 405–409. on feeding medically compromised children: Impli-
Buckley, S. D., Strunck, P. G., & Newchok, D. K. (2005). cations for occupational therapy. Australian Occupa-
A comparison of two multicomponent procedures to tional Therapy Journal, 50, 137–147.
increase food consumption. Behavioral Interventions, Freedman, D. S., Dietz, W. H., Srinivasan, S. R., & Be-
20, 139–146. renson, G. S. (1999). The relation of overweight to
Burklow, K. A., McGrath, A. M., Valerius, K. S., & cardiovascular risk factors among children and ado-
Rudolph, C. (2002). Relationships between feeding lescents: The Bogalusa Heart Study. Pediatrics, 103,
difficulties, medical complexity, and gestational age. 1175–1182.
Nutrition in Clinical Practice, 17, 373–378. Freeman, K. A., & Piazza, C. C. (1998). Combining
Casey, S. D., Cooper-Brown, L. J., Wacker, D. P., & stimulus fading, reinforcement, and extinction to
Rankin, B. E. (2006). The use of descriptive analysis treat food refusal. Journal of Applied Behavior Analy-
to identify and manipulate schedules of reinforce- sis, 31, 691–694.
ment in the treatment of food refusal. Journal of Be- Fucile, S., Gisel, E., & Lau, C. (2002). Oral stimula-
havioral Education, 15, 41–52. tion accelerates the transition from tube to oral
Coe, D. A., Babbitt, R. A., Williams, K. E., Hajimihalis, feeding in preterm infants. Journal of Pediatrics, 141,
C., Snyder, A. M., Ballard, C., et al. (1997). Use of 230–236.
extinction and reinforcement to increase food con- Girolami, P. A., Boscoe, J. H., & Roscoe, N. (2007).
sumption and reduce expulsion. Journal of Applied Decreasing expulsions by a child with a feeding dis-
Behavior Analysis, 30, 581–583. order: Using a brush to present and re-present food.
Cooper, L. J., Wacker, D. P., McComas, J. J., Brown, K., Journal of Applied Behavior Analysis, 40, 749–753.
A Behavior-Analytic Approach to Feeding Disorders 441
Girolami, P. A., & Scotti, J. R. (2001). Use of ana- multicomponent treatment for food refusal. Journal of
log functional analysis in assessing the function of Applied Behavior Analysis, 34, 93–96.
mealtime problem behavior. Education and Training Kelley, M. E., Piazza, C. C., & Fisher, W. W. (2003).
in Mental Retardation and Developmental Disabilities, Acquisition of cup drinking using previously refused
36, 207–223. foods as positive and negative reinforcement. Journal
Goto, A., Arah, O. A., Goto, M., Terauchi, Y., & Noda, of Applied Behavior Analysis, 36, 89–93.
M. (2013). Severe hypoglycemia and cardiovascular Kern, L., & Marder, T. J. (1996). A comparison of simul-
disease: Systematic review and meta-analysis with taneous and delayed reinforcement as treatments for
bias analysis. British Journal of Medicine, 347, 1–11. food selectivity. Journal of Applied Behavior Analysis,
Groff, R. A., Piazza, C. C., Volkert, V. M., & Jostad, 2, 243–246.
C. M. (2014). Syringe fading as treatment for feed- Kerwin, M. E. (1999). Empirically supported treatments
ing refusal. Journal of Applied Behavior Analysis, 47, in pediatric psychology: Severe feeding problems.
834–839. Journal of Pediatric Psychology, 24, 193–214.
Groff, R. A., Piazza, C. C., Zeleny, J. R., & Dempsey, J. Kerwin, M. E., Ahearn, W. H., Eicher, P. S., & Burd,
R. (2011). Spoon-to-cup fading as treatment for cup D. M. (1995). The costs of eating: A behavioral eco-
drinking in a child with intestinal failure. Journal of nomic analysis of food refusal. Journal of Applied Be-
Applied Behavior Analysis, 44, 949–954. havior Analysis, 28, 245–260.
Gulotta, C. S., Piazza, C. C., Patel, M. R., & Layer, S. A. Kirkwood, C. A., Piazza, C. C., & Peterson, K. M. (in
(2005). Using food redistribution to reduce packing press). A comparison of function- and nonfunction-
in children with severe food refusal. Journal of Ap- based extinction for inappropriate mealtime behav-
plied Behavior Analysis, 38, 39–50. ior. Journal of Applied Behavior Analysis.
Hadley, A. J., Krival, K. R., Ridgel, A. L., Hahn, E. C., & LaRue, R. H., Stewart, V., Piazza, C. C., Volkert, V. M.,
Tyler, D. J. (2015). Neural network pattern recogni- Patel, M. R., & Zeleny, J. (2011). Escape as reinforce-
tion of lingual–palatal pressure for automated detec- ment and escape extinction in the treatment of feed-
tion of swallow. Dysphagia, 30, 176–187. ing problems. Journal of Applied Behavior Analysis,
Hagopian, L. P., Farrell, D. A., & Amari, A. (1996). 44, 719–735.
Treating total liquid refusal with backward chaining Levin, D. S., Volkert, V. M., & Piazza, C. C. (2014). A
and fading. Journal of Applied Behavioral Analysis, 29, multi-component treatment to reduce packing in
573–575. children with feeding and autism spectrum disorders.
Hoch, T. A., Babbitt, R. L., Coe, D. A., Duncan, A., & Behavior Modification, 38, 940–963.
Trusty, E. M. (1995). A swallow induction avoidance Linscheid, T. R., Oliver, J., Blyler, E., & Palmer, S.
procedure to establish eating. Journal of Behavior (1978). Brief hospitalization for the behavioral treat-
Therapy and Experimental Psychiatry, 26, 41–50. ment of feeding problems in the developmentally
Hoch, T. A., Babbitt, R. L., Coe, D. A., Krell, D. M., & disabled. Journal of Pediatric Psychology, 12, 451–459.
Hackbert, L. (1994). Contingency contacting: Com- Luiselli, J. K., & Gleason, D. L. (1987). Combining sen-
bining positive reinforcement and escape extinction sory reinforcement and texture fading procedures to
procedures to treat persistent food refusal. Behavior overcome chronic food refusal. Journal of Behavior
Modification, 18, 106–128. Therapy and Experimental Psychiatry, 18, 149–155.
Ibañez, V. F., Piazza, C. C., & Peterson, K. M. (2019). Luiselli, J. K., Ricciardi, J. N., & Gilligan, K. (2005). Liq-
A translational evaluation of renewal of inappropri- uid fading to establish milk consumption by a child
ate mealtime behavior. Journal of Applied Behavior with autism. Behavioral Interventions, 20, 155–163.
Analysis, 52, 1005–1020. Lukens, C. T., & Silverman, A. H. (2014). Systematic
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & review of psychological interventions for pediatric
Richman, G. S. (1994). Toward a functional analysis feeding problems. Journal of Pediatric Psychology, 39,
of self-injury. Journal of Applied Behavior Analysis, 27, 903–917.
197–209. (Reprinted from Analysis and Intervention in Meier, A. E., Fryling, M. J., & Wallace, M. D. (2012).
Developmental Disabilities, 2, 3–20, 1982) Using high-probability foods to increase the accep-
Johnson, C. R., & Babbitt, R. L. (1993). Antecedent tance of low-probability foods. Journal of Applied Be-
manipulation in the treatment of primary solid food havior Analysis, 45, 149–153.
refusal. Behavior Modification, 17, 510–521. Mueller, M. M., Piazza, C. C., Moore, J. W., Kelley, M.
Kadey, H., Piazza, C. C., Rivas, K. M., & Zeleny, J. E., Bethke, S. A., Pruett, A., et al. (2003). Training
(2013). An evaluation of texture manipulations parents to implement pediatric feeding protocols.
to increase swallowing. Journal of Applied Behavior Journal of Applied Behavior Analysis, 36, 545–562.
Analysis, 46, 539–543. Mueller, M. M., Piazza, C. C., Patel, M. R., Kelley, M.
Kahng, S., Boscoe, J. H., & Byrne, S. (2003). The use E., & Pruett, A. (2004). Increasing variety of foods
of an escape contingency and a token economy to consumed by blending nonpreferred foods into pre-
increase food acceptance. Journal of Applied Behavior ferred foods. Journal of Applied Behavior Analysis, 37,
Analysis, 36, 349–353. 159–170.
Kahng, S., Tarbox, J., & Wilke, A. E. (2001). Use of a Munk, D. D., & Repp, A. C. (1994). Behavioral assess-
442 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
ment of feeding problems of individuals with severe Penrod, B., Gardella, L., & Fernand, J. (2012). An evalu-
disabilities. Journal of Applied Behavior Analysis, 27, ation of a progressive high-probability instructional
241–250. sequence combined with low-probability demand
Najdowski, A. C., Wallace, M. D., Doney, J. K., & Ghe- fading in the treatment of food selectivity. Journal of
zzi, P. M. (2003). Parental assessment and treatment Applied Behavior Analysis, 45, 527–537.
of food selectivity in natural settings. Journal of Ap- Peterson, K. M., Piazza, C. C., Ibañez, V. F., & Fisher, W.
plied Behavior Analysis, 36, 383–386. W. (2019). Randomized controlled trial of an applied
Najdowski, A. C., Wallace, M. D., Penrod, B., Tarbox, J., behavior analysis intervention for food selectivity in
Reagon, K., & Higbee, T. S. (2008). Caregiver-con- children with autism spectrum disorder. Journal of
ducted experimental functional analyses of inappro- Applied Behavior Analysis, 52, 895–917.
priate mealtime behavior. Journal of Applied Behavior Peterson, K. M., Piazza, C. C., & Volkert, V. M. (2016).
Analysis, 41, 459–465. Comparison of a modified sequential oral sensory ap-
Najdowski, A. C., Wallace, M. D., Reagon, K., Penrod, proach to an applied behavior analytic approach in
B., Higbee, T. S., & Tarbox, J. (2010). Utilizing a the treatment of food selectivity in children with au-
home-based parent training approach in the treat- tism spectrum disorders. Journal of Applied Behavior
ment of food selectivity. Behavioral Interventions, 25, Analysis, 49, 485–511.
89–107. Peterson, K. M., Volkert, V. M., & Zeleny, J. R. (2015).
Nebraska Legislature. (2009). Fiscal note: Legislative fis- Increasing self-drinking for children with feeding
cal analyst estimate. LB 243, Revision 01. disorders. Journal of Applied Behavior Analysis, 48,
Palmer, S., & Horn, S. (1978). Feeding problems in chil- 436–441.
dren. In S. Palmer & S. Ekvall (Eds.), Pediatric nutri- Piazza, C. C., Anderson, C., & Fisher, W. W. (1993).
tion in developmental disorders (pp. 107–129). Spring- Teaching self-feeding skills to patients with Rett syn-
field, IL: Charles C Thomas. drome. Developmental Medicine and Child Neurology,
Pangborn, M. M., Borrero, C. S. W., & Borrero, J. C. 35, 991–996.
(2013). Sequential application of caregiver training Piazza, C. C., & Carroll-Hernandez, T. A. (2004) As-
to implement pediatric feeding protocols. Behavioral sessment and treatment of pediatric feeding disor-
Interventions, 28, 107–130. ders. In R. E. Tremblay, R. G. Barr, & R. Peters (Eds.),
Patel, M. R., Piazza, C. C., Kelley, M. L., Ochsner, C. A., Encyclopedia on early childhood development [Online].
& Santana, C. M. (2001). Using a fading procedure Montreal: Centre of Excellence for Early Childhood
to increase fluid consumption in a child with feeding Development. Retrieved July 29, 2013, from www.ex-
problems. Journal of Applied Behavior Analysis, 34, cellence-earlychildhood.ca/documents/Piazza-Carroll-
357–360. HernandezANGxp.pdf.
Patel, M. R., Piazza, C. C., Layer, S. A., Coleman, R., & Piazza, C. C., Fisher, W. W., Brown, K. A., Shore, B. A.,
Swartzwelder, D. M. (2005). A systematic evaluation Patel, M. R., Katz, R. M., et al. (2003). Functional
of food textures to decrease packing and increase oral analysis of inappropriate mealtime behaviors. Journal
intake in children with pediatric feeding disorders. of Applied Behavior Analysis, 36, 187–204.
Journal of Applied Behavior Analysis, 38, 89–100. Piazza, C. C., Patel, M. R., Gulotta, C. S., Sevin, B. M.,
Patel, M. R., Piazza, C. C., Martinez, C. J., Volkert, V. & Layer, S. A. (2003). On the relative contribution
M., & Santana, C. M. (2002). An evaluation of two of positive reinforcement and escape extinction in
differential reinforcement procedures with escape the treatment of food refusal. Journal of Applied Be-
extinction to treat food refusal. Journal of Applied Be- havior Analysis, 36, 309–324.
havior Analysis, 35, 363–374. Piazza, C. C., Patel, M. R., Santana, C. M., Goh, H.,
Patel, M. R., Piazza, C. C., Santana, C. M., & Volkert, V. Delia, M. D., & Lancaster, B. M. (2002). An evalu-
M. (2002). An evaluation of food type and texture in ation of simultaneous and sequential presentation
the treatment of a feeding problem. Journal of Applied of preferred and nonpreferred food to treat food se-
Behavior Analysis, 35, 183–186. lectivity. Journal of Applied Behavior Analysis, 35,
Patel, M. R., Reed, G. K., Piazza, C. C., Bachmeyer, M. 259–270.
H., Layer, S. A., & Pabico, R. S. (2006). An evalua- Reed, G. K., Piazza, C. C., Patel, M. R., Layer, S. A.,
tion of a high-probability instructional sequence to Bachmeyer, M. H., Bethke, S. D., et al. (2004). On
increase acceptance of food and decrease inappro- the relative contributions of noncontingent rein-
priate behavior in children with pediatric feeding forcement and escape extinction in the treatment of
disorders. Research in Developmental Disabilities, 27, food refusal. Journal of Applied Behavior Analysis, 37,
430–442. 27–41.
Patel, M. R., Reed, G. K., Piazza, C. C., Mueller, M., Riordan, M. M., Iwata, B. A., Finney, J. W., Wohl, M. K.,
Bachmeyer, M. H., & Layer, S. A. (2007). Use of a & Stanley, A. E. (1984). Behavioral assessment and
high-probability instructional sequence to increase treatment of chronic food refusal in handicapped
compliance to feeding demands in the absence of children. Journal of Applied Behavior Analysis, 17,
escape extinction. Behavioral Interventions, 22, 305– 327–341.
310. Riordan, M. M., Iwata, B. A., Wohl, M. K., & Finney, J.
A Behavior-Analytic Approach to Feeding Disorders 443
W. (1980). Behavioral treatment of food refusal and to guide treatment of food refusal. Journal of Applied
selectivity in developmentally disabled children. Ap- Behavior Analysis, 45, 83–96.
plied Research in Mental Retardation, 1, 95–112. Shore, B. A., Babbitt, R. L., Williams, K. E., Coe, D.
Rivas, K. D., Piazza, C. C., Kadey, H. J., Volkert, V. M., A., & Snyder, A. (1998). Use of texture fading in the
& Stewart, V. (2011). Sequential treatment of a feed- treatment of food selectivity. Journal of Applied Be-
ing problem using a pacifier and flipped spoon. Jour- havior Analysis, 31, 621–633.
nal of Applied Behavior Analysis, 44, 387–391. Sicherer, S. H., & Sampson, H. A. (2010). Food aller-
Rivas, K. D., Piazza, C. C., Patel, M. R., & Bachmeyer, gy. Journal of Allergy and Clinical Immunology, 125,
M. H. (2010). Spoon distance fading with and with- S116–S125.
out escape extinction as treatment for food refusal. Singer, L. T., Sing, L. Y., Hill, B. P., & Jaffe, A. C. (1990).
Journal of Applied Behavior Analysis, 43, 673–683. Stress and depression in mothers of failure to thrive
Rocha, A. D., Moreira, M. E. L., Pimenta, H. P., Ramos, children. Journal of Pediatric Psychology, 15, 711–720.
J. R. M., & Lucena, S. L. (2007). A randomized study Stark, L. J., Mulvihill, M. M., Powers, S. W., Jelalian, E.,
of the efficacy of sensory–motor–oral stimulation Keating, K., Creveling, S., et al. (1996). Behavioral
and non-nutritive sucking in very low birthweight intervention to improve calorie intake of children
infant. Early Human Development, 83, 385–388. with cystic fibrosis: Treatment versus wait list con-
Rommel, N., DeMeyer, A. M., Feenstra, L., & Veere- trol. Journal of Pediatric Gastroenterology and Nutri-
man-Wauters, G. (2003). The complexity of feeding tion, 22, 240–253.
problems in 700 infants and young children present- Stark, L. J., Powers, S. W., Jelalian, E., Rape, R. N., &
ing to a tertiary care institution. Journal of Pediatric Miller, D. L. (1994). Modifying mealtime interac-
Gastroenterology and Nutrition, 37, 75–84. tions of children with cystic fibrosis and their parents
Roth, M. P., Williams, K. E., & Paul, C. M. (2010). via behavioral parent training. Journal of Pediatric
Treating food and liquid refusal in an adolescent Psychology, 19, 751–768.
with Asperger’s disorder. Clinical Case Studies, 9, Stein, B., Everhart, K. K., & Lacy, B. E. (2015). Gas-
260–272. troparesis: A review of current diagnosis and treat-
Schafe, G. E., & Bernstein, I. L. (1996). Taste aversion ment options. Journal of Clinical Gastroenterology, 49,
learning. In E. D. Capaldi (Ed.), Why we eat what we 550–558.
eat: The psychology of eating (pp. 31–51). Washington, Stubbs, K. H., Volkert, V. M., Rubio, E. K., & Ottinger,
DC: American Psychological Association. E. (2017). A comparison of flipped-spoon presenta-
Schwarz, S. M., Corredor, J., Fisher-Medina, J., Cohen, tion and redistribution to decrease packing in chil-
J., & Rabinowitz, S. (2001). Diagnosis and treatment dren with feeding disorders. Learning and Motivation,
of feeding disorders in children with developmental 62, 103–111.
disabilities. Pediatrics, 108, 671–676. Task Force on Promotion and Dissemination of Psy-
Sehgal, S. K., Prakash, O., Gupta, A., Mohan, M., & chological Procedures. (1995). Training in and dis-
Anand, N. K. (1990). Evaluation of beneficial effects semination of empirically validated treatments: Re-
of nonnutritive sucking in preterm infants. Indian Pe- port and recommendations. Clinical Psychologist, 48,
diatrics, 27, 263–266. 3–23.
Seiverling, L., Williams, K., Sturmey, P., & Hart, S. Thommessen, M., Heiberg, A., Kase, B. F., Larsen, S., &
(2012). Effects of behavioral skills training on paren- Riis, G. (1991). Feeding problems, height and weight
tal treatment of children’s food selectivity. Journal of in different groups of disabled children. Acta Paediat-
Applied Behavior Analysis, 45, 197–203. rica, 5, 527–533.
Sevin, B. M., Gulotta, C. S., Sierp, B. J., Rosica, L. A., & Tiger, J. H., & Hanley, G. P. (2006). Using reinforcer
Miller, L. J. (2002). Analysis of response covariation pairing and fading to increase the milk consump-
among multiple topographies of food refusal. Journal tion of a preschool child. Journal of Applied Behavior
of Applied Behavior Analysis, 35, 65–68. Analysis, 39, 399–403.
Shalev, R. A., Milnes, S. M., Piazza, C. C., & Kozisek, J. Toien, O., Blake, J., Edgar, D. M., Grahn, D. A., Heller,
M. (2018). Treating liquid expulsion in children with H. G., & Barnes, B. M. (2001). Hibernation in black
feeding disorders. Journal of Applied Behavior Analy- bears: Independence of metabolic suppression from
sis, 51, 70–79. body temperature. Science, 331, 906–909.
Sharp, W. G., Harker, S., & Jaquess, D. L. (2010). Com- Valdimarsdottir, H., Halldorsdottir, L. Y., & Sigurådót-
parison of bite-presentation methods in the treat- tir, Z. G. (2010). Increasing the variety of foods
ment of food refusal. Journal of Applied Behavior consumed by a picky eater: Generalization of effects
Analysis, 43, 739–743. across caregivers and settings. Journal of Applied Be-
Sharp, W. G., & Jaquess, D. L. (2009). Bite size and havior Analysis, 43, 101–105.
texture assessments to prescribe treatment for severe Vaz, P. C. M., Piazza, C. C., Stewart, V., Volkert, V. M.,
food selectivity in autism. Behavioral Interventions, & Groff, R. A. (2012). Using a chaser to decrease
24, 157–170. packing in children with feeding disorders. Journal of
Sharp, W. G., Odom, A., & Jaquess, D. L. (2012). Com- Applied Behavior Analysis, 45, 97–105.
parison of upright and flipped spoon presentations Volkert, V. M., Peterson, K. M., Zeleny, J. R., & Piazza,
444 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
C. C. (2014). A clinical protocol to increase chewing Wilder, D. A., Normand, M., & Atwell, J. (2005).
and assess mastication in children with feeding disor- Noncontingent reinforcement as treatment for food
ders. Behavior Modification, 38, 705–729. refusal and associated self-injury. Journal of Applied
Volkert, V. M., & Piazza, C. C. (2012). Pediatric feeding Behavior Analysis, 38, 549–553.
disorders. In P. Sturmey & M. Hersen (Eds.), Hand- Wilkins, J. W., Piazza, C. C., Groff, R. A., & Vaz, P. C.
book of evidence-based practice in clinical psychology: M. (2011). Chin prompt plus re-presentation as treat-
Vol. 1. Child and adolescent disorders (pp. 323–337). ment for expulsion in children with feeding disorders.
Hoboken, NJ: Wiley. Journal of Applied Behavior Analysis, 44, 513–522.
Volkert, V. M., Piazza, C. C., Vaz, P. C. M., & Frese, Wilkins, J. W., Piazza, C. C., Groff, R. A., Volkert, J.
J. (2013). A pilot study to increase chewing in chil- M., Kozisek, J. M., & Milnes, S. M. (2014). Utensil
dren with feeding disorders. Behavior Modification, manipulation during initial treatment of pediatric
37, 391–408. feeding problems. Journal of Applied Behavior Analy-
Volkert, V. M., Vaz, P. C. M., Piazza, C. C., Frese, J., & sis, 47, 1–16.
Barnett, L. (2011). Using a flipped spoon to decrease Williams, K. E., Riegel, K., Gibbons, B., & Field, D. G.
packing in children with feeding disorders. Journal of (2007). Intensive behavioral treatment for severe
Applied Behavior Analysis, 44, 617–621. feeding problems: A cost-effective alternative to tube
Werle, M. A., Murphy, T. B., & Budd, K. S. (1993). feeding? Journal of Developmental and Physical Dis-
Treating chronic food refusal in young children: abilities, 19, 227–235.
Home-based parent training. Journal of Applied Be- Winick, M. (1969). Malnutrition and brain develop-
havior Analysis, 26, 421–433. ment. Journal of Pediatrics, 74, 667–679.
CHAP TER 26
The power and precision with which early behav- Measuring responding accuracy and then fluency
ior analysts used operant-conditioning principles (responding accurately, quickly, and in different
to change behavior resulted in almost immediate contexts) allows the teacher to evaluate learning
applications to education. For example, Skinner’s outcomes in relation to instructional goals and
(1954) model of programmed instruction used provides the basis for varying instruction (Christ,
teaching machines to provide students with im- Zopluoglu, Monaghen, & Van Norman, 2013;
mediate feedback and reinforcement for correct re- Fuchs & Fuchs, 1986; Reschly, 2004). With a focus
sponding across many tightly sequenced tasks (e.g., on responding as the basic unit of analysis, the
arithmetic facts). Frequent reinforcement, self- student’s task becomes one of emitting gradually
paced practice, and prompting and fading—what more refined, coordinated, and effective response
Skinner originally called vanishing—produced repertoires under increasingly demanding and
learning that was enjoyable, occurred quickly, and naturalistic stimulus conditions (Martens, Daly, &
involved few errors (Skinner, 1984). Keller (1968) Ardoin, 2015). The teacher acts more as a trainer
showed how to apply these same principles to cre- or coach than as a lecturer or educator, to promote
ate a personalized system of instruction for college high rates of correct independent responding by
students. Curriculum materials for Keller’s person- the student (Keller, 1968). That is, the teacher fa-
alized system of instruction courses consisted of 30 cilitates learning by arranging repeated opportuni-
or more small content units, each associated with ties to respond in the presence of diverse curricular
its own set of study questions and exercises. Stu- materials; sometimes with modeling, prompting,
dents were free to study and attend lectures when- and error correction; and always with differential
ever they wanted, but the system prescribed brief reinforcement for correct or desired responding
examinations on the material in each “unit.” In- (Martens et al., 2015). From the perspective of
structors graphed performance on these unit tests both teacher and student, behavioral approaches
and required mastery before moving forward in the to skill training make learning active, fast-paced,
curriculum. relatively error-free, and more reinforcing than
The examples above suggest that behavioral ap- punishing (Miltenberger, 2016).
proaches to education are characterized by an em- The fundamentals of behavioral skill training
phasis on doing rather than knowing. The first and have remained unchanged since their inception:
most important feature of behavioral skill training (1) well-sequenced stimulus materials that are
is repeated measurement of student responding. linked horizontally across related skills and verti-
445
446 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
cally by difficulty; (2) brief, repeated opportunities ductive practice, and (3) the benefits of fluent re-
to respond in the presence of these stimuli; (3) sponding for maintaining performance under more
modeling, prompting, and feedback by the teacher demanding conditions. The next section discusses
to increase the likelihood of correct responding; challenges that behavioral educators face when at-
(4) differential reinforcement to establish and tempting to program for skill generalization, and
strengthen stimulus control over correct respond- ways of further refining instruction and perfor-
ing; and (5) frequent performance monitoring to mance assessment to produce generative-response
inform instructional planning and evaluate learn- repertoires. From a somewhat narrow perspective
ing outcomes (Fuchs, Fuchs, & Speece, 2002; on behavioral skill instruction, we then broaden
Martens & Witt, 2004). the focus of the chapter to discuss three examples
Professionals consider behavioral approaches of behaviorally oriented instruction programs/
to skill training standard practice when teaching systems: the Helping Early Literacy with Practice
daily living, communication, and other adaptive Strategies program, Direct Instruction, and the
behaviors to individuals with developmental dis- Morningside Model of Generative Instruction,
abilities (e.g., Carroll, Joachim, St. Peter, & Rob- including research supporting their effectiveness.
inson, 2015; Lerman, Hawkins, Hillman, Shire- We also address whether teachers use or receive
man, & Nissen, 2015; Marion, Martin, Yu, Buhler, training in these and other empirically supported
& Kerr, 2012), and when remediating deficits in techniques in this section. The last section of the
children’s basic academic skills (Daly, Martens, chapter describes how behavioral skill training
Barnett, Witt, & Olson, 2007; Daly, Neugebauer, and progress monitoring form the basis of a tiered
Chafouleas, & Skinner, 2015). Researchers have service delivery model known as response to inter-
applied behavioral instruction approaches to ath- vention, which teachers use widely in schools be-
letic performance (e.g., Brobst & Ward, 2002; fore they determine that children are eligible for
Kladopoulos & McComas, 2001; Tai & Milten- special education (e.g., Balu et al., 2015).
berger, 2017), in secondary and higher education
(Cavanaugh, Heward, & Donelson, 1996; Saville,
Zinn, Neef, Van Norman, & Ferreri, 2006; Stocco, STRATEGIES OF EFFECTIVE TEACHING
Thompson, Hart, & Soriano, 2017), during in-
The Learning/Instructional Hierarchy
service teacher training (DiGennaro, Martens,
as a Dynamic Teaching Model
& McIntyre, 2005; Hogan, Knez, & Kahng, 2015;
Luck, Lerman, Wu, Dupuis, & Hussein, 2018), and Haring et al. (1978) proposed the learning/instruc-
in teaching children important safety skills (Dick- tional hierarchy to describe how children’s perfor-
son & Vargo, 2017; Houvouras & Harvey, 2014). mance of skills improves over time with training
Although the effectiveness of these techniques and how teachers should modify instructional
has led to the development of comprehensive be- procedures as performance improves. Although
havioral instruction- and performance-monitor- often referred to as a stage model, the learning/
ing programs (e.g., Direct Instruction, Precision instructional hierarchy is perhaps better charac-
Teaching), public schools do not use the programs terized as a dynamic approach to teaching that
or their component techniques (Begeny & Mar- involves closely monitoring the proficiency with
tens, 2006; Lindsley, 1992; Saville et al., 2006). which a student performs a skill and then tailor-
We have organized the material in the present ing instruction and reinforcement to strengthen
chapter loosely around a model for the dynamic de- student responding. The teacher defines the tar-
velopment of proficient skill performance known get skill and its controlling stimuli. For example,
as the learning/instructional hierarchy (Daly, Lentz, a goal might be oral reading of 150-word passages
& Boyer, 1996; Haring, Lovitt, Eaton, & Hansen, at the third-grade level. The teacher tightly con-
1978; Martens & Witt, 2004). Consistent with trols training stimuli initially and measures how
this model, we begin with a discussion of stimu- the student performs the skills (Martens et al.,
lus control as the basis of proficient performance, 2015; Martens & Eckert, 2007). The goal is to in-
instructional strategies for initial skill acquisition, crease responding gradually along a continuum as
and the components of discrete-trial training. We accurate, rapid, sustained, and generalized reading
then discuss (1) ways to strengthen both respond- comes under stimulus control—that is, as reading
ing and stimulus control through fluency building, becomes accurate and fluent. During later stages
(2) ways to present curricular materials in a free- of training, when the teacher uses more diverse
operant format and arrange reinforcement for pro- and naturalistic stimuli to occasion responding, he
Behavioral Approaches to Education 447
or she measures proficiency by when the student efited more from practice combined with either
performs a skill or by the range of conditions under performance feedback or feedback plus reinforce-
which the student continues to exhibit proficient ment. These students had the weakest strength
performance. of responding, and the intervention likely was ef-
Once the teacher identifies a student’s strength fective because the additional intervention com-
of responding, the teacher uses the learning/in- ponents helped bring rapid and accurate reading
structional hierarchy to select training and rein- under stimulus control of the text.
forcement procedures that previous research has
shown best promote learning. For example, one
Acquisition
student may read tentatively with low accuracy,
and another student may read with high accu- Acquisition-level training focuses on providing
racy and fluency during brief performance runs. students with enough assistance to perform a skill
The learning/instructional hierarchy provides correctly that was not in their repertoire previous-
the framework for Direct Instruction (Gersten, ly, and then differentially reinforcing accurate per-
Carnine, & White, 1984), discussed later in the formance of the skill in the presence of its evok-
chapter, and researchers have used it to design, ing stimuli. Accuracy is the first and most basic
implement, and evaluate interventions for many performance criterion a student must meet when
academic performance problems (e.g., Szadokier- learning any skill. Promoting stimulus control
ski, Burns, & McComas, 2017; Daly & Martens, with reinforcement for correct responding may be
1994). as simple as praising a child for saying, “Red!” after
the teacher shows the child an apple and asks him
or her to say its color, or as complex as giving a
Stimulus Control
student a grade of A on a composition about the
Fundamental to proficient performance of any skill Civil War in response to a homework assignment
is the development of stimulus control, or bringing in history. Three-term acquisition trials (i.e., ante-
the student’s response under control of key stimuli cedent–behavior–consequence) may occur in the
in the environment. Failing to write the correct natural environment when opportunities to model
answer in response to a math problem or dictated and reinforce a target skill present themselves,
spelling word signifies a lack of stimulus control. such as in incidental teaching, or in isolation
Failing to write the correct answer when working under control of a trainer, such as in discrete-trial
in a noisy classroom suggests that stimulus con- training (Wolery, Bailey, & Sugai, 1988).
trol may not be strong enough for performance to Research has shown that discrete-trial training
remain stable in the face of distraction. The first is effective for teaching many skills and involves
goal of skill training, then, is to establish stimulus presentation of (1) a command and a discrimina-
control by providing the student with opportuni- tive stimulus in the presence of which we expect
ties to respond with modeling, prompting, and the student to emit the response, (2) a prompt to
feedback, plus differential reinforcement for cor- increase the likelihood of a correct response, (3)
rect responding. As practice trials accumulate, the an opportunity to emit the response, (4) corrective
student’s need for assistance decreases, stimulus feedback for incorrect responses, and (5) differen-
control increases, performance of the behavior be- tial reinforcement for correct responding. Many
comes more efficient, and the student better dis- children with developmental disabilities such as
criminates the conditions under which behavior autism spectrum disorder do not imitate adults,
produces reinforcement. exhibit repetitive or stereotypic behavior, pay
For example, Chafouleas, Martens, Dobson, little attention to social cues, have limited com-
Weinstein, and Gardner (2004) evaluated the ef- munication, and may not mand for (i.e., request)
fects of three interventions involving practice, information (Marion et al., 2012; Smith, 2001).
feedback, and reinforcement with children who As a result, these children may not acquire skills
exhibited different baseline levels of oral read- through exploring, playing, imitating, and talking
ing fluency. The two students whose oral reading in the same manner as their typically developing
was under stimulus control of the printed words peers. Behavior analysts often use discrete-trial
benefited most from practice alone via repeated training to teach these children important social,
readings. These were the students with the high- communication, and self-help skills (e.g., Valen-
est fluency and lowest error rates at baseline. The tino, LeBlanc, Veazey, Weaver, & Raetz, 2019).
students with the highest baseline error rates ben- Discrete-trial training includes teaching skills in
448 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
isolation; prompting, guiding, and reinforcing the creasingly efficient and fluent performance, which
correct response; and conducting many learning strengthens stimulus control. Research has sug-
trials per session. gested that productive skill practice should include
Once a student can perform a skill accurately (1) tasks, materials, or both to which the student
with assistance, the teacher withdraws or gradu- can respond with high accuracy and minimal as-
ally fades assistance to transfer control to the sistance (i.e., instructionally matched materials);
discriminative stimulus. Procedures for systemati- (2) brief, repeated practice opportunities with
cally fading prompts may involve intermittently feedback and reinforcement; (3) monitoring and
withdrawing or gradually reducing their intensity, charting performance; and (4) performance crite-
increasing the latency to their presentation, or ria for increasing the difficulty of material (Daly
using progressively less intrusive prompts (Ceng- et al., 2007).
her, Budd, Farrell, & Fienup, 2018; Wolery et al., A strategic concern with fluency building,
1988). One strategic concern during acquisition which involves exposure to material that increases
is how to fade or withdraw prompts in a way that in difficulty, is exactly where in a vertically linked
maintains correct responding (Cividini-Motta & curriculum sequence to begin practice. Depend-
Ahearn, 2013). Teachers typically address this ing on the skill, the teacher may determine the
concern by measuring correct responses and errors starting point when the student reaches a fluency
at each prompt level and by altering the prompt’s criterion on less difficult material, such as passage
intrusiveness based on the resulting performance reading (Daly, Martens, Kilmer, & Massie, 1996),
data. For example, if accuracy is high when the or material that differs in the ratio of known to
teacher provides assistance, then fading assistance unknown items, such as word-list training (Mac-
may be necessary to promote independent re- Quarrie, Tucker, Burns, & Hartman, 2002). As
sponding. If accuracy is low with no improvement an example, Martens et al. (2007) evaluated
over time, the student may need more assistance, the effects of a fluency-based after-school read-
or the teacher should teach the skill using easier ing program with 15 low-achieving second- and
materials (Wolery et al., 1988). If only errors occur, third-grade students. The curriculum included
then the teacher may need to teach prerequisite or four passages at each of six grade levels, sequenced
component skills before continuing. Conversely, if by difficulty both within and across grades. The
accuracy is high and stable in the absence of as- researchers required students to meet a retention
sistance, then the teacher may need to restructure criterion of reading 100 words correctly per min-
the training situation to build fluency. ute in the absence of practice 2 days after train-
ing to advance to a more difficult passage in the
curriculum. Training occurred 3 times a week
Fluency
and consisted of phrase drill–error correction for
Once the student has met the desired criterion for words missed at pretest, listening to a passage pre-
accuracy of responding with discrete-trial training view, and three repeated readings of the passage.
(e.g., 90% correct trials over three consecutive ses- Additional components of the program included
sions), training shifts to fluency building. Binder goal setting, charting, and token reinforcement for
(1996) defined fluency as “the fluid combination of meeting the fluency criterion. Results showed that
accuracy plus speed that characterizes competent after the equivalent of 5½ weeks of training, chil-
performance” (p. 164). By virtue of its emphasis dren advanced between two and three grade levels
on response rate, fluency building requires free-op- on average in the difficulty of passages they could
erant rather than discrete-trial training arrange- read above the retention criterion. The researchers
ments (Johnson & Layng, 1996). In a free-operant also observed significant pre- to postintervention
arrangement, the teacher gives students enough gains for children at each grade level on untrained,
stimulus materials, such as a worksheet of 30 ad- generalization passages.
dition problems, for them to practice timed per-
formance runs during which they are free to emit
Maintenance and Generalization
as many or as few responses as they wish. During
free-operant performance runs, accurate rate of re- Once the student reaches a desired level of fluency
sponding, such as words read correctly per minute, during relatively brief performance runs, such as 2
replaces percentage of correct trials as a measure minutes of math computation or 1 minute of oral
of student performance. A key feature of fluency reading, the teacher can strengthen performance
building is the differential reinforcement of in- further by modifying the training conditions to
Behavioral Approaches to Education 449
more closely approximate the natural environ- peers, and competent adults (Johnson & Layng,
ment. The progression from fluency building to 1996). The norm-referenced fluency aims were
maintenance to generalization involves strength- problematic because they did not always predict
ening stimulus control and programming stimulus generalization—fluent performance over time in
diversity. The teacher strengthens stimulus control different circumstances, with different materi-
through differential reinforcement of correct and als, or on more complex tasks (Binder, 1996). As
then rapid responding to discriminative stimuli a result, researchers developed functional fluency
(such as two-digit numbers and the sign for mul- aims, indicating performance levels above which
tiplication), over time and under many stimulus students would be likely to maintain fluency under
conditions (such as alone with the trainer, in class more demanding practice conditions. Summarized
during a group exercise, or at home for homework). by the acronym RESAA, researchers have used
The teacher also programs stimulus diversity by these aims to predict retention in the absence of
systematically varying features of the discrimina- practice; endurance over longer performance runs;
tive stimuli (such as presenting math computation stability in the face of distraction; application to
problems vertically in worksheets or horizontally more complex tasks; and adduction, which is spon-
in story problems). taneous emergence of new skill forms (Johnson &
A basic assumption of response-to-intervention Layng, 1996). For example, McDowell and Keenan
models is that students will show individual differ- (2001) trained a 9-year-old boy with attention-
ences in benefiting from instructional trials. Re- deficit/hyperactivity disorder (ADHD) to say pho-
search has shown that for some students, strength nemes displayed on flash cards at increasing levels
of responding following intervention will increase of fluency. Before performance reached a fluency
enough to produce generalized oral reading fluen- aim of 60–80 sounds correct per minute, probes
cy (Daly, Martens, et al., 1996), whereas for others for endurance showed decreases in fluency and on-
it may not (e.g., Daly, Martens, Hamler, Dool, & task behavior. The boy maintained on-task behav-
Eckert, 1999). Generalization is an important out- ior and fluency at high levels after he achieved the
come of intervention and often requires explicit fluency aim.
programming (Stokes & Baer, 1977). Researchers With respect to the second question, one im-
have examined several strategies for promoting plication of the learning/instructional hierarchy
generalized oral reading fluency, which include is that students should practice prerequisite or
training loosely with multiple low-word-overlap component skills to high levels of fluency before
passages; training with common stimuli, such as teachers require students to combine them into
high-word-overlap passages or color-coded rimes; more complex or composite skills (Binder, 1996).
training with passages containing multiple exem- This may be relatively easy to accomplish when
plars of key or frequent words; training to func- a teacher is training simple or basic skills the first
tional fluency aims, such as 100 words read cor- time they appear in a curriculum sequence, if the
rectly per minute; and training broadly applicable teacher can build sufficient time for practice into
skills such as phoneme blending (Ardoin, Mc- the instructional day. For building fluency in more
Call, & Klubnik, 2007; Bonfiglio, Daly, Martens, complex skills, however, the situation becomes
Lan-Hsiang, & Corsaut, 2004; Daly, Chafouleas, more challenging. Consider oral reading fluency,
Persampieri, Bonfiglio, & LaFleur, 2004; Daly, for example. Even after the teacher has identified
Martens, et al., 1996; Martens et al., 2007; Mar- a starting point for fluency building in the curricu-
tens, Werder, Hier, & Koenig, 2013; Martens et al., lum, deficits in one or more component skills may
2019; Mesmer et al., 2010). mitigate the effects of practice and reinforcement
Three questions arise with respect to fluency with connected text. For example, the teacher
building: (1) What rate of performance should we may identify a starting point as end-of-first-grade
require to promote maintenance and generaliza- passages for a third grader with significant read-
tion of a skill; (2) how should we design fluency- ing difficulties, but the student’s problems with
building activities for more complex or composite reading sight word vocabulary and decoding may
skills; and (3) how should we arrange reinforce- prevent her from achieving fluency. In such a case,
ment to support practice over time? Advocates the teacher may need to provide fluency- or even
of precision teaching adopted and subsequently acquisition-level training in isolation on the de-
abandoned three norm-referenced approaches to ficient component skills, along with opportuni-
answering the first question. These approaches ties to practice the composite skill. Research has
included comparisons to typical peers, competent shown that cumulative dysfluency is “perhaps the
450 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
single most important factor in long-term student academic behavior. Alessi’s (1987) conceptual-
failure” (Binder, 1996, p. 184), and cumulative flu- ization of the teacher’s task is most insightful for
ency accelerates the learning process and even understanding how to promote and analyze gener-
leads to the spontaneous emergence of new skills, alization of academic responding. An economical
as we discuss in the next section. and efficient approach to teaching is to train gen-
Previous research has shown that teachers can erative-response repertoires, which allow students to
maintain high rates of academic performance by respond and even combine trained responses in
using intermittent fixed-ratio schedules to support novel ways to differing configurations of stimuli
practice (McGinnis, Friman, & Carlyon, 1999), and task demands. These response repertoires are
multiple-ratio schedules where the magnitude of generative in the sense that a student is now ca-
reinforcement progressively increases as comple- pable of applying previously learned responses in
tion rate increases (Lovitt & Esveldt, 1970), and ways that a teacher has not instructed. These re-
even lottery schedules where chances of reinforce- sponse repertoires also may be recombinative in that
ment are as low as 50% (Martens et al., 2002). the student may combine responses in novel ways,
Preliminary evidence also suggests that students thereby generating a new response repertoire. The
may prefer different reinforcement contingencies results are new and more sophisticated behavioral
as their skill proficiency increases. For example, repertoires for responding adaptively to varying
Lannie and Martens (2004) gave students the op- stimulus conditions. The teacher, therefore, should
portunity to complete two sets of math problems, strive to teach a generative set of responses as a
either both easy or both difficult. The completion subset of all possible responses (the universal set
of problems from each set produced a different of all possible stimulus–response combinations for
reinforcement contingency. Students could earn the response class). Trained responses are genera-
rewards for on-task behavior at specified intervals tive and functional if they contribute to the stu-
while working one set of problems, or for the num- dent’s ability to respond appropriately in the pres-
ber of problems completed correctly from the other ence of untaught stimuli. After the teacher brings
set. Students in the study chose reinforcement for the student’s responding under stimulus control
time on task when working difficult problems, but for the generative set, stimulus generalization
switched to reinforcement for the number of prob- proceeds until the student reaches a threshold of
lems completed correctly when working easy prob- responding with the generative set that correlates
lems, as this contingency maximized the amount with increases in correct responding for items from
of reinforcement per session. the universal set. The teacher conducts measure-
ment with the generative-response set to assess
mastery and with samples from the universal set to
THE ROLE AND ANALYSIS assess generalization (Alessi, 1987).
OF GENERATIVE‑RESPONSE REPERTOIRES
Generative Repertoires in Oral Reading Fluency
As suggested in the previous section, student re-
sponding in a classroom should be highly predict- This conceptualization of generalization is per-
able to anyone familiar with the curriculum. In a haps most appropriate for the basic academic skills
well-sequenced curriculum, response demands will that serve as the foundation for all other skills in
be repetitive for many of the behavioral repertoires a curriculum. Oral reading fluency and phoneme
taught early in the curriculum, such as reading text blending are examples of basic academic skills that
or basic math calculations, as teachers incorporate have gained prominence in the wider educational
them into increasingly complex behavioral reper- community. Authoritative documents such as re-
toires that they presume will prepare students for ports from the National Reading Panel (2000) and
life beyond the classroom (e.g., preparing reports). the National Research Council (Snow, Burns, &
Although response demands may be predictable, Griffin, 1998) have established their critical role
the academic stimuli that occasion them are con- in reading development. In the case of oral read-
tinually changing as time, exemplars, and settings ing fluency, a student’s word reading should come
change. Some of these stimulus changes occur under the stimulus control of the text, which con-
naturally, but a teacher should program them. sists of varying configurations of letters separated
One limitation that teachers face, however, is by spaces and punctuation marks. Those letters, of
that they cannot teach all possible stimulus–re- course, are organized into words that the student
sponse relations as they prepare students for future must read or decode rapidly to understand or to
Behavioral Approaches to Education 451
provide a verbal report of the content of the text. condition. Next, they compared the effects of the
Words appear in different orders in texts, with individual interventions alone by alternating the
grammatical conventions constraining the order instructional strategy and the reward condition.
somewhat, and the student must be able to read Finally, they compared the intervention package
words fluently across texts for functional word with the individual intervention that produced the
reading. The curriculum defines the generative set highest level of responding when they evaluated
of response repertoires, and the teacher applies dif- the effects of the instructional strategy and reward
ferent strategies to bring student responding under conditions alone. In this type of brief experimental
the control of texts, which become progressively analysis, a behavior analyst compares performance
difficult throughout the curriculum. If the teacher in high-word-overlap passages to equal-difficulty-
simply measures student performance on the in- level, low-word-overlap passages and evaluates the
structional texts, he or she does not know whether differences across interventions. Low-word-overlap
responding for instructed words will occur in other passages serve as controls for difficulty level and
texts that contain the same words. If the student extraneous variables that might affect responding
fails to generalize word reading to other texts, over time. Assessment of oral reading fluency is
reading will not become a functional skill for other sufficiently sensitive that it can detect differences
tasks the student will need to perform (e.g., read- during rapidly alternating conditions. The result-
ing to prepare for a history exam). ing interventions are more robust and more likely
Researchers have analyzed generalization of to produce generalized improvements over time,
word reading by manipulating word overlap in two because the analysis is a direct measure of general-
passages and measuring student responding (Daly, ized responding (Daly, Persampieri, et al., 2005).
Martens, et al., 1996; Martens et al., 2019). Two Daly, Persampieri, et al. (2005) used the results
passages have high word overlap if many of the same of the brief experimental analysis to identify the
words appear in the two passages, but the stories most efficient intervention that produced the
are different and have different word orders (Daly, highest level of responding. The reward contin-
Martens, et al., 1996). Manipulating word over- gency was most effective for one student. The in-
lap between texts can facilitate the measurement tervention that combined instructional strategies
of generalized oral reading fluency. For example, with the reward contingency was most effective
Daly, Martens, et al. (1996) found greater gener- for the second student. The researchers taught the
alization for instruction to high-word-overlap pas- students to self-manage the reading intervention,
sages than to low-word-overlap passages of equal and they measured correctly read words per min-
difficulty. Word overlap interacted with difficulty ute during continuous monitoring of student per-
level, so that greater gains were observed for easier formance. Both students demonstrated substantial
materials than for harder materials. improvements in oral reading fluency in an inde-
Researchers have analyzed generalized reading pendent reading series.
fluency based on word overlap to identify poten- The prior conceptual analysis of generative re-
tially effective reading interventions through brief sponding suggests that word reading should gen-
experimental analyses (Daly, Bonfiglio, Mattson, eralize not only to untaught configurations of in-
Persampieri, & Foreman-Yates, 2005; Daly et al., structed words, such as the same words in novel
1999). For example, Daly, Persampieri, McCurdy, texts, but also to untaught words. The universal
and Gortmaker (2005) conducted brief experi- word set should include all words that might ap-
mental analyses using a single-case experimental pear in texts of the appropriate difficulty for a
design to evaluate whether the reading deficits of given level in the curriculum (e.g., first grade
two students were skill-based, performance-based, vs. second grade). Many of these words may not
or both. They used the results of the brief experi- share stimulus properties with the generative
mental analyses to identify individualized inter- set. Nonetheless, students should be able to read
ventions. One intervention was an instructional them. Teachers can sample this set by using equal-
strategy that included repeated readings, listening- difficulty-level but low-word-overlap passages. For
passage preview, phrase drill–error correction, and example, Daly, Persampieri, et al. (2005) used one
syllable segmentation–error correction. The other reading series for measurement over time and an
intervention was a reward contingency. Initially, independent series for instruction. They used the
they combined the instructional strategy and re- universal set to evaluate the effectiveness of the
ward conditions in an intervention package and empirically derived instructional interventions,
alternated the intervention package with a control which permitted subsequent conclusions regarding
452 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
reading difficulties (Archer & Hughes, 2011; Na- component and the proper sequence; (2) brief,
tional Reading Panel, 2000). Reading programs scripted instructions the educator provides to the
such as Open Court Reading (Adams et al., 2005) student(s); and (3) a tips and reminders sheet.
and Sound Partners (Vadasy et al., 2000) are two The flowchart and scripted instructions promote
examples that provide such instruction. Although strong implementation fidelity of the core proce-
researchers did not develop these programs solely dures, and the tips and reminders sheet assists the
from behavior-analytic principles, many key ele- educator with implementation quality (e.g., high-
ments of these programs, such as explicit instruc- quality implementation of each core procedure,
tion, systematic prompts, and frequent oppor- and use of behaviors that are most likely to en-
tunities to respond, undoubtedly map onto the gage and motivate each student). The HELPS pro-
characteristics of behavioral skill instruction. We grams also include a series of 100 systematically se-
describe three examples that combine behavioral quenced reading passages, a placement assessment
techniques with comprehensive instructional to match curriculum passages with the student’s
programs, curricula, models, or systems designed skill level, and other materials to assist educators
to improve students’ academic skills. We discuss with achieving high implementation fidelity.
the effectiveness of each program and provide es- Research shows that the HELPS program can
timates regarding educators’ training and use of improve oral reading fluency and comprehension
them. for many populations, including average readers,
students with reading difficulties or disabilities,
and English-language-learner students when a
Helping Early Literacy with Practice Strategies
teacher, a teacher assistant, a school volunteer, or
Helping Early Literacy with Practice Strategies a parent implements it approximately three times
(HELPS) offers two structured reading programs per week (e.g., Begeny, Braun, et al., 2012; Beg-
specifically designed to improve students’ reading eny et al., 2010; Begeny, Mitchell, Whitehouse,
fluency. Teachers can implement the programs Samuels, & Stage, 2011; Begeny, Ross, et al., 2012;
either in a one-on-one (Begeny, 2009) or a small- Mitchell & Begeny, 2014). Research also supports
group (Begeny, Ross, Greene, Mitchell, & White- the implementation fidelity (Begeny, Easton, Up-
house, 2012) context, and materials are currently right, Tunstall, & Ehrenbock, 2014; Begeny, Up-
available in English and Spanish. The HELPS right, Easton, Ehrenbock, & Tunstall, 2013) and
programs integrate each of the five characteris- progress-monitoring (Begeny et al., 2015) materi-
tics listed at the beginning of the chapter that are als that supplement or are built directly into the
most central to behavioral approaches to instruc- HELPS implementation procedures.
tion, and they integrate specific tools for monitor-
ing and promoting strong implementation fidelity.
Direct Instruction
Researchers based the instructional components
of HELPS on existing research on reading fluency As was the case for the Open Court Reading and
instruction and intervention (e.g., Chard, Vaughn, Sound Partners programs mentioned previously,
& Tyler, 2002; National Reading Panel, 2000; the researchers who developed Direct Instruction
Morgan & Sideridis, 2006; Therrien, 2004), and did not base it specifically on principles of behav-
the development process used iterative methods ior analysis, though this approach to instruction
and other principles of design-based research (e.g., clearly falls within a behavioral framework (Beck-
Anderson & Shattuck, 2012; Shernoff et al., 2011). er, 1992; Fredrick, Deitz, Bryceland, & Hummel,
The HELPS programs include repeated reading of 2000). According to Becker (1992), Direct In-
instructionally appropriate text, modeling, phrase struction is “a systematic approach to the design
drill–error correction, verbal cueing procedures and delivery of a range of procedures for building
for students to read with fluency and for compre- and maintaining basic cognitive skills” (p. 71).
hension, goal setting, performance feedback with Specifically, Direct Instruction is a skill-based in-
graphing, and a motivational reward system. structional curriculum in which teachers promote
Implementation of HELPS sessions requires the sequential development of student competen-
8–12 minutes apiece. The program includes a cies by following scripted instructional routines
teacher’s manual and an online training video, (Becker, 1992; Gersten et al., 1984). Teachers
together with three primary protocols to as- generally use small-group instruction and instruc-
sist educators with implementation fidelity: (1) a tional strategies, such as modeling, and positive
flowchart that visually displays each instructional reinforcement, such as praise, for accurate re-
454 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
sponding. Furthermore, Direct Instruction lessons instructional model developed from the work of
ensure that teachers allow students to obtain suf- Johnson and colleagues at Morningside Acad-
ficient practice with targeted material and receive emy in Seattle, WA. Founded by Kent Johnson in
frequent opportunities to respond with corrective 1980, Morningside Academy offers schooling for
feedback. elementary and middle school students who have
In Project Follow Through, one of the largest been academically unsuccessful at their previous
educational experiments ever conducted, Direct schools. Core instructional components at Morn-
Instruction was one of several instructional pro- ingside Academy include (1) student groupings
grams independently used with thousands of stu- according to their entering repertoires and levels
dents representing various socioeconomic levels of instructional achievement; (2) a carefully se-
and ethnicities throughout the United States. quenced curriculum of component and composite
This large-scale project aimed to assess each pro- foundational academic skills (reading, writing,
gram by comparing pre- and posttest scores on mathematics, reasoning, and problem solving); (3)
various measures to a similar control group (see explicit, direct instruction in components; and (4)
Watkins, 1997, for a detailed review of the study). daily measurement of performance, with a focus
Although Project Follow Through had impor- on building fluency in each of the skills outlined
tant limitations, such as an inconsistent use of in the instructional sequence (Johnson & Street,
experimental-design elements, data analyses sug- 2013). The Morningside Model of Generative
gested that students receiving Direct Instruction Instruction “hinges on the belief that complex
performed better than those receiving any other behavioral repertoires emerge without explicit
instructional program on basic skill measures, instruction when well selected component reper-
comprehension measures, and affective measures toires are appropriately sequenced, carefully in-
such as self-esteem (Becker, 1992; Watkins, 1997). structed, and well-rehearsed” (Johnson & Street,
Adams and Engelmann (1996) subsequently con- 2004, p. 26). In other words, the Morningside
ducted a meta-analytic review of 37 studies that Model of Generative Instruction explicitly seeks
examined the effectiveness of Direct Instruction to build generative-response repertoires through
after Project Follow Through concluded. The au- appropriate instructional sequencing, effective in-
thors reported that “Direct Instruction interven- structional practices, and student mastery of skills.
tions have been shown to produce superior perfor- Although conducting tightly controlled re-
mance with preschool, elementary, and secondary search on its educational programs is not a major
regular and special education students and adults. goal of Morningside Academy, educators routinely
Direct Instruction has also produced superior re- gather pre- and postoutcome data on the instruc-
sults with various minority populations, including tional package, including implementation of the
non-English speakers” (p. 3). Morningside Model of Generative Instruction in
A more recent review of Direct Instruction 128 partner schools and school districts across
by the U.S. Institute of Education’s What Works the United States and Canada. Data collection
Clearinghouse (WWC) found only two studies ranges from daily criterion-referenced measures to
that met their evidence standards (i.e., their re- yearly norm-referenced tests. Data collected over
search methodology standards) for review. The 35 years at Morningside Academy and its partner
WWC reported that one Direct Instruction study schools demonstrate that its educational model is
had “no discernible effects on the oral language, effective, especially compared to typical educa-
print knowledge, cognition, and math skills of spe- tional programs. For instance, in one rural public
cial education students” (WWC, 2007), and that school in British Columbia that implemented the
the other study, a review of the Direct Instruc- Morningside model of generative instruction, the
tion curriculum Reading Mastery, had “poten- percentage of students performing at grade level
tially positive effects on the reading achievement in writing rose from 39 to 80% in 9 months. In a
of English-language[-learner] students” (WWC, separate school in British Columbia, the number
2006). We describe educators’ use of and training of students reading in the below-average range de-
in Direct Instruction later. creased by 24% in one school year, and the number
of students reading in the above-average range in-
creased by 35%. Johnson and Street (2004, 2012)
Morningside Model of Generative Instruction
described similar results across each of the major
The Morningside Model of Generative Instruc- academic areas and across a range of grade levels
tion (Johnson & Street, 2004, 2012, 2013) is an and school types (e.g., rural, urban).
Behavioral Approaches to Education 455
Teachers’ Training in and Use of Behavioral tion most often of the eight practices in the survey,
Instruction Practices and Programs and they ranked applied behavior analysis fifth.
These findings suggest at least some evidence
A few publications offer insight into educators’ lev-
exists to indicate that behavioral instruction prac-
els of training and usage of behavioral instruction
tices are prevalent in many U.S. classrooms. Be-
practices. In the 1990s several authors observed
cause U.S. and state policies have emphasized the
that most teachers were not using such practices,
use of evidence-based practices, of which many
particularly educators in regular education class-
integrate core characteristics that are consistent
rooms (e.g., Axelrod, 1996; Binder, 1991; Fredrick
with the science of applied behavior analysis, con-
et al., 2000; Hall, 1991). The data presented earlier tinued or even more widespread use of and train-
about HELPS, Direct Instruction, and the Morn- ing in behavioral instruction practices in the near
ingside Model of Generative Instruction suggest future seem plausible.
that at least some behaviorally oriented programs
are widely used, are included in professional devel-
opment, or a combination. In addition, two more SCHOOL‑BASED
recent studies have addressed the topic through RESPONSE‑TO‑INTERVENTION MODELS
surveys.
Begeny and Martens (2006) asked master’s- Response-to-intervention models are becoming
level elementary, secondary, and special education more prevalent in today’s schools. Response to
teachers in training to estimate how much course- intervention is a multi-tiered prevention model in
work and applied training they received in various which the teacher matches evidence-based prac-
behavioral instruction and measurement practices tices to students’ instructional and behavioral
during their undergraduate and graduate training. needs (Ardoin, Wagner, & Bangs, 2016). These
Results indicated that students in each type of de- models are based on the notion that educators
gree program received little to no coursework or can order the severity of student problems along
applied training in most of the instructional prac- a continuum and deliver services in graduated
tices the survey listed (e.g., prompting, shaping, tiers (Tilly, 2008). A comprehensive description
fading). Moreover, teachers’ training in behavioral of response to intervention is beyond the scope
assessment practices (such as using graphs to make of this chapter; therefore, we discuss it solely from
instructional decisions) and instructional pro- a behavior-analytic perspective. In so doing, we
grams (such as Direct Instruction) was particularly describe (1) legislative influences on response to
low, even for special educators. intervention, (2) fundamental characteristics of
Burns and Ysseldyke (2009) surveyed special response-to-intervention models, and (3) the role
education teachers (N = 174) and school psycholo- that behavior-analytic principles and procedures
gists (N = 333) a few years later about their use play in response-to-intervention implementation.
of eight different evidence-based practices with
small, medium, and large effects in meta-analytic
research. Special educators used a 5-point Likert Legislative Influences
scale to rate how frequently they used each prac- The No Child Left Behind Act of 2001 and the
tice. The researchers surveyed school psycholo- Individuals with Disabilities Education Improve-
gists because they work closely with special educa- ment Act of 2004 have led to important reforms
tors and could offer observations about teachers’ in education practice by emphasizing the measure-
use of the eight evidence-based practices in the ment of student performance for high-stakes deci-
survey. As such, the survey asked school psycholo- sion making (Reschly & Bergstrom, 2009). Several
gists to rank-order the practices by how often they provisions in the No Child Left Behind Act have
observed the practices in classrooms for students aided in the development and implementation
with special needs. Findings showed that most of response to intervention (Burns & Gibbons,
special education teachers reported using applied 2008). Included among these are (1) frequent
behavior analysis almost every day (55% of the re- collection and review of data on student perfor-
spondents) or at least once per week (16%). Also, mance, (2) use of evidence-based instructional
83% reported using direct instruction (the general and intervention procedures, and (3) an empha-
approach, rather than the specific Direct Instruc- sis on prevention and early identification of aca-
tion program) almost every day. School psycholo- demic problems. These provisions have prompted
gists indicated that they observed direct instruc- schools to focus on student-learning outcomes by
456 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
collecting data that teachers then use to make protocol for assessment and educational deci-
educational decisions (Tilly, 2008). Most notably, sion making.
the No Child Left Behind Act has made schools 4. Educators use data-based decision rules for as-
accountable for student learning by creating con- sessing student progress and altering current
tingencies, such as rewards and sanctions, for edu- instruction, intervention, or both.
cational professionals for student outcomes. 5. Educators deliver evidence-based instruction
The Individuals with Disabilities Education Im- and interventions with high fidelity.
provement Act of 2004 introduced response to in- 6. A system exists to screen and to identify stu-
tervention as an alternative method for identifying dents who are not making adequate progress.
students with specific learning disabilities. Histor-
ically, a discrepancy between a student’s score on Educators provide students with universal in-
an individually administered measure of cognitive struction and behavior management strategies
ability and academic achievement identified the that they implement with high fidelity at the
student as having a learning disability. Educators prevention level, which is Tier 1. Approximately
have characterized this approach as a wait-to-fail 10–20% of students will not respond adequately
model, because the child does not receive ser- and will need more intensive intervention, given
vices until his or her achievement level falls sig- the continuum of student problems. Schools typi-
nificantly below that of same-grade or same-age cally conduct universal screening (1) to evaluate the
peers. Gresham (2009) suggested that delaying effectiveness of the instruction and interventions
services decreases their effectiveness. In addition at Tier 1, (2) to identify struggling students so that
to linking response to intervention with disabil- the school can provide additional services imme-
ity classification, the Individuals with Disabilities diately, and (3) to establish school-based norms to
Education Improvement Act promoted the use of aid in evaluating student performance. Universal
positive behavioral interventions and supports in screening involves the assessment of students’
the schools. Positive behavioral interventions and basic academic skills at least three times per year,
supports aligns with response to intervention to using standard curriculum-like probes (Buffman
prevent behavior problems by teaching and rein- et al., 2009; Burns & Gibbons, 2008; Erchul &
forcing appropriate behaviors with evidence-based Martens, 2010; Tilly, 2008). Schools identify and
interventions that teachers apply systematically to monitor at-risk students who score below a criteri-
students, based on the students’ demonstrated lev- on or percentile on a universal-screening measure,
els of need (Sugai & Horner, 2009). to determine whether these students are making
adequate progress in Tier 1.
Typically, schools use a dual-discrepancy ap-
Characteristics of Response to Intervention
proach to determine whether a student is making
Response to intervention is the practice of imple- adequate progress. A dual-discrepancy approach
menting evidence-based instruction and interven- compares the student’s performance level and
tions, systematically evaluating student progress, rate of progress to those of peers. As an example,
and altering instruction and intervention to align a school will identify a student as at risk if he or
with student needs (Buffman, Mattos, & Webber, she scores below the 25th percentile in oral read-
2009; Burns & Gibbons, 2008; Sugai & Horner, ing fluency during universal screening. The school
2009). Although the recommended structure of then must repeatedly measure this student’s read-
response to intervention has varied somewhat in ing skills over time. The school will label the stu-
the literature, a consensus has emerged around dent’s progress inadequate if he or she continues
a three-tiered model (Burns, Deno, & Jimerson, to score below a criterion and to display a rate of
2007) with six critical features (Fuchs & Fuchs, progress significantly below that of typical peers
1998; Fuchs et al., 2002; Sugai & Horner, 2009): (e.g., Burns & Senesac, 2005).
Students making inadequate progress at Tier 1
1. Interventions and instruction are evidence- receive Tier 2 interventions. About 5–15% of the
based. student population will need Tier 2 interventions
2. Educators match interventions to student to supplement Tier 1 supports. Note that Tier 2
needs along a graduated continuum that in- services supplement rather than replace Tier 1 sup-
creases in intensity, frequency, duration, and ports. Educators typically match Tier 2 interven-
individualization. tions to specific student problems and implement
3. Educators use a standardized problem-solving them in a small-group format with high fidelity.
Behavioral Approaches to Education 457
Ideally, these small groups consist of students with behavior if the goal is to decrease problem behav-
similar problems or instructional needs. Educators ior. The target behavior must not be more effortful
monitor students in Tier 2 at least monthly, and than problem behavior and must produce the same
they again use the dual-discrepancy approach to consequences with similar quality, rate, and delay.
evaluate intervention effectiveness and to deter- Selecting target academic behaviors is somewhat
mine whether a student needs more individual- more difficult. The goal of academic instruction
ized intervention, more intensive intervention, or is often for students to complete composite tasks
both. Educators remove Tier 2 interventions when that require several prerequisite skills. Therefore,
a student’s rates of skill acquisition and progress team members must collect data on the compos-
become equal to or exceed those of their peers ite task and on key prerequisite skills (Daly et al.,
(Buffman et al., 2009; Burns & Gibbons, 2008; 2007; Martens & Ardoin, 2010).
Erchul & Martens, 2010; Tilly, 2008). After selection of a target behavior, team mem-
Finally, educators reserve Tier 3 interventions bers must consider an appropriate evidence-based
for students who show inadequate progress at Tier intervention to improve student performance
2. Approximately 5% of the student population at Tier 2 or 3. They can make informed deci-
will receive Tier 3 interventions. An educator usu- sions about possible environmental modifications
ally delivers these in a one-on-one format, with through direct assessment of student skills and be-
high fidelity and with greater intensity, duration, havior in the educational environment (Barnett,
and frequency than Tier 2 interventions. Again, Daly, Jones, & Lentz, 2004; Gresham, Watson,
Tier 3 interventions supplement rather than re- & Skinner, 2001). Daly, Martens, Witt, and Dool
place Tier 1 support. Educators monitor students’ (1997) suggested that students’ deficits in aca-
progress weekly at Tier 3 and use the dual-discrep- demic performance may be related functionally to
ancy approach. Educators return students who instructional variables in several ways: (1) lack of
make adequate progress in Tier 3 to Tier 1 or 2, motivation, (2) insufficient opportunities to prac-
and provide more intensive interventions or make tice the skill, (3) inadequate assistance in how to
a referral for a special education evaluation for any perform the skill, (4) failure to vary curriculum
student who continues to show inadequate prog- materials to promote generalization, and (5) use
ress in Tier 3. of material that is too difficult for a student’s skill
level.
Researchers have implemented brief experi-
Applied Behavior Analysis
mental analyses to test academic interventions
and Response‑to‑Intervention Implementation
matched to these hypothesized functions (Daly
Response to intervention and applied behavior et al., 1997; Daly, Murdoch, Lillenstein, Web-
analysis are related closely, as both emphasize ber, & Lentz, 2002; Martens, Eckert, Bradley, &
student behavior, modification of instructional Ardoin, 1999). Single-case experimental designs
antecedents and consequences, and measurement may be uniquely suited to evaluating functional
of behavior change when educators manipulate relations between instructional variables and stu-
these variables systematically (Ardoin et al., 2016). dent academic performance when used as part of
Essentially, applied behavior analysis provides a a larger, data-based problem-solving model (Bar-
framework within which educators can identify nett et al., 2004). For example, Jones et al. (2009)
target behaviors, develop and match interventions described a systematic problem-solving approach
to those behaviors, and evaluate those interven- for improving the oral reading fluency of six
tions by measuring changes in behavior (Martens third- and fourth-grade students. Problem identi-
& Ardoin, 2010). In this section, we describe how fication used the schoolwide Dynamic Indicators
educators apply principles of applied behavior of Basic Early Literacy Skills passages as univer-
analysis in the context of a response-to-interven- sal-screening measures to identify children at risk
tion model in schools. for reading failure. Problem analysis applied each
Ardoin et al. (2016) outlined several consid- of four evidence-based instructional components
erations for student support team members to (rate-contingent reinforcement, repeated readings,
consider when selecting target behaviors for a listening-passage preview plus phrase drill–error
particular student. First, the team must consider correction, and training on easier material) to a
the social significance of the target behaviors to different passage of equivalent difficulty in a brief
the student, the school, and the family. Second, experimental analysis with a reversal. Researchers
the team must select an appropriate replacement conducted a 3- to 10-trial extended analysis after
458 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
they had identified one or more effective instruc- and the Individuals with Disabilities Education
tional components. The extended analysis evalu- Improvement Act of 2004 have created opportuni-
ated the effects of these strategies in combination. ties for behavior analysts to become more active in
Researchers then implemented the most effective promoting adoption of these evidence-based prac-
instructional package twice per week for the re- tices, and we hope the material in this chapter will
mainder of the school year as the final phase, prob- continue to prompt efforts in this direction.
lem evaluation. Students showed large increases in
oral reading fluency during the extended analysis
of intervention packages, and three of the six stu- REFERENCES
dents showed gains consistently above an aim line
of a one-word increase per week during problem Adams, G. L., & Engelmann, S. (1996). Research on
evaluation. Direct Instruction: 25 years beyond DISTAR. Seattle,
Finally, repeated measurement of intervention WA: Educational Achievement Systems.
Adams, M., Adcock, I., Bereiter, C., Brown, A., Campi-
effectiveness across tiers as part of a student’s cu-
one, J., Carruthers, I., et al. (2005). Open Court Read-
mulative intervention history is vital to the health ing. DeSoto, TX: SRA/McGraw-Hill.
of any response-to-intervention model (Daly et al., Alessi, G. (1987). Generative strategies and teaching for
2007). Teams should measure student behavior generalization. Analysis of Verbal Behavior, 5, 15–27.
continuously over time; otherwise, they may run Anderson, T., & Shattuck, J. (2012). Design-based re-
the risk of continuing ineffective interventions search: A decade of progress in education research?
(Ardoin et al., 2016). Teams should measure both Educational Researcher, 41, 16–25.
inappropriate and appropriate behavior in the Archer, A. L., & Hughes, C. A. (2011). Explicit instruc-
actual learning environment during behavioral tion: Effective and efficient teaching. New York: Guil-
interventions. Martens et al. (2015) suggests that ford Press.
teams measure direct effects of intervention on Ardoin, S. P., McCall, M., & Klubnik, C. (2007). Pro-
moting generalization of oral reading fluency: Pro-
trained material and generalized effects on global
viding drill versus practice opportunities. Journal of
outcome measures for academic performance. Behavioral Education, 16, 54–69.
Examining both sets of data can help determine Ardoin, S. P., Wagner, L., & Bangs, K. E. (2016). Ap-
whether students are not responding to interven- plied behavior analysis: A foundation for response
tion or are responding to intervention but failing to intervention. In S. R. Jimerson, M. K. Burns, &
to generalize what they learn. A. M. VanDerHeyden (Eds.), Handbook of response
to intervention: The science and practice of multi-tiered
systems of support (2nd ed., pp. 29–42). New York:
CONCLUSION Springer.
Axelrod, S. (1996). What’s wrong with behavior analy-
Large numbers of U.S. children continue to have sis? Journal of Behavioral Education, 6, 247–256.
significant difficulties with basic academic skills. Balu, R., Pei, Z., Doolittle, F., Schiller, E., Jenkins, J.,
& Gersten, R. (2015). Evaluation of response to inter-
The National Center for Education Statistics
vention practices for elementary school reading (NCEE
(2015) reported that 31% of fourth graders read 2016-4000). Washington, DC: National Center for
below the basic level, and the number of children Education Evaluation and Regional Assistance, In-
classified as having learning disabilities increased stitute of Education Sciences, U.S. Department of
by 351% from 1976 and 1977 to 1998 and 1999 (U.S. Education.
Department of Education, 2000). Many have cited Barnett, D. W., Daly, E. J., Jones, K. M., & Lentz, F. E.
the failure to adopt evidence-based practices for (2004). Response to intervention empirically based
the relative ineffectiveness of the American pub- special service decisions from single-case designs of
lic education system with low-achieving students increasing and decreasing intensity. Journal of Special
(e.g., Lindsley, 1992; Snow et al., 1998). As noted Education, 38(2), 66–79.
by Carnine (1992), dogma rather than science has Becker, W. C. (1992). Direct Instruction: A twenty year
review. In R. P. West & L. A. Hamerlynck (Eds.), De-
often dictated educational reform over the years,
signs for excellence in education: The legacy of B. F.
enabling fads to cycle through the schools with no Skinner (pp. 71–112). Longmont, CO: Sopris West.
demonstrable improvements in instruction. We Begeny, J. C. (2009). Helping Early Literacy with Practice
have shown in this chapter that effective teaching Strategies (HELPS): A one-on-one program designed to
methods based on the principles of behavior anal- improve students’ reading fluency. Raleigh, NC: Helps
ysis are available to educators and have been for Education Fund. Retrieved from www.helpsprogram.
some time. The No Child Left Behind Act of 2001 org.
Behavioral Approaches to Education 459
Begeny, J. C., Braun, L. M., Lynch, H. L., Ramsay, A. C., practices for teaching students with disabilities. New
& Wendt, J. M. (2012). Initial evidence for using the York: Guilford Press.
HELPS reading fluency program with small instruc- Buffman, A., Mattos, M., & Webber, C. (2009). Pyra-
tional groups. School Psychology Forum: Research in mid response to intervention: RTI, professional learning
Practice, 6, 50–63. communities, and how to respond when kids don’t learn.
Begeny, J. C., Easton, J. E., Upright, J. J., Tunstall, K. R., Bloomington, IN: Solution Tree.
& Ehrenbock, C. A. (2014). The reliability and user- Burns, M. K., Deno, S. L., & Jimerson, S. R. (2007).
feasibility of materials and procedures for monitoring Toward a unified response-to-intervention model.
the implementation integrity of a reading interven- In S. R. Jimerson, M. K. Burns, & A. M. VanDer-
tion. Psychology in the Schools, 51, 517–533. Heyden (Eds.), Handbook of response to intervention:
Begeny, J. C., Laugle, K. M., Krouse, H. E., Lynn, A. The science and practice of assessment and intervention
E., Tayrose, M. P., & Stage, S. A. (2010). A control- (pp. 428–440). New York: Springer.
group comparison of two reading fluency programs: Burns, M. K., & Gibbons, K. A. (2008). Implementing
The Helping Early Literacy with Practice Strategies response-to-intervention in elementary and secondary
(HELPS) program and the Great Leaps K–2 reading schools: Procedures to assure scientific-based practices.
program. School Psychology Review, 39, 137–155. New York: Routledge.
Begeny, J. C., & Martens, B. K. (2006). Assessing pre- Burns, M. K., & Senesac, B. V. (2005). Comparison of
service teachers’ training in empirically-validated dual discrepancy criteria to assess response to inter-
behavioral instruction practices. School Psychology vention. Journal of School Psychology, 43, 393–406.
Quarterly, 21, 262–285. Burns, M. K., & Ysseldyke, J. E. (2009). Reported preva-
Begeny, J. C., Mitchell, R. C., Whitehouse, M. H., lence of evidence-based instructional practices in
Samuels, F. H.., & Stage, S. A. (2011). Effects of the special education. Journal of Special Education, 43,
HELPS reading fluency program when implemented 3–11.
by classroom teachers with low-performing second- Carnine, D. (1992). Expanding the notion of teachers’
grade students. Learning Disabilities Research and rights: Access to tools that work. Journal of Applied
Practice, 26, 122–133. Behavior Analysis, 25, 13–19.
Begeny, J. C., Ross, S. G., Greene, D. J., Mitchell, R. C., Carroll, R. A., Joachim, B. T., St. Peter, C., & Robin-
& Whitehouse, M. H. (2012). Effects of the Help- son, N. (2015). A comparison of error-correction
ing Early Literacy with Practice Strategies (HELPS) procedures on skill acquisition during discrete-trial
Reading Fluency program with Latino English lan- instruction. Journal of Applied Behavior Analysis, 48,
guage learners: A preliminary evaluation. Journal of 257–273.
Behavioral Education, 21, 134–149. Cavanaugh, R. A., Heward, W. L., & Donelson, F.
Begeny, J. C., Upright, J. J., Easton, J. E., Ehrenbock, (1996). Effects of response cards during lesson closure
C. A., & Tunstall, K. R. (2013). Validity estimates on the academic performance of secondary students
and functionality of materials and procedures used in an earth science course. Journal of Applied Behav-
to monitor the implementation integrity of a reading ior Analysis, 29, 403–406.
intervention. Journal of Applied School Psychology, 29, Cengher, M., Budd, A., Farrell, N., & Fienup, D. M.
284–304. (2018). A review of prompt-fading procedures: Im-
Begeny, J. C., Whitehouse, M. H., Methe, S. A., Cod- plications for effective and efficient skill acquisition.
ding, R. S., Stage, S. A., & Neupert, S. (2015). Do Journal of Developmental and Physical Disabilities, 30,
intervention-embedded assessment procedures suc- 155–173.
cessfully measure student growth in reading? Psychol- Chafouleas, S. M., Martens, B. K., Dobson, R. J., Wein-
ogy in the Schools, 52, 578–593. stein, K. S., & Gardner, K. B. (2004). Fluent reading
Binder, C. V. (1991). Marketing measurably effective in- as the improvement of stimulus control: Additive ef-
structional methods. Journal of Behavioral Education, fects of performance-based interventions to repeated
1, 317–328. reading on students’ reading and error rates. Journal
Binder, C. (1996). Behavioral fluency: Evolution of a of Behavioral Education, 13, 67–81.
new paradigm. Behavior Analyst, 19, 163–197. Chard, D. J., Vaughn, S., & Tyler, B. J. (2002). A synthe-
Bonfiglio, C. M., Daly, E. J., Martens, B. K., Lan-Hsiang, sis of research on effective interventions for building
R. L., & Corsaut, S. (2004). An experimental analy- reading fluency with elementary students with learn-
sis of reading interventions: Generalization across ing disabilities. Journal of Learning Disabilities, 35,
instructional strategies, time, and passages. Journal of 386–406.
Applied Behavior Analysis, 37, 111–114. Christ, T. J., Zopluoglu, C., Monaghen, B. D., & Van
Brobst, B., & Ward, P. (2002). Effects of posting, goal Norman, E. R. (2013). Curriculum-based measure-
setting, and oral feedback on the skills of female soc- ment of oral reading: Multi-study evaluation of
cer players. Journal of Applied Behavior Analysis, 35, schedule, duration, and dataset quality on progress
247–257. monitoring outcomes. Journal of School Psychology,
Brownell, M. T., Smith, S. J., Crockett, J. B., & Grif- 51, 19–57.
fin, C. C. (2012). Inclusive instruction: Evidence-based Cividini-Motta, C., & Ahearn, W. H. (2013). Effects
460 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
of two variations of differential reinforcement on behavioral skills training. Journal of Applied Behavior
prompt dependency. Journal of Applied Behavior Analysis, 50, 407–412.
Analysis, 46, 640–650. DiGennaro, F. D., Martens, B. K., & McIntyre, L. L.
Daly, E. J., III, Bonfiglio, C. M., Mattson, T., Persamp- (2005). Increasing treatment integrity through nega-
ieri, M., & Foreman-Yates, K. (2005). Refining the tive reinforcement: Effects on teacher and student
experimental analysis of academic skill deficits: Part behavior. School Psychology Review, 34, 220–231.
I. An investigation of variables affecting generalized Erchul, W. P., & Martens, B. K. (2010). School consulta-
oral reading performance. Journal of Applied Behavior tion: Conceptual and empirical bases of practice. New
Analysis, 38, 485–498. York: Springer.
Daly, E. J., III, Chafouleas, S. M., Persampieri, M., Fredrick, L. D., Deitz, S. M., Bryceland, J. A., & Hum-
Bonfiglio, C. M., & LaFleur, K. (2004). Teaching mel, J. H. (2000). Behavior analysis, education, and ef-
phoneme segmenting and blending as critical early fective schooling. Reno, NV: Context Press.
literacy skills: An experimental analysis of minimal Fuchs, L. S., & Fuchs, D. (1986). Effects of systematic
textual repertoires. Journal of Behavioral Education, formative evaluation: A meta-analysis. Exceptional
13, 165–178. Children, 53, 199–208.
Daly, E. J., III, Lentz, F. E., & Boyer, J. (1996). The in- Fuchs, L. S., & Fuchs, D. (1998). Treatment validity: A
structional hierarchy: A conceptual model for under- unifying construct for reconceptualizing the identi-
standing the effective components of reading inter- fication of learning disabilities. Learning Disability
ventions. School Psychology Quarterly, 11, 369–386. Quarterly, 13, 204–219.
Daly, E. J., III, & Martens, B. K. (1994). A compari- Fuchs, L. S., Fuchs, D., & Speece, D. L. (2002). Treat-
son of three interventions for increasing oral read- ment validity as a unifying construct for identifying
ing performance: Application of the instructional learning disabilities. Learning Disability Quarterly,
hierarchy. Journal of Applied Behavior Analysis, 27, 25, 33–45.
459–469. Gersten, R., Carnine, D., & White, W. A. T. (1984).
Daly, E. J., III, Martens, B. K., Barnett, D., Witt, J. D., The pursuit of clarity: Direct instruction and applied
& Olson, S. C. (2007). Varying intervention delivery behavior analysis. In W. L. Heward, T. E. Heron, D.
in response to intervention: Confronting and resolv- S. Hill, & J. Trap-Porter (Eds.), Focus on behavior
ing challenges with measurement, instruction, and analysis in education (pp. 38–57). Columbus, OH:
intensity. School Psychology Review, 36, 562–581. Merrill.
Daly, E. J., III, Martens, B. K., Hamler, K. R., Dool, E. J., Gortmaker, V. J., Daly, E. J., III, McCurdy, M., Pers-
& Eckert, T. L. (1999). A brief experimental analysis ampieri, M. J., & Hergenrader, M. (2007). Improv-
for identifying instructional components needed to ing reading outcomes for children with learning dis-
improve oral reading fluency. Journal of Applied Be- abilities: Using brief experimental analysis to develop
havior Analysis, 32, 83–94. parent-tutoring interventions. Journal of Applied Be-
Daly, E. J., III, Martens, B. K., Kilmer, A., & Massie, havior Analysis, 40, 203–221.
D. (1996). The effects of instructional match and Gresham, F. M. (2009). Using response to intervention
content overlap on generalized reading performance. for identification of specific learning disabilities. In
Journal of Applied Behavior Analysis, 29, 507–518. A. Akin-Little, S. G. Little, M. A. Bray, & T. J. Kehle
Daly, E. J., III, Martens, B. K., Witt, J. C., & Dool, E. J. (Eds.), Behavioral interventions in schools: Evidence-
(1997). A model for conducting a functional analysis based positive strategies (pp. 205–220). Washington,
of academic performance problems. School Psychology DC: American Psychological Association.
Review, 26, 554–574. Gresham, F. M., Watson, T. S., & Skinner, C. H. (2001).
Daly, E. J., III, Murdoch, A., Lillenstein, L., Webber, L., Functional behavioral assessment: Principles, proce-
& Lentz, F. E. (2002). An examination of methods for dures, and future directions. School Psychology Re-
testing treatments: Conducting brief experimental view, 30(2), 156–172.
analyses of the effects of instructional components Hall, R. V. (1991). Behavior analysis and education: An
on oral reading fluency. Education and Treatment of unfulfilled dream. Journal of Behavioral Education, 1,
Children, 25, 288–316. 305–316.
Daly, E. J., III, Neugebauer, S. R., Chafouleas, S. M., & Haring, N. G., Lovitt, T. C., Eaton, M. D., & Hansen,
Skinner, C. H. (2015). Interventions for reading prob- C. L. (1978). The fourth R: Research in the classroom.
lems: Designing and evaluating effective strategies (2nd Columbus, OH: Merrill.
ed.). New York: Guilford Press. Hogan, A., Knez, N., & Kahng, S. (2015). Evaluat-
Daly, E. J., III, Persampieri, M., McCurdy, M., & Gort- ing the use of behavioral skills training to improve
maker, V. (2005). Generating reading interventions school staffs’ implementation of behavior interven-
through experimental analysis of academic skills: tion plans. Journal of Behavioral Education, 24, 242–
Demonstration and empirical evaluation. School Psy- 254.
chology Review, 34, 395–414. Houvouras, A. J., IV, & Harvey, M. T. (2014). Establish-
Dickson, M. J., & Vargo, K. K. (2017). Training kin- ing fire safety skills using behavioral skills training.
dergarten students lockdown drill procedures using Journal of Applied Behavior Analysis, 47, 420–424.
Behavioral Approaches to Education 461
Individuals with Disabilities Education Improvement rell (Eds.), Practical handbook of school psychology: Ef-
Act of 2004, Pub. L. No. 108-446, 20 U.S.C. 1400 et fective practices for the 21st century (pp. 157–174). New
seq. (2004). York: Guilford Press.
Johnson, K. R., & Layng, T. V. J. (1996). On terms and Martens, B. K., Ardoin, S. P., Hilt, A., Lannie, A. L.,
procedures: Fluency. Behavior Analyst, 19, 281–288. Panahon, C. J., & Wolfe, L. (2002). Sensitivity of
Johnson, K., & Street, E. M. (2004). The Morningside children’s behavior to probabilistic reward: Effects
Model of Generative Instruction: What it means to of a decreasing-ratio lottery system on math per-
leave no child behind. Concord, MA: Cambridge Cen- formance. Journal of Applied Behavior Analysis, 35,
ter for Behavioral Studies. 403–406.
Johnson, K., & Street, E. M. (2012). From the laboratory Martens, B. K., Daly, E. J., & Ardoin, S. P. (2015). Ap-
to the field and back again: Morningside Academy’s plications of applied behavior analysis to school-
32 years of improving students’ academic perfor- based instructional intervention. In H. S. Roane, J.
mance. Behavior Analyst Today, 13, 20–40. L. Ringdahl, & T. S. Falcomata (Eds.), Clinical and
Johnson, K., & Street, E. M. (2013). Response to inter- organizational applications of applied behavior analysis
vention and precision teaching: Creating synergy in the (pp. 125–150). New York: Elsevier.
classroom. New York: Guilford Press. Martens, B. K., & Eckert, T. L. (2007). The instructional
Jones, K. M., Wickstrom, K. F., Noltemeyer, A. L., hierarchy as a model of stimulus control over student
Brown, S. M., Schuka, J. R., & Therrien, W. J. (2009). and teacher behavior: We’re close but are we close
An experimental analysis of reading fluency. Journal enough? Journal of Behavioral Education, 16, 83–91.
of Behavioral Education, 18, 35–55. Martens, B. K., Eckert, T. L., Begeny, J. C., Lewandows-
Keller, F. S. (1968). “Good-bye teacher . . . ” Journal of ki, L. J., DiGennaro, F., Montarello, S., et al. (2007).
Applied Behavior Analysis, 1, 79–89. Effects of a fluency-building program on the reading
Kladopoulos, C. N., & McComas, J. J. (2001). The ef- performance of low-achieving second and third grade
fects of form training on foul-shooting performance students. Journal of Behavioral Education, 16, 39–54.
in members of a women’s college basketball team. Martens, B. K., Eckert, T. L., Bradley, T. A., & Ardoin,
Journal of Applied Behavior Analysis, 34, 329–332. S. P. (1999). Identifying effective treatments from a
Lannie, A. L., & Martens, B. K. (2004). Effects of task brief experimental analysis: Using single-case design
difficulty and type of contingency on students’ allo- elements to aid decision making. School Psychology
cation of responding to math worksheets. Journal of Quarterly, 14(2), 163–181.
Applied Behavior Analysis, 37, 53–65. Martens, B. K., Werder, C. S., Hier, B. O., & Koenig,
Lerman, D. C., Hawkins, L., Hillman, C., Shireman, E. A. (2013). Fluency training in phoneme blending:
M., & Nissen, M. A. (2015). Adults with autism A preliminary study of generalized effects. Journal of
spectrum disorder as behavior technicians for young Behavioral Education, 22, 16–36.
children with autism: Outcomes of a behavioral skills Martens, B. K., & Witt, J. C. (2004). Competence, per-
training program. Journal of Applied Behavior Analy- sistence, and success: The positive psychology of be-
sis, 48, 233–256. havioral skill instruction. Psychology in the Schools,
Lindsley, O. R. (1992). Why aren’t effective teaching 41, 19–30.
tools widely adopted? Journal of Applied Behavior Martens, B. K., Young, N. D., Mullane, M. P., Baxter,
Analysis, 25, 21–26. E. B., Sallade, S. J., Kellen, D., et al. (2019). Effects
Lovitt, T. C., & Esveldt, K. A. (1970). The relative ef- of word overlap on generalized gains from a repeated
fects on math performance of single- versus multiple- readings intervention. Journal of School Psychology,
ratio schedules: A case study. Journal of Applied Be- 74, 1–9.
havior Analysis, 3, 261–270. McDowell, C., & Keenan, M. (2001). Developing flu-
Luck, K. M., Lerman, D. C., Wu, W. L., Dupuis, D. L., ency and endurance in a child diagnosed with atten-
& Hussein, L. A. (2018). A comparison of written, tion deficit hyperactivity disorder. Journal of Applied
vocal, and video feedback when training teachers. Behavior Analysis, 34, 345–348.
Journal of Behavioral Education, 27, 124–144. McGinnis, J. C., Friman, P. C., & Carlyon, W. D. (1999).
MacQuarrie, L. L., Tucker, J. A., Burns, M. K., & Hart- The effect of token rewards on “intrinsic” motivation
man, B. (2002). Comparison of retention rates using for doing math. Journal of Applied Behavior Analysis,
traditional, drill sandwich, and incremental rehears- 32, 375–379.
al flash card methods. School Psychology Review, 31, Mesmer, E. M., Duhon, G. J., Hogan, K., Newry, B.,
584–595. Hommema, S., Fletcher, C., et al. (2010). Generaliza-
Marion, C., Martin, G. L., Yu, C. T., Buhler, C., & Kerr, tion of sight word accuracy using a common stimulus
D. (2012). Teaching children with autism spectrum procedure: A preliminary investigation. Journal of
disorder to mand “Where?” Journal of Behavioral Behavior Education, 19, 47–61.
Education, 21, 273–294. Miltenberger, R. G. (2016). Behavior modification: Prin-
Martens, B. K., & Ardoin, S. P. (2010). Assessing disrup- ciples and procedures (6th ed.). Boston: Cengage
tive behavior within a problem-solving model. In G. Learning.
G. Peacock, R. A. Ervin, E. J. Daly, III, & K. W. Mer- Mitchell, C., & Begeny, J. C. (2014). Improving student
462 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
reading through parents’ implementation of a struc- intervention and school-wide positive behavior sup-
tured reading program. School Psychology Review, 43, ports: Integration of multi-tiered system approaches.
41–58. Exceptionality, 17(4), 223–237.
Morgan, P. L., & Sideridis, G. D. (2006). Contrasting Szadokierski, I., Burns, M. K., & McComas, J. J. (2017).
the effectiveness of fluency interventions for students Predicting intervention effectiveness from reading
with or at risk for learning disabilities: A multilevel accuracy and rate measures through the instructional
random coefficient modeling meta-analysis. Learning hierarchy: Evidence for a skill-by-treatment interac-
Disabilities: Research and Practice, 21, 191–210. tion. School Psychology Review, 46, 190–200.
National Center for Education Statistics. (2015). The Tai, S. S., & Miltenberger, R. G. (2017). Evaluating be-
nation’s report card: Mathematics and reading as- havioral skills training to teach safe tackling skills
sessments. Retrieved February 9, 2015, from www. to youth football players. Journal of Applied Behavior
nationsreportcard.gov/reading_math_2015/#?grade=4. Analysis, 50, 849–855.
National Reading Panel. (2000). Report of the National Therrien, W. J. (2004). Fluency and comprehension
Reading Panel. Teaching children to read: An evidence- gains as a result of repeated reading: A meta-analysis.
based assessment of the scientific research literature Remedial and Special Education, 25, 252–261.
on reading and its implications for reading instruction Tilly, W. D. (2008). The evolution of school psychol-
(NIH Publication No. 00-4769). Washington, DC: ogy to science-based practice: Problem solving and
National Institute of Child Health and Human De- the three-tiered model. In A. Thomas & J. Grimes
velopment. (Eds.), Best practices in school psychology V (pp. 17–
No Child Left Behind Act of 2001, Pub. L. No. 107-110, 35). Bethesda, MD: National Association of School
20 U.S.C. 6301 (2002). Psychologists.
Reschly, D. J. (2004). Paradigm shift, outcomes, criteria, U.S. Department of Education. (2000). Twenty-second
and behavioral interventions: Foundations for the fu- annual report to Congress on the implementation of the
ture of school psychology. School Psychology Review, Individual with Disabilities Education Act. Washing-
33, 408–416. ton, DC: U.S. Government Printing Office.
Reschly, D. J., & Bergstrom, M. K. (2009). Response Vadasy, P. F., Wayne, S., O’Conner, R., Jenkins, J., Fire-
to intervention. In T. B. Gutkin & C. R. Reynolds baugh, M., & Peyton, J. (2000). Sound Partners [Cur-
(Eds.), The handbook of school psychology (4th ed., riculum program]. Longmont, CO: Cambium Learn-
pp. 434–460). Hoboken, NJ: Wiley. ing Group/Sopris West.
Saville, B. K., Zinn, T. E., Neef, N. A., Van Norman, R., Valentino, A. L., LeBlanc, L. A., Veazey, S. E., Weaver,
& Ferreri, S. J. (2006). A comparison of interteach- L. A., & Raetz, P. B. (2019). Using a prerequisite skills
ing and lecture in the college classroom. Journal of assessment to identify optimal modalities for mand
Applied Behavior Analysis, 39, 49–61. training. Behavior Analysis in Practice, 12, 22–32.
Shernoff, E. S., Mariñez-Lora, A. M., Frazier, S. L., Ja- Vellutino, F. R., Scanlon, D. M., & Tanzman, V. S.
kobsons, L. J., Atkins, M. S., & Bonner, D. (2011). (1998). The case for early intervention in diagnosing
Teachers supporting teachers in urban schools: What specific reading disability. Journal of School Psychol-
iterative research designs can teach us. School Psy- ogy, 36, 367–397.
chology Review, 40, 465–485. Watkins, C. L. (1997). Project Follow Through: A case
Skinner, B. F. (1954). The science of learning and the art study of contingencies influencing instructional prac-
of teaching. Harvard Educational Review, 24, 86–97. tices of the educational establishment. Concord, MA:
Skinner, B. F. (1957). Verbal behavior. Acton, MA: Co- Cambridge Center for Behavioral Studies.
pley. What Works Clearinghouse (WWC). (2006). WWC
Skinner, B. F. (1984). A matter of consequences. New intervention report: Reading Mastery/SRA/McGraw
York: New York University Press. Hill. Retrieved February 2016, from http://ies.ed.gov/
Smith, T. (2001). Discrete trial training in the treat- ncee/wwc/pdf/intervention_reports/WWC_Reading_
ment of autism. Focus on Autism and Other Develop- Mastery_092806.pdf.
mental Disabilities, 16, 86–92. What Works Clearinghouse (WWC). (2007). WWC
Snow, C. E., Burns, M. S., & Griffin, P. (1998). Prevent- intervention report: Direct Instruction, DISTAR,
ing reading difficulties in young children. Washington, and Language for Learning. Retrieved February 2016,
DC: National Academy Press. from http://ies.ed.gov/ncee/wwc/pdf/intervention_re-
Stocco, C. S., Thompson, R. H., Hart, J. M., & Soriano, ports/WWC_Direct_Instruction_052107.pdf.
H. L. (2017). Improving the interview skills of col- Wolery, M., Bailey, D. B., & Sugai, G. M. (1988). Ef-
lege students using behavioral skills training. Journal fective teaching: Principles and procedures of applied
of Applied Behavior Analysis, 50, 495–510. behavior analysis with exceptional students. Boston:
Stokes, T. F., & Baer, D. M. (1977). An implicit tech- Allyn & Bacon.
nology of generalization. Journal of Applied Behavior Wood, K. D., & Blanton, W. E. (2009). Literacy instruc-
Analysis, 10, 349–367. tion for adolescents: Research-based practice. New
Sugai, G., & Horner, R. H. (2009). Responsiveness-to- York: Guilford Press.
CHAP TER 27
Florence D. DiGennaro Reed, Lisa M. Hagermoser Sanetti,
and Robin S. Codding
The Individuals with Disabilities Education Im- already face. In fact, 4.2% of children 17 years
provement Act (IDEA) of 2004 requires states or younger in the United States were diagnosed
and local education agencies to offer students with with a disability in 2013 (Houtenville, Brucker,
disabilities a free appropriate public education in & Lauer, 2014); this percentage represents more
the least restrictive environment, and it empha- than 3 million public school students who may
sizes the right to a high-quality education. These require specialized services. Teachers often are
and other educational reforms, such as the No primarily responsible for implementing preventa-
Child Left Behind Act of 2001, mandate accom- tive practices, such as effective classroom man-
modations and modified instruction to ensure that agement, and academic and behavior interven-
students with disabilities participate in the general tion plans (Forman, Olin, Hoagwood, Crowe, &
education curriculum to the maximum extent ap- Saka, 2009; Fahmie & Luczynski, 2018; Forman
propriate with fellow students without disabilities. et al., 2013; Long et al., 2016). Nevertheless, they
Such mandates oblige teachers to support the edu- would benefit from the expertise of a professional
cational needs of an increasingly diverse group of who has the training and experience to assess
students with individualized needs (Putnam, Han- and treat academic and behavioral challenges
dler, Rey, & McCarty, 2005). Teachers already have in the classroom. Subsequently, the growing de-
many responsibilities for student learning, such as mands on educators and increased accountability
providing core curricular instruction; identifying for high-quality education have facilitated the de-
individual learning needs; differentiating instruc- velopment of school consultation as a service one
tion; and creating a positive, safe, and engaging or more professionals offer to educators (Luiselli
classroom environment for students. Teachers also & Diament, 2002).
assume an important role in promoting and sup- Behavioral consultation—a model that relies on
porting the use of evidence-based prevention and the principles of behavior analysis—has decades
intervention strategies. of empirical support and scientific evidence dem-
We cannot expect teachers to master the skills onstrating its effectiveness and popularity (e.g.,
necessary to address the unique needs of every Sheridan et al., 2012; Sheridan, Welch, & Orme,
student they will educate during their career 1996). Behavioral consultation is the most widely
given the many responsibilities and pressures they evaluated type of consultation (Sheridan et al.,
463
464 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
if the consultee does not implement it. Finally, carello, 2011). The first stage, establishing relation-
some consultees will be responsible for supervision ships, focuses on the importance of developing a
of other individuals who assist with intervention productive relationship between the consultant
implementation. For example, a teacher may su- and consultee. Research consistently demonstrates
pervise a paraprofessional who implements the in- that trust, openness, flexibility, genuineness, and
tervention that a consultant and the teacher have positive communication are essential to facilitat-
developed for a student. ing collaboration during the consultation process
(Dinnebeil, Hale, & Rule, 1996; Gilles et al., 2011;
Kratochwill et al., 2014). Although competence
Client
in problem identification and analysis are neces-
Clients typically have one responsibility in the sary to successful consultation, they may not be
consultation process, which is to respond to the sufficient. The integration of technical expertise
intervention, thereby informing the goals of con- in consultation and behavioral theory with posi-
sultation. For example, if a client is improving, the tive interpersonal skills is essential to maximizing
goal of consultation may be to continue the inter- consultation outcomes (Kratochwill et al., 2014).
vention for a specified period and then develop a The second stage, problem identification, focuses
data-based, systematic plan for fading intervention on operationally defining the problem behavior,
or consultation supports. Alternatively, if a client the expected behavior, and the discrepancy be-
is not progressing, or if his or her rate of progress tween these. Researchers consider this the most
is not acceptable, these data will inform the con- critical step, as accurate problem identification is
sultant and consultee that they need to reengage predictive of effective planning and implementa-
in the problem-solving process. Consultants and tion (Bergan & Tombari, 1975, 1976). The third
consultees may decide to engage clients who are stage, problem analysis, includes analyzing the en-
able in the process of setting and evaluating their vironmental conditions occasioning the problem
goals, which is an effective intervention technique behavior, developing hypotheses regarding the
(e.g., O’Leary & Dubey, 1979). function(s) of that behavior, and designing an in-
tervention plan. The consultant may recommend
additional data collection, typically using behav-
Consultation Models
ioral assessment techniques (e.g., antecedent–be-
Researchers have developed and evaluated numer- havior–consequence [A-B-C] recording, time
ous models of behavioral consultation over the sampling, direct observation, functional analysis).
past 40-plus years. In addition, behavior analysts The fourth stage, plan implementation, focuses on
have published numerous texts outlining consulta- the consultee’s implementation of the interven-
tion strategies; perhaps the most popular of these tion. The consultant (1) ensures the availability of
is Bailey and Burch (2010). We have not reviewed materials, (2) ensures the consultee has the skills
these texts here, as they do not describe empiri- to implement the intervention, (3) monitors con-
cally evaluated consultation models, even though sultee implementation and client progress, (4) con-
they may be informative. Rather, we note overlap ducts regular check-ins or interacts with the con-
between strategies and consultation models when sultee, (5) regularly analyzes data, and (6) makes
applicable. We briefly describe three empirically data-based decisions about the need to revise the
supported consultation models that researchers intervention. The fifth stage, plan evaluation, fo-
have implemented most commonly. cuses on data-based evaluation of plan effective-
ness and goal attainment. Plan evaluation is not
necessarily the end of the consultation process,
Problem‑Solving Consultation
because the consultant may develop new goals for
Historically, problem-solving consultation fol- maintenance, generalization, other behaviors, or a
lowed a four-stage process (Bergan & Kratochwill, combination of these.
1990); however, researchers have identified a fifth
stage as a best practice (Kratochwill et al., 2014).
Integrated Model of School Consultation
Although we present the stages in a linear order
here, they are fluid in practice; stages often over- The following sources inform the integrated
lap, and consultants and consultees may return model of school consultation: (1) Bergan and Kra-
to a previous stage to meet the consultation goals tochwill’s (1990) model of behavioral consulta-
(Gilles, Kratochwill, Felt, Schienebeck, & Vac- tion, (2) Caplan’s (1963) model of mental health
466 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
consultation, (3) research on relational commu- Throughout the consultation process, the in-
nication (Erchul & Chewning, 1990), and (4) tegrated model of school consultation emphasizes
social-psychology research on social power and the importance of supporting consultees’ efforts as
influence (Erchul & Martens, 2010). This model teachers and intervention agents, while simulta-
includes three stages: precursors to consultation, neously empowering them to become capable and
the consultation process, and consultation out- independent problem solvers (Witt & Martens,
comes. During the precursors stage, a consultant 1988). Doing so may include connecting con-
must have or develop a general understanding of sultees with resources, providing emotional sup-
the operations of the consultation setting (i.e., ports, and developing their professional skills (e.g.,
schools and classrooms) and must fully and sys- assessment and intervention training).
tematically enter the consultation setting. The
consultant must understand the intervention and
Conjoint Behavioral Consultation
referral processes in the consultation setting, the
consultee’s expectations of the consultative rela- Youth who demonstrate challenging behaviors
tionship, previously implemented problem-solving often do so across multiple environments and con-
efforts and interventions, and the consultee’s per- texts (e.g., home, school, community; Sheridan
ception of the factors that are responsible for the et al., 2012). Conjoint behavioral consultation is
problem behavior (Erchul & Martens, 2010). grounded in ecological-systems theory (Bronfen-
The consultation process begins after the con- brenner, 1979), in which the consultant actively
sultant has gained entry to the consultation set- involves caregivers as co-consultees with teach-
ting. In this model, the process centers on three ers, thereby addressing the unique perspectives
interrelated tasks: problem solving, social influ- and contributions of the primary environments
ence, and support and development. The problem- in which youth develop. Conjoint behavioral
solving process is based on Bergan and Kratoch- consultation fosters positive caregiver–teacher
will’s model previously described. This model of relationships, incorporates data-based problem
consultation emphasizes the consultant’s use of so- solving and collaboration, and implements evi-
cial influence during the problem-solving process dence-based interventions across home and school
because consultants often need to alter consultees’ settings (Sheridan et al., 2012).
attitudes and beliefs to benefit clients. As outlined Conjoint behavioral consultation includes four
in the French and Raven (1959) model, there are six stages implemented in a collaborative manner
bases of social power: (1) coercive power, in which among the consultant and caregiver and teacher
the consultee perceives that the consultant can consultees (Sheridan & Kratochwill, 2008). In the
punish the consultee if he or she doesn’t comply; first stage, conjoint needs identification, the con-
(2) reward power, in which the consultee perceives sultant and consultees work together to (1) iden-
that the consultant can reward the consultee for tify the client’s most pressing needs in home and
compliance; (3) legitimate power, in which the school settings, (2) select and operationally define
consultee believes that the consultant has a legiti- target behaviors, (3) choose the specific settings
mate right to influence the consultant’s beliefs or for and goals of consultation, and (4) collect base-
attitudes based on his or her position; (4) expert line data across settings. The consultant’s goals
power, in which the consultee perceives that the in this stage of conjoint behavioral consultation
consultant has expertise in the area of interest to include developing the relationship between the
the consultee; (5) referent power, in which the con- caregiver(s) and teachers, and identifying the
sultant is able to influence the consultee via the strengths and capacities of the client, family, and
consultee’s real or perceived identification with the school to promote goal attainment. In the sec-
consultant; and (6) informational power, in which ond stage, conjoint needs analysis, the consultant
the consultant is able to influence the consultee reviews data to identify environmental variables
by providing logical information to the consultee across settings that may influence target behavior,
about the need for change. Researchers have used with a specific focus on setting events, ecological
these bases of social power to develop numerous conditions, and cross-system variables. The con-
strategies for social influence, and consultants may sultant develops hypotheses about the function of
use these strategies in the problem-solving process. the target behavior, and the consultant and con-
A full review of social power and social influence is sultees engage in a collaborative, strengths-based
beyond the scope of this chapter; interested readers approach to developing an intervention plan
should see Erchul and Martens (2010). across home and school settings. In the third stage,
Teacher Consultation 467
cross-system plan implementation, caregiver(s) and ment, graphing student performance, single-case
teachers implement the intervention plan in the design, use of timed trials for repeated practice of
home and school, with ongoing support from the academic skills, strategies for promoting general-
consultant. Simultaneous implementation across ization of skills, or use of guided notes.
settings increases the likelihood of generalization Most general education teachers reported a lack
and maintenance as an outcome of conjoint be- of familiarity with evidence-based behavioral in-
havioral consultation (Sheridan, Clarke, & Burt, terventions and programs, such as the Good Be-
2008). In the fourth stage, conjoint plan evaluation, havior Game (Barrish, Saunders, & Wolf, 1969) or
the behavioral data form the basis of a discussion the Olweus Bullying Prevention Program (Olweus,
among the consultant, caregiver(s), and teacher 1991). Only 57% knew whether professionals in
regarding next steps (e.g., identifying new goals, their schools conducted functional behavioral as-
continuing intervention). sessments and implemented behavioral interven-
tions (Stormont et al., 2011a). Many teachers also
reported a need for more training in classroom
TEACHER INVOLVEMENT WITH BEHAVIORAL management and behavioral interventions (Re-
ASSESSMENT AND INTERVENTION inke, Stormont, Herman, Puri, & Goel, 2011).
Across surveys, teachers and teachers in train-
Despite a mutually compatible interest in helping ing in special education (1) reported significantly
children become independent and effective learn- more coursework and applied training in academic
ers, teachers encounter many barriers that consul- assessment, (2) agreed at significantly higher rates
tants should consider. Increasing accountability that using evidence-based practices in behavioral
standards and adoption of the Common Core State intervention is important, and (3) rated non-ev-
Standards Initiative have led to growing demands idence-based practices (e.g., having a discussion
on teachers’ time and have intensified and added with a child following misbehavior) as less ac-
teaching responsibilities (Maras, Splett, Reinke, ceptable than teachers and teachers in training in
Stormont, & Herman, 2014). These changes are general education rated them (Begeny & Martens,
occurring during a climate of persistent decreases 2006; Stormont et al., 2011b).
in school budgets, leading to an ever-present no- Despite low levels of training on behavioral
tion of having to do more with less. Simultane- practices and limited familiarity with specific ev-
ously, time allocated in the school day for plan- idence-based programs, overwhelmingly teachers
ning and collaboration with other professionals is agree on the importance of using evidence-based
often minimal, even though teachers report desir- behavioral intervention practices such as (1) iden-
ing more opportunities for collaboration (Long et tifying triggers and reinforcers for problem behav-
al., 2016; Maras et al., 2014). Bosworth Gingiss, ior; (2) teaching skills using examples, practice,
Potthoff, and Roberts-Gray (1999) have identified and feedback; (3) reinforcing and practicing be-
these diminishing resources (i.e., materials, staff, havioral expectations; (4) adapting instructional
funding, facilities, and time) as some of the most strategies to increase engagement and opportu-
important predictors of effective implementation nities for success; and (5) observing and record-
of school-based interventions. ing behavior (Stormont et al., 2011b). Moreover,
Researchers have conducted surveys to assess teachers largely agree that they should not use
teachers’ training in behaviorally oriented prac- non-evidence-based practices, such as grade re-
tices (Begeny & Martens, 2006; Stormont, Rein- tention and suspension. Therefore, determining
ke, & Herman, 2011a, 2011b). General education teachers’ baseline knowledge of behavioral prin-
teachers in training at both primary and second- ciples, concepts, and skills is important when the
ary levels report low levels of coursework and ap- consultant is establishing his or her own and the
plied training in most instructional concepts, teachers’ responsibilities for intervention plan-
strategies, programs, and assessment practices ning, development, and implementation. For ex-
that are behaviorally oriented. As many as 43% ample, given the limited opportunities teachers
of teacher trainees indicated no training on the have had to graph data, this responsibility might
items assessed in one study (Begeny & Martens, be better suited for a consultant. Considering the
2006). For example, 40% or more of teachers re- barriers in the school, such as limited time and
ported receiving no coursework or applied training resources, working with teachers to mitigate those
opportunities with direct instruction, personalized barriers through consultative support is impor-
systems of instruction, curriculum-based measure- tant.
468 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
tion; (2) modeling of the intervention; (3) role this step, having more than one consultant to as-
plays and practice implementing the intervention, sist with providing feedback may be useful. Teach-
with immediate feedback and error correction; ers benefit from receiving behavior-specific praise
and (4) ongoing follow-up support facilitate suc- for correct intervention implementation, and cor-
cessful intervention implementation. rective feedback with instruction and additional
Behavioral skills training is one approach to modeling for incorrect implementation.
this training model (Parsons, Rollyson, & Reid, The final aspect of behavioral skills training
2012; Sarokoff & Sturmey, 2004). During such is to ensure that teachers meet a criterion level
training, the instructor (1) describes the target in- established by the consultant, trainer, or both.
tervention, (2) provides a written protocol or brief For example, having each teacher implement
description of the intervention, (3) demonstrates the behavioral intervention with 100% accuracy
the intervention, (4) requires the teacher (or con- would be useful (Parsons et al., 2012). The con-
sultee) to practice the intervention, (5) provides sultant can arrange tests during training to as-
feedback during practice, and (6) repeats Steps 4 sess performance in which each teacher practices
and 5 until the teacher has achieved mastery. We implementing the intervention with the trainer
recommend that a consultant tell a teacher the or consultant, after two opportunities to practice
rationale for implementing an intervention. The the intervention with a partner. An alternative
consultant should tell the teacher how the inter- strategy is to video-record teachers practicing the
vention will help students achieve the expected intervention with one another and score their per-
goals and how the intervention is consistent with formance after the training (Sarokoff & Sturmey,
the school context and mission. The written pro- 2004). The consultant can conduct additional
tocol should include only the essential steps that a training and proficiency tests with teachers who
teacher needs to follow. Extraneous information or do not meet the mastery criterion.
extensive explanations and examples may reduce Although these training procedures have im-
the likelihood that teachers will use these written proved intervention integrity, data from two meta-
protocols. A posttraining test may be useful to as- analyses suggest that training alone does not pro-
sess the teacher’s comprehension of the interven- duce sustained implementation (Noell et al., 2014;
tion (DiGennaro Reed et al., 2010). Slider et al., 2006; Solomon, Klein, & Politylo,
The consultant should model implementation 2012). Noell et al. (2014) found that mean con-
of the intervention for the teacher after introduc- sultee intervention implementation was 36%, re-
ing the verbal explanation and written protocol. gardless of whether the consultant used behavioral
The consultant can model the intervention con- skills training or other standard training proce-
ventionally or in vivo, using role plays during which dures. These data are consistent with recommen-
the consultant serves as the implementer and the dations by Joyce and Showers (2002) indicating
teacher serves as the confederate. The consultant that teachers need ongoing follow-up support. Two
can also use self-created or web-accessed videos novel antecedent strategies, intervention choice
(DiGennaro Reed et al., 2010; Slider et al., 2006). and implementation planning, have demonstrated
A postmodeling test may be useful to assess the promise for sustaining intervention adherence
teacher’s comprehension of the intervention. The after training (Dart, Cook, Collins, Gresham,
test might describe role-play scenarios and ask the & Chenier, 2012; Sanetti, Kratochwill, & Long,
teacher to identify errors of omission and commis- 2013). We describe these two approaches below.
sion (Slider et al., 2006).
The consultant should schedule time for the “Test-Driving” Intervention Options and Choosing
teacher to practice implementing the intervention One. Dart et al. (2012) conducted a study where-
during role plays, taking turns in the roles of imple- by teachers used one of three interventions (e.g.,
menter and student. The consultant can conduct self-monitoring, check-in/check-out, and response
these practice opportunities simultaneously with cost). Teachers used each intervention once dur-
pairs or groups of teachers, or can have one pair or ing a 30-minute interval that the teacher and con-
group model the intervention for the larger group sultant identified during a meeting. Teachers sam-
participating in the training (Parsons et al., 2012). pled each intervention over the course of 2 days
During this time, the consultant should circulate with a target student in their classroom. Teachers
around the room to each pair or group and provide then rank-ordered each intervention from least to
immediate error correction and feedback on inter- most acceptable. Each teacher ultimately imple-
vention implementation. Given the importance of mented the intervention he or she chose as most
470 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
acceptable. The opportunity to try different inter- ing its effectiveness at facilitating adherence to
ventions and select the most acceptable produced academic and behavioral intervention plans
sustained accurate intervention implementation. (Noell et al., 2014). In fact, researchers have es-
tablished performance feedback as an evidence-
Implementation Planning for Logistics and Barri- based practice (Fallon, Collier-Meek, Maggin,
ers. Sanetti et al. (2013) have developed a model Sanetti, & Johnson, 2015). Implementation out-
of training that emphasizes preimplementation comes with performance feedback alone are supe-
planning. First, the teacher and consultant de- rior (Noell et al., 2014) to those of self-monitoring
velop an action and a coping plan. During the or variations of performance feedback (e.g., meet-
action-planning process, the teacher and consul- ing cancelation, directed rehearsal). Consultants
tant (1) define the intervention steps; (2) adapt can deliver performance feedback on any sched-
these steps to match the teacher’s classroom con- ule (e.g., daily, weekly) after direct observation of
text; (3) establish the logistical details of inter- teachers’ intervention integrity. Researchers have
vention delivery (i.e., when, how often, for how used performance feedback only when interven-
long, where, by whom); and (4) identify resources tion integrity reaches a threshold (e.g., <75%).
needed to facilitate implementation (Sanetti et The interval between direct observations of in-
al., 2013). Action planning crystallizes the roles tervention integrity and feedback to teachers can
of the consultant and teacher in a way that also be as short as a few hours or up to 1 week without
ensures greater implementation feasibility. Coping affecting intervention integrity outcomes (Solo-
planning requires the teacher and consultant to mon et al., 2012).
identify up to four barriers and develop strategies Consultants most often deliver performance
to address or mitigate each barrier. A consultant feedback verbally during individual meetings with
can conduct action and coping planning during a teacher, but a consultant also can deliver per-
one 20-minute meeting (Sanetti et al., 2015). Em- formance feedback during team-based meetings
pirical findings suggest that implementation plan- or via email or written notes (DiGennaro, Mar-
ning, when provided after standard consultation tens, & Kleinmann, 2007; Fallon et al., 2015).
and training, results in higher intervention ad- Performance feedback includes (1) review of in-
herence and quality (Sanetti et al., 2015; Sanetti, tervention adherence data with teachers, provided
Collier-Meek, Long, Kim, & Kratochwill, 2014; graphically, verbally, or both; (2) praise for inter-
Sanetti et al., 2013). vention steps implemented accurately and with
high quality; (3) corrective feedback for errors of
omission and commission; (4) review of the inter-
Ongoing Follow‑Up Supports
vention plan; and (5) an opportunity for teach-
The standard model of behavioral consultation as- ers to ask questions and problem-solve barriers to
sumes that consultants meet with teachers weekly implementation. The materials required for per-
and conduct interviews to evaluate how the plan is formance feedback include the intervention plan
going (Bergan & Kratochwill, 1990). During this or protocol and observational data on interven-
interview, a consultant asks a teacher whether he tion integrity. Consultants often review graphic
or she is implementing the plan and if the target displays of teacher implementation data, student
student or students are improving. The consultant data, or both. Performance feedback meetings can
schedules time for teacher questions. The con- also include directed rehearsal, in which a teacher
sultant does not review intervention protocols or practices intervention components that he or she
data. This form of follow-up meeting is not effec- omitted or implemented incorrectly, and cancel-
tive for enhancing intervention implementation lation of performance feedback meetings contin-
(Noell et al., 2005, 2014). Yet a robust finding in gent on accurate implementation that meets a
the literature is the benefit of ongoing consulta- criterion (e.g., DiGennaro et al., 2007; DiGenn-
tive support after formal training. Three specific aro et al., 2005). Researchers have also embedded
strategies that improve teachers’ intervention goal setting with performance feedback meetings
implementation include performance feedback, (Codding & Smyth, 2008; Duchaine, Jolivette, &
self-monitoring, and coaching. We describe these Fredrick, 2011; Hall & Macvean, 1997; Hawkins
strategies below. & Heflin, 2011; Myers, Simonsen, & Sugai, 2011).
tion of discrete-trial instruction (Belfiore, Fritts, & feedback, after the consultant removes follow-up
Herman, 2008), token economies (Pelletier, Mc- supports (Noell et al., 2014). Consultants can pro-
Namara, Braga-Kenyon, & Ahearn, 2010; Petscher mote maintenance by fading intervention support
& Bailey, 2006; Plavnick, Ferreri, & Maupin, or intensity. Dynamic fading is a common strategy
2010) and behavior support plans (Mouzakitis et that researchers have used (e.g., DiGennaro et
al., 2015). Self-monitoring requires teachers to al., 2007; DiGennaro et al., 2005; Kaufman et al.,
record their own behavior; however, researchers 2013); it is a thinning schedule of follow-up support
have provided considerable support and training contingent on teachers’ intervention implementa-
to ensure that teachers record implementation tion. For example, let’s assume that a consultant
accuracy correctly. In these studies, teacher self- initially provides follow-up support twice weekly
monitoring occurred (1) immediately after a pager after each observation of a teacher. After the
signaled that a teacher implemented an interven- teacher demonstrates 100% accurate intervention
tion step, (2) after the teacher viewed a video of implementation for three consecutive sessions, the
him- or herself implementing an intervention, consultant might provide follow-up support once
(3) via completion of a checklist on the same day weekly. The consultant could thin the schedule
the teacher implemented the intervention, or (4) to once every other week when the teacher meets
with a combination of these strategies. Research- the criterion for a second time. Follow-up support
ers observed higher levels of implementation and returns to a richer schedule if the teacher cannot
greater consistency when they included additional maintain high levels of accurate implementation.
supports, such as using a pager, prompting teachers Gross, Duhon, and Doerksen-Klopp (2014) used a
to perform steps in the intervention plan, viewing changing-criterion design to evaluate a variation
a video of themselves, or following written perfor- of dynamic fading that the authors called fading
mance feedback illustrating comparisons between with indiscriminable contingencies. The consultant
the consultant’s and the teacher’s integrity data. initially provided follow-up support daily. The
consultant decreased follow-up support to every
Coaching. The two common models of coach- other day and then to weekly each time the teach-
ing are supervisory and side-by-side (i.e., in vivo; er’s implementation integrity was 100% for 2 con-
Blakely, 2001; Joyce & Showers, 1995). Supervi- secutive days.
sory coaching involves an observation of a teacher Another option may be to fade the intensity of
implementing a strategy or intervention followed follow-up support. Mouzakitis et al. (2015) removed
by descriptive feedback to the teacher regard- consultant-provided performance feedback after
ing strengths and challenges. During side-by-side pairing performance feedback with self-monitor-
coaching, the coach (1) models the intervention ing. Two of three teachers maintained acceptable
in the classroom; (2) provides immediate in vivo levels of intervention adherence when the consul-
praise and corrective feedback after the teacher tant used self-monitoring only. The third teacher
implements the intervention; (3) prompts the in- required both components of support to be suc-
tervention steps when needed; and (4) provides cessful.
additional modeling when needed. The consul-
tant can implement coaching in any frequency
or duration (e.g., one session, weekly). Systematic CONCLUSION
reviews suggest that coaching improves teachers’
intervention implementation. Less evidence ex- In this chapter, we have summarized empirically
ists on whether these improvements translate into supported models of behavioral consultation and
improved student outcomes, and only a small per- have outlined the roles and responsibilities of in-
centage of existing studies have analyzed fidelity dividuals in the consultation process. We have
of the coaching process (Kretlow & Bartholomew, described the various barriers teachers encounter
2010; Stormont, Reinke, Newcomer, Marchese, & in the consultative relationship and advocate that
Lewis, 2015). consultants work alongside teachers to mitigate
those barriers through consultative support. We
have concluded the chapter with research docu-
Fading Consultation Support
menting the effective use of teachers in behavioral
by using a Data‑Based Process
assessment and intervention, which requires an
Teachers can maintain high levels of implemen- evidence-based approach to training and ongoing
tation integrity, particularly with performance follow-up support.
472 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
tematic review and evaluation. Exceptional Children, through staff development. Alexandria, VA: Associa-
81, 227–246. tion for Supervision and Curriculum Development.
Forman, S. G., Olin, S. S., Hoagwood, K. E., Crowe, M., Kaufman, D., Codding, R. S., Markus, K. A., Tryon, G.
& Saka, N. (2009). Evidence-based intervention in S., & Kyse, E. N. (2013). Effects of verbal and written
schools: Developers’ views of implementation barri- performance feedback on intervention adherence:
ers and facilitators. School Mental Health, 1, 26–36. Practical application of two delivery formats. Jour-
Forman, S. G., Shapiro, E. S., Codding, R. S., Gonzales, nal of Educational and Psychological Consultation, 23,
J. E., Reddy, L. A., Rosenfield, S. A., et al. (2013). Im- 264–299.
plementation science and school psychology. School Kratochwill, T. R., Altschaefl, M. R., & Bice-Urbach,
Psychology Quarterly, 28(2), 77–100. B. (2014). Best practices in problem-solving consul-
French, J. R. P., & Raven, B. H. (1959). The bases of tation: Applications in prevention and intervention
social power. In D. Carwright (Ed.), Studies in social systems. In P. L. Harrison & A. Thomas (Eds.), Best
power (pp. 159–167). Ann Arbor, MI: Institute for practices in school psychology: Data-based and collab-
Social Research. orative decision making (pp. 461–482). Bethesda, MD:
Gilbertson, D., Witt, J. C., Singletary, L. L., & VanDer- National Association of School Psychologists.
Heyden, A. (2007). Supporting teacher use of inter- Kratochwill, T. R., & Bergan, J. R. (1990). Behavioral
vention: Effects of response dependent performance consultation in applied settings: An individual guide.
feedback on teacher implementation of a math inter- Boston: Kluwer Academic.
vention. Journal of Behavioral Education, 16, 311–326. Kratochwill, T. R., Elliott, S. N., & Busse, R. T. (1995).
Gilles, C. N. R., Kratochwill, T. R., Felt, J. N., Schiene- Behavior consultation: A five-year evaluation of con-
beck, C. J., & Vaccarello, C. A. (2011). Problem solv- sultant and client outcomes. School Psychology Quar-
ing consultation: Applications in evidence-based terly, 10(2), 87–117.
prevention and intervention. In M. A. Bray & T. J. Kretlow, A. G., & Bartholomew, C. C. (2010). Using
Kehle (Eds.), The Oxford handbook of school psychol- coaching to improve the fidelity of evidence-based
ogy (pp. 666–667). New York: Oxford University practices: A review of studies. Teacher Education and
Press. Special Education, 33, 279–299.
Gormley, M. J., & DuPaul, G. J. (2015). Teacher-to- Kunnavatana, S. S., Bloom, S. E., Samaha, A. L., &
teacher consultation: Facilitating consistent and ef- Dayton, E. (2013). Training teachers to conduct trial-
fective intervention across grade levels for students based functional analyses. Behavior Modification, 37,
with ADHD. Psychology in the Schools, 52(2), 124– 707–722.
138. Lerman, D. C., Tetreault, A., Hovanetz, A., Strobel,
Gross, T., Duhon, G. J., & Doerksen-Klopp, B. (2014). M., & Garro, J. (2008). Further evaluation of a brief,
Enhancing intervention integrity maintenance intensive teacher-training model. Journal of Applied
through fading with indiscriminable contingencies. Behavior Analysis, 41, 243–248.
Journal of Behavioral Education, 23, 108–131. Loman, S. L., & Horner, R. H. (2014). Examining the
Hall, L. J., & Macvean, M. L. (1997). Increases in the efficacy of a basic functional behavioral assessment
communicative behaviours of students with cerebral training package for school personnel. Journal of Posi-
palsy as a result of feedback to, and the selection of tive Behavior Interventions, 16, 18–30.
goals by, paraprofessionals. Behaviour Change, 14, Long, A. C. J., Hagermoser Sanetti, L. M., Collier-Meek,
174–184. M. A., Gallucci, J., Altschaefl, M., & Kratochwill, T.
Hawkins, S. M., & Heflin, L. J. (2011). Increasing sec- R. (2016). An exploratory investigation of teachers’
ondary teachers’ behavior-specific praise using a intervention planning and perceived implementa-
video self-modeling and visual performance feedback tion barriers. Journal of School Psychology, 55, 1–26.
intervention. Journal of Positive Behavior Interven- Luiselli, J. K., & Diament, C. (Eds.). (2002). Behav-
tions, 1, 97–108. ior psychology in the schools: Innovations in evalua-
Houtenville, A. J., Brucker, D. L., & Lauer, E. A. (2014). tion, support, and consultation. West Hazleton, PA:
Annual compendium of disability statistics: 2014. Dur- Haworth Press.
ham: University of New Hampshire, Institute on Dis- Maag, J. W., & Larson, P. J. (2004). Training a general
ability. education teacher to apply functional assessment.
Individuals with Disabilities Education Improvement Education and Treatment of Children, 27, 26–36.
Act (IDEA) of 2004, Pub. L. No. 108-446, 20 U.S.C. Machalicek, W., O’Reilly, M., Chan, J. M., Rispoli, M.,
1400 et seq. (2004). Lang, R., Davis, T., et al. (2009). Using videocon-
Jenkins, L. N., Floress, M. T., & Reinke, W. (2015). Rates ferencing to support teachers to conduct preference
and types of teacher praise: A review and future di- assessments with students with autism and develop-
rections. Psychology in the Schools, 52, 463–476. mental disabilities. Research in Autism Spectrum Dis-
Joyce, B., & Showers, B. (1995). The evolution of peer orders, 3, 32–41.
coaching. Improving Professional Practice, 53(6), 12– Maras, M. A., Splett, J. W., Reinke, W. M., Stormont,
16. M., & Herman, K. C. (2014). School practitioners’
Joyce, B., & Showers, B. (2002). Student achievement perspectives on planning, implementing, and evalu-
474 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
ating evidence-based practices. Children and Youth Petscher, E. S., & Bailey, J. S. (2006). Effects of training,
Services Review, 47, 314–322. prompting, and self-monitoring on staff behavior in
Moore, J. W., Edwards, R. P., Sterling-Turner, H. E., a classroom for students with disabilities. Journal of
Riley, J., DuBard, M., & McGeorge, A. (2002). Applied Behavior Analysis, 39, 215–226.
Teacher acquisition of functional analysis methodol- Plavnick, J. B., Ferreri, S. J., & Maupin, A. N. (2010).
ogy. Journal of Applied Behavior Analysis, 35, 73–77. The effects of self-monitoring on the procedural in-
Mouzakitis, A., Codding, R. S., & Tryon, G. (2015). The tegrity of a behavioral intervention for young chil-
effects of self-monitoring and performance feedback dren with developmental disabilities. Journal of Ap-
on the intervention integrity of behavior support plied Behavior Analysis, 43, 315–320.
plan implementation. Journal of Positive Behavior In- Putnam, R. F., Handler, M. W., Rey, J., & McCarty, J.
terventions, 17, 223–234. (2005). The development of behaviorally based pub-
Myers, D. M., Simonsen, B., & Sugai, G. (2011). In- lic school consultation services. Behavior Modifica-
creasing teachers’ use of praise with a response-to- tion, 29, 521–537.
intervention approach. Education and Treatment of Reinke, W. M., Stormont, M., Herman, K. C., Puri,
Children, 34, 35–59. R., & Goel, N. (2011). Supporting children’s men-
Newton, S. J., Horner, R. H., Algozzine, B., Todd, A. tal health in schools: Teacher perceptions of needs,
W., & Algozzine, K. (2012). A randomized wait-list roles, and barriers. School Psychology Quarterly, 26,
controlled analysis of the implementation integrity 1–12.
of team-initiated problem solving processes. Journal Rispoli, M., Ninci, J., Burke, M. D., Zaini, S., Hatton, H.,
of School Psychology, 50, 421–441. & Sanchez, L. (2015). Evaluating the accuracy of re-
No Child Left Behind Act of 2001, Pub. L. No. 107-110, sults for teacher implemented trial-based functional
20 U.S.C. 6301 (2002). analyses. Behavior Modification, 39, 627–653.
Noell, G. H., Gansle, K. A., Mevers, J. L., Knox, R. M., Sanetti, L. M. H., Collier-Meek, M. A., Long, A. C. J.,
Mintz, J. C., & Dahir, A. (2014). Improving inter- Byron, J., & Kratochwill, T. R. (2015). Increasing
vention plan implementation in schools: A meta- teacher intervention integrity of behavior support
analysis of single subject design studies. Journal of plans through consultation and implementation
Behavioral Education, 23, 168–191. planning. Journal of School Psychology, 53, 209–229.
Noell, G. H., Witt, J. C., Slider, N. J., Connell, J. E., Sanetti, L. M. H., Collier-Meek, M. A., Long, A. C. J.,
Gatti, S. L., Williams, K. L., et al. (2005). Treatment Kim, J., & Kratochwill, T. R. (2014). Using imple-
implementation following behavioral consultation mentation planning to increase teachers’ adherence
in schools: A comparison of three follow-up studies. and quality to behavior support plans. Psychology in
School Psychology Review, 34, 87–107. the Schools, 51, 879–895.
O’Leary, S. G., & Dubey, D. R. (1979). Applications of Sanetti, L. M. H., Kratochwill, T. R., & Long, A. C. J.
self-control procedures by children: A review. Journal (2013). Applying adult behavior change theory to
of Applied Behavior Analysis, 12(3), 449–465. support mediator-based intervention implementa-
Oliver, R. M., Wehby, J. H., & Nelson, J. R. (2015). tion. School Psychology Quarterly, 28, 47–62.
Helping teachers maintain classroom management Sarokoff, R. A., & Sturmey, P. (2004). The effects of
practices using a self-monitoring checklist. Teaching behavioral skills training on staff implementation
and Teacher Education, 51, 113–120. of discrete-trial teaching. Journal of Applied Behavior
Olweus, D. (1991). Bully/victim problems among school- Analysis, 37, 535–538.
children: Basic facts and effects of a school-based in- Sheridan, S. M., Bovaird, J. A., Glover, T. A., Garbacz,
tervention program. In D. J. Pepler & K. H. Rubin A., Witte, A., & Kwon, K. (2012). A randomized
(Eds.), The development and intervention of childhood trial examining the effects of conjoint behavioral
aggression (pp. 411–448). Hillsdale, NJ: Erlbaum. consultation and the mediating role of the parent-
Parsons, M. B., Rollyson, J. H., & Reid, D. H. (2012). teacher relationship. School Psychology Review, 41,
Evidence-based staff training: A guide for practitio- 23–46.
ners. Behavior Analysis in Practice, 5(2), 2–11. Sheridan, S. M., Clarke, B. L., & Burt, J. D. (2008).
Pelletier, K., McNamara, B., Braga-Kenyon, P., & Conjoint behavioral consultation: What do we know
Ahearn, W. H. (2010). Effect of video self-monitor- and what do we need to know? In W. P. Erchul &
ing on procedural integrity. Behavioral Interventions, S. M Sheridan (Eds.), Handbook of research in school
25, 261–274. consultation (pp. 171–202). Mahwah, NJ: Erlbaum.
Pence, S. T., St. Peter, C. C., & Giles, A. F. (2014). Sheridan, S. M., & Kratochwill, T. R. (2008). Con-
Teacher acquisition of functional analysis methods joint behavioral consultation: Promoting family–school
using pyramidal training. Journal of Behavioral Edu- connections and interventions (2nd ed.). New York:
cation, 23, 132–149. Springer.
Pence, S. T., St. Peter, C. C., & Tetreault, A. S. (2012). Sheridan, S. M., Welch, M., & Orme, S. F. (1996). Is
Increasing accurate preference assessment implemen- consultation effective?: A review of outcome re-
tation through pyramidal training. Journal of Applied search. Remedial and Special Education, 17, 341–354.
Behavior Analysis, 45, 345–359. Slider, N. J., Noell, G. H., & Williams, K. L. (2006). Pro-
Teacher Consultation 475
viding practicing teachers classroom management outcomes: A review of the literature. Journal of Posi-
professional development in a brief study format. tive Behavior Interventions, 17, 69–82.
Journal of Behavioral Education, 15, 215–228. Suess, A. N., Wacker, D. P., Schwartz, J. E., Lustig, N., &
Solomon, B. G., Klein, S. K., & Politylo, B. C. (2012). Detrick, J. (2016). Preliminary evidence on the use of
The effect of performance feedback on teachers’ in- telehealth in an outpatient behavior clinic. Journal of
tervention integrity: A meta-analysis of single-case Applied Behavior Analysis, 49, 686–692.
literature. School Psychology Review, 41, 160–175. Wallace, M. D., Doney, J. K., Mintz-Resudek, C. M., &
Stormont, M., Reinke, W., & Herman, K. (2011a). Tarbox, R. S. F. (2004). Training educators to imple-
Teachers’ characteristics and ratings for evidence- ment functional analyses. Journal of Applied Behavior
based behavioral interventions. Behavioral Disorders, Analysis, 37, 89–92.
37, 19–29. Watson, T. S., Ray, K. P., Sterling-Turner, H., & Logan,
Stormont, M., Reinke, W., & Herman, K. (2011b). P. (1999). Teacher-implemented functional analysis
Teachers’ knowledge of evidence-based interventions and intervention: A method for linking assessment
and available school resources for children with emo- to intervention. School Psychology Review, 28, 292–
tional and behavioral problems. Journal of Behavioral 302.
Education, 20, 138–147. Witt, J. C., & Martens, B. K. (1988). Problems with
Stormont, M., Reinke, W. M., Newcomer, L., Marchese, problem-solving consultation: A re-analysis of as-
D., & Lewis, C. (2015). Coaching teachers’ use of sumptions, methods, and goals. School Psychology
social behavior interventions to improve children’s Review, 17, 211–226.
CHAP TER 28
Children may encounter numerous threats to per- on safety skills for low-incidence but life-threaten-
sonal safety in their lives. Safety experts divide ing safety threats. In particular, we focus on teach-
these threats into two categories: (1) frequently ing appropriate safety skills in the presence of an
occurring situations in which a child has repeat- unattended firearm (Maxfield, Miltenberger, &
ed opportunities to engage in safe behavior, and Novotny, 2019). We review the problem of firearm
(2) low-incidence but life-threatening situations in injuries and deaths, and discuss two approaches to
which the child may have only one opportunity to prevention of these: modifying parent behavior to
engage in safe behavior successfully (Miltenberger, promote safe storage of firearms (Violano et al.,
2008). Examples of frequently occurring situations 2018), and teaching children appropriate safety
that require safe behavior include riding in an au- skills in the presence of an unattended firearm,
tomobile, riding a bike, and crossing the street. Ex- with an emphasis on behavioral skills training.
amples of safe behavior in these situations include
using a seatbelt, wearing a helmet, and looking
both ways before crossing, respectively. Examples FIREARM INJURIES AND DEATHS
of low-incidence but life-threatening situations
Prevalence of Firearm Injuries and Deaths
include attempted abduction, home fires (e.g.,
Garcia, Dukes, Brady, Scott, & Wilson, 2016), Firearms injure or kill hundreds of children in
and finding an unattended firearm. Examples of the United States (Parikh, Silver, Patel, Iqbal, &
safe behavior in these situations include saying, Goyal, 2017). Between 2009 and 2013, an average
“No,” leaving the area, and telling an adult; im- of 3,061 children ages 0–19 years survived unin-
mediately evacuating the house; and not touching tentional firearm injuries, and 951 children ages
the firearm, leaving the area, and telling an adult, 0–19 years died from suicide or unintentional
respectively. A child’s use of safe behavior in such shootings (Brady Campaign to Prevent Gun Vi-
situations could save his or her life. Many safety olence, n.d.-a). Childhood firearm injuries and
threats require a child to emit multiple responses deaths are often not deliberate. The American
to maintain his or her safety, and we refer to those Academy of Pediatrics (AAP, 2000) reported that
responses as safety skills. the percentages of unintentional deaths caused by
Although use of safety skills during frequently firearms for children under age 5, ages 5–9, ages
occurring safety threats is important for preventing 10–14, and ages 15–19 were 24%, 26%, 21%, and
accidental injury or death, we focus in this chapter 5%, respectively. A firearm accidentally killed over
476
Teaching Safety Skills to Children 477
1,500 children between 1996 and 2001 (Common children are unsafe storage practices (i.e., the
Sense about Kids and Guns, n.d.). Only vehicular owner stores the firearm unlocked, loaded, near
accidents caused more unintentional deaths than ammunition) and children’s tendencies to play
firearms did (Kellerman, 1993; Zavoski, Lapidus, with firearms they find (Himle & Miltenberger,
Lerer, & Banco, 1995). 2004). Several investigators have documented
Handguns, which are firearms designed to be that many firearm owners fail to store their fire-
held and fired with one hand, cause more unin- arms safely (e.g., Azreal, Cohen, Salhi, & Miller,
tentional firearm injuries and deaths than do 2018; Azrael, Miller, & Hemenway, 2000; Crifasi,
shotguns and rifles, which are firearms designed Doucette, McGinty, Webster, & Barry, 2018),
to be held and fired with both hands (AAP, 2000; and that children often play with firearms when
Grossman, Reay, & Baker, 1999; Knight-Bohnhoff they find them (Hardy, 2002; Hardy, Armstrong,
& Harris, 1998; Zavoski et al., 1995). Easy access Martin, & Strawn, 1996; Jackman, Farah, Keller-
to firearms increases the risk of accidental firearm man, & Simon, 2001).
injuries and fatalities (Hemenway & Solnick, 2015;
Miller, Azrael, & Hemenway, 2001; Miller, Azrael,
Hemenway, & Vriniotis, 2005; Ordog et al., 1988). Parental Behavior and Beliefs about Firearms
In fact, most incidents occur in the homes of the Firearm Storage Behavior
victims or of friends or family members of the vic-
tims (Common Sense about Kids and Guns, n.d.; Many parents do not store their firearms safely,
DiScala & Sege, 2004; Eber, Annest, Mercy, & even though access to firearms is so closely associ-
Ryan, 2004; Grossman et al., 1999; Kellerman & ated with childhood firearm injuries and deaths.
Reay, 1986; Wintemute, Teret, Kraus, Wright, & Fewer than half of parents reported storing their
Bradfield, 1987). firearms in the safest manner—that is, locked, un-
loaded, and separate from ammunition (Farah &
Simon, 1999; Stennies, Ikeda, Leadbetter, Hous-
Prevalence of Gun Ownership ton, & Sacks, 1999; Wiley & Casey, 1993). More
Reported rates of firearm ownership vary, with importantly, 13–30% of parents reported storing
20–40% of households owning at least one fire- their firearms both unlocked and loaded, which is
arm (Common Sense about Kids and Guns, n.d.; the most unsafe storage practice (Common Sense
Haught, Grossman, & Connell, 1995; National about Kids and Guns, n.d.; Farah & Simon, 1999;
Opinion Research Center, 2014; Schuster, Franke, Hemenway, Solnick, & Azrael, 1995; Miller et al.,
Bastian, Sor, & Halfon, 2000; Senturia, Christof- 2005; Schuster et al., 2000; Senturia et al., 1994;
fel, & Donovan, 1994). Owning one firearm in- Stennies et al., 1999). Many firearm owners re-
creases the likelihood of owning another firearm ported storing their firearms in a manner between
fivefold (Senturia et al., 1994), and households these two extremes (Stennies et al., 1999). Nearly
containing a male member are more likely to have half of firearm owners reported storing ammuni-
a firearm than those without a male (Knight- tion separately from firearms (Haught et al., 1995;
Bohnhoff & Harris, 1998; Schuster et al., 2000). Hendricks & Reichert, 1996).
More than half of firearm-owning households own Investigators have shown that childproof safety
a handgun. Parents are more likely to have a rifle, devices for firearms prevent firearm injuries and
a shotgun, and a handgun, respectively (Schus- deaths. The General Accounting Office (1991) re-
ter et al., 2000). Handgun owners cite protection viewed medical examiners’ and coroners’ reports
most often as the reason for keeping the firearm in and concluded that childproof safety devices and
the home (AAP, 2000; Dresang, 2001; Haught et loading indicators could have prevented about 8%
al., 1995; Wiley & Casey, 1993). However, a fire- and 23% of unintentional injuries and deaths, re-
arm in the home is far more likely to kill a family spectively. Vernick et al. (2003) determined that
member or friend than to protect the family from one or more of three safety devices—personalized
a stranger (AAP, 2000; Dresang, 2001; Kellerman, firearms (Crifasi, O’Dwyer, McGinty, Webster,
1993; Kellerman & Reay, 1986). & Barry, 2019), loaded chamber indicators, and
magazine safeties—could have prevented 45% of
unintentional deaths. Unfortunately, fewer than
Risk Factors for Firearm Injuries and Deaths half of firearm-owning parents use such safety de-
Two important risk factors associated with un- vices (Common Sense about Kids and Guns, n.d.;
intentional firearm injuries to and deaths of Haught et al., 1995; Schuster et al., 2000).
478 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Firearm Safety Beliefs were also more likely to think that their own chil-
dren could make this judgment reliably (Farah &
Results of surveys have shown that firearm-own-
Simon, 1999; Webster et al., 1992).
ing and non-firearm-owning parents differ in
Many parents, regardless of firearm ownership
their firearm safety beliefs. Knight-Bohnhoff and
status, were confident that their children would
Harris (1998) interviewed firearm-owning and
not touch or play with a firearm if the opportunity
non-firearm-owning parents about their knowl-
arose (Common Sense about Kids and Guns, n.d.;
edge and beliefs regarding unintentional firearm
Farah & Simon, 1999). Parents reported talking
injuries and deaths. Firearm-owning parents re-
to their children about firearm safety (Farah &
ported that education-based safety training was Simon, 1999; Knight-Bohnhoff & Harris, 1998),
a sufficient strategy for avoiding unintentional but most reported not discussing firearm owner-
firearm injuries and deaths (Knight-Bohnhoff & ship with parents of their children’s friends (Brady
Harris, 1998), whereas non-firearm-owning par- Campaign to Prevent Gun Violence, n.d.-b).
ents were more likely to believe that keeping fire- Of those who had not discussed the issue with
arms out of the house was the best way to avoid the friends’ parents, most said that they had not
such injuries and deaths (Farah & Simon, 1999). thought about it; some assumed that there were no
Firearm-owning parents preferred to obtain safety firearms in the households of the friends’ parents,
information from a firearm organization, whereas or that the friends’ parents stored their firearms
non-firearm-owning parents preferred to obtain safely (Brady Campaign to Prevent Gun Violence,
safety information from the police. Firearm-own- n.d.-b).
ing parents stated that they (1) would be willing to
talk to their pediatrician about firearms, and (2)
believed statistics regarding the greater risks than Child Behavior
benefits of owning firearms. Nevertheless, these Although parents tend to believe that their chil-
parents did not remove firearms from their house- dren will not touch or play with a firearm, this is
hold, although most reported that they would fol- often not the case. Hardy et al. (1996) found that
low advice to keep firearms locked and unloaded 19% of children whose parents owned a firearm
(Webster, Wilson, Duggan, & Pakula, 1992). reported playing with it without their parents’
Non-firearm-owning parents stated that informa- knowledge. In addition, 24% more children than
tion from their pediatrician would make them less parents verified that the children were aware that
likely to buy a firearm in the future. their parents kept a firearm in the house, and 67%
Webster et al. (1992) surveyed parents in a pedi- of children stated that they knew where the fire-
atrician’s office and found that firearm owners were arm was located and had access to it (Hardy et al.,
more willing than nonowners to trust children 1996). Hardy (2002) asked parents and children
with firearms at a much younger age. Fourteen per- similar questions about firearm safety. When in-
cent of firearm owners were willing to trust a child vestigators asked parents if their children would
under 12 years of age with a firearm, and 39% were play with a firearm, 41% said yes, 32% said no, and
willing to trust a child between 12 and 15 years 27% were unsure. Interestingly, 40% of children
of age with a firearm. Twenty-six percent of fire- whose parents answered no had played with a fire-
arm owners said that they would never trust a child arm, as did 59% of children of the unsure parents.
with a firearm. By contrast, only 3% of nonowners Furthermore, when investigators placed children
were willing to trust a child under 12 years of age in a room with various toys, toy firearms, and real
with a firearm, and 9% were willing to trust a child but disabled firearms, the children often touched
between 12 and 15 years of age with a firearm. For- and played with the real firearms (Hardy, 2002;
ty-two percent of nonowners said that they would Jackman et al., 2001).
never trust a child with a firearm (Webster et al., Jackman et al. (2001) sent pairs or trios of 8- to
1992). Firearm-owning and non-firearm-owning 12-year-old boys into a room in which the investi-
parents combined said that they would trust their gators had placed a firearm in a drawer. Jackman et
own children with a firearm at an average age of 9, al. found that at least one child handled the fire-
but that they would not trust other children until arm in 76% of the groups, and at least one child
age 21 (Farah & Simon, 1999). Firearm-owning pulled the trigger in 48% of the groups. Only once
parents were more likely than non-firearm-owning (5%) did a child leave the room and report finding
parents to believe that children could discriminate the firearm to an adult. Nearly all children who
between real and toy firearms at an earlier age, and touched the firearm (93%) and pulled the trigger
Teaching Safety Skills to Children 479
(94%) had received firearm safety information at Webster and Starnes (2000) also evaluated
some point. Hardy (2002) obtained similar results: the effectiveness of the Child Access Prevention
53% of children played with the firearm in the Law in decreasing firearm deaths. They found no
room, and only 1 child out of 70 reported find- change in firearm death rates in states where the
ing the firearm to an adult. Furthermore, children penalty was a misdemeanor, but a decrease in such
were not able to discriminate between real and toy deaths in states where the penalty was a felony.
firearms as well as parents believed. In fact, about However, when they excluded Florida’s data from
half the children who found the firearm were not the analysis, the decrease in firearm death rates
sure whether it was real or a toy (Hardy, 2002; after the implementation of the Child Access Pre-
Jackman et al., 2001). Children were more likely to vention Law was not statistically significant. Flor-
identify a firearm as a toy when it was real than to ida may have been so successful because it was the
identify it as real when it was a toy (Hardy, 2002). first state to implement the law; therefore, the law
received much publicity. Florida also had the most
severe penalty, and the death rate before imple-
PREVENTION OF CHILDHOOD FIREARM INJURIES mentation was quite high, so there was more room
AND DEATHS for change. Researchers continue to hypothesize
about reasons for the success in Florida but not in
The two major risk factors for childhood firearm other states (Webster & Starnes, 2000).
injuries and deaths are accessible firearms and
children’s tendencies to play with firearms they
Parent Education
find. Therefore, two approaches to preventing fire-
arm injuries are (1) to promote safe storage of fire- Grossman, Mang, and Rivara (1995) evaluated
arms by parents and (2) to assess and teach child family physicians’ and pediatricians’ beliefs and
safety skills. practices regarding firearm safety and safe storage
practices. They found that many family physicians
and pediatricians agreed that they should be re-
Promoting Safe Storage Practices
sponsible for counseling families about firearms,
Investigators and community members have used yet it was low on their list of priorities. About
various strategies to promote safe storage of fire- one-third stated that they did not know what to
arms. These strategies have included legislation tell parents, and one-half admitted that they had
making storing a firearm where it is accessible to never counseled a family on firearm safety. Family
children a felony, as well as efforts by pediatricians, physicians were more likely to promote teaching
physicians, the mass media, and others to educate children safe firearm use when the children were
parents about the dangers of firearms and the need “old enough,” and pediatricians were more likely
for safe storage (e.g., Himle & Miltenberger, 2004; to agree that individuals should not keep firearms
Jostad & Miltenberger, 2004). in homes with children. Pediatricians were more
willing to suggest removal of firearms to parents,
yet both groups doubted that families would fol-
Gun Legislation
low this advice. Both family physicians and pedia-
Florida was the first state to implement a law that tricians said that they would tell parents to store
punished those who stored or left a loaded firearm ammunition and firearms separately, and thought
where a child could access it. The Child Access parents would be receptive to this suggestion
Prevention Law had positive results in Florida dur- (Grossman et al., 1995).
ing its first year of implementation, with a 50% Grossman et al. (2000) evaluated the effec-
drop in firearm deaths. As of 1997, 19 other states tiveness of physician education, written materi-
had adopted similar laws (Cummings, Grossman, als about safe storage, and a discount coupon for
Rivara, & Koepsell, 1997). Cummings et al. (1997) purchasing safe storage devices on parents’ firearm
evaluated whether these laws were effective in de- ownership and storage behavior. The program did
creasing the death rate of children under 15 years not produce any significant changes in ownership
of age. Overall, the death rate was 23% lower than or storage behavior compared to that of controls.
expected in states that adopted safe storage laws. Sidman et al. (2005) evaluated the effects of a
The study also found that the decrease in deaths media campaign and discount coupons for lock
was greater for children under 10 years of age than boxes on the storage practices of firearm owners in
for those between 10 and 14 years of age. King County, Washington. The authors conduct-
480 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
ed telephone surveys to assess storage practices in get behaviors are the same across assessment and
the intervention county and nine control counties teaching procedures: “Stop what you are doing, do
in different states, and found that the intervention not touch the firearm, leave the area, and tell an
did not produce statistically significant changes in adult.” Most investigators have used a 0–3 rating
safe storage practices. scale (0 = touched the firearm, regardless of complet-
Coyne-Beasley, Schoenbach, and Johnson ing further steps; 1 = did not touch the firearm; 2
(2001) evaluated the effectiveness of the Love Our = did not touch the firearm and immediately left the
Kids, Lock Your Guns program on participants’ room; and 3 = did not touch the firearm, immediately
safe storage practices. The investigators gave a sur- left the room, and told an adult that he or she found
vey, individualized counseling, safety information, the firearm). Across assessment methods, a teacher
and a firearm lock with instructions at no expense (or other evaluator) does not provide feedback to
to individuals in the parking lot of a mall. Coyne- a child about his or her responses during the as-
Beasley et al. evaluated the effectiveness of the sessment, as the purpose is to observe and measure
program after 6 months. Almost all participants what the child does either before or after teaching.
thought the program was helpful. At follow-up,
77% of participants said that they locked their fire- Self-Report. Self-report is a method in which
arms, compared to 48% at baseline; 72% said that a child states what he or she would do when en-
they used the firearm lock, compared to none at countering an unattended firearm. The teacher
baseline; and only 7% stored their firearms loaded describes a situation in which the child encoun-
and unlocked, compared to 18% at baseline. After ters an unattended firearm (e.g., “Imagine that you
the program, participants who had children were are at your friend’s house playing in the basement,
more likely than those without children to lock while your friend’s parents are upstairs. You are
their firearms. playing hide and seek, and when you open a closet
Although the research by Coyne-Beasley et al. to hide in, you see a firearm in the closet. What
(2001) illustrates a program that was successful in would you do?”). The child then responds by tell-
promoting safe storage of firearms, the investiga- ing the teacher what he or she would do (e.g., “I
tors made personal contact with individual firearm would stop playing and not touch the firearm;
owners and gave away firearm locks; this is a time- then I would go upstairs and tell my friend’s par-
and resource-intensive practice that communities ents right away”).
are not likely to apply on a wide scale. In general, Although self-report assessments are time-
research on changing firearm storage practices efficient relative to assessments in which a child
has produced mixed results, with many programs performs the behavior, their limitations outweigh
producing no changes. Furthermore, this research their benefits. That is, a child may describe the
used self-reports of storage practices, so we need safety skill vocally, but may not demonstrate the
to interpret the results cautiously. McGee, Coyne- skills when he or she encounters a firearm (e.g.,
Beasley, and Johnson (2003) concluded that Gatheridge et al., 2004; Himle, Miltenberger,
what interventions or intervention components Gatheridge, & Flessner, 2004). Given the poten-
increase the likelihood of safe firearm storage is tial of self-reports to produce false positives in
not clear. Clearly, the field needs more research to which the child states that he or she would per-
evaluate strategies for promoting safe storage prac- form the correct skills but does not perform them
tices by firearm owners. Safe firearm storage would in the actual situation, we do not recommend self-
decrease the threat to child safety and the need to report to evaluate firearm safety skills.
teach children firearm safety skills.
Role Play. During role-play assessments, the
teacher presents the child with a scenario and asks
Assessing and Teaching Child Safety Skills him or her to demonstrate what he or she would
in that situation. For example, the teacher might
Assessing Child Safety Skills
place a disabled firearm or a toy firearm on top of a
Given that methods for promoting safe firearm desk and say to the child, “Pretend that you walked
storage have not been successful, an alternative ap- into your parents’ office and found a firearm on
proach is to teach firearm safety skills to children. top of their desk. Now pretend that I’m your dad,
Investigators have used three methods to evaluate and show me what you would do.” The teacher
whether teaching approaches are effective: self- then would walk to another room and wait for the
report, role play, and in situ assessments. The tar- child to demonstrate what he or she would do.
Teaching Safety Skills to Children 481
Researchers consider role-play assessments supe- formance and gives corrective feedback for incor-
rior to self-reports, as the child must demonstrate rect performance (Gatheridge et al., 2004; Himle,
what he or she would do in the presence of a fire- Miltenberger, Flessner, & Gatheridge, 2004; Milt-
arm. However, research shows that the behavior enberger et al., 2004).
the child demonstrates during role play may not
match what he or she does when he or she encoun- Informational Training. Hardy et al. (1996) eval-
ters a firearm (e.g., Gatheridge et al., 2004; Himle, uated an informational-training approach to teach
Miltenberger, Gatheridge, et al., 2004). We do not firearm safety skills by surreptitiously videotaping
recommend role-play assessments, given the dan- children as they found a firearm in a playroom be-
ger of a child’s finding an unattended firearm, and fore and after receiving firearm safety information.
the false positives associated with role-play assess- The pre- and postteaching assessments consisted
ments. of two children playing in a room that contained a
disabled firearm, a toy firearm, and other toys. The
In Situ Assessments. In situ assessments repre- intervention was a 30-minute session in which a
sent the most valid method for evaluating safety police officer told children and parents that a
skills (Miltenberger, Sanchez, & Valbuena, 2015). child should never touch a firearm without per-
Three essential features must be present during an mission from parents; a child should tell an adult if
in situ assessment of firearm safety skills. First, the he or she finds a firearm or if another child is play-
child must encounter an unattended firearm in ing with a firearm; and all firearms are dangerous
the natural environment. Second, the child must unless an adult says otherwise. Hardy et al. found
not know that someone is observing him or her. that children in the training group were just as
Third, the child must not be in the presence of likely to touch and play with the real firearm after
an adult when he or she finds the firearm. Impor- training as children in the control group were.
tantly, investigators and clinicians should use a Hardy (2002) reported similar results.
disabled firearm or realistic replica to ensure the The Eddie Eagle GunSafe Program is a com-
child’s safety. In situ assessments simulate a real- monly used training program for children, distrib-
life situation and capture whether child behavior uted by the National Rifle Association (Himle,
is under the stimulus control of the firearm rather Miltenberger, Gatheridge, et al., 2004). This pro-
than the observer. We recommend that investiga- gram uses various materials and activities (such
tors and clinicians use in situ assessments of fire- as posters, coloring books, videos, cutting materi-
arm safety skills, because of the shortcomings of als, sequencing cards with the safety motto, and
self-report and role-play assessments as described certificates and stickers to use as rewards) to pro-
above (e.g., Himle, Miltenberger, Gatheridge, et vide information about firearm safety skills. Dur-
al., 2004; Gatheridge et al., 2004). ing training, children receive information, recite
the safety motto, and verbally respond with the
safety motto to “what if” scenarios during five
Teaching Safety Skills to Children
15-minute training sessions. We consider this an
Investigators have used two approaches to teach informational approach because the child does
firearm safety skills to children. One is informa- not actually perform the target behaviors (stop, do
tional, and the other is active learning. In an in- not touch, leave the area, and tell an adult) in a
formational approach, the teacher talks about the situation in which he or she finds an unattended
dangers of firearms and describes the safety skills firearm. The Eddie Eagle GunSafe Program has
associated with firearms, often with the child vo- trained well over 15 million children, yet there are
cally rehearsing the safety skills. Live or video few studies evaluating its effectiveness.
modeling is often part of the informational ap- Himle, Miltenberger, Gatheridge, et al. (2004)
proach. The safety skills associated with a firearm found that 4- and 5-year-old children trained with
are not to touch it, to get away from it, and to re- the Eddie Eagle GunSafe Program could tell the
port its presence to an adult (e.g., Gatheridge et investigator what they were supposed to do when
al., 2004; Himle, Miltenberger, Gatheridge, et al., they found a firearm during a self-report assess-
2004; Kelso, Miltenberger, Waters, Egemo-Helm, ment. However, they did not perform better than
& Bagne, 2007). In an active-learning approach, untrained children when the investigator asked
the teacher provides instructions and modeling, them to role-play what they would do if they found
and the child practices the safety skills in simu- an unattended firearm. Furthermore, the trained
lated situations. The teacher praises correct per- children did not perform the correct safety skills
482 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
during in situ assessments. Gatheridge et al. (2004) child with a range of simulated safety threats (e.g.,
also evaluated the Eddie Eagle GunSafe Program’s Himle & Miltenberger, 2004; Miltenberger, 2008).
effectiveness with 6- and 7-year-old children. As a result, the child should engage in the safety
These children performed well on the self-report skills when faced with a real-life situation involv-
assessment and did slightly better on role-play ing a similar safety threat.
assessments than children in the control group. Research evaluating behavioral skills training
However, trained children did not perform the for teaching safety skills to children and adults
safety skills during in situ assessments. with intellectual disabilities has produced many
Research has shown that the informational ap- findings. First, behavioral skills training is effec-
proaches reviewed above may not be effective for tive in producing skill acquisition, although the
teaching children firearm safety skills. This find- skills are not always generalized to other settings
ing is consistent with research showing that infor- or maintained over time (Marchand-Martella &
mational approaches are not effective for teaching Huber, 1996; Miltenberger et al., 2004; Poche et
safety skills for other safety threats, such as abduc- al., 1981). Second, behavioral skills training ap-
tion and sexual abuse (e.g., Beck & Miltenberger, pears to be more effective with individuals rather
2009; Miltenberger et al., 2013; Miltenberger & than groups of children (Carroll-Rowan & Milt-
Hanratty, 2013; Poche, Yoder, & Miltenberger, enberger, 1994; Johnson et al., 2005; Miltenberger
1988). & Olsen, 1996). Third, behavioral skills training
is not effective in some instances until investiga-
Behavioral Skills and In Situ Training. The inef- tors add in situ training; that is, children learn the
fectiveness of informational approaches for teach- skills with behavioral skills training, but do not use
ing safety skills should not be surprising, because them until training occurs in the natural setting
they do not require children to perform the safety (Johnson et al., 2005; Miltenberger et al., 1999).
skill correctly during training (Gatheridge et al., Fourth, behavioral skills training is more effective
2004; Hardy, 2000; Hardy et al., 1996; Himle, Milt- than informational approaches that do not involve
enberger, Gatheridge, et al., 2004). On the other performing the behavior in the context in which it
hand, behavioral skills training is an approach may occur (Gatheridge et al., 2004; Himle, Milt-
that requires active child rehearsal until the child enberger, Gatheridge, et al., 2004). Finally, behav-
masters the skills across a range of simulated situ- ioral skills training can be time-intensive, because
ations. Research has shown that behavioral skills some children require numerous sessions before
training is effective for training abduction preven- acquiring and generalizing the skills (Johnson et
tion skills (Carroll-Rowan & Miltenberger, 1994; al., 2005; Miltenberger et al., 1999).
Johnson et al., 2005, 2006; Marchand-Martella & Researchers have used in situ training with be-
Huber, 1996; Olsen-Woods, Miltenberger, & For- havioral skills training in situations where behav-
man, 1998; Poche, Brouwer, & Swearingen, 1981; ioral skills training alone has not been effective at
Poche et al., 1988), sexual abuse prevention skills teaching firearm safety skills. In situ training has
(Lumley, Miltenberger, Long, Rapp, & Roberts, the same components as behavioral skills training
1998; Miltenberger et al., 1999; Miltenberger & (i.e., instructions, modeling, rehearsal, and feed-
Thiesse-Duffy, 1988; Miltenberger, Thiesse-Duffy, back), but the investigator or clinician conducts
Suda, Kozak, & Bruellman, 1990), pedestrian safe- the training immediately after a failed in situ as-
ty skills (Yeaton & Bailey, 1978), and fire safety sessment. The moment the child fails to perform
skills (Jones, Kazdin, & Haney, 1981; Jones, Ollen- the safety skill, the investigator enters the area,
dick, McLaughlin, & Williams, 1989). terminates the assessment, and begins the in situ
Behavioral skills training involves instruc- training. The investigator discusses the safety
tions, modeling, rehearsal, and feedback. The threat, reviews the correct and incorrect aspects
teacher gives instructions describing the safety of the performance, and asks the child to rehearse
threat and the safety skills to use in response to the skills while providing additional feedback (e.g.,
the threat. Next, the teacher models the safety Miltenberger et al., 2004, 2005, 2009, 2013).
skills in simulated situations of the safety threat. Miltenberger and colleagues have evaluated be-
Then the child rehearses the safety skills during havioral skills training for teaching firearm safety
a role play. The teacher provides praise for correct skills to children, extending the existing literature
performance and corrective feedback for incorrect on behavioral skills training in threat situations
performance. Rehearsal and feedback continue such as abduction and sexual abuse (Gatheridge
until the child performs the safety skills correctly et al., 2004; Himle, Miltenberger, Flessner, et
and immediately when the teacher presents the al., 2004; Himle, Miltenberger, Gatheridge, et
Teaching Safety Skills to Children 483
al., 2004; Miltenberger et al., 2004, 2005, 2009). firearm safety skills from training sessions to in situ
Himle, Miltenberger, Gatheridge, et al. (2004) assessments with 4- and 5- year-olds. The investi-
compared the firearm safety skills of 4- and 5-year- gators trained eight children individually in two
olds assigned to behavioral skills training, the 30-minute behavioral skills training sessions. In
Eddie Eagle GunSafe Program, or a control group. each session, the trainer gave the child instructions
The investigators kept training time consistent and modeled the safety skills. The child rehearsed
across the two types of training by conducting the skills in various scenarios in which he or she
training during five brief sessions. Investigators found a firearm. The trainer provided praise and
conducted the Eddie Eagle program with small corrective feedback until the child performed the
groups, consistent with the instructions in the skills correctly five consecutive times. Investigators
Eddie Eagle training materials. Similarly, investi- then assessed safety skills during an in situ assess-
gators provided instructions and modeling to small ment. Investigators conducted up to three booster
groups of children assigned to the behavioral skills training sessions if a child did not perform the skills
training group. Children in the behavioral skills correctly during the in situ assessment. The booster
training group rehearsed the skills and received sessions were like the initial training sessions. In-
individual feedback in scenarios in which they vestigators implemented in situ training if a child
found a real but disabled firearm in the home to still did not engage in the correct safety skills after
promote generalization. After training, the in- the third booster session. Immediately after the
vestigators conducted self-report, role-play, and in child did not engage in the correct firearm safety
situ assessments and found that children in both skills during an in situ assessment, the investigator
training groups scored significantly higher than entered the assessment situation, vocally repeated
children in the control group on the self-report and modeled the safety skills, and prompted the
measure. That is, children in both training groups child to rehearse the safety skill until he or she per-
could say what to do when they found a firearm. formed it correctly five consecutive times. No chil-
Children in the behavioral skills training group dren performed the correct safety skills during base-
scored significantly higher than children in the line. Three children performed the correct skills
Eddie Eagle and control groups on the role-play after behavioral skills training and booster sessions.
assessment. In fact, all children in the behavioral The other five children performed the correct skills
skills training group did not touch the firearm, left only after the addition of in situ training. Follow-up
the room, and reported finding the firearm to an assessments occurred in each child’s home between
adult. Finally, the skills did not generalize to in situ 2 weeks and 2 months after training. All children
assessments for children in any group; the in situ performed the correct skills during follow-up, ex-
assessment scores were similar for all three groups. cept for one child who did not report the firearm to
This study highlights the importance of multi- an adult. This child exhibited the correct skills dur-
ple assessment procedures to examine the breadth ing a subsequent assessment (Himle, Miltenberger,
of skill acquisition and generalization (Himle, Flessner, et al., 2004).
Miltenberger, Gatheridge, et al., 2004). The re- In a similar study, Miltenberger et al. (2004)
sults also suggest that the failure of children in the evaluated individual behavioral skills training
training groups to perform the skills either in the with in situ training as needed for teaching safety
role-play assessments (the Eddie Eagle group) or skills to six children ages 6 and 7 years. Three of
the in situ assessments (both training groups) may the children performed the correct skills during
have reflected a performance rather than a skill in situ assessments after two to four behavioral
deficit. That is, children in both training groups skills training sessions, but the other three chil-
could self-report the correct safety skills, and chil- dren required in situ training before exhibiting the
dren in the behavioral skills training group could skills consistently during in situ assessments. All
role-play the correct safety skills. By contrast, chil- children generalized the skills to their homes and
dren in the Eddie Eagle group did not perform the maintained them 5 months after training. The re-
safety skills during either role-play or in situ assess- sults of Himle, Miltenberger, Flessner, et al. (2004)
ments, and children in the behavioral skills train- and Miltenberger et al. (2004) demonstrated the
ing group did not perform the skills during the in importance of in situ training for teaching safety
situ assessment. Other studies of safety skills train- skills to children: In situ training produced cor-
ing have reported similar findings (Lumley et al., rect performance for all children when behavioral
1998; Miltenberger et al., 1999). skills training alone was not effective.
Himle, Miltenberger, Flessner, et al. (2004) Gatheridge et al. (2004) compared the effec-
evaluated procedures to promote generalization of tiveness of the Eddie Eagle GunSafe Program and
484 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
behavioral skills training implemented in small children performed the skills after receiving two
groups of 6- and 7-year-olds. In addition, they eval- behavioral skills training sessions followed by an
uated in situ training with children who did not in situ assessment, and the skills were generalized
exhibit the skills after the Eddie Eagle GunSafe and maintained over a 3-month follow-up period.
Program or behavioral skills training. The results Furthermore, 5 of the children participated in a
showed that children in the Eddie Eagle and be- dyad in situ assessment in which they found the
havioral skills training groups had higher safety firearm while accompanied by a peer, and all five
skills scores than children in the control group on engaged in the correct safety skills. These results
a self-report assessment; that is, children in both suggest that the inclusion of in situ training earlier
training groups could tell the investigator what to in training may improve the effectiveness and ef-
do upon finding a firearm. Children in both train- ficiency of behavioral skills training (Miltenberger
ing groups also performed significantly better than et al., 2005). An alternative explanation is that in
those in the control group during role-play assess- situ training alone would have been effective.
ments, in which the investigator described a sce- Although no study has evaluated in situ training
nario where a child finds an unattended firearm alone to teach firearm safety skills, Miltenberger
and told the child to act out what he or she would et al.’s (2009) results may provide an indication
do in that scenario. Children in the behavioral of the effects of in situ training alone. In this re-
skills training group performed significantly bet- search, one group received behavioral skills train-
ter than children in the Eddie Eagle group dur- ing; one group received simulated in situ training
ing role-play assessments, because all children in with the same components as behavioral skills
the behavioral skills training group engaged in training, plus discussion of real-life scenarios dur-
the correct safety skills. These findings were con- ing instruction, video clips of children encounter-
sistent with those of Himle, Miltenberger, Gath- ing firearms alone and with a peer challenge, and
eridge, et al. (2004) for 4- and 5-year-olds in the rehearsal with a peer challenge; and one group
Eddie Eagle group relative to the behavioral skills received no training. In the first posttraining as-
training groups. Children in the behavioral skills sessment, investigators found no differences in the
training group in the Gatheridge et al. study also scores among the three groups, suggesting that
performed significantly better than children in the neither training procedure was effective. The in-
Eddie Eagle and control groups during in situ as- vestigators then conducted one in situ training ses-
sessments. Most children in the behavioral skills sion for those children in the three groups who did
training group in Gatheridge et al. performed the not engage in the correct skills. The children in all
correct safety skills during in situ assessments. By three groups demonstrated the safety skills, with
contrast, the 4- and 5-year-olds in Himle, Milten- no significant differences between groups after
berger, Gatheridge et al. did not perform the cor- in situ training. These results suggest that in situ
rect safety skills during in situ assessments. Almost training alone may be sufficient for teaching safety
all children in the Eddie Eagle group in Gather- skills, as the children in the control group acquired
idge et al. did not demonstrate correct safety skills the skills after in situ training alone. Miltenberger
until they participated in in situ training, whereas et al. (2013) obtained similar results, in that in situ
most children in the behavioral skills training training alone was effective for teaching abduc-
group demonstrated correct safety skills without in tion prevention skills to children who were part
situ training. of a no-treatment control group. Although these
Research results suggest that children often do results are promising, the field needs more research
not demonstrate correct safety skills during in situ to evaluate in situ training as a stand-alone inter-
assessment after behavioral skills training, but do vention.
demonstrate generalization of safety skills after the
addition of in situ training (Himle, Miltenberger,
Flessner, et al., 2004; Miltenberger et al., 2004). CONCLUSIONS
Consequently, Miltenberger et al. (2005) imple-
mented two behavioral skills training sessions, fol- We can draw several conclusions from the results
lowed by an in situ assessment within 30 minutes of of the studies evaluating behavioral skills training
the second training session, with 10 children ages for teaching firearm safety skills to children. First,
4 and 5 years. If a child did not perform the safety behavioral skills training, an approach in which a
skills during the in situ assessment, a novel trainer child practices the target safety skills and receives
entered the assessment situation and conducted feedback, is superior to the Eddie Eagle GunSafe
in situ training. The results showed that all 10 Program, an informational approach without
Teaching Safety Skills to Children 485
practice or feedback components. This finding TABLE 28.1. Steps in Teaching Firearm Safety Skills
is consistent with other research demonstrating to Children
that practice of skills with feedback is necessary 1. Provide instructions.
for training programs to be effective (e.g., Beck & a. Describe the dangers of playing with firearms and
Miltenberger, 2009; Poche et al., 1988). Table 28.1 the safety skills to use when finding an unattended
shows the sequence of steps in behavioral skills firearm (“Don’t touch it, get away, and tell an
training for teaching firearm safety skills to chil- adult”).
dren. b. Give examples to illustrate the safety skills in
different situations.
Second, the number of behavioral skills train-
ing sessions required for children to perform the 2. Model the safety skills.
safety skills varies. In some cases, children per- a. Simulate a situation in which you find a firearm
formed the skills after two training sessions; in (using a disabled firearm or a replica of a real
firearm), and demonstrate the safety skills: “Don’t
other cases, children did not perform the skills
touch it, run away from the firearm, and tell an
after five training sessions. This finding was con- adult about the firearm.”
sistent across 4- to 7-year-old children and suggests b. Describe the importance of the skills after
that investigators must repeat safety skills assess- modeling them.
ments to determine how many training sessions
3. Provide the opportunity for rehearsal.
children require to demonstrate generalized use of
a. Set up a scenario in which the child could find a
skills. firearm (e.g., on a shelf in the parents’ bedroom),
Third, some children do not perform safety skills and place a firearm (a disabled firearm or a replica
until they participate in in situ training. All chil- of a real firearm) in the simulated situation.
dren who participated in in situ training performed b. Ask the child to show you the safety skills.
safety skills after one to three training sessions c. During the rehearsal, set up the situation so that
(most after just one session), regardless of how the child has to run out of the room and tell an
many behavioral skills training sessions preceded adult in another room about finding the firearm.
the in situ training. The behavioral mechanism 4. Provide praise and feedback.
that underlies the effectiveness of in situ training is a. Provide descriptive praise for correct rehearsal of
not clear. One possibility is that it simply involves the safety skills or for any aspect of the skills that
reinforcing instances of generalization—a known the child executed correctly.
strategy for promoting generalization (Stokes & b. Provide further instruction for improvement
(feedback) if the child executed any aspect of the
Baer, 1977). Another possibility is that being ob-
safety skills incorrectly.
served exhibiting the incorrect behavior and hav-
ing to rehearse the safety skills repeatedly func- 5. Repeat with a variety of scenarios.
tions as punishment for incorrect behavior, and a. Have the child rehearse the safety skills with praise
avoidance of observation and rehearsal functions and feedback in a variety of different scenarios.
as negative reinforcement of safety skills. b. Create each scenario to represent a situation in
which the child could find a firearm in his or her
Fourth, children’s responses to different assess- home or the home of a friend.
ments were not consistent, showing that different
repertoires are involved in a child’s (1) vocally 6. Conduct in situ assessment.
describing the appropriate safety skills, (2) dem- a. Conduct in situ assessment by placing a disabled
firearm or replica in a location where the child will
onstrating the skills in the presence of the inves- find it, and recording the child’s behavior.
tigator, and (3) using the skills during in situ as- b. In situ assessment must occur without the child’s
sessments. Results of research show that children knowledge.
may demonstrate safety skills in role plays, yet fail
7. Conduct in situ training as needed.
to perform safety skills during in situ assessments.
a. If the child fails to use the skills during the in situ
As such, we view the failure to perform the skills assessment, enter the room and conduct training.
during in situ assessments as a performance defi- b. Review the safety skills.
cit, not a skills deficit. We hypothesize that in situ c. Have the child rehearse three to five times
training functions as contingency management correctly in the situation.
rather than as skills training. Therefore, behav-
ioral skills training may be most appropriate for
teaching skills, whereas in situ training may be
most appropriate for reinforcing use of skills, but
investigators should evaluate this hypothesis in
future studies.
486 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
training to teach safety skills to her 3- and 4-year- settings: Recommendations for prevention. Journal of
old students. This training was not effective. In- the American Board of Family Practice, 14, 107–115.
vestigators then implemented in situ training and Eber, G. B., Annest, J. L., Mercy, J. A., & Ryan, G. W.
incentives, including access to leisure activities, (2004). Nonfatal and fatal firearm-related injuries
among children aged 14 years and younger: United
and the children demonstrated the skills. Addi-
States, 1993–2000. Pediatrics, 113, 1686–1692.
tional research is needed to develop and evaluate Farah, M. M., & Simon, H. K. (1999). Firearms in the
the effectiveness of such programs, with the ulti- home: Parental perceptions. Pediatrics, 104, 1059–
mate goal of promoting wide-scale dissemination. 1063.
Garcia, D., Dukes, C., Brady, M. P., Scott, J., & Wilson,
C. L. (2016). Using modeling and rehearsal to teach
REFERENCES
fire safety to children with autism. Journal of Applied
American Academy of Pediatrics (AAP), Committee Behavior Analysis, 49, 699–704.
on Injury and Poison Prevention. (2000). Firearm- Gatheridge, B. J., Miltenberger, R. G., Huneke, D. F.,
related injuries affecting the pediatric population. Satterlund, M. J., Mattern, A. R., Johnson, B. M.,
Pediatrics, 105, 888–895. et al. (2004). Comparison of two programs to teach
Azrael, D., Cohen, J., Salhi, C., & Miller, M. (2018). firearm injury prevention skills to 6- and 7-year-old
Firearm storage in gun-owning households with chil- children. Pediatrics, 114, 294–299.
dren: Results of a 2015 national survey. Journal of General Accounting Office. (1991). Accidental shoot-
Urban Health, 95, 295–304. ings: Many deaths and injuries caused by firearms
Azrael, D., Miller, M., & Hemenway, D. (2000). Are could be prevented. Retrieved August 6, 2020, from
household firearms stored safely?: It depends whom www.gao.gov/assets/160/150353.pdf.
you ask. Pediatrics, 106, e31–e36. Gross, A., Miltenberger, R., Knudson, P., Bosch, A., &
Beck, K., & Miltenberger, R. (2009). Evaluation of a com- Brower-Breitwieser, C. (2007). Preliminary evalua-
mercially available abduction prevention program. tion of a parent training program to prevent gun play.
Journal of Applied Behavior Analysis, 42, 761–772. Journal of Applied Behavior Analysis, 40, 691–695.
Brady Campaign to Prevent Gun Violence. (n.d.-a). Grossman, D. C., Cummings, P., Koepsell, T. D., Mar-
Sensible gun laws save lives. Retrieved July 28, 2015, shall, J., D’Ambrosio, L., Thompson, R. S., et al.
from www.bradycampaign.org. (2000). Firearm safety counseling in primary care
Brady Campaign to Prevent Gun Violence. (n.d.-b). pediatrics: A randomized, controlled trial. Pediatrics,
Firearm facts. Retrieved June 1, 2015, from www.bra- 106, 22–26.
dycampaign.org. Grossman, D. C., Mang, K., & Rivara, F. P. (1995). Fire-
Carroll-Rowan, L., & Miltenberger, R. (1994). A com- arm injury prevention counseling by pediatricians
parison of procedures for teaching abduction preven- and family physicians. Archives of Pediatrics and Ado-
tion to preschoolers. Education and Treatment of Chil- lescent Medicine, 149, 973–977.
dren, 17, 113–128. Grossman, D. C., Reay, D. T., & Baker, S. A. (1999).
Common Sense about Kids and Guns. (n.d.). Fact file. Self-inflicted and unintentional firearm injuries
Retrieved July 28, 2015, from www.kidsandguns.org. among children and adolescents: The source of the
Coyne-Beasley, T., Schoenbach, V. J., & Johnson, R. firearm. Archives of Pediatrics and Adolescent Medi-
M. (2001). “Love Our Kids, Lock Your Guns”: A cine, 1538, 875–882.
community-based firearm safety counseling and gun Hanratty, L., Miltenberger, R., & Florentino, S. (2016).
lock distribution program. Archives of Pediatrics and Evaluating the effectiveness of a teaching package
Adolescent Medicine, 155, 659–664. utilizing behavioral skills training and in situ training
Crifasi, C. K., Doucette, M. L., McGinty, E. E., Web- to teach gun safety skills in a preschool classroom.
ster, D. W., & Barry, C. L. (2018). Storage practice Journal of Behavioral Education, 25, 310–323.
of US gun owners in 2016. American Journal of Public Hardy, M. S. (2002). Teaching firearm safety to chil-
Health, 108, 532–537. dren: Failure of a program. Developmental and Behav-
Crifasi, C. K., O’Dwyer, J. K., McGinty, E. E., Webster, ioral Pediatrics, 23, 71–76.
D. W., & Barry, C. L. (2019). Desirability of personal- Hardy, M. S., Armstrong, F. D., Martin, B. L., & Strawn,
ized guns among current gun owners. American Jour- K. N. (1996). A firearm safety program for children:
nal of Preventive Medicine, 57, 191–196. They just can’t say no. Journal of Developmental and
Cummings, P., Grossman, D. C., Rivara, F. P., & Ko- Behavioral Pediatrics, 17, 216–221.
epsell, T. D. (1997). State gun safe storage laws and Haught, K., Grossman, D., & Connell, F. (1995). Par-
child mortality due to firearms. Journal of the Ameri- ents’ attitudes toward firearm injury prevention
can Medical Association, 278, 1084–1086. counseling in urban pediatric clinics. Pediatrics, 96,
DiScala, C., & Sege, R. (2004). Outcomes in children 649–653.
and young adults who are hospitalized for firearms- Hemenway, D., & Solnick, S. J. (2015). Children and
related injuries. Pediatrics, 113, 1306–1312. unintentional firearm death. Injury Epidemiology, 2,
Dresang, L. T. (2001). Gun deaths in rural and urban 26.
488 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Hemenway, D., Solnick, S. J., & Azrael, D. R. (1995). & Bagne, A. (2007). Teaching skills to second and
Firearm training and storage. Journal of the American third grade children to prevent gun play: A compari-
Medical Association, 273, 46–50. son of procedures. Education and Treatment of Chil-
Hendricks, C. M., & Reichert, A. (1996). Parents’ self- dren, 30, 29–48.
reported behaviors related to health and safety of Knight-Bohnhoff, K., & Harris, M. B. (1998). Parent’s
very young children. Journal of School Health, 66, behaviors, knowledge, and beliefs related to uninten-
247–251. tional firearm injuries among children and youth in
Himle, M., & Miltenberger, R. (2004). Preventing un- the Southwest. Journal of Pediatric Health Care, 12,
intentional firearm injury in children: The need for 139–146.
behavioral skills training. Education and Treatment of Lumley, V. A., Miltenberger, R. G., Long, E. S., Rapp,
Children, 27, 161–177. J. T., & Roberts, J. A. (1998). Evaluation of a sexual
Himle, M. B., Miltenberger, R. G., Flessner, C., & Gath- abuse prevention program for adults with mental re-
eridge, B. (2004). Teaching safety skills to children to tardation. Journal of Applied Behavior Analysis, 31,
prevent gun play. Journal of Applied Behavior Analy- 91–101.
sis, 37, 1–9. Marchand-Martella, N., & Huber, G. (1996). Assessing
Himle, M. B., Miltenberger, R. G., Gatheridge, B., & the long-term maintenance of abduction prevention
Flessner, C. (2004). An evaluation of two procedures skills by disadvantaged preschoolers. Education and
for training skills to prevent gun play in children. Pe- Treatment of Children, 19, 55–58.
diatrics, 113, 70–77. Maxfield, T. C., Miltenberger, R. G., & Novotny, M. A.
Jackman, G. A., Farah, M. M., Kellerman, A. L., & (2019). Evaluating small-scale simulation for train-
Simon, H. K. (2001). Seeing is believing: What do ing firearm safety skills. Journal of Applied Behavior
boys do when they find a real gun? Pediatrics, 107, Analysis, 52, 491–498.
1247–1250. McGee, K. S., Coyne-Beasley, T., & Johnson, R. M.
Johnson, B. M., Miltenberger, R. G., Egemo-Helm, K., (2003). Review of evaluations of educational ap-
Jostad, C. J., Flessner, C., & Gatheridge, B. (2005). proaches to promote safe storage of firearms. Injury
Evaluation of behavioral skills training for teaching Prevention, 9, 108–111.
abduction-prevention skills to young children. Jour- Miller, M., Azrael, D., & Hemenway, D. (2001). Firearm
nal of Applied Behavior Analysis, 38, 67–78. availability and unintentional firearm deaths. Acci-
Johnson, B. M., Miltenberger, R. G., Knudson, P., Ege- dent Analysis and Prevention, 33, 477–484.
mo-Helm, K., Kelso, P., Jostad, C., et al. (2006). A Miller, M., Azrael, D., Hemenway, D., & Vriniotis, M.
preliminary evaluation of two behavioral skills train- (2005). Firearm storage practices and rates of unin-
ing procedures for teaching abduction prevention tentional firearm deaths in the United States. Acci-
skills to school-age children. Journal of Applied Be- dent Analysis and Prevention, 37, 661–667.
havior Analysis, 39, 25–34. Miltenberger, R. (2008). Teaching safety skills to chil-
Jones, R. T., Kazdin, A. E., & Haney, J. I. (1981). Social dren: Prevention of firearm injury as an exemplar of
validation and training of emergency fire safety skills best practice in assessment, training, and generaliza-
for potential injury prevention and life saving. Jour- tion of safety skills. Behavior Analysis in Practice, 1,
nal of Applied Behavior Analysis, 14, 249–260. 30–36.
Jones, R. T., Ollendick, T. H., McLaughlin, K. J., & Miltenberger, R. G., Flessner, C., Gatheridge, B., John-
Williams, C. E. (1989). Elaborative and behavioral son, B., Satterlund, M., & Egemo, K. (2004). Evalu-
rehearsal in the acquisition of fire emergency skills ation of behavioral skills training procedures to pre-
and the reduction of fear of fire. Behavior Therapy, vent gun play in children. Journal of Applied Behavior
20, 93–101. Analysis, 37, 513–516.
Jostad, C. M., & Miltenberger, R. G. (2004). Firearm Miltenberger, R., Fogel, V., Beck, K., Koehler, S., Graves,
injury prevention skills: Increasing the efficiency of R., Noah, J., et al. (2013). Examining the efficacy of
training with peer tutoring. Child and Family Behav- the Stranger Safety abduction prevention program
ior Therapy, 26, 21–35. and parent conducted in situ training. Journal of Ap-
Jostad, C. M., Miltenberger, R. G., Kelso, P., & Knudson, plied Behavior Analysis, 46, 817–820.
P. (2008). Peer tutoring to prevent gun play: Acquisi- Miltenberger, R. G., Gatheridge, B. J., Satterlund, M.,
tion, generalization, and maintenance of safety skills. Egemo-Helm, K. R., Johnson, B. M., Jostad, C., et al.
Journal of Applied Behavior Analysis, 41, 117–123. (2005). Teaching safety skills to prevent gun play:
Kellerman, A. L. (1993). Preventing firearm injuries: A An evaluation of in situ training. Journal of Applied
review of epidemiologic research. American Journal of Behavior Analysis, 38, 395–398.
Preventive Medicine, 9(Suppl. 1), 12–15. Miltenberger, R., Gross, A., Knudson, P., Jostad, C.,
Kellerman, A. L., & Reay, D. T. (1986). Protection or Bosch, A., & Brower Breitwieser, C. (2009). Evaluat-
peril?: An analysis of firearm-related deaths in the ing behavioral skills training with and without simu-
home. New England Journal of Medicine, 314, 1557– lated in situ training for teaching safety skills to chil-
1560. dren. Education and Treatment of Children, 32, 63–75.
Kelso, P., Miltenberger, R., Waters, M., Egemo-Helm, K., Miltenberger, R., & Hanratty, L. (2013). Teaching sexu-
Teaching Safety Skills to Children 489
al abuse prevention skills to children. In D. S. Brom- Senturia, Y. D., Christoffel, K. K., & Donovan, M.
berg & W. T. O’Donohue (Eds.), Handbook of child (1994). Children’s household exposure to guns: A pe-
and adolescent sexuality: Developmental and forensic diatric practice-based survey. Pediatrics, 93, 469–475.
psychology (pp. 419–447). London: Elsevier. Sidman, E. A., Grossman, D. C., Koepsell, T. D.,
Miltenberger, R. G., & Olsen, L. A. (1996). Abduction D’Ambrosio, L., Britt, J., Simpson, E. S., et al. (2005).
prevention training: A review of findings and issues Evaluation of a community-based handgun safe-stor-
for future research. Education and Treatment of Chil- age campaign. Pediatrics, 115, e654–e661.
dren, 19(1), 69–82. Stennies, G., Ikeda, R., Leadbetter, S., Houston, B., &
Miltenberger, R. G., Roberts, J. A., Ellingson, S., Galen- Sacks, J. (1999). Firearm storage practices and chil-
sky, T., Rapp, J. T., Long, E. S., et al. (1999). Training dren in the home, United States, 1994. Archives of
and generalization of sexual abuse prevention skills Pediatrics and Adolescent Medicine, 153, 586–590.
for women with mental retardation. Journal of Ap- Stokes, T. F., & Baer, D. M. (1977). An implicit tech-
plied Behavior Analysis, 32, 385–388. nology of generalization. Journal of Applied Behavior
Miltenberger, R., Sanchez, S., & Valbuena, D. (2015). Analysis, 10, 349–367.
Teaching safety skills to children. In H. S. Roane, J. Tarasenko, M. A., Miltenberger, R. G., Brower-Breitwie-
E. Ringdahl, & T. S. Falcomata (Eds.), Clinical and ser, C., & Bosch, A. (2010). Evaluation of peer train-
organizational applications of applied behavior analysis ing for teaching abduction prevention skills. Child
(pp. 477–499). London: Elsevier. and Family Behavior Therapy, 32, 219–230.
Miltenberger, R., & Thiesse-Duffy, E. (1988). Evalua- Vanselow, N. R., & Hanley, G. P. (2014). An evalua-
tion of home-based programs for teaching personal tion of computerized behavior skills training to teach
safety skills to children. Journal of Applied Behavior safety skills to young children. Journal of Applied Be-
Analysis, 21, 81–88. havior Analysis, 47, 51–69.
Miltenberger, R. G., Thiesse-Duffy, E., Suda, K. T., Vernick, J. S., O’Brien, M., Hepburn, L. M., Johnson, S.
Kozak, C., & Bruellman, J. (1990). Teaching preven- B., Webster, C. W., & Hargarten, S. W. (2003). Unin-
tion skills to children: The use of multiple measures tentional and undetermined firearm related deaths:
to evaluate parent versus expert instruction. Child A preventable death analysis for three safety devices.
and Family Behavior Therapy, 12, 65–87. Injury Prevention: Journal of the International Society
National Opinion Research Center. (2014). General for Child and Adolescent Injury, 9, 307–311.
Social Survey final report: Trends in gun owner- Violano, P., Bonne, S., Duncan, T., Pappas, P., Christ-
ship in the United States, 1972–2014. Retrieved mas, A. B., Goldberg, S., et al. (2018). Prevention of
August 1, 2014, from www.norc.org/PDFs/GSS%20 firearm injuries with gun safety devices and storage:
Reports/GSS_Trends%20in%20Gun%20Ownership_ An Eastern Association for the Surgery of Trauma
US_1972-2014.pdf. systemic review. Journal of Trauma and Acute Care
Olsen-Woods, L. A., Miltenberger, R. G., & Forman, G. Surgery, 84, 1003–1011.
(1998). Effects of correspondence training in an ab- Webster, D. W., & Starnes, M. (2000). Reexamining the
duction prevention training program. Child and Fam- association between child access prevention gun laws
ily Behavior Therapy, 20, 15–34. and unintentional shooting deaths of children. Pedi-
Ordog, G. J., Wasserberger, J., Schatz, I., Owens-Collins, atrics, 106, 1466–1469.
D., English, K., Balasubramanian, S., et al. (1988). Webster, D. W., Wilson, M. E. H., Duggan, A. K., &
Gunshot wounds in children under 10 years of age: A Pakula, L. C. (1992). Parents’ beliefs about prevent-
new epidemic. American Journal of Diseases of Chil- ing gun injuries to children. Pediatrics, 89, 908–914.
dren, 142, 618–622. Wiley, C. C., & Casey, R. (1993). Family experiences,
Parikh, K., Silver, A., Patel, S. J., Iqbal, S. F., & Goyal, attitudes, and household safety practices regarding
M. (2017). Pediatric firearm-related injuries in the firearms. Clinical Pediatrics, 32, 71–76.
United States. Hospital Pediatrics, 7, 303–312. Wintemute, G. J., Teret, S. P., Kraus, J. F., Wright, M. A.,
Poche, C., Brouwer, R., & Swearingen, M. (1981). & Bradfield, G. (1987). When children shoot chil-
Teaching self-protection to young children. Journal dren: 88 unintended deaths in California. Journal of
of Applied Behavior Analysis, 14, 169–176. the American Medical Association, 257, 3107–3109.
Poche, C., Yoder, P., & Miltenberger, R. (1988). Teach- Yeaton, W. H., & Bailey, J. S. (1978). Teaching pedes-
ing self-protection to children using television tech- trian safety skills to young children: An analysis and
niques. Journal of Applied Behavior Analysis, 21, one-year follow-up. Journal of Applied Behavior Analy-
253–261. sis, 11, 315–329.
Schuster, M. A., Franke, T. M., Bastian, A. M., Sor, S., Zavoski, R. W., Lapidus, G. D., Lerer, T. J., & Banco,
& Halfon, N. (2000). Firearm storage patterns in L. I. (1995). A population-based study of severe fire-
US homes with children. American Journal of Public arm injury among children and youth. Pediatrics, 96,
Health, 90, 588–594. 278–282.
CHAP TER 29
490
Behavior Analysis and Treatment of Drug Addiction 491
for drug addiction treatment (Silverman, Kamin- ness (Benishek et al., 2014; Davis et al., 2016).
ski, et al., 2011). Three research programs on this First, the duration of abstinence reinforcement
method stand out. Milby et al. (1996) investigated interventions is relatively short, with a mean of 12
the use of abstinence-contingent housing and weeks in both Davis et al. (2016) and Benishek et
work therapy to promote abstinence in home- al. (2014), and a median of 8 and 12 weeks in Davis
less cocaine-dependent adults. Ries et al. (2004) et al. and Benishek et al., respectively. Davis et al.
used U.S. Social Security Disability benefits in a reviewed voucher- and money-based contingency
contingent fashion to promote drug abstinence in management interventions for substance use dis-
adults with severe mental illness. Finally, Silver- orders from 2009 through 2014; they reported that
man (2004) used abstinence-contingent access to 43 of 51 studies (84%) showed significant effects of
employment in a series of studies to initiate and the abstinence reinforcement interventions, and
maintain drug abstinence in unemployed adults the rest (16%) did not produce significant effects.
with long histories of drug addiction. Of the 22 studies that produced significant effects
while the abstinence reinforcement interventions
were in effect and assessed posttreatment absti-
RECENT ADVANCES IN RESEARCH nence, only 7 (32%) showed a statistically signifi-
ON ABSTINENCE REINFORCEMENT cant effect at follow-up. Benishek et al. conducted
a meta-analysis of prize-based abstinence rein-
At the time the earlier chapter was written, conclu- forcement interventions from 2000 through 2013.
sions about research on abstinence reinforcement Results were like those Davis et al. reported. Eigh-
interventions were relatively simple (Silverman, teen of the 19 studies (95%) reported a significant
Kaminski, et al., 2011): (1) These interventions effect while the prize-based interventions were in
promoted abstinence from most commonly abused effect. The effects of these interventions decreased
drugs and in diverse populations and settings; (2) in the 6 months after discontinuation of the in-
a need existed to develop procedures to increase terventions. Six of the nine studies (66%) with a
the proportion of patients who achieved absti- 3-month follow-up assessment showed a significant
nence when exposed to these interventions; (3) a effect of the prize-based abstinence reinforcement
need existed to develop procedures that promoted interventions. Two of the six studies (33%) with
long-term abstinence and prevented relapse; and a 6-month follow-up assessment showed a signifi-
(4) a need existed to develop practical applica- cant intervention effect. An analysis that com-
tions of abstinence reinforcement interventions bined data from the 6-month follow-up showed
that ensured the widespread application of these that the overall effects of the prize-based interven-
procedures in society. In the remainder of this tions were not detectable at that time.
chapter, we review studies conducted since we The more recent reviews and meta-analyses
completed the earlier chapter, with special atten- document the effectiveness of operant-condi-
tion to whether recent research has confirmed and tioning principles in the treatment of drug ad-
extended our earlier conclusions. diction. Abstinence reinforcement interventions
significantly increase drug abstinence while the
interventions are in effect and occasionally after
General Utility of Abstinence
they end (Benishek et al., 2014; Cahill et al., 2015;
Reinforcement Interventions
Castells et al., 2009; Davis et al., 2016). However,
Reviews and meta-analyses published in the last these reviews and meta-analyses summarize the
several years have confirmed that abstinence re- literature by statistical significance and effect
inforcement interventions are highly effective sizes, but they do not fully highlight the clinical
in promoting abstinence from most commonly importance and limitations of abstinence rein-
abused drugs and in diverse populations (Benishek forcement interventions: What proportion of pa-
et al., 2014; Cahill, Hartmann-Boyce, & Perera, tients fail to respond to such interventions? Do
2015; Castells et al., 2009; Davis et al., 2016). these interventions serve a clinically useful role if
These reviews and meta-analyses also confirmed they do not maintain abstinence over time? Has
that such interventions are not always effective recent research identified methods of promoting
and rarely produce effects that are evident after abstinence in treatment-resistant patients and pro-
the interventions end. moting long-term abstinence? In this section, we
Two analyses of these interventions highlight summarize the results of selected studies to shed
key and consistent findings about their effective- some light on these questions.
494 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
stinence, or a no-treatment control group. Partici- disorder who were enrolled in an outpatient treat-
pants in the voucher intervention could earn up to ment program were randomly assigned to receive
$1,418 in vouchers for providing cocaine-free urine usual care or usual care plus prize-based absti-
samples every Monday, Wednesday, and Friday nence reinforcement (Killeen, McRae-Clark, Wal-
over 16 weeks. Although the study examined the drop, Upadhyaya, & Brady, 2012). After a 2-week
effects of flexible methadone dosing and vouch- washout period, participants in the prize-based ab-
er-based reinforcement of cocaine abstinence on stinence reinforcement group could earn the op-
opiate and cocaine use, we review only the ef- portunity to draw prizes for providing alcohol-free
fects of the voucher-based intervention here. The breath samples and urine samples that were nega-
group exposed to the voucher-based intervention tive for marijuana, cocaine, opiates, methamphet-
had significantly higher rates of cocaine-negative amine, and amphetamine; control participants
urine samples during treatment than the group earned two prize draws for providing samples, in-
that received neither intervention. Surprisingly, dependent of the breath or urinalysis results. The
the group exposed to both flexible dosing and prize intervention was in effect for 10 weeks. The
voucher-based cocaine abstinence reinforcement study showed no difference in urinalysis results
did not provide significantly more cocaine-neg- between the two groups on any measure of drug
ative urine samples than the group that received use. Although this study included other drug use
neither intervention. As in other studies, although in the abstinence reinforcement intervention and
the voucher-based intervention increased cocaine in the outcome measures, marijuana was the pre-
abstinence, not all participants responded to the dominant drug participants used in this study, and
intervention. Only about half the urine samples of there were very low levels of drug use other than
participants in the group exposed to the voucher- marijuana in both groups.
based intervention alone were cocaine-negative. In a second study, 59 adolescents between the
This study did not assess posttreatment results. ages of 14 and 18 who met Diagnostic and Statistical
Manual of Mental Disorders, fourth edition (DSM-
IV) criteria for cannabis abuse or dependence and
Marijuana
showed recent evidence of marijuana use, but no
Litt, Kadden, and Petry (2013) assigned 215 mar- use of substances other than alcohol and tobacco,
ijuana-dependent adults to one of three 9-week were randomly assigned to one of two conditions:
outpatient treatments: a case management (con- cognitive-behavioral therapy and voucher-based
trol) condition; a condition combining motiva- abstinence reinforcement, or cognitive-behavioral
tional-enhancement therapy, cognitive-behavioral therapy and noncontingent vouchers. The voucher
therapy, and prize-based reinforcement for home- condition started in Week 3 and continued until
work completion; and a condition combining Week 10, and participants could earn up to $242
motivational-enhancement therapy, cognitive- in vouchers for providing drug-free urine samples.
behavioral therapy, and prize-based reinforcement Researchers tested urine samples for cannabis, co-
for abstinence from marijuana. The study evalu- caine, opioids, benzodiazepines, amphetamines,
ated the effects of those interventions on self-re- and methamphetamine. Participants exchanged
ported marijuana use in the months after the voucher earnings for gift cards. There were no sig-
intervention ended. Neither of the experimental nificant differences between groups on measures
interventions that included prize-based reinforce- of marijuana use either during or after treatment.
ment increased self-reported marijuana abstinence In a final study, 153 adolescents between the
relative to the control condition; however, the ages of 12 and 18 who met DSM-IV criteria for
intervention that included prize-based reinforce- cannabis abuse or dependence and showed recent
ment for marijuana abstinence produced transient evidence of marijuana use, but were not dependent
increases in marijuana abstinence in Months 5–8, on another substance, were randomly assigned
compared to prize-based reinforcement of home- to one of three groups (Stanger, Ryan, Scherer,
work completion. There were no significant differ- Norton, & Budney, 2015). All groups received
ences between groups at the latest follow-up time motivational-enhancement therapy and cogni-
at Months 11–14. tive-behavioral therapy for 14 weeks. Two groups
Three controlled studies evaluated the effects received an additional contingency management
of abstinence reinforcement interventions on intervention, one with a parent-training compo-
marijuana use in adolescents. In the first of these nent and one without parent training. After a
studies, 31 adolescents with primary marijuana use 2-week washout period, participants in the contin-
Behavior Analysis and Treatment of Drug Addiction 497
gency management groups began the contingency experimental design to evaluate its effectiveness
management intervention (Weeks 3–14). During (Barnett, Tidey, Murphy, Swift, & Colby, 2011).
this intervention, participants could earn voucher- The second study used a within-participant
based reinforcement for urine samples collected cross-over design to evaluate the effectiveness of
twice per week that were negative for all drugs a contingency management intervention, includ-
assessed (cannabis, cocaine, opioids, benzodiaz- ing use of the SCRAM bracelet, in promoting
epines, amphetamine, and methamphetamine). alcohol abstinence (Dougherty et al., 2014). In
Participants in the voucher condition could earn that study, participants (N = 26) were exposed
up to $590 in vouchers that were exchangeable for to $0 and $25 contingency management condi-
gift cards for providing drug-free urine samples. In tions in counterbalanced order, followed by a $50
addition, participants received a prize-based absti- contingency management condition. Each condi-
nence reinforcement intervention in which they tion was in effect for 4 weeks. In the $0 condition,
could earn up to about $135 in prizes for provid- researchers did not give participants any instruc-
ing drug-free urine samples during Weeks 1–4. tions about alcohol consumption. During the $25
Finally, parents could earn prize draws for devel- and $50 conditions, participants could earn $25
oping and using a substance-abuse-monitoring and $50 per week, respectively, if their transder-
contract. The contingency management interven- mal alcohol concentration never exceeded 0.03
tion produced significant increases in abstinence grams per decaliter (g/dl; this concentration cor-
from marijuana during treatment and at the end responded to light to moderate drinking) on any
of treatment. Forty-eight percent of participants in day during the week. The $25 incentive condition
the two groups exposed to the contingency man- significantly reduced drinking episodes and heavy
agement intervention achieved 4 or more weeks of drinking, relative to the $0 condition. In addition,
continuous abstinence during treatment, whereas participants exposed to the $25 condition followed
only 30% of participants in the group not exposed by the $0 condition had less frequent drinking and
to the contingency management intervention did less heavy drinking in the $0 condition than par-
so. Abstinence rates decreased between the end ticipants exposed to the $0 condition followed by
of treatment and the 3-month follow-up, and the the $25 condition. This result demonstrated that
three groups had similar rates of abstinence at the the $25 condition produced some lasting carry-
3-month follow-up. over effects on drinking that persisted after the
$25 condition ended. There were few differences
on drinking outcomes between the $25 and $50
Alcohol
conditions, except that the $50 condition was
Researchers have conducted limited research on more effective in reducing heavy weekend drink-
abstinence reinforcement interventions for alco- ing than the $25 condition.
hol use, at least in part because measures have A third study using the SCRAM technology
not been available to detect alcohol use reliably enrolled 80 alcohol-dependent adults in a three-
beyond a relatively brief window of several hours phase study that included an observation phase,
since the last drink. Researchers can use breath a contingency management phase, and a follow-
alcohol tests, but participants can test alcohol- up phase (Dougherty, Karns, et al., 2015). At each
negative on these by remaining abstinent from weekly clinic visit during the observation and con-
alcohol for several hours before providing a breath tingency management phases and monthly during
sample. the follow-up phase, participants completed Time-
Since our earlier chapter appeared, three stud- line Follow-Back interviews about drinking. Dur-
ies have evaluated the effectiveness of using a ing the initial observation phase, each participant
transdermal alcohol sensor bracelet as a part of wore a SCRAM bracelet with no explicit conse-
an abstinence reinforcement intervention for al- quences for drinking. During the contingency
cohol use. A participant wears this bracelet, called management phase, a participant earned $50 each
the Secure Continuous Remote Alcohol Monitor week if the SCRAM bracelet showed that the
(SCRAM), locked on the ankle. The participant participant’s transdermal alcohol concentration
can wear the bracelet continuously, including in stayed below 0.03 g/dl every day of the week. Dur-
the shower. The bracelet includes features that ing the follow-up phase, participants returned to
detect removal and tampering. One of the studies the clinic every month for 3 months and complet-
used the bracelet in a contingency management ed Timeline Follow-Back interviews about drink-
intervention, but did not include an adequate ing in the past 28 days. Transdermal alcohol con-
498 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
centration data showed that heavy drinking was cohol, <0.02 g/dl) under an escalating schedule
significantly lower during the contingency phase for sustained alcohol abstinence. The voucher in-
than during the observation phase (Dougherty, tervention was in effect for 4 weeks, and partici-
Karns, et al., 2015). Self-reported days of drinking pants could earn up to $340 in vouchers for pro-
and heavy drinking were significantly lower dur- viding alcohol-negative breath tests. Both groups
ing the contingency phase and follow-up phases provided comparable rates of breath samples, but
than during the observation phase (Dougherty, participants in the monitoring-plus-contingency-
Lake, et al., 2015). Although the study appeared to management group provided significantly more
show that the effects of the contingency manage- alcohol-negative breath samples (87%) than par-
ment intervention on participants’ drinking were ticipants in the monitoring-only group (67%). Re-
maintained in the follow-up period, the study did searchers did not assess rates of alcohol use after
not include a rigorous experimental design that voucher-based abstinence reinforcement ended.
would have allowed clear demonstration of its ef-
fectiveness.
Polydrug Use
A small study evaluated the use of ethyl gluc-
uronide, a metabolite of alcohol, in a voucher- Petry, Weinstock, and Alessi (2011) evaluated the
based abstinence reinforcement intervention effects of a group contingency management in-
(McDonell et al., 2012). Laboratories can measure tervention on use of multiple substances. In that
ethyl glucuronide in urine samples. This measure study, 239 outpatients enrolled in two community-
may be useful in abstinence reinforcement inter- based substance abuse treatment clinics were ran-
ventions, because it provides a longer window of domly assigned to standard care or to contingency
detection than breath alcohol samples do. This management. Both groups provided two urine and
study used a within-participant reversal design to breath samples per week for 12 weeks. Participants
evaluate the effectiveness of the voucher-based in- in the contingency management group received a
tervention in promoting abstinence from alcohol prize-based reinforcement intervention in which
in 10 adults who met DSM-IV criteria for alcohol they could earn draws for prizes for attending the
dependence. Four of the 10 participants did not clinic and for providing urine and breath samples
complete the return-to-baseline condition, which that were negative for cocaine, methamphetamine,
limits the value of the research design. Given that opioids, and alcohol. The researchers drew slips of
limitation, the study provided some preliminary paper with the participants’ names from a hat to
evidence of the effectiveness of reinforcing ethyl- identify participants who could draw for prizes.
glucuronide-negative urine samples in promoting The number of each participant’s slips increased
abstinence from alcohol in study participants. as the duration the participant attended clinic and
Another study evaluated the effectiveness of provided drug-negative samples increased. The
using cellphone technology to observe, record, first five participants whose names were selected
and reinforce negative breath alcohol tests on a drew for one prize each; the sixth person whose
breathalyzer (Alessi & Petry, 2013). In this study, name was selected drew for five prizes. This con-
30 participants who regularly used alcohol were tingency management intervention significantly
randomly assigned to monitoring only or to moni- increased the number of days attended, continu-
toring plus contingency management. Researchers ous weeks of attendance, and longest duration of
gave cellphones and breathalyzers to participants abstinence from all drugs and from each of the
in both groups, and instructed each participant drugs tested. There was no effect on the percent-
to use the cellphone to video-record him- or her- age of drug-negative samples during or after treat-
self providing a breath sample in the breathalyzer ment.
within an hour of receiving a text message to pro- In another study, 170 adults living with HIV
vide a sample. Researchers told participants they who met DSM-IV criteria for cocaine or opioid
would receive a text three times per day, and paid abuse or dependence, and who were members of
participants for providing valid (i.e., with the an HIV drop-in center, were randomly assigned
breathalyzer reading recorded on the video) breath to a Twelve-Step facilitation group or to a con-
samples within an hour of receiving a text. In ad- tingency management group (Petry, Weinstock,
dition, participants in the monitoring-plus-con- Alessi, Lewis, & Dieckhaus, 2010). Both groups
tingency-management group received vouchers for provided weekly breath samples tested for alcohol,
providing valid on-time samples indicating recent and urine samples tested for opioids and cocaine,
abstinence from alcohol (i.e., negative breath al- for 24 weeks. Participants in the contingency
Behavior Analysis and Treatment of Drug Addiction 499
management condition received a prize-based had to enroll in methadone treatment to gain ac-
reinforcement intervention in which they could cess to the workplace, and they also had to provide
draw for prizes contingent on completion of health urine samples negative for opiates and cocaine
activities and for submitting drug-free breath and to maintain maximum pay in the workplace. An
urine samples. Participants in the contingency intent-to-treat analysis showed that not all partici-
management condition achieved significantly pants responded to the abstinence reinforcement
longer durations of sustained abstinence from all contingencies, but that abstinence, methadone,
drugs and from cocaine in particular during treat- and work reinforcement participants provided sig-
ment. There was no effect on the percentage of nificantly more urine samples negative for opiates
drug-negative samples during or after treatment. and for cocaine than work reinforcement partici-
One study randomly assigned 305 adolescents pants did during treatment. Participants did not
with alcohol and other substance use disorders maintain those effects at the 6-month follow-up
in a residential treatment program to one of four (Holtyn et al., 2014a). A within-group analysis of
conditions: a usual-care control condition, a con- abstinence, methadone, and work reinforcement
tingency management condition, an assertive participants confirmed that abstinence from opi-
continuing-care condition, or a condition that ates and cocaine increased when abstinence rein-
combined contingency management and assertive forcement contingencies for those two drugs were
continuing care (Godley et al., 2014). Adolescents applied sequentially (Holtyn et al., 2014b).
in the contingency management condition earned
prize-based reinforcement for completing proso-
Improving Outcomes
cial activities and for providing urine and breath
samples negative for alcohol, amphetamine/meth- At the time we wrote the earlier chapter, re-
amphetamine, marijuana, cocaine, and opiates searchers clearly needed to develop procedures to
over 12 weeks. Participants assigned to the contin- increase the proportion of patients who achieve
gency management group achieved significantly abstinence when exposed to abstinence reinforce-
higher percentages of self-reported days abstinent ment interventions and procedures that promote
from drugs and alcohol during the 12 months after long-term abstinence and prevent relapse. The
discharge from the residential treatment program studies published since that earlier review and
(65%) than participants assigned to the usual-care described above confirm the need to increase the
control condition (53%). Oddly, the group ex- proportion of such patients. This section reviews
posed to both the contingency management and selected studies that researchers have conducted
assertive continuing-care interventions did not since our earlier chapter to address these issues.
differ from the usual-care control condition.
Holtyn et al. (2014a, 2014b) evaluated the ef-
Increasing Effectiveness
fects of an employment-based abstinence reinforce-
ment intervention in promoting abstinence from Changing Reinforcement Parameters. Festinger,
heroin and cocaine in out-of-treatment adults who Dugosh, Kirby, and Seymour (2014) evaluated the
injected drugs. Researchers invited unemployed potential benefits of offering cash as opposed to
out-of-treatment adults who injected drugs to at- vouchers for cocaine abstinence reinforcement.
tend a therapeutic workplace and referred them to They randomly assigned 222 consecutive admis-
methadone treatment. After a 4-week induction sions to a Philadelphia methadone treatment
period, researchers invited participants to attend program who met DSM-IV criteria for cocaine de-
the workplace for 26 weeks and randomly assigned pendence to a usual-care control, a cash-based ab-
them to a work reinforcement group; a methadone stinence reinforcement intervention, or a vouch-
and work reinforcement group; or an abstinence, er-based abstinence reinforcement intervention.
methadone, and work reinforcement group. Work They asked participants to provide urine samples
reinforcement participants could work in the three times per week for 12 weeks. Participants in
therapeutic workplace independently of whether the cash-based and voucher-based abstinence re-
they enrolled in methadone treatment or pro- inforcement groups could earn up to about $1,000
vided drug-free urine samples. Methadone and for providing cocaine-free urine samples over the
work reinforcement participants had to enroll in 12-week period in cash or vouchers, respectively.
methadone treatment to gain access to the work- Participants in both abstinence reinforcement
place and to maintain maximum pay. Abstinence, groups achieved significantly longer durations of
methadone, and work reinforcement participants cocaine abstinence (6 weeks for each group) than
500 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
participants in the usual-care control group did reinforcement increased early in treatment. Both
(4 weeks), but the two abstinence reinforcement abstinence reinforcement interventions increased
groups did not differ significantly from each other. smoking cessation during treatment, compared to
This study showed that cash-based and voucher- the standard-care condition; however, the two re-
based reinforcement were both effective in pro- inforcement groups did not differ from each other
moting cocaine abstinence, but cash did not pro- in smoking cessation during or after treatment.
vide any benefit over vouchers. Neither abstinence reinforcement intervention
Two studies examined the potential benefit produced significant effects on smoking cessation,
of increasing the magnitude of reinforcement at compared to standard care, that were evident at
the beginning of an abstinence reinforcement in- 2-month or 6-month follow-up after the interven-
tervention, to increase the overall percentage of tions had ended.
participants that achieve sustained abstinence. In another study, 100 adults who injected drugs
Higgins et al. (2014) evaluated the benefit of in- and smoked cigarettes were randomly assigned to
creasing the magnitude of reinforcement at the a usual-care control group, to a contingency man-
beginning of an intervention promoting smoking agement group, to a lung age group, or to a group
cessation in pregnant cigarette-smoking women. that received both a contingency management
They randomly assigned 130 women to receive the and lung age intervention (Drummond et al.,
revised voucher-based abstinence reinforcement 2014). Participants in the contingency manage-
intervention, the usual abstinence reinforcement ment intervention received monetary incentives
intervention, or noncontingent vouchers. Par- for CO samples that indicated recent abstinence
ticipants in both abstinence reinforcement groups from smoking; however, this intervention used a
received vouchers under the same schedule of es- very low overall incentive magnitude and infre-
calating reinforcement for sustained abstinence quent CO testing. Researchers told participants
that previous studies have used. Participants in the in the lung age intervention the age equivalence
revised voucher-based intervention received large of their lungs, based on spirometric assessments.
bonus vouchers for smoking cessation during the No intervention affected the 6-month biologically
first 6 weeks of the intervention. To keep the total verified rates of smoking cessation.
earnings comparable between the two abstinence In a study of individuals who were interested in
reinforcement groups, participants in the revised quitting smoking, Lamb, Kirby, Morral, Glabicka,
voucher-based condition received less frequent as- and Iguchi (2010) compared a fixed-reinforcement
sessments and vouchers late in treatment. Both schedule in which the criterion for reinforcement
voucher-based abstinence reinforcement interven- was a constant CO value (CO < 4 parts per million
tions increased point-prevalence smoking cessa- [ppm]) to a percentile schedule in which the re-
tion early in pregnancy, the percentage of nega- searchers adjusted the criterion for reinforcement
tive samples, and the mean number of consecutive gradually and individually to require progressively
negative samples, compared to the control condi- lower CO values for reinforcement. Overall, the
tion; however, the two interventions did not differ two schedules did not produce differences in the
from each other on those measures. Overall, the overall rate of participants who were abstinent at
revised voucher-based intervention did not appear the end of treatment or in the overall number of
to improve outcomes over the usual voucher-based CO-negative breath samples. There was some evi-
intervention. Neither abstinence reinforcement dence that the percentile schedule was more effec-
intervention significantly increased smoking ces- tive for a subset of hard-to-treat smokers.
sation at the 24-week postpartum assessment after Romanowich and Lamb conducted a series of
the interventions ended. studies to evaluate parametrically different absti-
Similarly, Ledgerwood, Arfken, Petry, and nence reinforcement interventions to promote
Alessi (2014) attempted to improve the effec- smoking cessation in adults who did (Romano-
tiveness of prize-based reinforcement for smok- wich & Lamb, 2014, 2015) and did not (Romano-
ing cessation by increasing the probability of re- wich & Lamb, 2010, 2013) plan to stop smoking.
inforcement early in the treatment period. They One of the studies compared a fixed-reinforcement
randomly assigned cigarette-smoking participants schedule in which the criterion for reinforcement
(N = 81) to receive standard care, traditional prize- was a constant CO value (CO < 3 ppm) to a per-
based reinforcement for smoking cessation, or an centile schedule in which the researchers adjusted
enhanced prize-based reinforcement intervention the criterion for reinforcement gradually and indi-
for smoking cessation in which the probability of vidually to require progressively lower CO values
Behavior Analysis and Treatment of Drug Addiction 501
for reinforcement (Romanowich & Lamb, 2014). intervention, and the $900 voucher intervention
Contrary to expectation, the study did not show produced 36%, 56%, 55%, and 59% drug-negative
an overall benefit of the percentile schedule. An- urine samples, respectively. As in most other stud-
other study compared the framing of incentives ies, none of the groups differed from each other at
as either losses or gains (Romanowich & Lamb, the 12-month follow-up assessment.
2013). Framing incentives as losses may have im-
proved abstinence more than framing incentives Combining Abstinence Reinforcement with Phar-
as gains did, but framing incentives as losses did macotherapy. Researchers have designed sev-
not produce statistically significant increases in eral randomized controlled studies to determine
overall abstinence rates compared to framing in- whether pharmacotherapy can enhance the effects
centives as gains. Another study compared escalat- of an abstinence reinforcement intervention. Two
ing versus descending schedules of reinforcement, studies evaluated the potential benefit of combin-
in which the value of the monetary incentives for ing bupropion with abstinence reinforcement to
smoking cessation (CO < 3 ppm) either increased promote smoking cessation. Tidey, Rohsenow, Ka-
or decreased, respectively, with consecutive visits plan, Swift, and Reid (2011) evaluated the effects
(Romanowich & Lamb, 2010). Missed samples or of bupropion and abstinence reinforcement alone
samples that did not meet the abstinence criterion and in combination over 4 weeks in promoting
did not affect the magnitude of the incentive, as smoking cessation in 57 adults with schizophrenia
they do in typical abstinence reinforcement sched- who were not seeking treatment for smoking ces-
ules. The schedule of reinforcement (escalating sation. Participants in the abstinence reinforce-
vs. descending) did not affect the overall rates of ment intervention earned cash under an escalat-
smoking cessation. Finally, one study compared ing schedule for sustained abstinence for providing
fixed magnitude of reinforcement for abstinence urine samples three times per week showing a 25%
(CO < 3 ppm), escalating magnitude of reinforce- reduction in cotinine levels from the previous
ment for sustained abstinence, and a control con- sample or a value below 80 nanograms per milli-
dition (Romanowich & Lamb, 2015). The two liter (ng/ml; i.e., cotinine-negative). The value of
abstinence reinforcement conditions produced the incentive started at $25 for the first cotinine-
significantly more overall rates of smoking absti- reduced or -negative urine sample and increased
nence than the control group, but did not differ by $5 for each consecutive reduced or negative
from each other on overall smoking abstinence. sample. The study showed that the abstinence re-
The escalating-reinforcement schedule, however, inforcement intervention increased smoking ces-
did produce longer periods of sustained abstinence sation compared to a noncontingent condition,
than the fixed schedule. but that bupropion did not increase smoking ces-
Petry, Alessi, Barry, and Carroll (2015) com- sation compared to placebo and did not increase
pared the effectiveness of different magnitudes of the effects of the abstinence reinforcement inter-
prize and voucher reinforcement in promoting co- vention.
caine abstinence in methadone-treated patients. Gray et al. (2011) evaluated the effects of bupro-
That study randomly assigned 240 methadone- pion and abstinence reinforcement alone and in
maintained patients who used cocaine to a usual- combination over 6 weeks in promoting smoking
care control group or to one of three abstinence cessation in 134 adolescents between the ages of
reinforcement interventions: a $300 prize con- 12 and 21 who were seeking treatment for smoking
tingency management intervention, a $900 prize cessation. Participants in the abstinence reinforce-
contingency management intervention, or a $900 ment intervention could earn cash payments for
voucher reinforcement intervention. All three providing objective evidence of smoking cessation
abstinence reinforcement interventions increased at the end of the first week and then two times
cocaine abstinence (percentage of drug-negative per week for the remaining 5 weeks. At the end of
urine samples and longest duration of abstinence), the first week, CO-negative breath samples (CO
compared to the usual-care control group; howev- ≤ 7 ppm) confirmed smoking cessation. Thereaf-
er, the three interventions did not differ from each ter, abstinence was based on self-reported smok-
other. As in other studies, no abstinence reinforce- ing cessation and urine cotinine (≤100 ng/ml).
ment intervention promoted abstinence in all par- The value of the cash payments started at $10 and
ticipants. The usual-care control intervention, the increased by $3 for each consecutive sample that
$300 prize contingency management interven- met the abstinence criteria. Participants could
tion, the $900 prize contingency management earn $275 in total. Control participants (those not
502 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
assigned to the abstinence reinforcement inter- bining fluoxetine with voucher-based reinforce-
vention) earned $10 for attending each visit. All ment of cocaine abstinence in cocaine-dependent
participants also were randomly assigned to re- adults enrolled in methadone treatment. Par-
ceive either bupropion SR or a placebo. The study ticipants in the voucher intervention condition
failed to show any effect of abstinence reinforce- could earn up to $1,155 in vouchers for providing
ment alone compared to the control condition cocaine-negative urine samples every Monday,
(when combined with placebo) or of bupropion Wednesday, and Friday for 12 weeks. Participants
alone compared to placebo (when combined with were randomly assigned to receive voucher-based
control). Bupropion combined with abstinence reinforcement of cocaine abstinence, fluoxetine,
reinforcement produced higher rates of smoking both voucher-based reinforcement of cocaine ab-
cessation than did placebo combined with absti- stinence and fluoxetine, or neither. Voucher-based
nence reinforcement during Weeks 3 and 4, but reinforcement alone appeared to increase cocaine
not during Weeks 5 and 6. As with other studies, abstinence, but the addition of fluoxetine did not
results did not show any effects of the abstinence increase cocaine abstinence and may have de-
reinforcement intervention at the 12-week follow- creased abstinence. The group receiving placebo
up assessment. plus voucher-based reinforcement of cocaine absti-
Three studies assessed the potential benefit of nence achieved the highest percentage of cocaine-
combining medication with abstinence reinforce- negative urine samples; this group provided 69%
ment to promote cocaine abstinence. Schmitz, cocaine-negative urine samples, compared to 35%
Lindsay, Stotts, Green, and Moeller (2010) com- cocaine-negative urine samples in the placebo
pared levodopa–carbidopa to placebo under three group.
different voucher-based reinforcement conditions: Umbricht et al. (2014) evaluated the effective-
reinforcement for attendance, reinforcement for ness of topiramate and voucher-based abstinence
medication adherence, and reinforcement for ab- reinforcement in promoting cocaine abstinence
stinence. The participants were adults who met in cocaine-dependent methadone-maintained
DSM-IV criteria for cocaine dependence and en- patients. Participants in the voucher intervention
rolled in outpatient treatment. Participants in all could earn up to $1,155 in vouchers for providing
three conditions could earn vouchers every Mon- cocaine-negative urine samples every Monday,
day, Wednesday, and Friday for the assigned target Wednesday, and Friday for 12 weeks. Participants
behavior in their condition (attendance, medica- were randomly assigned to receive voucher-based
tion adherence, or cocaine abstinence, respective- reinforcement of cocaine abstinence, topiramate,
ly). Participants could earn vouchers worth $2.50 both voucher-based reinforcement of cocaine ab-
initially. The value of the vouchers increased by stinence and topiramate, or neither. Although
$1.25 for every consecutive occurrence of the tar- study retention was high, neither topiramate nor
get behavior. Participants also could earn a $10 the voucher intervention affected cocaine use in
bonus for three consecutive occurrences of the this study.
target behavior. The voucher intervention was in
effect for 12 weeks, and participants could earn Group Contingencies. Dallery and colleagues
up to $997.50. Levodopa–carbidopa did not affect conducted a series of studies to evaluate the poten-
cocaine-negative urine samples in participants tial benefit of arranging contingencies for smoking
who earned vouchers for attendance or medica- cessation in which each participant’s earnings were
tion adherence; however, the medication appeared contingent on the behavior of the group (Dallery,
to increase cocaine abstinence relative to placebo Meredith, Jarvis, & Nuzzo, 2015; Meredith & Dal-
in participants who earned vouchers for cocaine- lery, 2013; Meredith, Grabinski, & Dallery, 2011).
negative urine samples. Although these results They used this contingency to recruit social rein-
suggest that levodopa–carbidopa may enhance the forcement and to test whether the combination of
effectiveness of voucher reinforcement in promot- financial and social reinforcement would increase
ing cocaine abstinence, there were high rates of abstinence rates. The researchers gave participants
attrition in this study (only 35% of participants access to online forums and encouraged them to
were still in treatment by the end of the 12 weeks), offer support to their group members to foster
which limit the conclusions we can draw from the communication and social reinforcement.
study. In one study, participants assigned to a no-
Winstanley, Bigelow, Silverman, Johnson, and voucher control condition provided significantly
Strain (2011) evaluated the effectiveness of com- fewer CO-negative breath samples (35%) than
Behavior Analysis and Treatment of Drug Addiction 503
when they were exposed to a condition in which weekly assessments. The abstinence reinforce-
they could earn vouchers for CO-negative breath ment intervention alone did not produce better
samples, independent of whether the voucher re- outcomes than cognitive-behavioral therapy alone
inforcement was independent (56%) or dependent did. Also, cognitive-behavioral therapy did not
(53%) on the CO results of other participants in improve outcomes when added to the abstinence
their group (Meredith & Dallery, 2013). The addi- reinforcement intervention, either during treat-
tion of social contingencies by arranging financial ment or during the year-long follow-up period. In
reinforcement for CO-negative samples based on fact, the addition of cognitive-behavioral therapy
the group results did not increase the percent- may have had a deleterious effect on cannabis use
age of participants who stopped smoking, nor outcomes.
did having access to the online forum. Another Krishnan-Sarin et al. (2013) conducted a
study randomly assigned participants to receive all 4-week randomized controlled trial that compared
vouchers contingent on smoking cessation (CO- cognitive-behavioral therapy alone, abstinence
negative samples) of all members of the group (full reinforcement alone, and cognitive- behavioral
group), or to receive some vouchers contingent therapy plus abstinence reinforcement in 72 high
on each participant’s smoking cessation and some school students who smoked cigarettes. Research-
vouchers contingent on the group’s smoking cessa- ers determined abstinence with CO-negative
tion (mixed group). Arranging all reinforcement breath samples and cotinine-negative urine sam-
contingent on the group’s smoking cessation did ples, daily during Weeks 1 and 2 and every other
not affect the smoking cessation outcomes either day in Weeks 3 and 4. The CO cutoff was <7 ppm.
during the intervention (full group, 18% CO-neg- The cutoff for cotinine initially required a decrease
ative; mixed group, 29% CO-negative) or after it from the last reading or <100 ng/ml and then
ended (full group, 9% CO-negative; mixed group, switched to the fixed requirement of <100 ng/ml.
14% CO-negative). Another study (Halpern et al., During treatment, the groups receiving abstinence
2015), described in detail below, also failed to show reinforcement alone and cognitive-behavioral
any benefit of a group-based incentive interven- therapy plus abstinence reinforcement achieved
tion. higher rates of smoking cessation than the group
receiving cognitive-behavioral therapy alone, but
did not differ from each other. The combination of
Preventing Relapse
cognitive-behavioral therapy with the abstinence
Combining Abstinence Reinforcement with Coun- reinforcement intervention did not increase ab-
seling Interventions. Some researchers have hy- stinence during the follow-up period. In fact, par-
pothesized that combining abstinence reinforce- ticipants exposed to the abstinence reinforcement
ment interventions with cognitive-behavioral intervention alone had the highest rates of post-
relapse prevention therapy might prevent or re- treatment abstinence of all three groups.
duce relapse after the abstinence reinforcement Ling, Hillhouse, Ang, Jenkins, and Fahey
interventions end. Although early studies failed to (2013) randomly assigned 202 opioid-dependent
demonstrate such effects (Silverman, Kaminski, et adults enrolled in treatment with sublingual bu-
al., 2011), several studies published since our earlier prenorphine to four different 16-week treatment
chapter appeared have investigated this possibility. conditions: cognitive-behavioral therapy, prize-
Carroll et al. (2012) randomly assigned 127 adults based reinforcement of opioid abstinence, both
who met DSM-IV criteria for current cannabis cognitive-behavioral therapy and prize-based rein-
dependence to four groups: cognitive-behavioral forcement of opioid abstinence, or neither (medi-
therapy alone, prize-based abstinence reinforce- cal management). The study failed to show any
ment alone, cognitive-behavioral therapy with difference in the rates of opioid abstinence among
reinforcement for treatment adherence (attending any of the conditions, either during treatment or
counseling sessions and completing homework), at the follow-up assessments after the 16-week con-
or cognitive-behavioral therapy plus prize-based ditions ended.
abstinence reinforcement. In all prize-based rein- McKay et al. (2010) randomly assigned 100
forcement interventions, participants could earn adults who met DSM-IV criteria for current co-
up to about $250 in prizes over 12 weeks. In the caine dependence and who maintained 2 weeks of
abstinence reinforcement interventions, partici- regular attendance in intensive outpatient treat-
pants could earn prize draws for providing urine ment to four treatment conditions: treatment as
samples that tested negative for cannabis at 12 usual, voucher-based reinforcement of cocaine
504 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
ers required participants to provide cocaine-free groups in cocaine abstinence during the initial 12
and opiate-free urine samples to gain and maintain weeks, when both groups received voucher-based
access to the workplace. Participants who initiated reinforcement. The participants exposed to the
cocaine and opiate abstinence and acquired need- extended intervention achieved significantly more
ed skills during this initial 6-month phase were cocaine abstinence during the following 24 weeks
hired in a model data-entry business for 1 year and than those in the 12-week program who were no
randomly assigned to an employment-only group longer receiving vouchers. Importantly, the two
or an abstinence-contingent employment group. groups did not differ in cocaine abstinence dur-
As employees, participants could work 30 hours ing a 12-week follow-up period when neither group
per week and were paid biweekly. Participants as- received voucher-based reinforcement.
signed to the employment-only condition could
work regardless of their urinalysis results, as in typ-
Dissemination
ical employment. By contrast, researchers required
participants assigned to abstinence-contingent Dissemination of abstinence reinforcement inter-
employment to provide drug-free urine samples ventions has remained relatively limited since the
in order to work and to maintain maximum pay. publication of our earlier chapter. However, two
Analyses of monthly urine samples collected from large-scale applications deserve mention. First, the
participants in both groups throughout the year U.S. Department of Veterans Affairs (VA) dissem-
of employment in the data-entry business showed inated prize-based abstinence reinforcement as an
that abstinence-contingent-employment partici- evidence-based practice for substance use disorders
pants provided significantly more cocaine-nega- and implemented it throughout the VA treatment
tive samples than employment-only participants. system (Petry, DePhilippis, Rash, Drapkin, &
In a more recent report, DeFulio and Silverman McKay, 2014). The VA implemented prize-based
(2011) reported the abstinence outcomes for the contingency management interventions in over 70
year during and the year after employment in the substance abuse treatment clinics in the United
model data-entry business, based on urine samples States. This program illustrates the adoption and
collected at 6-month intervals. During the year of dissemination of abstinence reinforcement into
employment, abstinence-contingent-employment real-world clinical settings.
participants provided significantly more cocaine- Two studies by Halpern et al. (2015, 2018)
negative urine samples than employment-only represent important large-scale applications of
controls (83% vs. 54%). However, during the incentives in real-world settings. The study by
follow-up year when the employment-based absti- Halpern et al. (2015) evaluated the effective-
nence reinforcement contingency ended, the two ness of financial incentives in promoting smok-
groups provided similar rates of cocaine-negative ing cessation in 2,538 CVS Caremark employees
urine samples (44% vs. 50%). This study showed or their relatives and friends across the United
that researchers could use abstinence reinforce- States. Participants were randomly assigned to a
ment to maintain cocaine abstinence for as long usual-care control group or to one of four absti-
as a year, but that the effects of the abstinence nence reinforcement interventions to promote
reinforcement intervention did not persist after it smoking cessation: an individual-rewards group, a
ended. collaborative-rewards group, an individual-deposit
Three other studies confirmed that researchers group, or a competitive-deposit group. Participants
could extend the effects of an abstinence reinforce- in the collaborative-rewards and the competitive-
ment intervention by extending the duration of the deposit groups were placed in groups of six each
abstinence reinforcement contingencies (Aklin and could communicate with other group mem-
et al., 2014; Kirby et al., 2013; Roll et al., 2013). bers through a web-based chat room. Participants
The study by Kirby et al. (2013) is particularly im- in the individual-deposit and competitive-deposit
portant, because it experimentally evaluated the groups had to submit a $150 deposit to partici-
effect of duration of voucher-based abstinence re- pate in their respective conditions. Participants
inforcement on abstinence outcomes. That study in the two individual-incentives groups could
randomly assigned 130 cocaine-dependent, meth- earn $200 for providing biochemically confirmed
adone-maintained patients to standard 12-week evidence of smoking cessation at 14 days, 30 days,
cocaine abstinence reinforcement or to extended and 6 months, plus a bonus of $200 for evidence
36-week cocaine abstinence reinforcement. There of smoking cessation at 6 months. Participants
was no significant difference between the two in the collaborative-rewards group who provided
506 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
the biochemically confirmed measure of smoking suggests that deposit contracts may be useful for
cessation at the 14-day, 30-day, and 6-month as- a subset of the population. Nevertheless, based on
sessments earned $100 for each of six members in an intent-to-treat analysis, deposit contracts were
the group who met the cessation criterion at that not a particularly useful way to fund abstinence re-
time (i.e., up to $600 per time). In addition, par- inforcement interventions.
ticipants in this condition earned a bonus of $200
for providing evidence of smoking cessation at 6
months. Participants in the competitive-deposit CONCLUSIONS
group could earn up to $1,200 at each time for
providing the biochemically confirmed measure of Reviews, meta-analyses, and detailed summaries
smoking cessation, but the amount earned was di- of individual studies published in the last several
vided by the number of participants who provided years have confirmed that abstinence reinforce-
the biochemically confirmed measure of smoking ment interventions are highly effective in pro-
cessation. moting abstinence from most commonly abused
The study by Halpern et al. (2015) illustrates drugs and in diverse populations. Yet we still must
several key points. First, the individual-rewards improve these interventions. As detailed in this
and collaborative-rewards interventions promoted chapter, studies have not shown consistently that
significantly more smoking cessation than the abstinence reinforcement interventions are effec-
usual-care control group, including the 12-month tive. Studies that showing that these interventions
follow-up assessment that researchers conducted 6 are effective also show that they are not effective
months after the incentive interventions ended. for every participant, and the proportion of treat-
These results confirm the general finding that ab- ment failures is frequently large. Moreover, many
stinence reinforcement interventions can be effec- studies show that the effects of abstinence rein-
tive. Second, although the individual-rewards and forcement interventions do not persist after the
collaborative-rewards interventions were effective interventions end. Even in studies finding that
in promoting smoking cessation, they were effec- these interventions have detectable posttreatment
tive in only about one-quarter of the participants effects, the proportion of patients remaining absti-
at most, and the percentage of participants who nent in the posttreatment period decreases reliably
sustained smoking cessation decreased over time. over time.
This result confirms the general finding that al- Efforts to increase the effectiveness of absti-
though abstinence reinforcement can be effective, nence reinforcement interventions or to produce
many people do not respond to such interven- more lasting effects after the interventions end—
tions. Third, although the rates of smoking ces- by manipulating parameters of the interventions,
sation for participants in the individual-rewards combining the interventions with pharmacother-
and collaborative-rewards groups were still signifi- apies, or combining the interventions with psy-
cantly higher than those for participants in the chosocial treatments—have met with very little
usual-care control group at the 12-month follow- success. As detailed in our earlier chapter (Silver-
up, the percentage of participants who were still man, Kaminski, et al., 2011) and in the current
abstinent in the two rewards groups had decreased one, increasing the magnitude of reinforcement
substantially by that follow-up, and fewer than 9% can substantially increase the effectiveness of ab-
of participants in those groups were still abstinent stinence reinforcement interventions, and arrang-
at that time. This result confirms the finding that ing extended exposure to such interventions can
even when abstinence reinforcement interven- prevent relapse and promote long-term abstinence,
tions produce long-term effects after the interven- at least while the intervention continues.
tions end, many people continue to relapse into The limitations of abstinence reinforcement
drug use in the subsequent weeks and months. interventions identified in this chapter raise ques-
Finally, although researchers have viewed deposits tions about their clinical utility: What proportion
as a potentially viable way to fund incentive inter- of patients fail to respond to these interventions?
ventions, fewer than 20% of participants agreed to Do such interventions serve a clinically useful role
pay deposits, and the overall rates of smoking ces- if they do not maintain abstinence over time? Not
sation were low and generally not significantly dif- much has changed in our understanding of the
ferent from those for the usual-care control group. benefits and limitations of abstinence reinforce-
A secondary analysis showed that those who paid ment interventions in recent years. Overall, they
a deposit were likely to become abstinent, which can be effective in promoting abstinence in some
Behavior Analysis and Treatment of Drug Addiction 507
patients, but they fail to address the needs of a Silverman, Holtyn, & Morrison, 2016), but other
substantial proportion of patients, and they almost researchers have not adopted this method. Physi-
always fail to promote abstinence that does not re- cian health programs require impaired physicians
verse when the intervention ends. to undergo random drug testing for 5 years and
Given that the costs of abstinence reinforce- stay abstinent throughout that time to continue
ment interventions can be minimal in patients practicing medicine. This model shows some real
who do not initiate abstinence, the failure to initi- promise in addressing the chronic nature of drug
ate abstinence in some patients is not a great cost addiction (DuPont, McLellan, White, Merlo, &
to treatment providers and does not diminish the Gold, 2009). Abstinence reinforcement interven-
value of these interventions to patients who do tions have great potential and could prove most
respond. Finding ways to increase the proportion useful in addressing the chronic, persistent prob-
of patients who respond to these interventions is lem of drug addiction if researchers can develop
an important focus for the scientific community, effective and practical methods that ensure long-
but in the absence of those improvements, exist- term and lifelong abstinence ultimately.
ing abstinence reinforcement interventions can
still have value.
In some applications, abstinence reinforcement ACKNOWLEDGMENTS
interventions that promote short-term abstinence
can have profound effects on health (e.g., Higgins Joseph V. Brady contributed to the version of this chap-
et al., 2010). Except in those instances in which ter in the first edition of the Handbook of Applied Be-
havior Analysis.
short-term exposure to an intervention has clear
The National Institute on Drug Abuse and the
health benefits, it is not clear that abstinence re-
National Institute of Allergy and Infectious Diseases
inforcement interventions are clinically useful if of the National Institutes of Health, under Award
they provide only a short-term break in drug use. Nos. R01DA037314, R01DA019497, R01AI117065, and
Many view drug addiction as a chronic problem T32DA07209, supported preparation of this chapter.
(McLellan et al., 2000). Relapse to drug use is The content is solely our responsibility and does not
common after most treatments end (McLellan et necessarily represent the official views of the National
al., 2000), and abstinence reinforcement is obvi- Institutes of Health.
ously no exception to this pattern. Recognizing
the chronic relapsing nature of drug addiction, Du-
Pont, Compton, and McLellan (2015) have recom- REFERENCES
mended a 5-year recovery as a standard for assess-
ing treatments for drug addiction for researchers Aklin, W. M., Wong, C. J., Hampton, J., Svikis, D. S.,
and clinicians. A 5-year evaluation period would Stitzer, M. L., Bigelow, G. E., et al. 2014). A thera-
peutic workplace for the long-term treatment of drug
obviously place sizable financial and practical de-
addiction and unemployment: Eight-year outcomes
mands on researchers and funding agencies to ar-
of a social business intervention. Journal of Substance
range and support such long-term studies. Typical Abuse Treatment, 47, 329–338.
studies apply interventions that last 12 weeks and Alessi, S. M., & Petry, N. M. (2013). A randomized
conduct 6- and 12-month follow-up assessments study of cellphone technology to reinforce alcohol
to examine the long-term effects of the interven- abstinence in the natural environment. Addiction,
tion. Thus a 5-year evaluation period represents a 108, 900–909.
significant challenge to the scientific community. Alessi, S. M., & Petry, N. M. (2014). Smoking reduc-
At this point, arranging abstinence reinforce- tions and increased self-efficacy in a randomized
ment as a maintenance intervention appears to be controlled trial of smoking abstinence-contingent
the most likely means of maintaining long-term incentives in residential substance abuse treatment
abstinence (DeFulio et al., 2009; Kirby et al., 2013; patients. Nicotine and Tobacco Research, 16, 1436–
1445.
Silverman, DeFulio, & Sigurdsson, 2012; Silver-
Barnett, N. P., Tidey, J., Murphy, J. G., Swift, R., &
man et al., 2004). Researchers have not identified
Colby, S. M. (2011). Contingency management for
practical methods of arranging long-term exposure alcohol use reduction: A pilot study using a trans-
to abstinence reinforcement interventions, how- dermal alcohol sensor. Drug and Alcohol Dependence,
ever. Silverman and colleagues have attempted to 118, 391–399.
use employment as a vehicle for maintaining long- Benishek, L. A., Dugosh, K. L., Kirby, K. C., Matejkows-
term employment-based abstinence reinforce- ki, J., Clements, N. T., Seymour, B. L., et al. (2014).
ment (Silverman, 2004; Silverman et al., 2012; Prize-based contingency management for the treat-
508 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
ment of substance abusers: A meta-analysis. Addic- ment as a maintenance intervention for the treat-
tion, 109, 1426–1436. ment of cocaine dependence: A randomized con-
Businelle, M. S., Kendzor, D. E., Kesh, A., Cuate, E. L., trolled trial. Addiction, 104, 1530–1538.
Poonawalla, I. B., Reitzel, L. R., et al. (2014). Small DeFulio, A., & Silverman, K. (2011). Employment-based
financial incentives increase smoking cessation in abstinence reinforcement as a maintenance interven-
homeless smokers: A pilot study. Addictive Behaviors, tion for the treatment of cocaine dependence: Post-
39, 717–720. intervention outcomes. Addiction, 106, 960–967.
Cahill, K., Hartmann-Boyce, J., & Perera, R. (2015). In- Dougherty, D. M., Hill-Kapturczak, N., Liang, Y., Karns,
centives for smoking cessation. Cochrane Database of T. E., Cates, S. E., Lake, S. L., et al. (2014). Use of
Systematic Reviews, 5, CD004307. continuous transdermal alcohol monitoring during a
Carroll, K. M., Nich, C., Lapaglia, D. M., Peters, E. N., contingency management procedure to reduce exces-
Easton, C. J., & Petry, N. M. (2012). Combining cog- sive alcohol use. Drug and Alcohol Dependence, 142,
nitive behavioral therapy and contingency manage- 301–306.
ment to enhance their effects in treating cannabis Dougherty, D. M., Karns, T. E., Mullen, J., Liang, Y.,
dependence: Less can be more, more or less. Addic- Lake, S. L., Roache, J. D., et al. (2015). Transder-
tion, 107, 1650–1659. mal alcohol concentration data collected during a
Castells, X., Kosten, T. R., Capella, D., Vidal, X., Colom, contingency management program to reduce at-risk
J., & Casas, M. (2009). Efficacy of opiate mainte- drinking. Drug and Alcohol Dependence, 148, 77–84.
nance therapy and adjunctive interventions for opi- Dougherty, D. M., Lake, S. L., Hill-Kapturczak, N.,
oid dependence with comorbid cocaine use disorders: Liang, Y., Karns, T. E., Mullen, J., et al. (2015). Using
A systematic review and meta-analysis of controlled contingency management procedures to reduce at-
clinical trials. American Journal of Drug and Alcohol risk drinking in heavy drinkers. Alcoholism: Clinical
Abuse, 35, 339–349. and Experimental Research, 39, 743–751.
Chen, W., Hong, Y., Zou, X., McLaughlin, M. M., Xia, Drummond, M. B., Astemborski, J., Lambert, A. A.,
Y., & Ling, L. (2013). Effectiveness of prize-based Goldberg, S., Stitzer, M. L., Merlo, C. A., et al.
contingency management in a methadone mainte- (2014). A randomized study of contingency manage-
nance program in China. Drug and Alcohol Depen- ment and spirometric lung age for motivating smok-
dence, 133, 270–274. ing cessation among injection drug users. BMC Pub-
Christensen, D. R., Landes, R. D., Jackson, L., Marsch, lic Health, 14, 761.
L. A., Mancino, M. J., Chopra, M. P., et al. (2014). Dunn, K. E., Saulsgiver, K. A., & Sigmon, S. C. (2011).
Adding an Internet-delivered treatment to an ef- Contingency management for behavior change: Ap-
ficacious treatment package for opioid dependence. plications to promote brief smoking cessation among
Journal of Consulting and Clinical Psychology, 82, opioid-maintained patients. Experimental and Clini-
964–972. cal Psychopharmacology, 19, 20–30.
Dallery, J., & Glenn, I. M. (2005). Effects of an Internet- DuPont, R. L., Compton, W. M., & McLellan A. T.
based voucher reinforcement program for smoking (2015). Five-year recovery: A new standard for assess-
abstinence: A feasibility study. Journal of Applied Be- ing effectiveness of substance use disorder treatment.
havior Analysis, 38, 349–357. Journal of Substance Abuse Treatment, 58, 1–5.
Dallery, J., Meredith, S., Jarvis, B. P., & Nuzzo, P. A. DuPont, R. L., McLellan, A. T., White, W. L., Merlo,
(2015). Internet-based group contingency manage- L. J., & Gold, M. S. (2009). Setting the standard for
ment to promote smoking abstinence. Experimental recovery: Physicians’ health programs. Journal of Sub-
and Clinical Psychopharmacology, 23, 176–183. stance Abuse Treatment, 36, 159–171.
Dallery, J., Raiff, B. R., & Grabinski, M. J. (2013). In- Dutra, L., Stathopoulou, G., Basden, S. L., Leyro, T.
ternet-based contingency management to promote M., Powers, M. B., & Otto, M. W. (2008). A meta-
smoking cessation: A randomized controlled study. analytic review of psychosocial interventions for sub-
Journal of Applied Behavior Analysis, 46, 750–764. stance use disorders. American Journal of Psychiatry,
Dallery, J., Raiff, B. R., Kim, S. J., Marsch, L. A., Stitzer, 165, 179–187.
M., & Grabinski, M. J. (2017). Nationwide access to Elliott, R., & Tighe, T. (1968). Breaking the cigarette
an Internet-based contingency management inter- habit: Effects of a technique involving threatened
vention to promote smoking cessation: A random- loss of money. Psychological Record, 18, 503–513.
ized controlled trial. Addiction, 112, 875–883. Festinger, D. S., Dugosh, K. L., Kirby, K. C., & Seymour,
Davis, D. R., Kurti, A. N., Skelly, J. M., Redner, R., B. L. (2014). Contingency management for cocaine
White, T. J., & Higgins, S. T. (2016). A review of the treatment: Cash vs. vouchers. Journal of Substance
literature on contingency management in the treat- Abuse Treatment, 47, 168–174.
ment of substance use disorders, 2009–2014. Preven- Glasgow, R. E., Hollis, J. F., Ary, D. V., & Boles, S. M.
tive Medicine, 92, 36–46. (1993). Results of a year-long incentives-based work-
DeFulio, A., Donlin, W. D., Wong, C. J., & Silverman, site smoking-cessation program. Addictive Behaviors,
K. (2009). Employment-based abstinence reinforce- 18, 455–464.
Behavior Analysis and Treatment of Drug Addiction 509
Glover, M., Kira, A., Walker, N., & Bauld, L. (2015). et al. (2011). Effects of a randomized contingency
Using incentives to encourage smoking abstinence management intervention on opiate abstinence and
among pregnant Indigenous women?: A feasibility retention in methadone maintenance treatment in
study. Maternal and Child Health Journal, 19, 1393– China. Addiction, 106, 1801–1809.
1399. Jiang, H., Du, J., Wu, F., Wang, Z., Fan, S., Li, Z., et
Godley, M. D., Godley, S. H., Dennis, M. L., Funk, R. al. (2012). Efficacy of contingency management in
R., Passetti, L. L., & Petry, N. M. (2014). A random- improving retention and compliance to methadone
ized trial of assertive continuing care and contingen- maintenance treatment: A random controlled study.
cy management for adolescents with substance use Shanghai Archives of Psychiatry, 24, 11–19.
disorders. Journal of Consulting and Clinical Psychol- Kendzor, D. E., Businelle, M. S., Poonawalla, I. B.,
ogy, 82, 40–51. Cuate, E. L., Kesh, A., Rios, D. M., et al. (2015). Fi-
Gray, K. M., Carpenter, M. J., Baker, N. L., Hartwell, K. nancial incentives for abstinence among socioeco-
J., Lewis, A. L., Hiott, D. W., et al. (2011). Bupropion nomically disadvantaged individuals in smoking ces-
SR and contingency management for adolescent sation treatment. American Journal of Public Health,
smoking cessation. Journal of Substance Abuse Treat- 105, 1198–1205.
ment, 40, 77–86. Kennedy, A. P., Phillips, K. A., Epstein, D. H., Reamer,
Halpern, S. D., French, B., Small, D. S., Saulsgiver, K., D. A., Schmittner, J., & Preston, K. L. (2013). A ran-
Harhay, M. O., Audrain-McGovern, J., et al. (2015). domized investigation of methadone doses at or over
Randomized trial of four financial-incentive pro- 100 mg/day, combined with contingency manage-
grams for smoking cessation. New England Journal of ment. Drug and Alcohol Dependence, 130, 77–84.
Medicine, 372, 2108–2117. Killeen, T. K., McRae-Clark, A. L., Waldrop, A. E.,
Halpern, S. D., Harhay, M. O., Saulsgiver, K., Brophy, Upadhyaya, H., & Brady, K. T. (2012). Contingency
C., Troxel, A. B., & Volpp, K. G. (2018). A pragmatic management in community programs treating ado-
trial of e-cigarettes, incentives, and drugs for smok- lescent substance abuse: A feasibility study. Journal
ing cessation. New England Journal of Medicine, 378, of Child and Adolescent Psychiatric Nursing, 25, 33–41.
2302–2310. Kirby, K. C., Carpenedo, C. M., Dugosh, K. L., Rosen-
Heil, S. H., Higgins, S. T., Bernstein, I. M., Solomon, wasser, B. J., Benishek, L. A., Janik, A., et al. (2013).
L. J., Rogers, R. E., Thomas, C. S., et al. (2008). Ef- Randomized clinical trial examining duration of
fects of voucher-based incentives on abstinence from voucher-based reinforcement therapy for cocaine ab-
cigarette smoking and fetal growth among pregnant stinence. Drug and Alcohol Dependence, 132, 639–645.
women. Addiction, 103, 1009–1018. Krishnan-Sarin, S., Cavallo, D. A., Cooney, J. L., Sche-
Higgins, S. T., Bernstein, I. M., Washio, Y., Heil, S. H., pis, T. S., Kong, G., Liss, T. B., et al. (2013). An
Badger, G. J., Skelly, J. M., et al. (2010). Effects of exploratory randomized controlled trial of a novel
smoking cessation with voucher-based contingency high-school-based smoking cessation intervention
management on birth outcomes. Addiction, 105, for adolescent smokers using abstinence-contingent
2023–2030. incentives and cognitive behavioral therapy. Drug
Higgins, S. T., Delaney, D. D., Budney, A. J., Bickel, W. and Alcohol Dependence, 132, 346–351.
K., Hughes, J. R., Foerg, F., et al. (1991). A behavioral Lamb, R. J., Kirby, K. C., Morral, A. R., Galbicka, G., &
approach to achieving initial cocaine abstinence. Iguchi, M. Y. (2010). Shaping smoking cessation in
American Journal of Psychiatry, 148, 1218–1224. hard-to-treat smokers. Journal of Consulting and Clini-
Higgins, S. T., Washio, Y., Lopez, A. A., Heil, S. H., cal Psychology, 78, 62–71.
Solomon, L. J., Lynch, M. E., et al. (2014). Examin- Ledgerwood, D. M., Arfken, C. L., Petry, N. M., &
ing two different schedules of financial incentives for Alessi, S. M. (2014). Prize contingency management
smoking cessation among pregnant women. Preven- for smoking cessation: A randomized trial. Drug and
tive Medicine, 68, 51–57. Alcohol Dependence, 140, 208–212.
Holtyn, A. F., Koffarnus, M. N., DeFulio, A., Sigurdsson, Ling, W., Hillhouse, M., Ang, A., Jenkins, J., & Fahey,
S. O., Strain, E. C., Schwartz, R. P., et al. (2014a). J. (2013). Comparison of behavioral treatment condi-
The therapeutic workplace to promote treatment tions in buprenorphine maintenance. Addiction, 108,
engagement and drug abstinence in out-of-treatment 1788–1798.
injection drug users: A randomized controlled trial. Litt, M. D., Kadden, R. M., & Petry, N. M. (2013).
Preventive Medicine, 68, 62–70. Behavioral treatment for marijuana dependence:
Holtyn, A. F., Koffarnus, M. N., DeFulio, A., Sig- Randomized trial of contingency management and
urdsson, S. O., Strain, E. C., Schwartz, R. P., et al. self-efficacy enhancement. Addictive Behaviors, 38,
(2014b). Employment-based abstinence reinforce- 1764–1775.
ment promotes opiate and cocaine abstinence in out- Lussier, J. P., Heil, S. H., Mongeon, J. A., Badger, G. J.,
of-treatment injection drug users. Journal of Applied & Higgins, S. T. (2006). A meta-analysis of voucher-
Behavior Analysis, 47, 681–693. based reinforcement therapy for substance use disor-
Hser, Y. I., Li, J., Jiang, H., Zhang, R., Du, J., Zhang, C., ders. Addiction, 101, 192–203.
510 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
McDonell, M. G., Howell, D. N., McPherson, S., Cam- Petry, N. M., Alessi, S. M., Marx, J., Austin, M., & Tar-
eron, J. M., Srebnik, D., Roll, J. M., et al. (2012). dif, M. (2005). Vouchers versus prizes: Contingency
Voucher-based reinforcement for alcohol abstinence management treatment of substance abusers in com-
using the ethyl-glucuronide alcohol biomarker. Jour- munity settings. Journal of Consulting and Clinical
nal of Applied Behavior Analysis, 45, 161–165. Psychology, 73, 1005–1014.
McDonell, M. G., Srebnik, D., Angelo, F., McPherson, Petry, N. M., Alessi, S. M., & Rash, C. J. (2013). A
S., Lowe, J. M., Sugar, A., et al. (2013). Randomized randomized study of contingency management in
controlled trial of contingency management for stim- cocaine-dependent patients with severe and persis-
ulant use in community mental health patients with tent mental health disorders. Drug and Alcohol De-
serious mental illness. American Journal of Psychiatry, pendence, 130, 234–237.
170, 94–101. Petry, N. M., DePhilippis, D., Rash, C. J., Drapkin, M.,
McKay, J. R., Lynch, K. G., Coviello, D., Morrison, R., & McKay, J. R. (2014). Nationwide dissemination of
Cary, M. S., Skalina, L., et al. (2010). Randomized contingency management: The Veterans Adminis-
trial of continuing care enhancements for cocaine- tration initiative. American Journal on Addictions,
dependent patients following initial engagement. 23, 205–210.
Journal of Consulting and Clinical Psychology, 78, Petry, N. M., Martin, B., Cooney, J. L., & Kranzler, H.
111–120. R. (2000). Give them prizes, and they will come:
McLellan, A. T., Lewis, D. C., O’Brien, C. P., & Kleber, Contingency management for treatment of alcohol
H. D. (2000). Drug dependence, a chronic medical dependence. Journal of Consulting and Clinical Psy-
illness: Implications for treatment, insurance, and chology, 68, 250–257.
outcomes evaluation. Journal of the American Medical Petry, N. M., Tedford, J., Austin, M., Nich, C., Carroll,
Association, 284, 1689–1695. K. M., & Rounsaville, B. J. (2004). Prize reinforce-
Menza, T. W., Jameson, D. R., Hughes, J. P., Colfax, G. ment contingency management for treating cocaine
N., Shoptaw, S., & Golden, M. R. (2010). Contin- users: How low can we go, and with whom? Addic-
gency management to reduce methamphetamine use tion, 99, 349–360.
and sexual risk among men who have sex with men: Petry, N. M., Weinstock, J., & Alessi, S. M. (2011). A
A randomized controlled trial. BMC Public Health, randomized trial of contingency management deliv-
10, 774. ered in the context of group counseling. Journal of
Meredith, S. E., & Dallery, J. (2013). Investigating group Consulting and Clinical Psychology, 79, 686–696.
contingencies to promote brief abstinence from ciga- Petry, N. M., Weinstock, J., Alessi, S. M., Lewis, M. W.,
rette smoking. Experimental and Clinical Psychophar- & Dieckhaus, K. (2010). Group-based randomized
macology, 21, 144–154. trial of contingencies for health and abstinence in
Meredith, S. E., Grabinski, M. J., & Dallery, J. (2011). HIV patients. Journal of Consulting and Clinical Psy-
Internet-based group contingency management to chology, 78, 89–97.
promote abstinence from cigarette smoking: A fea- Pilling, S., Strang, J., Gerada, C., & National Institute
sibility study. Drug and Alcohol Dependence, 118, for Clinical Excellence (NICE). (2007). Psychoso-
23–30. cial interventions and opioid detoxification for drug
Milby, J. B., Schumacher, J. E., Raczynski, J. M., misuse: Summary of NICE guidance. British Medical
Caldwell, E., Engle, M., Michael, M., et al. (1996). Journal, 335, 203–205.
Sufficient conditions for effective treatment of sub- Preston, K. L., Umbricht, A., & Epstein, D. H. (2002).
stance abusing homeless persons. Drug and Alcohol Abstinence reinforcement maintenance contingen-
Dependence, 43, 39–47. cy and one-year follow-up. Drug and Alcohol Depen-
Peirce, J. M., Petry, N. M., Stitzer, M. L., Blaine, J., Kel- dence, 67, 125–137.
logg, S., Satterfield, F., et al. (2006). Effects of lower- Ries, R. K., Dyck, D. G., Short, R., Srebnik, D., Fisher,
cost incentives on stimulant abstinence in metha- A., & Comtois, K. A. (2004). Outcomes of manag-
done maintenance treatment: A national drug abuse ing disability benefits among patients with substance
treatment clinical trials network study. Archives of dependence and severe mental illness. Psychiatric
General Psychiatry, 63, 201–208. Services, 55, 445–447.
Petitjean, S. A., Dursteler-MacFarland, K. M., Krokar, Rohsenow, D. J., Tidey, J. W., Martin, R. A., Colby, S.
M. C., Strasser, J., Mueller, S. E., Degen, B., et al. M., Sirota, A. D., Swift, R. M., et al. (2015). Con-
(2014). A randomized, controlled trial of combined tingent vouchers and motivational interviewing
cognitive-behavioral therapy plus prize-based con- for cigarette smokers in residential substance abuse
tingency management for cocaine dependence. Drug treatment. Journal of Substance Abuse Treatment, 55,
and Alcohol Dependence, 145, 94–100. 29–38.
Petry, N. M., Alessi, S. M., Barry, D., & Carroll, K. M. Roll, J. M., Chudzynski, J., Cameron, J. M., Howell,
(2015). Standard magnitude prize reinforcers can D. N., & McPherson, S. (2013). Duration effects in
be as efficacious as larger magnitude reinforcers in contingency management treatment of metham
cocaine-dependent methadone patients. Journal of phetamine disorders. Addictive Behaviors, 38, 2455–
Consulting and Clinical Psychology, 83, 464–472. 2462.
Behavior Analysis and Treatment of Drug Addiction 511
Romanowich, P., & Lamb, R. J. (2010). Effects of es- Silverman, K., DeFulio, A., & Sigurdsson, S. O. (2012).
calating and descending schedules of incentives on Maintenance of reinforcement to address the chron-
cigarette smoking in smokers without plans to quit. ic nature of drug addiction. Preventive Medicine, 55,
Journal of Applied Behavior Analysis, 43, 357–367. 46–53.
Romanowich, P., & Lamb, R. J. (2013). The effect of Silverman, K., Holtyn, A. F., & Morrison, R. (2016).
framing incentives as either losses or gains with con- The therapeutic utility of employment in treating
tingency management for smoking cessation. Addic- drug addiction: Science to application. Translational
tive Behaviors, 38, 2084–2088. Issues in Psychological Science, 2, 203–212.
Romanowich, P., & Lamb, R. J. (2014). The effects of Silverman, K., Kaminski, B. J., Higgins, S. T., & Brady,
percentile versus fixed criterion schedules on smok- J. V. (2011). Behavior analysis and treatment of drug
ing with equal incentive magnitude for initial absti- addiction. In W. W. Fisher, C. C. Piazza, & H. S.
nence. Experimental and Clinical Psychopharmacol- Roane (Eds.), Handbook of applied behavior analysis
ogy, 22, 348–355. (pp. 451–471). New York: Guilford Press.
Romanowich, P., & Lamb, R. J. (2015). The effects of Silverman, K., Robles, E., Mudric, T., Bigelow, G.
fixed versus escalating reinforcement schedules on E., & Stitzer, M. L. (2004). A randomized trial of
smoking abstinence. Journal of Applied Behavior long-term reinforcement of cocaine abstinence in
Analysis, 48, 25–37. methadone-maintained patients who inject drugs.
Schmitz, J. M., Lindsay, J. A., Stotts, A. L., Green, C. E., Journal of Consulting and Clinical Psychology, 72,
& Moeller, F. G. (2010). Contingency management 839–854.
and levodopa–carbidopa for cocaine treatment: A Stanger, C., Ryan, S. R., Scherer, E. A., Norton, G. E.,
comparison of three behavioral targets. Experimental & Budney, A. J. (2015). Clinic- and home-based con-
and Clinical Psychopharmacology, 18, 238–244. tingency management plus parent training for ado-
Schottenfeld, R. S., Moore, B., & Pantalon, M. V. lescent cannabis use disorders. Journal of the Ameri-
(2011). Contingency management with community can Academy of Child and Adolescent Psychiatry, 54,
reinforcement approach or Twelve-Step facilitation 445–453.
drug counseling for cocaine dependent pregnant Tevyaw, T. O., Colby, S. M., Tidey, J. W., Kahler, C. W.,
women or women with young children. Drug and Al- Rohsenow, D. J., Barnett, N. P., et al. (2009). Contin-
cohol Dependence, 118, 48–55. gency management and motivational enhancement:
Secades-Villa, R., Garcia-Rodriguez, O., Lopez-Nunez, A randomized clinical trial for college student smok-
C., Alonso-Perez, F., & Fernandez-Hermida, J. R. ers. Nicotine and Tobacco Research, 11, 739–749.
(2014). Contingency management for smoking cessa- Tidey, J. W., Rohsenow, D. J., Kaplan, G. B., Swift, R.
tion among treatment-seeking patients in a communi- M., & Reid, N. (2011). Effects of contingency man-
ty setting. Drug and Alcohol Dependence, 140, 63–68. agement and bupropion on cigarette smoking in
Silverman, K. (2004). Exploring the limits and utility of smokers with schizophrenia. Psychopharmacology,
operant conditioning in the treatment of drug addic- 217, 279–287.
tion. Behavior Analyst, 27, 209–230. Umbricht, A., DeFulio, A., Winstanley, E. L., Tompkins,
Silverman, K., Chutuape, M. A., Bigelow, G. E., & D. A., Peirce, J., Mintzer, M. Z., et al. (2014). Topi-
Stitzer, M. L. (1999). Voucher-based reinforcement ramate for cocaine dependence during methadone
of cocaine abstinence in treatment-resistant metha- maintenance treatment: A randomized controlled
done patients: Effects of reinforcement magnitude. trial. Drug and Alcohol Dependence, 140, 92–100.
Psychopharmacology, 146, 128–138. Winstanley, E. L., Bigelow, G. E., Silverman, K., John-
Silverman, K., DeFulio, A., & Everly, J. J. (2011). Be- son, R. E., & Strain, E. C. (2011). A randomized
havioral aspects. In P. Ruiz & E. C. Strain (Eds.), controlled trial of fluoxetine in the treatment of
Lowinson and Ruiz’s substance abuse: A comprehensive cocaine dependence among methadone-maintained
textbook (5th ed, pp. 88–106). Philadelphia: Lippin- patients. Journal of Substance Abuse Treatment, 40,
cott Williams & Wilkins. 255–264.
CHAP TER 30
Behavioral Gerontology
Global and national demographics have shifted declines in skills observed in older adults are re-
steadily toward an older population, with adults 65 versible (Baltes & Barton, 1977). We use the term
years or older projected to account for 20% of the behavioral gerontology to refer to the application
total U.S. population by the year 2030 and 24% by of behavior analysis to older adults in areas such
2060 (Colby & Ortman, 2014). The group ages 85 as basic behavioral research, clinical applications,
and older is growing faster than any other group, and organizational issues in service delivery agen-
with a disproportionate increase for women and cies (Adkins & Mathews, 1999; Burgio & Burgio,
minority groups (Colby & Ortman, 2014). These 1986).
adults will need a range of medical and psycho- Since the mid-1980s, a small, stable number of
logical services, because they are likely to experi- publications on aging have appeared in behavioral
ence declining health, sensory deficits (e.g., hear- journals (e.g., Baker & LeBlanc, 2011; Buchan-
ing loss), and cognitive impairments that will be an, Husfeldt, Berg, & Houlihan, 2011; Oleson &
costly, debilitating, and potentially socially isolat- Baker, 2014; Pachis & Zonneveld, 2018; Raetz,
ing (Belsky, 1999). However, the infrastructure for LeBlanc, Baker, & Hilton, 2013; Trahan, Donald-
providing these services is projected to be woefully son, McNabney, & Kahng, 2014; Virues-Ortega,
inadequate as these individuals age and demand Iwata, Nogales-González, & Frades, 2012), and
alternatives to traditional nursing homes (Moli- a growing number of behavior-analytic studies
nari et al., 2003). have appeared in multidisciplinary aging journals
Due to the growing discrepancy between needs (e.g., Altus, Engelman, & Mathews, 2002a, 2002b;
and infrastructure, enormous research and prac- Burgio et al., 2002; Feliciano, LeBlanc, & Feeney,
tice opportunities exist for behavior analysts in- 2010; Hussian & Brown, 1987; Ilem, Feliciano, &
terested in working with older adults. Behavior LeBlanc, 2015; Noguchi, Kawano, & Yamanaka,
analysts have long advocated the use of environ- 2013). Despite the enormous potential for positive
mental modifications to enhance the lives of older social impact, however, the field of behavioral ger-
adults (Lindsley, 1964). Prominent behavior ana- ontology has not grown at a rate commensurate
lysts have suggested that (1) natural contingencies with its potential (Burgio & Burgio, 1986).
for older adults support ineffective behavior (Skin- Behavior analysts who publish in multidisci-
ner, 1983), (2) basic operant principles account for plinary aging journals have attempted to illustrate
aging-related phenomena, and (3) many of the the advantages of a behavior-analytic approach to
512
Behavioral Gerontology 513
age-related differences with trace conditioning but human participants showed not only that the per-
not with delay conditioning. Second, Graves and formance of older participants was not like that of
Solomon noted that increases in the complexity younger participants, but also that the combina-
of procedures are related inversely to the acquisi- tion resulted in almost no conditioned responses
tion of conditioned responses for older organisms. (Cheng et al., 2010). These results highlight that
Research on trace- and delay-conditioning proce- specific preparations can have an impact on re-
dures have shown that the trace procedure, where sponding in older adults. Respondent conditioning
there is a gap between the end of the presentation involves stimulus–stimulus pairing. The results of
of the conditioned stimulus and the beginning of these studies indicate that researchers examining
the unconditioned stimulus, is a more complex any procedures using stimulus–stimulus pairing
procedure than the delay procedure, where the with older adults need to be careful about draw-
two stimuli overlap temporally and can affect the ing conclusions without a complete understanding
acquisition of conditioned responses. of how those preparations can affect responding.
Graves and Solomon (1985) examined the dif- These results also suggest that simple procedural
ferences in trace and delay conditioning on the ac- variations could interact with age-related changes
quisition of conditioned responses in 6-month-old in performance and could affect findings from
(young) and 36- to 60-month-old (old) albino rab- clinical approaches that seek to use stimulus–
bits. They defined acquisition of the conditioned stimulus pairing, such as establishing conditioned
response as those responses occurring during 4 reinforcement.
trials of a 5-trial block or conditioned responses Researchers have evaluated differences among
occurring during 8 trials of a 10-trial block. The younger adults, older adults, and older adults with
older rabbits acquired the conditioned response neurocognitive disorders (such as neurocognitive
between a mean of 700 and 800 trials, and the disorder due to Alzheimer’s disease). As noted
younger rabbits acquired the conditioned response above, changes in classically conditioned response
between a mean of 430 and 470 trials in trace con- acquisition appear in later middle age (typically
ditioning, but the investigators observed no differ- early 40s to early 60s) and progress into old age
ences during delay conditioning. Although this (over age 65) (Finkbiner & Woodruff-Pak, 1991;
study did not differentiate between age-related Woodruff-Pak & Jaeger, 1998; Woodruff-Pak &
differences associated with hippocampus damage Thompson, 1988), with even greater changes ob-
versus complexity, it showed that a simple gap be- served for individuals with neurocognitive disor-
tween the presentation of the conditioned stimu- ders. Acquisition of conditioned eyeblink respons-
lus and the presentation of the unconditioned es and overall percentage of conditioned responses
stimulus resulted in older rabbits’ requiring twice across trial blocks reliably differentiate typically
as many trials to acquire the response as when aging individuals from those with vascular neuro-
the two stimuli overlapped temporally. Other re- cognitive disorder and those with probable neuro-
searchers have demonstrated similar differences in cognitive disorder due to Alzheimer’s (Woodruff-
trace but not delay conditioning in older humans Pak, 2001; Woodruff-Pak, Papka, Romano, & Li,
(e.g., Finkbiner & Woodruff-Pak, 1991), once 1996).
again highlighting how a simple procedural differ- Taken together, the research described above
ence can affect older but not younger organisms’ supports that older adults acquire fewer condi-
acquisition of conditioned responses. tioned responses or engage in fewer conditioned
Younger humans have acquired conditioned responses across trial blocks than younger adults
responses when researchers combine delay and do, and that older adults with neurocognitive dis-
trace conditioning, which has led researchers to order acquire, engage in, or acquire and engage in
evaluate whether the same would be true for older even fewer conditioned responses than healthy
humans. The question is whether interspersing older adults. Researchers have posited that the
delay conditioning with trace conditioning can differences in respondent conditioning between
improve the acquisition of conditioned responses younger and older adults are related to changes
for older organisms, so that it is more like that of in brain structure (Cheng et al., 2010; Woodruff-
younger organisms (Cheng, Faulkner, Disterhoft, Pak, 2001), particularly cerebellar cortical atrophy,
& Desmond, 2010). Researchers using a combina- rather than to the mere passage of time. Further-
tion of trace and delay conditioning to evaluate more, researchers have hypothesized that disease-
the acquisition of conditioned responses of young- related changes in the same cerebellar structure
er (i.e., 20–25 years) and older (i.e., 60–68 years) may be responsible for differences between healthy
Behavioral Gerontology 515
older adults and older adults with neurocogni- ules, and found that older adults’ behavior did not
tive disorders (Cheng et al., 2010; Woodruff-Pak, respond to schedules of reinforcement, regard-
2001). Researchers have found that repeated expo- less of exposure. Plaud, Plaud, and von Duvillard
sure to conditioning procedures improves the per- (1999) manipulated reinforcement magnitude on
formance of older adults. For example, Graves and a computer-based task and found that older adults
Solomon (1985) showed that older adults could ac- altered their responding accordingly, indicating
quire conditioned responses in trace conditioning that they were sensitive to changes in reinforce-
procedures eventually, and that procedural varia- ment schedules. That is, unlike Tripp and Alsop
tions could also minimize age-related differences or Fisher and Noll, Plaud et al. found that older
in procedures using delay versus trace condition- adults’ behavior did change with schedule chang-
ing. es. Plaud et al. exposed older adults to extinction
Behavior analysts working in clinical settings and found that they demonstrated persistent re-
with aging adults might not use respondent-con- sponse bias (i.e., resistance to extinction), even
ditioning procedures in their clinical practice, but when the researchers changed the contingencies.
much applied research in behavioral gerontology Taken together, this line of research demon-
includes procedures to enhance stimulus control. strates that older adults’ behavior is sensitive to
These studies might inform clinical interventions changes in reinforcement. However, researchers
that aim (1) to establish stimuli as reinforcement, classified the older adults’ behavior as less sensi-
such as pairing stimuli using a temporal overlap tive to reinforcement than that of younger adults
rather than attempting to present paired stimuli in almost every study. This research has several
with temporal gaps; and (2) to ensure that dis- clinical implications, though clearly more research
criminative stimuli are present throughout con- is needed before we can make definitive recom-
ditioning, rather than simply at the beginning of mendations. Behavior analysts should take basic
conditioning blocks. Using a trace-conditioning operant research findings into consideration when
procedure directly, or expecting a stimulus to ac- designing and implementing interventions for
quire reinforcing or evocative properties after a few socially significant issues with aging adults. For
pairings, may set up clinical procedures for failure. example, basic research suggests that older adults
Until further research can evaluate whether the may require increased exposure to intervention
age-related effects found in respondent condi- contingencies for socially appropriate behaviors
tioning do not apply to operant conditioning, we to develop and be maintained when social stim-
advise behavior analysts to consider whether the uli maintain challenging behaviors, and when ap-
literature on respondent conditioning might en- propriate responses have produced limited or no
hance their procedures. reinforcement. Behavior analysts might provide
higher-quality or higher-magnitude reinforcement
for socially appropriate behaviors. Finally, based
Operant Conditioning
on the results of Plaud et al. (1999), behavior ana-
Perhaps the two most studied areas in age-related lysts should anticipate resurgence during periods
basic research on operant conditioning are sensi- of extinction when staff members implement in-
tivity to reinforcement and stimulus equivalence. terventions with poor integrity.
Several studies have examined age-related sensitiv- Several studies have examined age-related dif-
ity to reinforcement, because previous researchers ferences in responding in stimulus equivalence
have suggested that older adults’ behavior was less preparations and found generally weaker for-
sensitive to environmental changes than that of mation of equivalence classes for older adults.
younger adults (Tripp & Alsop, 1999). Research in Stimulus equivalence preparations typically in-
this area has produced interesting and sometimes volve researchers’ arbitrarily assigning stimuli to
conflicting results. Fisher and Noll (1996) found classes (e.g., Class A, Class B, Class C) and to
that when schedules of reinforcement varied, groups within those classes (e.g., A1, A2, A3).
older adults’ behavior was slower to change (i.e., to Researchers then use match-to-sample prepara-
match the new schedule of reinforcement). How- tions to teach relations (e.g., when a participant is
ever, their behavior matched relative reinforce- presented with stimulus A1, the correct response
ment rates with increased exposure to reinforce- out of an array of C2, A3, and B1 is B1). Wilson
ment contingencies. Tripp and Alsop (1999) later and Milan (1995) compared groups of older and
compared responding of children, young adults, younger adults, and found slower response times
and older adults on different reinforcement sched- and poorer performance on posttests of equiva-
516 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
lence relations for older adults. As a result, re- cesses, such as the establishment of conditioned
searchers have attempted to evaluate whether responses, sensitivity to changes in schedules of re-
such a difference is due to the procedures or is inforcement, and formation of equivalence classes.
age-related. Perez-Gonzalez and Moreno-Sierra As in classical-conditioning research, seemingly
(1999) found that older adults’ performance on minor procedural variations can have an impact
equivalence preparations was slightly better than on responding in operant conditioning (e.g.,
that of Wilson and Milan’s participants, but still changes in the stimulus delay in match-to-sample
found impaired formation of equivalence classes preparations). Behavior analysts who conduct re-
in older adults. Other studies have evaluated sev- search with or provide clinical services for older
eral methods for training equivalence class for- adults should evaluate whether procedural aspects
mation in older adults, to determine whether of their interventions might have an impact on re-
modifications in preparations could improve this sponsiveness.
formation. Saunders, Chaney, and Marquis (2005)
found no difference in equivalence class formation
when they used different numbers of samples (e.g., CLINICAL BEHAVIOR‑ANALYTIC RESEARCH
two-, three-, and four-choice match-to-sample ON AGING
configurations) and different training structures
(e.g., linear series, many-to-one, and one-to-ma- Clinical research in behavioral gerontology has
ny training structures). Interestingly, their data focused primarily on mental health problems (e.g.,
suggested equivalence class formation in 75% of depression, anxiety), health maintenance, and
participants, compared to 45% of participants in problems associated with neurocognitive disor-
Wilson and Milan. Researchers continue to evalu- ders. The medical rather than the psychosocial
ate the variables that are responsible for varying model has been dominant, perhaps because 85%
performance across age groups in match-to-sample of older individuals have health concerns warrant-
preparations. For example, Steingrimsdottir and ing regular medical visits (Butler, Finkel, Lewis,
Arntzen (2014) showed that healthy older adults Sherman, & Sunderland, 1992). By contrast, few
(ages 70–86) had shorter trials to criterion when older adults have regular contact with mental
(1) identity matching (i.e., classes were based on health professionals (Belsky, 1999). However, evi-
physical properties of stimuli) preceded arbitrary dence exists in each of these areas that change in
matching (i.e., classes were arbitrary), and (2) the the environment can produce change in behavior
procedure incorporated a 0-second delay (i.e., the even when medical interventions cannot alter
sample stimulus disappeared as the response op- physical or cognitive status. For example, when re-
tions appeared). Saunders et al. also found that a searchers hold psychotropic medication constant,
0-second delay improved performance. The 0-sec- behavioral interventions can produce changes in
ond delay is contrasted with preparations in which behavior (Baker, Hanley, & Mathews, 2006). In
the sample stimulus is present at the same time as this section, we review the effectiveness of behav-
the response options. The effectiveness of insert- ioral interventions for mental and physical health
ing a break between the sample stimulus and the issues, as well as recent advances in assessment,
response options seems to contradict the find- function-based intervention for behavioral excess-
ings for effective manipulations in the literature es, non-function-based intervention for behavioral
on basic respondent conditioning. As noted ear- excesses, and intervention for behavioral deficits.
lier, researchers on respondent conditioning have
shown that the trace preparation, where the con-
ditioned stimulus ended before the unconditioned Mental and Physical Health Issues
stimulus began, produced poorer acquisition of Mental Health
conditioned responses than the delay condition,
where the conditioned and unconditioned stimu- Depression and anxiety are common men-
lus overlapped. tal health problems for older adults (Sorocco,
Although further research is needed in both Kinoshita, & Gallagher-Thompson, 2005) and
basic respondent and operant conditioning, re- have been the subjects of investigation in exten-
searchers have identified age-related differences in sive behavioral and cognitive-behavioral therapy
behavior and procedural variations that can miti- research. Symptoms of depression include sadness,
gate those differences. Basic research has uncov- feelings of worthlessness and guilt, lethargy, sleep
ered age-related changes in basic behavioral pro- and appetite disturbances, and loss of interest in
Behavioral Gerontology 517
activities. Major depression occurs in 1–6% of that evaluating and establishing pretherapy skills
older adults living in the community (Mojtabai early in intervention are critical to the therapy
& Olfson, 2004), 5–10% of medically ill or frail process. Charlesworth et al. suggested that older
individuals (Dick & Gallagher-Thompson, 1996), adults with comorbid anxiety and neurocognitive
20–30% of individuals with neurocognitive disor- disorders often lack pretherapy skills and may not
ders (Steinberg et al., 2003; Zarit & Zarit, 1998), be able to describe their emotional difficulties.
and more than 50% of those living in nursing Charlesworth et al. suggested modifying short-
homes (Pellegrin, Peters, Lyketsos, & Marano, term cognitive-behavioral therapy to “focus more
2013). An additional 9–30% of older adults living on developing the missing pre-therapy skills than
in the community report symptoms of a duration to attempt to ‘rush through’ to goal-focused change
or severity that do not meet the diagnostic criteria material” (p. 392). Although Charlesworth et al.
for major depression, but that still decrease quality did not provide data on the effects of the approach,
of life significantly (Blazer, 1993; Thompson, Fut- Tonga et al. (2015) presented three case examples
terman, & Gallagher, 1988). Generalized anxiety to illustrate the challenges of using and modifying
disorder, the most common anxiety disorder, oc- manual-based cognitive-behavioral therapy with
curs in approximately 3–17% of older adults and is older adults with neurocognitive disorders. The re-
characterized by worry (Ladouceur, Leger, Dugas, searchers tailored therapy to each client’s present-
& Freeston, 2004; Stein, 2004; Wild et al., 2010); ing skills (e.g., use of technology-based memory
an additional 15–43% of healthy older adults re- aids, modified homework, slowed pace of progress
port anxiety symptoms that do not meet the di- through the manual); therefore, Tonga et al. also
agnostic criteria for an anxiety disorder (Mehta et recommended that behavior analysts use individu-
al., 2003). alized approaches.
Cognitive-behavioral therapy is an empirically
supported intervention for depression and anxi-
Physical Health Issues
ety for older adults (Arean, 2004; Stanley, Dief-
enbach, & Hopko, 2004) when administered in- Common physical health issues for older adults
dividually (Arean, 2004; Gallagher-Thompson & include diet, hydration, and continence. Other
Thompson, 1996) or in groups (DeVries & Coon, issues, such as compliance with health or medi-
2002). Cognitive-behavioral therapy generally in- cation recommendations, can affect health and
volves education about depression and anxiety; create additional health issues. Compliance with
self-monitoring of negative or anxious thoughts health or medication recommendations is referred
and emotion states; replacement of dysfunctional to as intervention regimen adherence, and the degree
beliefs and self-statements with functional ones; of compliance is typically a primary determinant
scheduling pleasant events; and training in such of overall health status (Meichenbaum & Turk,
skills as problem solving, coping, and relaxation 1987). Nonadherence to medical recommenda-
(Dick & Gallagher-Thompson, 1996; Dick, Galla- tions can have a direct impact on diet, hydration,
gher-Thompson, & Thompson, 1996; Gatz et al., and continence, which can then contribute to
1998). risk for cancer, diabetes, heart disease, delirium,
Researchers have provided guidelines for how to urinary tract infections, and medication toxicity
use cognitive-behavioral therapy to address anxiety (Ho, Lee, & Meyskens, 1991; Kannel, 1986; San-
among older adults with neurocognitive disorders servo, 1997; Warren et al., 1994). Nonadherence
(Charlesworth, Sadek, Schepers, & Spector, 2015). (i.e., noncompliance) estimates for medication
Charlesworth et al. (2015) provided a description regimens by older adults range from 43 to 62%,
of an individually tailored approach to cognitive- despite evidence that the adults are aware of the
behavioral therapy adapted for older adults with benefits of the medication and the potentially dire
mild to moderate neurocognitive disorders. They consequences for nonadherence (Meichenbaum
described a 10-session approach including educa- & Turk, 1987). Indeed, research suggests that
tion, development of pretherapy skills (i.e., behav- education alone consistently produces little to no
iors that research has shown must be present for the change in compliance (Sands & Holman, 1985).
approach to work, such as awareness of emotions Interventions involving prompts and behavioral
and ability to label emotions), self-monitoring, col- contingencies for adherence to diet and hydration
laboration and goal setting, progressive muscle re- recommendations have proven effective.
laxation, and perspective taking. As an example of Stock and Milan (1993) compared the effects
these modifications, Charlesworth et al. suggested of two sets of behavioral interventions on dietary
518 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
practices of older adults in a retirement commu- voiding schedules, and the use of discriminative
nity. Baseline involved prompts (i.e., researchers stimuli. Behavioral continence training consists
helped identify healthy foods), and the first inter- of education about mechanisms of bladder control
vention involved additional prompts (e.g., media, and specific recommendations, in-session practice
buttons, verbal reminders) with feedback (i.e., of contracting and identifying relevant muscles,
graphed percentage of healthy items) and praise and assigned practice in contraction exercises
when an individual reported selecting a healthy (Burton, Pearce, Burgio, Engle, & Whitehead,
food. The second intervention incorporated a to- 1988). This intervention produced an 82% reduc-
ken-based lottery system with immediate and de- tion in incontinence, compared to a 79% reduc-
layed reinforcement when the individual selected tion for community-dwelling participants who re-
a healthy food. The additional prompts increased ceived behavioral training plus bladder sphincter
healthy selections from a baseline mean of 24% to biofeedback for practicing muscle contractions—
an intervention mean of 65%, and the addition of an intrusive intervention in which a medical pro-
the lottery system produced a marginal increase to fessional inserts a bladder into the body and fills
68% healthy selections. it to simulate the feeling of an actual full bladder.
Prompt systems have proven effective in increas- Burgio et al. (2002) showed that reductions in in-
ing fluid consumption for older adults in nursing continence for a group that received behavioral
homes, but no studies have targeted community- training exceeded reductions for a group that re-
dwelling elders. Nursing aides in Spangler, Risley, ceived behavioral training plus bladder sphincter
and Bilyew (1984) presented a cup and offered biofeedback and for a group that received only
nursing home residents a choice of beverages every written instructions (69%, 63%, and 58%, respec-
1.5 hour. This intervention produced clinically tively).
and statistically significant improvements in hy- Prompted-voiding schedules involve education,
dration measured through urine. Simmons, Alessi, scheduled toilet visits with assistance, positive re-
and Schnelle (2001) manipulated the frequency of inforcement for dry intervals and continent voids,
prompts and beverage choices and found that 80% and encouragement to resist urinary urges be-
of nursing home residents increased their mean tween scheduled visits (Fantol, Wyman, Harkins,
daily fluid intake with systematic prompts alone. & Hadley, 1990; Jeffcoate, 1961). Perhaps one of
The addition of choice of beverage produced an the most prolific researchers in the area, Schnelle
additional 21% increase in fluid intake and a de- and his colleagues have published many demon-
crease in the number of beverage refusals. strations of the efficacy of the intervention. For
As described earlier, Feliciano et al. (2010) example, Schnelle et al. (2010) demonstrated that
showed how behavior analysts could develop an an intervention package with prompted voiding
assessment that informs intervention, and demon- could decrease urinary incontinence and could
strated the efficacy of that intervention. Feliciano increase physical activity and fluid intake for more
et al. developed a functional assessment interview, than 100 participants in a randomized controlled
the Hydration Interview, in which hydration was trial.
the target behavior. The interview evaluated risks In summary, researchers have demonstrated
of dehydration and then directed researchers to that behavioral assessments and interventions can
hypothesis-based interventions. Feliciano et al. effectively address mainstream gerontology issues.
showed that interventions produced increases in In addition to mental and physical health, a large
hydration measured through urine and increases body of behavioral gerontology research has fo-
in healthy fluid intake, with concomitant decreas- cused on behavioral disturbances related to what
es in unhealthy fluid intake. are now called neurocognitive disorders (and were
Urinary incontinence becomes more common formerly known as dementia).
with age, due to muscle weakness, decreased mo-
bility, memory loss, and communication difficul-
Advances in Assessment
ties (Burgio & Locher, 1996). Older adults may
and Assessment‑Informed Interventions
restrict fluid intake to avoid accidents (Simmons
et al., 2001), leading to dehydration, potential so- Behavior analysts should be well informed about
cial stigma, and increased care requirements in aging populations and proper preintervention as-
nursing homes. Several interventions have proven sessments, which is similar to the expectations for
effective in targeting incontinence, including psy- behavior analysts working in other clinical spe-
choeducation and behavioral training, prompted cialties. Effective clinical approaches to behavioral
Behavioral Gerontology 519
gerontology must also incorporate population-spe- more preferred stimuli increases target behaviors
cific behavioral assessments, such as those evaluat- more than presenting less preferred stimuli. Be-
ing cognitive functioning, health, and diet. cause therapeutic aging-care centers usually do
not provide prompts or reinforcement for activity
engagement, several studies on preference assess-
Stimulus Preference Assessment
ments with older adults with neurocognitive dis-
The primary goals in care settings for older adults, orders have used engagement analyses rather than
such as day programming, assisted living, and neu- reinforcement analyses to evaluate the utility of
rocognitive care, are usually increased participa- preference assessments. Some studies have used
tion in and enjoyment of activities. Thus applied selection-based reinforcement assessments. For
research on aging has focused on identifying strat- the purposes of this review, an engagement-based
egies and stimuli for increasing engagement in lei- reinforcement assessment measures duration of
sure activities (LeBlanc, Raetz, & Feliciano, 2011; unprompted activity engagement, and a selection-
Williamson & Ascione, 1983). Increasing activity based reinforcement assessment measures a partici-
engagement helps to improve quality of life, main- pant’s selection of one activity over another in a
tain functioning, and prevent depression (Engel- concurrent-operants arrangement to identify the
man, Altus, & Mathews, 1999; Engstrom, Mud- relative reinforcing efficacy of stimuli.
ford, & Brand, 2015; Garcia, Feliciano, & Ilem, LeBlanc et al. (2006) compared different meth-
2018; LeBlanc et al., 2011; Teri, 1991). For older ods of paired-stimulus presentation, such as tangi-
adults with neurocognitive disorders, increased ble presentations, pictorial representations, print-
engagement in leisure activities also reduces agita- ed text names, and vocally presented names, to
tion and aggression (Feliciano, Steers, Elite-Mar- evaluate the preferences of older adults with neu-
candonatou, McLane, & Arean, 2009). rocognitive disorders or aphasia. Tangible presen-
Historically, gerontologists have used surveys tations for one participant and vocal presentations
to identify preferred activities or have not at- for the other three participants produced higher
tempted to identify preferences. The Pleasant subsequent independent engagement. LeBlanc et
Events Schedule—Alzheimer’s Disease (Logsdon al. then used that optimal format for each par-
& Teri, 1997; Teri & Logsdon, 1991) is a survey ticipant to present choices between two activities,
assessment that prompts caregivers to identify ex- and the participant could engage in the chosen
amples of potentially preferred activities for older activity for up to 15 minutes (i.e., an engagement-
adults. Clinicians and researchers also have used based reinforcement assessment). Allowing the
the shortened version of this schedule (Logsdon participants to select from activities identified in
& Teri, 1997) to assess preferences by asking the their prior preference assessment produced more
older adults with neurocognitive disorders a series engagement than when the researchers presented
of yes–no questions (e.g., “Do you enjoy reading a standard list of leisure items.
magazines?”). One problem with the schedule is Raetz et al. (2013) used an engagement-based
that the adults may respond, “Yes, I would like to reinforcement assessment to examine whether hi-
do that,” but may not engage in the activities they erarchies of item preferences identified via a brief
have endorsed when opportunities arise (LeBlanc, multiple stimulus without replacement assessment
Raetz, Baker, Strobel, & Feeney, 2008). (i.e., three presentations of multiple-item arrays)
Because of the limitations of surveys, research- would predict the subsequent engagement of older
ers have begun to use direct observations to assess adults with neurocognitive disorders. Results of
preferences. Both paired-stimulus (Fisher et al., the multiple stimulus without replacement assess-
1992) and multiple stimulus without replacement ment generally predicted the level of subsequent
(DeLeon & Iwata, 1996) preference assessments engagement, although some participants engaged
produce preference hierarchies that can predict with items that were lower in the preference hier-
the effects of stimuli as reinforcers for the behavior archy. In addition, a single-array presentation from
of older adults (e.g., LeBlanc, Cherup, Feliciano, & the multiple stimulus without replacement proce-
Sidener, 2006; Raetz et al., 2013). Most research dure (i.e., one presentation of the multiple-item
with individuals with intellectual disabilities em- array) correlated reasonably well with the results
ploys a fixed-ratio (FR) 1 schedule to examine the of the mean of the three arrays (i.e., standard brief
effects of contingent access to an activity on a multiple stimulus without replacement). The pref-
target behavior (Hagopian, Long, & Rush, 2004). erences identified for more than half the partici-
This research investigates whether presenting pants remained stable over a 3- to 5-month period.
520 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Beattie, Wagner, and Baker (2019) compared produced a false positive for the other participant:
a multiple stimulus without replacement assess- The item the participant engaged with most did
ment, which is a selection-based preference assess- not show a reinforcing effect.
ment (i.e., preference hierarchies based on order Quick, Baker, and Ringdahl (2018) directly
of selection), to a free-operant assessment (Roane, compared the multiple stimulus without replace-
Vollmer, Ringdahl, & Marcus, 1998), which is ment and free-operant assessments with older
an engagement-based preference assessment (i.e., adults with neurocognitive disorders, to repli-
preference hierarchies based on relative duration cate and extend the Beattie et al. (2019) results.
of engagement), with older adults with neurocog- Quick et al.’s results suggested that although both
nitive disorders. Beattie et al. also conducted two assessments identified preferred items, only the
forms of reinforcement assessment. One was an multiple stimulus without replacement assessment
engagement-based reinforcement assessment in produced a hierarchy of preferred items that was
which two items were available concurrently. The useful for reestablishing functional skills. Further-
second was a selection-based reinforcement assess- more, they found that both selection-based and
ment in which four items were available, but a par- engagement-based assessments confirmed the re-
ticipant had to move to the area of the item. That sults of the free-operant (for two of the three) and
is, after the participant selected an item, he or she the multiple stimulus without replacement (for
had to leave the chosen item and move to engage all three) assessments, though engagement-based
with other items. The preference hierarchies of reinforcement assessments yielded quicker results
the multiple stimulus without replacement and across participants.
the free-operant assessments positively correlated Researchers also have used the paired-stimulus
with the level of subsequent engagement during and multiple stimulus without replacement assess-
the engagement-based reinforcement assessment. ment procedures to identify items to reestablish
In some cases, the results of the free-operant pref- functional skills, using a traditional contingent-
erence assessment better predicted subsequent en- access arrangement (an FR 1 schedule to examine
gagement during the engagement-based reinforce- the effects of contingent access to an activity on
ment assessment. In other cases, the results of the a target behavior; Hagopian et al., 2004). Virues-
multiple stimulus without replacement assessment Ortega et al. (2012) showed that paired-stimulus
better predicted subsequent selections during the assessments could identify leisure activities and
selection-based reinforcement assessment. These edible items for use as reinforcement for arbitrary
results indicate that behavior analysts should at- responses, such as pressing a lever, and significant
tend to the types of preference assessments they responses, such as stacking blocks during physi-
are using, as different preference assessments pro- cal therapy, for older adults with neurocognitive
duced different accuracy levels. Beattie et al. sug- disorders. Although some research has shown that
gested that the differential results may have been individuals with intellectual and developmental
due to the fact that the free-operant preference disabilities and autism spectrum disorder will se-
assessment is an engagement-based preference as- lect edible items over preferred leisure items when
sessment, whereas the multiple stimulus without both are available, Virues-Ortega et al. found that
replacement assessment is a selection-based assess- older adults selected leisure items over edible items
ment. when both were available. Ritchie, Reuter-Yuill,
Oleson and Baker (2014) used a free-operant Perez, and Baker (2019) obtained similar results,
preference assessment before using a concurrent- in that older adults selected leisure items over
operants (i.e., selection-based) reinforcement edibles when both were available. Although this
assessment. During the concurrent-operants as- difference in preference of leisure versus edible
sessment, the researchers presented three items items might appear to be an age-related difference,
concurrently, provided each individual with the a more parsimonious explanation might be related
selected item, and removed the other items. They to the environment of and health-related changes
found that the free-operant assessment produced in older adults. Food tends to be available readily
false negatives for one participant: One item the in most aging-care settings, providing more than
participant did not engage with showed reinforcing the necessary calories, given the lower levels of
effects similar to those for an item the participant activity and reduced metabolism of older adults.
did engage with during the free-operant assess- As adults age, changes in olfactory senses and me-
ment in subsequent selection-based reinforcement tabolism combine to decrease sensations of hun-
assessments. The free-operant assessment also ger and decrease enjoyable aspects of food. That
Behavioral Gerontology 521
is, their inability to taste and smell food decreases Moniz-Cook, Stokes, & Agar, 2003; Moniz-Cook,
the value of the food. However, leisure activities Woods, & Richards, 2001). Functional assess-
are typically much less available (e.g., McClan- ments have identified the most common socially
nahan & Risley, 1975). As such, the difference mediated categories of reinforcement for the prob-
in the preference literature between older adults lematic behaviors of older adults with neurocogni-
and individuals with intellectual and developmen- tive disorders, such as attention (e.g., Buchanan &
tal disabilities and autism spectrum disorder may Fisher, 2002; Dwyer-Moore & Dixon, 2007), access
be due to physiological changes that abolish the to tangibles (e.g., Heard & Watson, 1999), or es-
value of food reinforcement, or may be due to the cape from staff proximity (e.g., Baker et al., 2006).
discriminability (Virues-Ortega et al., 2012) and Researchers also have developed hypothesis-based
availability (J. E. Ringdahl, personal communica- interventions to address problematic behaviors
tion, 2014) of different items in the older adults’ maintained by nonsocially mediated consequenc-
environments. es, such as providing items that produce sensory
Finally, researchers have evaluated prefer- experiences like those produced by engaging in
ence assessments as an antecedent intervention the problem behavior (Baker et al., 2011; Burgio
to address behavioral excesses (Feliciano et al., et al., 1996).
2009; Fisher & Buchanan, 2018). Feliciano et al. Researchers have shown that function-based
(2009) showed that providing items identified via interventions are effective for older adults with
a paired-stimulus choice assessment decreased de- neurocognitive disorders. One intervention is
pression and agitation in older adults with neuro- the noncontingent delivery of reinforcement at
cognitive disorders. In some cases, the researchers set intervals (e.g., Baker et al., 2006, Buchanan
identified when the target behavior, such as wan- & Fisher, 2002). Other interventions with older
dering, was likely to occur. They then developed a adults with neurocognitive disorders have in-
schedule that prompted staff members to provide cluded teaching socially appropriate responses to
preferred items before a target behavior occurred, produce reinforcement (i.e., differential reinforce-
to reduce the likelihood of occurrence. In other ment of alternative behaviors), while withholding
cases, the staff provided the preferred items during reinforcement following problematic behaviors
an activity choice-based redirection following the (e.g., Dwyer-Moore & Dixon, 2007). Function-
target behavior. In both cases, staff provided the based interventions have reduced many prob-
preferred items as part of a multicomponent inter- lematic behaviors effectively, including wander-
vention plan that reduced target behaviors. ing (e.g., Dwyer-Moore & Dixon, 2007; Heard &
Watson, 1999), aggression (Baker et al., 2006), dis-
ruptive vocalizations (Buchanan & Fisher, 2002),
Functional Assessment
hoarding (Baker et al., 2011), and noncompliance
Researchers have reported that function-based (Moniz-Cook et al., 2003).
interventions reduce problematic or unsafe be- In addition to demonstrating the applicability
haviors more effectively than non-function-based of functional assessment in aging-care settings, re-
interventions do (Iwata, Pace, Cowdery, & Milt- searchers have conducted component analyses to
enberger, 1994). More recently, researchers have identify the crucial components and experimental
used functional assessments and function-based manipulations to include in functional-analytic
interventions (i.e., interventions that involve the conditions. Larrabee, Baker, and O’Neill (2018)
manipulation of the response–reinforcement rela- evaluated the impact of researcher-programmed
tion that maintains the target behavior) to address discriminative stimuli (which were stimuli that
problematic or unsafe behaviors in older adults the researcher added to signal the condition, such
with neurocognitive disorders. as colored clothing and poster boards) in func-
Several studies used functional assessments, tional analyses for disruptive vocalizations among
such as interviewing staff members, direct obser- older adults with neurocognitive disorders. Lar-
vation, or experimental functional analysis, to rabee et al. conducted multielement functional
guide the development of function-based interven- analyses with two versions of each condition, one
tions for older adults with neurocognitive disorders with and one without programmed discrimina-
(e.g., Baker et al., 2006; Baker, LeBlanc, Raetz, & tive stimuli. They found that conditions without
Hilton, 2011; Buchanan & Fisher, 2002; Burgio, programmed discriminative stimuli produced lev-
Scilley, Hardin, Hsu, & Yancey, 1996; Dwyer- els of undifferentiated responding during test and
Moore & Dixon, 2007; Heard & Watson, 1999; control conditions similar to those reported in
522 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
previous published studies (cf. Beaton, Peeler, & from another individual) repertoires among five
Harvey, 2006, who did not report any programmed older adults with varying levels of cognitive im-
stimuli associated with conditions and were unable pairment. Although the overall effects of the in-
to obtain differentiated results in their functional tervention were limited, Henry and Horne argued
analyses). Larrabee et al. found that conditions that replications were needed. Even though the
with programmed discriminative stimuli produced participants had very few and weak remaining
differentiated responding during test and control repertoires (e.g., one participant engaged only in
conditions, which led to function-based interven- echoic behavior), interventions produced increas-
tions. Thus including programmed discrimina- es in responding.
tive stimuli is important in functional analyses Baker et al. (2008) provided a behavioral con-
with older adults with cognitive impairments, to ceptualization of aphasia to explain unusual pat-
enhance these adults’ ability to discriminate the terns of skill loss common in older adults. Al-
experimental conditions; this should improve the though not experimental in nature, this paper
efficacy and efficiency of the functional analysis. provided a behavioral framework for assessing
language deficits and abilities (cf. Henry & Horne,
2000) that behavior analysts could use to program
Assessment and Interventions for Verbal Behavior interventions. Gross, Fuqua, and Merritt (2013)
Language disruptions among older adults are typi- expanded upon Baker et al. by developing and
cally the results of stroke or neurocognitive de- using an assessment to show evidence of verbal-
generation (Baker, LeBlanc, & Raetz, 2008). Al- operant deficits among older adults with neuro-
though the broader field of gerontology has studied cognitive disorders. They also demonstrated the
changes in verbal behavior among older adults, functional independence of those deficits. For ex-
Sundberg (1991) was among the first to point out ample, an older adult could not say, “Unicorn,” in
the potential to understand, assess, and address the presence of a picture of a horse with a spiraled
changes in verbal behavior among older adults horn on its head (a tact), but could say the word
from a behavior-analytic perspective. Sundberg when asked to read the letters U-N-I-C-O-R-N (a
argued: textual). Such results are important, as functional-
ly independent deficits mean that researchers can
It is a known fact that the verbal repertoires of elderly use responses that are still occurring (e.g., the tex-
individuals tend to weaken. However, it is unclear tual response in the unicorn example) to prompt
what the crucial variables are. It is often assumed a response when it normally does not occur (e.g.,
that biological deterioration is responsible for verbal the tact in the unicorn example) and then trans-
problems, when quite possibly, it is environment de- fer stimulus control from one response to another.
terioration that is the key variable. (p. 84) Oleson and Baker (2014) used a combination of
the Baker et al. and Gross et al. assessments to
Several studies in behavioral gerontology have evaluate mand repertoires. The assessments that
provided evidence that supports Sundberg’s claim. Baker et al. and Gross et al. proposed constitute
For example, Blackman, Howe, and Pinkston an important advancement in the science of be-
(1976), Carroll (1978), Quattrochi-Tubin and havioral gerontology. The behavior-analytic as-
Jason (1980), Carstensen and Erickson (1986), sessments incorporate ways to evaluate age-related
and Bourgeois (1993) provided evidence that changes that the larger field of gerontology and
simple antecedent environmental manipulations basic behavior-analytic research has identified,
(e.g., providing refreshments, rearranging rooms, such as changes in stimulus control, and can in-
providing memory books) resulted in increased form intervention directly.
interaction and verbal statements among older Several investigators have evaluated strate-
adults. However, Carstensen and Erickson argued gies to remediate language deficits, such as those
that simply increasing the rate of interaction was observed in adults with aphasia and neurocogni-
not enough. tive disorders. Most research has focused on es-
Henry and Horne (2000) were the first to apply tablishing either echoic repertoires (e.g., Dixon,
Skinner’s analysis to address language changes Baker, & Sadowski, 2011; Henry & Horne, 2000)
among older adults with neurocognitive disorders. or mand repertoires (i.e., requesting an item; e.g.,
The authors attempted to reestablish echoic (re- Løkke, Granmo, Leirvick, & Lund, 2013; Oleson
peating what someone said), tact (labeling items), & Baker, 2014; Trahan et al., 2014). Baker et al.
and mand compliance (complying with requests (2008) suggested that procedures for transfer of
Behavioral Gerontology 523
stimulus control might prove an effective way to primarily with children and young adults with
reteach verbal behavior among older adults with autism spectrum disorder, with intellectual and
language deficits. Oleson and Baker (2014) dem- developmental disabilities, or with both, rather
onstrated that such an approach could be effec- than with older adults. Though the concepts and
tive. Most recently, Ritchie et al. (2019) evaluated principles of behavior analysis are the same, nu-
different transfer-of-stimulus-control procedures merous important clinical considerations involved
that might prove most useful. They compared in working with older adults necessitate specific
time delays where they presented the prompt ini- training and supervision before marketing services
tially and subsequently delayed its presentation to to a new client base. LeBlanc, Heinicke, and Baker
a prompt–prompt–probe method. (2012) describe two tasks behavior analysts might
Verbal behavior remains an important area for focus on when considering expanding into aging
future research. Changes in verbal behavior are of services: (1) increasing professional competence
great interest in the broader field of gerontology. with older adults, and (2) identifying and man-
However, most research has focused on assessing aging employment opportunities. LeBlanc et al.
deficits that gerontologists believe have a biologi- broke each task into component activities, such as
cal etiology. Behavior analysis is poised well to reading the literature, pursuing supervision, and
provide a technology for intervening on and reme- obtaining professional credentials, and provided
diating verbal deficits among older adults where no guidance for engaging in the activities. The soci-
other options exist. etal need is great, and we hope that at least some
behavior analysts will become interested in pursu-
ing a new area and will expand the impact of be-
SUMMARY AND CONCLUSIONS havior analysis into aging services, making effec-
tive interventions available to more older adults.
The broader field of gerontology highly values ap- Behavior analysts must continue to publish
plications targeted toward the many socially im- studies in aging journals and journals specific to
portant aging-related issues that researchers need other disciplines, such as nursing and occupa-
to address. Current topics in mainstream gerontol- tional therapy, to introduce these groups to the
ogy research include health and diet, poor fluid in- benefits of the behavioral approach. However,
take, hospital readmission for acute health issues, behavior analysts also must continue to publish
ambulation, activities of daily living, disengage- enough studies on behavioral gerontology in flag-
ment in aging-care facilities, minimizing the use ship behavior-analytic outlets to ensure that new
of physical and chemical restraints, and increasing behavior analysts become interested in the field.
the amount of meaningful activities that nursing The field needs additional basic operant studies
homes offer residents (Drossel & Trahan, 2015; with humans for virtually every aspect of respond-
Fisher, Drossel, Yury, & Cherup, 2007; Kales, Git- ing that age-related changes might impact. The
lin, & Lyketsos, 2015; Tolson et al., 2011). Howev- applied area needs more studies that incorporate
er, most mainstream research is descriptive (Mont- functional assessment and address health and
gomery, 1996) or focused only on assessment, with mental health issues in community-dwelling older
little research on efficacious interventions for adults. There have been relatively few functional-
these issues. Behavior analysts interested in aging analytic studies of older adults with neurocogni-
have almost unlimited potential for research and tive disorders, compared to the thriving literature
clinical opportunity, due to the growing demand on functional analysis with individuals with devel-
by older adults for a range of services. opmental disabilities (Hanley, Iwata, & McCord,
Although behavioral gerontology has enormous 2003), even though the adults with neurocogni-
potential for positive social impact, several promi- tive disorders are just as likely to have behaviors
nent behavior analysts have suggested that behav- that staff members identify as challenging or dif-
ioral gerontology has not flourished as a subfield ficult to manage. Though research suggests that
(Burgio & Burgio, 1986; Carstensen, 1988) and function-based interventions are promising, little
has not fully explored all potential applications research has evaluated whether effective function-
of behavior analysis to aging (Derenne & Baron, based interventions for problem behavior can re-
2002). Interested behavior analysts may have dif- duce the need for medications, which is a concern
ficulty expanding into services for aging adults, in aging research now (Kales et al., 2015). Behav-
due to the scope of their training and experience. ior analysts would do well to become familiar with
Many applied behavior analysts receive training the effects prescription medications can have on
524 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
behavioral processes. Laraway, Snycerski, Michael, Beaton, S., Peeler, C. M., & Harvey, T. (2006). A func-
and Poling (2003) highlighted the varying effects tional analysis and treatment of the irrational and ra-
that pharmaceutical interventions can have on tional statements of an elderly woman with Alzheim-
er’s disease. Behavioral Interventions, 21(1), 1–12.
motivating operations, particularly on establish-
Beattie, S., Wagner, S., & Baker, J. C. (2019). Evalua-
ing and abolishing operations for reinforcers and tion of preference and subsequent stimulus engagement
punishers. Additionally, Valdovinos and Kennedy among older adults with neurocognitive disorder. Manu-
(2004) noted that pharmacological studies primar- script in preparation.
ily focus on unintended or unwanted side effects Belsky, J. (1999). The psychology of aging: Theory, re-
such as sweating, diarrhea, or tremors, but seldom search, and interventions. Belmont, CA: Brooks/Cole.
discuss their effects on behavioral processes. Side Birren, J. E., & Schaie, K. W. (Eds.). (2001). Handbook
effects of psychotropic medications could influ- of the psychology of aging (5th ed.). San Diego, CA:
ence motivating variables, disrupt stimulus con- Academic Press.
trol, or establish new discriminative stimuli (Val- Blackman, D. K., Howe, M., & Pinkston, E. M. (1976).
Increasing participation in social interaction of the
dovinos & Kennedy, 2004).
institutionalized elderly. Gerontologist, 16, 69–76.
Blazer, D. (1993). Depression in late life (2nd ed.). St.
Louis, MO: Mosby.
REFERENCES Bourgeois, M. (1993). Effects of memory aids on the
dyadic conversations of individuals with dementia.
Adkins, V. K., & Mathews, R. M. (1999). Behavioral Journal of Applied Behavior Analysis, 26, 77–87.
gerontology: State of the science. Journal of Clinical Buchanan, J. A., & Fisher, J. E. (2002). Functional as-
Geropsychology, 5, 39–49. sessment and noncontingent reinforcement in the
Altus, D. E., Engelman, K. K., & Mathews, R. M. treatment of disruptive vocalization in elderly de-
(2002a). Increasing mealtime participation and com- mentia patients. Journal of Applied Behavior Analysis,
munication of persons with dementia. Journal of Ge- 35, 99–103.
rontological Nursing, 28, 47–53. Buchanan, J., Husfeldt, J., Berg, T., & Houlihan, D.
Altus, D. E., Engelman, K. K., & Mathews, R. M. (2011). Publication trends in behavioral gerontology
(2002b). Finding a practical method to increase en- in the past 25 years: Are the elderly still an under-
gagement of residents of a dementia care unit. Ameri- studied population in behavioral research? Behavioral
can Journal of Alzheimer’s Disease and Related Disor- Interventions, 23, 65–74.
ders, 17, 245–248. Burgio, K., & Locher, J. (1996). Urinary incontinence.
Arean, P. A. (2004). Psychosocial treatments for depres- In L. Carstensen, B. Edelstein, & L. Dornbrand
sion in the elderly. Primary Psychiatry, 11, 48–53. (Eds.), The practical handbook of clinical gerontology
Baker, J. C., Fairchild, K., & Seefeldt, D. (2015). Behav- (pp. 349–373). Thousand Oaks, CA: Sage.
ioral gerontology: Research and clinical consider- Burgio, L. D., & Burgio, K. L. (1986). Behavioral ger-
ations. In H. S. Roane, J. E. Ringdahl, & T. S. Fal- ontology: Application of behavioral methods to the
comata (Eds.), Clinical and organizational applications problems of older adults. Journal of Applied Behavior
of applied behavior analysis (pp. 425–450). San Diego, Analysis, 19, 321–328.
CA: Academic Press/Elsevier. Burgio, L., Scilley, K., Hardin, J. M., Hsu, C., & Yancey,
Baker, J., Hanley, G. P., & Mathews, R. M. (2006). Staff J. (1996). Environmental “white noise”: An interven-
administered functional analysis and treatment of tion for verbally agitated nursing home residents.
aggression by an elder with dementia. Journal of Ap- Journal of Gerontology, 51B, 364–373.
plied Behavior Analysis, 39, 469–474. Burgio, L., Stevens, A., Burgio, K., Roth, D., Paul, P., &
Baker, J. C., & LeBlanc, L. A. (2011). Acceptability of Gerstle, J. (2002). Teaching and maintaining behav-
interventions for aggressive behavior in long-term ior management in the nursing home. Gerontologist,
care settings: Comparing ratings and hierarchical 42, 487–496.
selection. Behavior Therapy, 42, 30–41. Burton, J. R., Pearce, K. L., Burgio, K. L., Engel, B. T.,
Baker, J. C., LeBlanc, L. A., & Raetz, P. B. (2008). A & Whitehead, W. E. (1988). Behavioral training for
behavioral conceptualization of aphasia. Analysis of urinary incontinence in the elderly ambulatory pa-
Verbal Behavior, 24, 147–158. tient. Journal of the American Geriatrics Society, 36,
Baker, J. C., LeBlanc, L. A., Raetz, P. B., & Hilton, L. 693–698.
C. (2011). Assessment and treatment of hoarding in Butler, R. N., Finkel, S. I., Lewis, M. I., Sherman, F. T.,
an individual with dementia. Behavior Therapy, 42, & Sunderland, T. (1992). Aging and mental health:
135–142. Primary care of the healthy older adult. A roundtable
Baltes, M., & Barton, E. (1977). Behavioral analysis of discussion: Part I. Geriatrics, 47, 54, 56, 61–65.
aging: A review of the operant model and research. Carroll, P. S. (1978). The social hour for geropsychiatric
Educational Gerontology: An International Quarterly, patients. Journal of American Geriatrics Society, 11,
2, 383–405. 32–35.
Behavioral Gerontology 525
Carstensen, L. L. (1988). The emerging field of behav- adults in long-term care settings. Journal of Applied
ioral gerontology. Behavior Therapy, 19, 259–281. Behavior Analysis, 40, 679–684.
Carstensen, L. L., & Erickson, R. J. (1986). Enhancing Engelman, K. K., Altus, D. E., & Mathews, R. M. (1999).
the social environments of elderly nursing home resi- Increasing engagement in daily activities by older
dents: Are high rates of interaction enough? Journal adults with dementia. Journal of Applied Behavior
of Applied Behavior Analysis, 19, 349–355. Analysis, 32, 107–110.
Charlesworth, G., Sadek, S., Schepers, A., & Spector, Engstrom, E., Mudford, O. C., & Brand, D. (2015).
A. (2015). Cognitive behavior therapy for anxiety Replication and extension of a check-in procedure
in people with dementia: A clinician guideline for a to increase activity engagement among people with
person-centered approach. Behavior Modification, 39, severe dementia. Journal of Applied Behavior Analysis,
390–412. 48, 460–465.
Cheng, D. T., Faulkner, M. L., Disterhoft, J. F., & Des- Fantol, J. A., Wyman, J. F., Harkins, S. W., & Hadley,
mond, J. E. (2010). The effects of aging in delay and E. C. (1990). Bladder training in the management of
trace human eyeblink conditioning. Psychology and lower urinary tract dysfunction in women: A review.
Aging, 25, 684–690. Journal of the American Geriatrics Society, 38, 329–332.
Cherry, K. E., & Smith, A. D. (1998). Normal memo- Feliciano, L., LeBlanc, L. A., & Feeney, B. J. (2010). As-
ry aging. In M. Hersen & V. B. Van Hasselt (Eds.), sessment and management of barriers to fluid intake
Handbook of clinical geropsychology (pp. 87–110). New in community dwelling older adults. Journal of Behav-
York: Plenum Press. ioral Health and Medicine, 1, 3–14.
Colby, S. L., & Ortman, J. M. (2014). Projections of the Feliciano, L., Steers, M. E., Elite-Marcandontou, A.,
size and composition of the U.S. population: 2014 to McLane, M., & Areán, P. A. (2009). Applications
2060 (Current Population Reports, No. P25-1143). of preference assessment procedures in depression
Washington, DC: U.S. Census Bureau. Retrieved and agitation management in elders with dementia.
from www.census.gov/content/dam/Census/library/ Clinical Gerontologist, 32, 239–259.
publications/2015/demo/p25-1143.pdf. Finkbiner, R., & Woodruff-Pak, D. (1991). Classical
DeLeon, I. G., & Iwata, J. C. (1996). Evaluation of a eye-blink conditioning in adulthood: Effects of age
multiple-stimulus presentation format for assessing and interstimulus interval on acquisition in the trace
reinforcer preferences. Journal of Applied Behavior paradigm. Psychology and Aging, 6, 109–117.
Analysis, 29, 519–533. Fisher, J. E., & Buchanan, J. A. (2018). Presentation of
Derenne, A., & Baron, A. (2002). Behavior analysis preferred stimuli as an intervention for aggression in
and the study of human aging. Behavior Analyst, 25, a person with dementia. Behavior Analysis: Research
151–160. and Practice, 18, 33–40.
DeVries, H. M., & Coon, D. W. (2002). Cognitive/be- Fisher, J. E., Drossel, C., Yury, C., & Cherup, S. (2007).
havioral group therapy with older adults. In F. W. A contextual model of restraint-free care for persons
Kaslow & T. Patterson (Eds.), Comprehensive hand- with dementia. In P. Sturmey (Ed.), Functional analy-
book of psychotherapy: Cognitive-behavioral approaches sis in clinical treatment (pp. 211–237). Burlington,
(Vol. 2, pp. 547–576). New York: Wiley. MA: Academic Press/Elsevier.
Dick, L. P., & Gallagher-Thompson, D. (1996). Late-life Fisher, J. E., & Noll, J. (1996). Age-associated differenc-
depression. In M. Hersen & V. B. Van Hasselt (Eds.), es in sensitivity to reinforcement frequency. Journal
Psychological treatment of older adults: An introductory of Clinical Geropsychology, 2, 297–306.
text (pp. 181–208). New York: Plenum Press. Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L.
Dick, L. P., Gallagher-Thompson, D., & Thompson, L. P., Owens, J. C., & Slevin, I. (1992). A comparison of
W. (1996). Cognitive-behavioral therapy. In R. T. two approaches for identifying reinforcers for persons
Woods (Ed.), Handbook of the clinical psychology of with severe and profound disabilities. Journal of Ap-
aging (pp. 509–544). Chichester, UK: Wiley. plied Behavior Analysis, 25, 491–498.
Dixon, M., Baker, J. C., & Sadowski, K. A. (2011). Ap- Gallagher-Thompson, D., & Thompson, L. W. (1996).
plying Skinner’s analysis of verbal behavior to per- Applying cognitive-behavioral therapy to the psy-
sons with dementia. Behavior Therapy, 42, 120–126. chological problems of later life. In S. H. Zarit & B.
Drossel, C., & Trahan, M. A. (2015). Behavioral inter- G. Knight (Eds.), A guide to psychotherapy and aging:
ventions are the first-line treatments for managing Effective clinical interventions in a life-stage context
changes associated with cognitive decline. Behavior (pp. 61–82). Washington, DC: American Psychologi-
Therapist, 38(5), 126–131. cal Association.
Dupree, L., & Schonfeld, L. (1998). The value of be- Garcia, S., Feliciano, L., & Ilem, A. A. (2018). Prefer-
havioral perspectives in treating older adults. In M. ence assessments in older adults with dementia. Be-
Hersen & V. B. Van Hasselt (Eds.), Handbook of havior Analysis: Research and Practice, 18, 78–91.
clinical geropsychology (pp. 51–70). New York: Ple- Gatz, M., Fiske, A., Fox, L. S., Kaskie, B., Kasl-Godsley,
num Press. J. E., McCallum, T. J., et al. (1998). Empirically vali-
Dwyer-Moore, K. J., & Dixon, M. R. (2007). Functional dated psychological treatments for older adults. Jour-
analysis and treatment of problem behavior of elderly nal of Mental Health and Ageing, 4, 9–46.
526 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Graves, C. A., & Solomon, P. R. (1985). Age-related dis- them: Some further refinements. Journal of Applied
ruption of trace but not delay classical conditioning Behavior Analysis, 36, 407–414.
of the rabbit’s nictitating membrane response. Behav- Larrabee, D., Baker, J. C., & O’Neill, D. (2018). Effects
ioral Neuroscience, 99, 88–96. of programmed discriminative stimuli in a functional
Gross, A. C., Fuqua, R. W., & Merritt, T. A. (2013). analysis on language disruptions in older adults with
Evaluation of verbal behavior in older adults. Analy- neurocognitive disorders. Behavior Analysis: Research
sis of Verbal Behavior, 29, 85–99. and Practice, 18(1), 16–32.
Hagopian, L. P., Long, E. S., & Rush, K. S. (2004). Pref- LeBlanc, L. A., Cherup, S. M., Feliciano, L., & Sidener,
erence assessment procedures for individuals with T. M. (2006). Using choice-making opportunities to
developmental disabilities. Behavior Modification, 28, increase activity engagement in individuals with de-
668–677. mentia. American Journal of Alzheimer’s Disease and
Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003). Other Dementias, 21, 318–325.
Functional analysis of problem behavior: A review. LeBlanc, L. A., Heineke, M., & Baker, J. C. (2012). Ex-
Journal of Applied Behavior Analysis, 36, 147–185. panding your consumer base for behavior analytic
Heard, K., & Watson, T. S. (1999). Reducing wander- services. Behavior Analysis in Practice, 5, 4–14.
ing by persons with dementia using differential rein- LeBlanc, L. A., Raetz, P. B., Baker, J. C., Strobel, M. J.,
forcement. Journal of Applied Behavior Analysis, 32, & Feeney, B. J. (2008). Assessing preference in elders
381–384. with dementia using multimedia and verbal Pleas-
Henry, L. M., & Horne, P. J. (2000). Partial remedia- ant Events Schedules. Behavioral Interventions, 23,
tion of speaker and listener behaviors in people with 213–225.
severe dementia. Journal of Applied Behavior Analysis, LeBlanc, L. A., Raetz, P. B., & Feliciano, L. (2011). Be-
33, 631–634. havioral gerontology. In W. W. Fisher, C. C. Piazza,
Ho, E. E., Lee, F. C., & Meyskens, F. L. (1991). An ex- & H. S. Roane (Eds.), Handbook of applied behavior
ploratory study of attitudes, beliefs and practices re- analysis (pp. 472–488). New York: Guilford Press.
lated to the interim dietary guidelines for reducing Lindsley, O. R. (1964). Geriatric behavioral prosthet-
cancer in the elderly. Journal of Nutrition for the El- ics. In R. Kastenbaum (Ed.), New thoughts in old age
derly, 10, 31–49. (pp. 41–61). New York: Springer.
Hussian, R. A., & Brown, D. (1987). Use of two-dimen- Logsdon, R. G., & Teri, L. (1997). The Pleasant Events
sional grid patterns to limit hazardous ambulation Schedule—AD: Psychometric properties and rela-
in demented patients. Journal of Gerontology, 42, tionship to depression and cognition in Alzheimer’s
558–560. disease patients. Gerontologist, 37, 40–45.
Ilem, A., Feliciano, L., & LeBlanc L. A. (2015). Recog- Løkke, J. A., Granmo, S., Leirvick, S. E. S., & Lund,
nition of self-referent stimuli in people with demen- P. A. (2013). Prompting and influence on mand-
tia: Names and pictures as prosthetic memory aids. ing in participants with Alzheimer’s disease. Norsk
Clinical Gerontologist, 38(2), 157–169. Tidsskrift for Atferdsanalyse, 40, 189–195.
Iwata, B. A., Pace, G. M., Cowdery, G. E., & Milten- McClannahan, L. E., & Risley, T. R. (1975). Design of
berger, R. G. (1994). What makes extinction work: living environments for nursing-home residents: In-
An analysis of procedural form and function. Journal creasing participation in recreation activities. Journal
of Applied Behavior Analysis, 27, 131–144. of Applied Behavior Analysis, 8, 261–268.
Jeffcoate, T. N. (1961). Functional disturbances of the Mehta, K. M., Simonsick, E. M., Penninx, B., Schulz, R.,
female bladder and urethra. Journal of the Royal Col- Rubin, S., Scatterfield, S., et al. (2003). Prevalence
lege of Surgeons of Edinburgh, 7, 28–47. and correlates of anxiety symptoms in well-function-
Kales, H. C., Gitlin, L. N., & Lyketsos, C. G. (2015). ing older adults: Findings from the Health Aging and
Assessment and management of behavioral and psy- Body Composition Study. Journal of the American
chological symptoms of dementia. British Medical Geriatrics Society, 51, 499–504.
Journal, 350, h369. Meichenbaum, D., & Turk, D. (1987). Facilitating treat-
Kannel, W. B. (1986). Nutritional contributors to car- ment adherence: A practitioner’s guidebook. New York:
diovascular disease in the elderly. Journal of the Plenum Press.
American Geriatrics Society, 34, 27–36. Mojtabai, R., & Olfson, M. (2004). Cognitive deficits
Keller, H., Beck, A. M., & Namasivayam, A. (2015). Im- and the course of major depression in a cohort of
proving food and fluid intake for older adults living middle-aged and older community-dwelling adults.
in long-term care: A research agenda. Journal of the Journal of the American Geriatrics Society, 52, 1060–
American Medical Directors Association, 16, 93–100. 1069.
Ladouceur, R., Leger, E., Dugas, M., & Freeston, M. H. Molinari, V., Karel, M., Jones, S., Zeiss, A., Cooley, S.
(2004). Cognitive-behavioral treatment of general- G., Wray, L., et al. (2003). Recommendations about
ized anxiety disorder (GAD) for older adults. Inter- the knowledge and skills required of psychologists
national Psychogeriatrics, 16, 195–207. working with older adults. Professional Psychology:
Laraway, S., Syncerski, S., Michael, J., & Poling, A. Research and Practice, 34, 435–443.
(2003). Motivating operations and terms to describe Moniz-Cook, E., Stokes, G., & Agar, S. (2003). Difficult
Behavioral Gerontology 527
behaviour and dementia in nursing homes: Five cases Sanservo, A. C. (1997). Dehydration in the elderly:
of psychosocial intervention. Clinical Psychology and Strategies for prevention and management. Nurse
Psychotherapy, 10, 197–208. Practitioner, 22, 41–70.
Moniz-Cook, E., Woods, R. T., & Richards, K. (2001). Saunders, R. R., Chaney, L., & Marquis, J. G. (2005).
Functional analysis of challenging behaviour in de- Equivalence class establishment with two-, three-,
mentia: The role of superstition. International Journal and four-choice matching to sample by senior citi-
of Geriatric Psychiatry, 16, 45–56. zens. Psychological Record, 55, 539–559.
Montgomery, R. J. V. (1996). Advancing caregiver re- Schnelle, J. F., Leung, F. W., Rao, S. S. C., Beusher, L.,
search: Weighing efficacy and feasibility of interven- Keeler, M., Clift, J. W., et al. (2010). A controlled
tions. Journal of Gerontology: Social Sciences, 51(3), trial of an intervention to improve urinary/fecal in-
S109–S110. continence and constipation. Journal of the American
Noguchi, D., Kawano, Y., & Yamanaka, K. (2013). Care Geriatrics Society, 58, 1504–1511.
staff training in residential homes for managing be- Simmons, S. F., Alessi, C., & Schnelle, J. F. (2001).
havioral and psychological symptoms of dementia An intervention to increase fluid intake in nursing
based on different reinforcement procedures of ap- home residents: Prompting and preference compli-
plied behavior analysis: A process research. Psycho- ance. Journal of the American Geriatrics Society, 49,
geriatrics, 13, 108–117. 926–933.
Oleson, C. R., & Baker, J. C. (2014). Teaching mands to Skinner, B. F. (1983). Intellectual self-management in
older adults with dementia. Analysis of Verbal Behav- old age. American Psychologist, 38, 239–244.
ior, 30, 113–127. Sorocco, K., Kinoshita, L., & Gallagher-Thompson, D.
Pachis, J. A., & Zonneveld, K. L. M. (2019). Compari- (2005). Mental health and aging: Current trends and
son of prompting procedures to teach Internet skills future directions. In J. E. Maddux & B. A. Winstead
to older adults. Journal of Applied Behavior Analysis, (Eds.), Psychopathology: Foundations for a contempo-
52, 173–187. rary understanding (pp. 393–419). Mahwah, NJ: Erl-
Pellegrin, L. D., Peters, M. E., Lyketsos, C. G., & Ma- baum.
rano, C. M. (2013). Depression in cognitive impair- Spangler, P. F., Risley, T. R., & Bilyew, D. D. (1984). The
ment. Current Psychiatry Reports, 15, 383–391. management of dehydration and incontinence in
Perez-Gonzalez, L. A., & Moreno-Sierra, V. (1999). nonambulatory geriatric patients. Journal of Applied
Equivalence class formations in elderly persons. Psi- Behavior Analysis, 17, 397–401.
cothema, 11, 325–336. Stanley, M. A., Diefenbach, G. J., & Hopko, D. R.
Plaud, J. J., Plaud, D. M., & von Duvillard, S. (1999). (2004). Cognitive behavioral treatment for older
Human behavioral momentum in a sample of older adults with generalized anxiety disorders: A therapist
adults. Journal of General Psychology, 126, 165–175. manual for primary care settings. Behavior Modifica-
Quattrochi-Tubin, S., & Jason, L. A. (1980). Enhanc- tion, 28, 73–117.
ing social interactions and activity among the elderly Stein, M. B. (2004). Public health perspectives on gen-
through stimulus control. Journal of Applied Behavior eralized anxiety disorder. Journal of Clinical Psychia-
Analysis, 13, 159–169. try, 65, 3–7.
Quick, M. J., Baker, J. C., & Ringdahl, J. E. (2018). As- Steinberg, M., Sheppard, J., Tschanz, J. T., Norton, M.
sessing the validity of engagement-based and selec- C., Steffens, D. C., Breitner, J. C. S., et al. (2003). The
tion-based preference assessments in elderly individ- incidence of mental and behavioral disturbances in
uals with dementia. Behavior Analysis: Research and dementia: The Cache County study. Journal of Neu-
Practice, 18(1), 92–102. ropsychiatry and Clinical Neurosciences, 15, 340–345.
Raetz, P. B., LeBlanc, L. A., Baker, J. C., & Hilton, L. Steingrimsdottir, H. S., & Arntzen, E. (2014). Per-
(2013). Utility of the multiple stimulus without re- formance by older adults on identity and arbitrary
placement procedure and stability of preferences of matching-to-sample tasks. Psychological Record, 64,
older adults with dementia. Journal of Applied Behav- 827–839.
ior Analysis. 46, 765–780. Stock, L. Z., & Milan, M. A. (1993). Improving dietary
Ritchie, H., Reuter-Yuill, L., Perez, A., & Baker, J. C. practices of elderly individuals: The power of prompt-
(2019). Assessment and treatment of verbal behavior ing, feedback, and social reinforcement. Journal of
deficits with an older adult diagnosed with aphasia and Applied Behavior Analysis, 26, 379–387.
a comparison of transfer of stimulus control procedures. Sundberg, M. L. (1991). 301 research topics from Skin-
Manuscript accepted for publication. ner’s book Verbal Behavior. Analysis of Verbal Behav-
Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, ior, 9, 81–96.
B. A. (1998). Evaluation of a brief stimulus preference Teri, L. (1991). Behavioral assessment and treatment of
assessment. Journal of Applied Behavior Analysis, 31, depression in older adults. In P. Wisocki (Ed.), Hand-
605–620. book of clinical behavior therapy with the elderly client
Sands, D., & Holman, E. (1985). Does knowledge en- (pp. 225–240). New York: Plenum Press.
hance patient compliance? Journal of Gerontological Teri, L., & Logsdon, R. G. (1991). Identifying pleas-
Nursing, 11, 23–29. ant activities for Alzheimer’s disease patients: The
528 S u b s p e c i a lt ie s in A p p l ie d B e h av i o r A n a ly s i s
Pleasant Events Schedule—AD. Gerontologist, 31, Foley, D. J., & Phillips, C. (1994). The burden and
124–127. outcomes associated with dehydration among U.S.
Thompson, L. W., Futterman, A., & Gallagher, D. elderly, 1991. American Journal of Public Health, 84,
(1988). Assessment of late-life depression. Psycho- 1265–1269.
pharmacology Bulletin, 24, 577–586. Wild, B., Eckl, A., Herzog, W., Niehoff, D., Lechner, S.,
Tolson, D., Rolland, Y., Andrieu, S., Aquino, J. P., Beard, Maatouk, I., et al. (2010). Assessing generalized anxi-
J., Benetos, A., et al. (2011). International association ety disorder in elderly people using the GAD-7 and
of gerontology and geriatrics: A global agenda for GAD-2 scales: Results of a validation study. Ameri-
clinical research and quality of life in nursing homes. can Journal of Geriatric Psychiatry, 22, 1029–1039.
Journal of the American Medical Directors Association, Williamson, P. N., & Ascione, F. R. (1983). Behavioral
12, 184–189. treatment of the elderly: Implications for theory and
Tonga, J. B., Karlsoeen, B. B., Arnevik, E. A., Werheid, therapy. Behavior Modification, 7, 583–610.
K., Korsnes, M. S., & Ulstein, I. D. (2015). Chal- Wilson, K. M., & Milan, M. A. (1995). Age differences
lenges with manual-based multimodal psychotherapy in the formation of equivalence classes. Journal of
for people with Alzheimer’s disease: A case study. Gerontology: Series B: Psychological Sciences and So-
American Journal of Alzheimer’s Disease and Other cial Sciences, 50B, 212–218.
Dementias, 31(4), 311–317. Woodruff-Pak, D. (2001). Eyeblink classical condition-
Trahan, M. A., Donaldson, J. M., McNabney, M. K., & ing differentiates normal aging from Alzheimer’s dis-
Kahng, S. (2014). Training and maintenance of a pic- ease. Integrative Physiological and Behavioral Science,
ture-based communication response in older adults 36, 87–108.
with dementia. Journal of Applied Behavior Analysis, Woodruff-Pak, D., & Jaeger, M. (1998). Predictors of
47, 404–409. eyeblink classical conditioning over the life span.
Tripp, G., & Alsop, B. (1999). Age-related changes in Psychology and Aging, 13, 193–205.
sensitivity to relative reward frequency. New Zealand Woodruff-Pak, D., Papka, M., Romano, S., & Li, Y.
Journal of Psychology, 28, 30–36. (1996). Eyeblink classical conditioning in Alzheim-
Valdovinos, M. G., & Kennedy, C. H. (2004). A be- er’s disease and cerebrovascular dementia. Neurobiol-
havior-analytic conceptualization of the side effects ogy of Aging, 17, 505–512.
of psychotropic medication. Behavior Analyst, 27, Woodruff-Pak, D., & Thompson, R. (1988). Classical
231–238. conditioning of the eyeblink response in the delay
Virues-Ortega, J., Iwata, B. A., Nogales-González, C., & paradigm in adults aged 18–23. Psychology and Aging,
Frades, B. (2012). Assessment of preference for edible 3, 219–229.
and leisure items in individuals with dementia. Jour- Zarit, S., & Zarit, J. (1998). Mental disorders in older
nal of Applied Behavior Analysis, 45, 839–844. adults: Fundamentals of assessment and treatment.
Warren, J. L., Bacon, W. E., Harris, T., McBean, A. M., New York: Guilford Press.
CHAP TER 31
Applied behavior analysts at the Center for Dis- exchange information from one site to another
abilities and Development and the Department via electronic communications to improve health
of Pediatrics at The University of Iowa have used outcomes, lower costs, and provide better care
telehealth to assess and treat problem behavior (American Telemedicine Association, 2016; Cen-
since the late 1990s (see Barretto, Wacker, Hard- ter for Connected Health Policy, 2016; Committee
ing, Lee, & Berg, 2006). The Center for Disabilities on Pediatric Workforce, 2015). Throughout this
and Development has had a telehealth center for chapter, we use the term telehealth interchangeably
over 10 years. Federal grants and the Department with the traditional term, telemedicine. Much of
of Pediatrics have supported our involvement in the telehealth literature demonstrates its applica-
telehealth, which is part of a national movement tion to healthcare in four broad categories: consul-
to use telehealth services across all subdisciplines tation, diagnosis, training, and intervention (see
in medicine. Our hospital recently developed an Figure 31.1). According to the Center for Connect-
e-health department, whose specific function is ed Health Policy (2016), the electronic communi-
to advance telehealth services throughout Iowa cations most commonly used in telehealth include
while addressing issues ranging from reimburse- synchronous interactions, asynchronous interactions,
ment to technology. In this chapter, we first de- remote patient monitoring, and mobile health. Syn-
fine telehealth and briefly describe how pediatric chronous interactions (aka: live video or real time)
subspecialties and applied behavior analysts have are live, two-way interactions between a client
used telehealth services nationally. We specifically (i.e., a patient, care provider, or professional) and a
describe the development of telehealth services provider, whereas asynchronous interactions (aka:
at Iowa and current applications, and we discuss store and forward) are transmissions of prerecord-
questions and concerns regarding further expan- ed health information to a provider who evaluates
sion of telehealth services. We conclude the chap- the information later. Providers use remote patient
ter by suggesting some future directions that ap- monitoring and mobile health to track and moni-
plied behavior analysts might consider. tor patient health outcomes and to provide health
education and practice to promote healthy behav-
ior, respectively.
DEFINITION OF TELEHEALTH Telehealth is not considered a replacement for
in-person health care delivery, but rather a tool or
Telehealth is an umbrella term that encompasses supplement to deliver the same or an enhanced
many health-related services using technology to level of care typically provided at a distance. The
529
FIGURE 31.1. Example of applied behavior analysis services conducted via telehealth for problem behavior.
530
Telehealth and Applied Behavior Analysis 531
about the patient and uploading any still or mo- telehealth (Heffner, Lyon, Brousseau, Holland, &
tion images to a secure, encrypted website. Sub- Yen, 2009).
specialists receive notification of the consultation
submission; review the patient’s information; and
Training
provide recommendations on diagnosis, inter-
vention, or both on the website. Results of this Telehealth training refers to using technology to
study demonstrated that primary care physicians conduct specialized training on specific topics
received consultations from 33 different pediatric and to engage in online discussions with groups
subspecialists, and that the median consult re- to provide peer-to-peer support and education
sponse time was 12 hours. In addition, the qual- (American Telemedicine Association, 2016). The
ity of care improved (as shown by changes in di- primary example of training in medicine has been
agnosis, the diagnostic plan, or the intervention the Extension for Community Healthcare Out-
plan) for 15%, 21%, and 24% of cases, respectively. comes project or Project ECHO, which was “de-
Finally, costs decreased because patients avoided signed to improve patient care by developing and
travel to the medical center (typically a 5-hour supporting the competence of primary care pro-
plane flight and 1-week hospital stay) for 12% of viders in underserved areas to manage complex
cases. Similar peer-to-peer consultations conduct- disorders” (Arora et al., 2007, p. 154). Research-
ed in real time have demonstrated changes in di- ers first developed Project ECHO at the Univer-
agnosis, diagnostic plans, or intervention plans for sity of New Mexico to address high rates of the
children with obesity (Shaikh, Cole, Marcin, & Hepatitis C virus, which is amenable to treatment
Nesbitt, 2008), and have produced decreased pa- but time-consuming to treat and often beyond
tient transfers and costs for children presenting to the scope of practice for many primary care physi-
rural emergency departments with acute illness or cians. The Project ECHO training model focuses
injury that required immediate physician involve- on case-based, disease-specific learning during
ment (Yang et al., 2015). weekly 2-hour virtual visits in which primary
care physicians present relevant cases from their
practices, and specialists instruct and guide these
Diagnosis
physicians through appropriate care management
During telehealth diagnosis, a specialist makes or as- (Arora et al., 2011). The benefits of this training
sists another provider to make a diagnosis for a pa- model are twofold. First, primary care physicians
tient (American Telemedicine Association, 2016), develop content knowledge and skills for manag-
using synchronous and asynchronous formats. An ing specific illnesses, as well as a network of col-
off-site pediatric cardiologist, for example, might leagues to consult with on arising issues (Arora et
watch and listen to a live heart study in real time al., 2011). Second, patients receive safer and more
while guiding the on-site provider to obtain ad- comprehensive care for illnesses that typically re-
ditional views of the heart during synchronous quire specialty services often not available readily
diagnostic consultation. By contrast, an off-site in the patients’ communities (Arora et al., 2007).
cardiologist using asynchronous diagnostic con- After the initial success of Project ECHO, others
sultation might receive still, motion, or both kinds have applied the program to many specialty health
of images to review and interpret later (Sable, care needs, including pediatric obesity (Arora
2003). Research has shown that these telehealth et al., 2007), early detection of autism spectrum
diagnostic methods in pediatric cardiology are ef- disorder (ASD) (Arora et al., 2011), treatment
fective for many pediatric populations, including for ASD symptoms (Mazurek, Brown, Curran, &
fetuses and neonates (Gomes et al., 2010), and for Sohl, 2017), and children and youth with epilepsy
diagnoses ranging from no heart disease to patho- (American Academy of Pediatrics, 2016).
logical heart disease (Gomes et al., 2010; Mahnke
et al., 2008). Specifically, results of these studies
Intervention
indicated that health care providers made or con-
firmed accurate diagnoses via telehealth, which in Telehealth intervention refers to an off-site profes-
turn guided treatment plan initiations or changes sional providing intervention services, often indi-
and contributed to decisions regarding the neces- rectly via remote patient monitoring (American
sity of emergency patient transfers. Research has Telemedicine Association, 2016). For example,
produced similar results in pediatric dermatology, Ghio et al. (2002) used telehealth with two chil-
in which health care professionals provided accu- dren who were receiving automated peritoneal
rate diagnoses for many dermatological issues via dialysis treatment. The system transmitted and
Telehealth and Applied Behavior Analysis 533
stored the peritoneal dialysis data from the chil- fied applied behavior analysts in 2016 (Behavior
dren’s peritoneal dialysis cycler to a computer in Analysis Certification Board, 2016). Currently,
the dialysis lab. Thus the physician could remotely our center continues to be the only provider of
monitor skipped or shortened treatment, chang- function-based assessment and treatment services
es to the therapy parameters, skipped phases or in Iowa. The current wait time for services is 8
cycles, or reduced fill volumes, which indicated months from the time we schedule the clinic visit
whether the patients and families were comply- to the time when we conduct the evaluation. Most
ing with the dialysis treatment. In addition, real- families receiving the services live within 100
time telehealth visits occurred at least every 3 days miles of the tertiary-level hospital where the clinic
to follow up on collected data, compliance with is located, leaving most of Iowa without access to
treatment, additional clinical problems, and tech- these services.
nical problems. Results of the remote-monitoring We launched our telehealth services in the mid-
data demonstrated that the children in this study to late 1990s to increase access to ABA services
always complied with the dialysis treatment. Over- for Iowans (Lee et al., 2014). Our outpatient clinic,
all, these results demonstrated the effectiveness of the Biobehavioral Service, initiated telehealth
telehealth for home dialysis patients. Izquierdo et services as part of a much larger telehealth grant
al. (2009) produced similar results for children to The University of Iowa Hospitals and Clinics
with Type 1 diabetes, in which the nurses at the (Kienzle, 2000). We conducted only consultation
children’s school connected with the diabetes services for the first several years, meaning that we
center via telehealth to obtain consultation and consulted with local pediatricians, nurse practitio-
education. Researchers observed decreases in he- ners, and school teams regarding the problem be-
moglobin A1c and emergency visits to the nurses. havior they were encountering in their clinics and
classrooms. These were “talking heads” conversa-
tions, and we never saw the clients or practitioners
Selected Applications across ABA Services
in action. The local service providers appreciated
Clinical behavior-analytic researchers have inte- the functional approach and expertise we brought
grated telehealth into ongoing delivery of services, to managing behavior.
such as those in pediatric subspecialties. Wacker The telehealth center at the Center for Disabili-
et al. (2016) conceptualized the research and ap- ties and Development currently has four computer
plication of telehealth in ABA in a generational workstations devoted to telehealth services. Par-
framework. The first generation included research titions seclude these spaces (see Lee et al., 2014,
and applications on consultation, diagnosis, train- for a description), and our telehealth providers
ing, and intervention (see bottom panel of Figure use headphones to maintain patient privacy. Our
31.1), and the second generation of research and biobehavioral telehealth service assesses and treats
applications extended these four broad areas to problem behavior displayed by persons with devel-
further identify best practices for ABA telehealth. opmental disabilities. Consultation continues to
be the most common type of telehealth service
this clinic provides. We have scheduled appoint-
Consultation
ments routinely to consult via telehealth with
We began providing outpatient services to chil- specific school teams, ASD centers, and group
dren and adults with developmental disabilities homes. Most consultations involve the discussion
who displayed severe problem behavior at the of clients that we have never seen in the clinic,
Center for Disabilities and Development in Iowa and we advise professionals how to proceed with
during the mid-1980s (see Wacker et al., 2016, and on-site assessment and treatment. Increasingly, we
Wacker, Schieltz, & Romani, 2015, for descrip- are using telehealth to follow up with patients in
tions). When we initiated the outpatient clinic, their homes after initial evaluations in the clinic.
we scheduled roughly two children or adults every We advise caregivers on proceeding with treat-
other week, and this was sufficient to meet the ment, on incorporating treatment into their daily
need for this service. Over time, however, the de- routines, and on promoting the generalization and
mand for the service continued to grow, just as the maintenance of treatment effects. We have not
demand for ABA services has grown throughout obtained reimbursement from insurance compa-
the United States. The ability of universities and nies for these services. We recently received infor-
training programs to produce trained applied be- mation suggesting that we may be able to charge
havior analysts has not matched the demand for for these services soon, including clinic-to-home
these services. For example, Iowa had 77 certi- consultations. We also have begun to expand
534 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
real-time didactic training and real-time coaching. training with young children with ASD. They
Researchers have developed and tested web-based conducted a randomized clinical trial, with partic-
tutorials to train users on intervention techniques. ipants assigned to a telehealth or an in vivo group.
For example, Kobak et al. (2011) developed a web- Participants practiced implementing the discrete-
based educational program for caregivers of young trial training procedures with the children and re-
children with ASD to increase their children’s so- ceived real-time coaching (e.g., instructions, mod-
cial communication skills and to manage problem eling) and feedback on their performance. Results
behaviors. Behavior management strategies target- showed improvements in discrete-trial training
ed in the web-based modules included reinforce- implementation skills across both groups, with no
ment, modeling, and prompting. The modules significant difference between groups on fidelity
contained videos demonstrating child behavior of implementation, suggesting that telehealth was
and caregiver models implementing the behavior an effective modality for providing staff training.
management strategies. Caregivers completed the Studies using telehealth to train novice gradu-
web-based modules at their own pace. Pre- and ate students to conduct preference assessments
posttest measures demonstrated an increase in the (Machalicek, O’Reilly, Chan, Rispoli, et al., 2009),
caregivers’ knowledge of behavior-analytic prin- and novice teachers to conduct functional analysis
ciples, as well as high caregiver satisfaction with (Frieder et al., 2009; Machalicek et al., 2010) and
technology-based training. functional communication training (Gibson et al.,
An alternative to online instruction is provid- 2010), with children with ASD in school settings
ing real-time didactic training via telehealth. The have produced similar findings.
trainer connects remotely with the user (usually Researchers have used telehealth to train
a caregiver, therapist, or teacher) to provide real- caregivers to implement intervention programs.
time instruction. Xie et al. (2013) conducted a McDuffie et al. (2013) used real-time coaching
randomized clinical trial of didactic training via via telehealth to train caregivers of children with
telehealth with caregivers of young children with ASD to implement a play-based intervention to
attention-deficit/hyperactivity disorder (ADHD). facilitate language development. The researchers
The investigators assigned caregivers to either the first provided the caregivers with in vivo didactic
telehealth or face-to-face training groups. Train- training on component skills of the interven-
ing targeted basic behavior management strategies tion, followed by in vivo coaching in which the
(e.g., establishing a token economy, using time caregivers practiced the skills with their children
out effectively, delivering demands appropriately). while a therapist coached them. The researchers
The trainers provided instruction only on the tar- then coached the caregivers via telehealth while
geted skills, and they did not incorporate real-time the caregivers practiced the procedures in their
coaching with the children. The results demon- homes. Results showed that caregivers’ use of the
strated that the caregivers’ disciplinary practices targeted skills increased after the in vivo coaching,
improved, and that ADHD and problem behavior and that skills were maintained when researchers
symptoms decreased on standardized rating scales. provided coaching via telehealth.
Treatment effects did not differ between training More recently, telehealth studies have com-
groups. Reese, Slone, Soares, and Sprang (2015) bined modalities (e.g., online instruction with re-
produced similar results when they trained care- mote coaching) into one training program. Fisher
givers of children with ADHD in their homes via et al. (2014) developed a virtual program to train
telehealth to establish positive parenting practices. participants on ABA principles and procedures in
Although online instruction and real-time di- discrete-trial and play-based formats. Participants
dactic training allow users to acquire new skills, first completed online modules with videos that
they have some limitations, such as the amount described and demonstrated ABA principles and
and quality of feedback received during training procedures. Each participant then practiced the
if researchers do not incorporate real-time coach- procedures with a confederate while a therapist
ing with confederates or children in the program. provided coaching and feedback via telehealth.
Researchers have used real-time coaching via tele- Participants significantly improved their knowl-
health to train caregivers, teachers, and therapists edge and implementation of ABA principles and
on specific skills, and the trainees have practiced procedures after the completion of the virtual
the skills with confederates or children while re- training program, again suggesting that telehealth
ceiving immediate feedback from the trainers. For is an effective way to provide training. A study con-
example, Hay-Hansson and Eldevik (2013) used ducted by Heitzman-Powell et al. (2014) also dem-
telehealth to train staff to implement discrete-trial onstrated the effectiveness of a telehealth training
536 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
program that incorporated online instruction and was 94%. Caregivers rated telehealth delivery of
real-time coaching to train caregivers of children FCT very highly on the Treatment Acceptability
with ASD to implement ABA-based procedures. Rating Form—Revised. Ratings were comparable
We have combined modalities for our training to those obtained when trained applied behavior
project for school professionals, in which we pro- analysts traveled to the caregivers’ homes and
vide real-time didactic instruction via telehealth on worked directly with them (Reimers, Wacker,
behavioral principles and assessment procedures as Cooper, & DeRaad, 1992).
a supplement to the in-person training each pro- In a more recently completed project (Lindgren
fessional receives from our team (Bassingthwaite & Wacker, 2011, 2015), we implemented these pro-
& Wacker, 2015). Over the years of this project, cedures via telehealth, but in caregivers’ homes.
this didactic instruction has evolved from the Our previous clinic-to-clinic results had established
experts’ providing the instruction to the school that we could conduct the procedures effectively
professionals to the school professionals’ providing via telehealth, which was important for establish-
the instruction to each other, which may be like ing insurance reimbursement and expanding ac-
the premise of Project ECHO mentioned earlier. cess to care. However, determining whether we
In addition to didactic instruction, we provide re- could obtain the same results in caregivers’ homes
al-time training. Project staff members observe the with no in vivo assistance, and when the caregivers
performance of school teams while they evaluate had never met the applied behavior analysts work-
a child. As with the didactic training, school staff ing with them, was important. As summarized in
members that our research team initially trained Lindgren et al. (2016), we worked with 30 families
are now training other school staffers while our and achieved a 97% mean reduction in levels of
project staff members monitor them in real time problem behavior. As in prior studies, caregivers
via telehealth. In one case, a school team conduct- rated the acceptability of the procedures very high-
ed a functional analysis at a school site while proj- ly. Statistical analyses showed that the in-home in
ect staffers observed and coached the school team, vivo, clinic telehealth, and home telehealth groups
as needed, from a remote location (Bassingthwaite showed (1) no significant differences in reductions
et al., 2018). of problem behavior, (2) equivalent improvement
in adaptive behavior such as task completion, and
(3) equivalent caregiver acceptability ratings. Thus
Intervention
these results indicated that functional analysis and
The treatment procedure that we conduct most FCT conducted via telehealth in caregivers’ homes
often via telehealth is functional communication were as effective as those conducted in vivo and in
training (FCT), for several reasons (Wacker et regional outpatient clinics, respectively, in at least
al., 2013a). First, when we initiated our in-home some cases. In addition, the costs associated with
(Wacker et al., 1998), outpatient (Northup et al., the in-home telehealth services were lower than
1991), and inpatient (Wacker et al., 1990) treat- those of other treatment delivery methods.
ment programs, we most often used FCT for treat- Monitoring procedural fidelity is a critical as-
ment. As research by our team and others has pect of behavioral treatments delivered via tele-
shown (Greer, Fisher, Saini, Owen, & Jones, 2016; health. Applied behavior analysts have begun
Tiger, Hanley, & Bruzek, 2008), FCT is often ef- systematically evaluating the impact of different
fective in reducing problem behavior, and caregiv- levels of fidelity on the effects of behavioral treat-
ers have rated it as highly acceptable. The goal of ments, and the results to date suggest that treat-
replacing problem behavior with an appropriate ment fidelity can play a large role in the initial
communicative response has made good sense to effects of treatment and in the maintenance of
caregivers, and we based the treatment directly on treatment effects over time (Bergmann, Kodak, &
the results of the functional analysis. Given both LeBlanc, 2017; Carroll, Kodak, & Fisher, 2013; St.
the reductions in problem behavior and the high Peter Pipkin, Vollmer, & Sloman, 2010; Volkert,
acceptability ratings, we have continued to use Lerman, Call, & Trosclair-Lasserre, 2009). Tele-
FCT as our treatment of choice. health severely restricts the way behavior ana-
We conducted our first systematic effort to im- lysts can implement and model the recommended
plement FCT via telehealth in regional outpatient treatments; therefore, close monitoring of treat-
clinics with 17 families of young children with ment implementation fidelity by caregivers and
ASD who displayed problem behavior (Wacker et other service providers is critical.
al., 2013a). Seventeen children received FCT, and Suess et al. (2014) evaluated caregiver fidelity
the mean decrease in levels of problem behavior during FCT in the home with applied behavior
Telehealth and Applied Behavior Analysis 537
analysts conducting remote coaching. Suess et al. For the telehealth visits, the caregivers and chil-
showed that caregivers could conduct the proce- dren traveled to a regional ASD center near their
dure in their homes with good integrity, and that homes. The caregivers implemented the functional
they could achieve clinically significant reductions analysis and the FCT procedures, with an applied
in their children’s problem behavior. Although behavior analyst providing coaching via Skype™.
Suess et al. produced good fidelity in the home via The applied behavior analysts were in the tele-
telehealth, the results of other studies suggest that health center located at the University of Iowa
errors in implementation integrity can negatively Children’s Hospital, which was approximately 170
affect treatment (Schieltz et al., 2018; St. Peter miles from the regional ASD center (Wacker et al.,
Pipkin et al., 2010). Researchers are conducting 2013b). The director of the regional ASD center
studies to identify the best procedures for in-home also participated in the telehealth visits by helping
caregiver training via telehealth (e.g., Fisher et al., the caregivers as needed (e.g., making sure that ap-
2014), and the results of those studies will be im- propriate toys and materials were available).
portant to the success of future applications and During the visit for the functional analysis,
extensions of telehealth services. our team completed a behavioral diagnosis. That
is, we conducted tests for positive-reinforcement
functions (tangibles and attention) and negative-
Current Applications
reinforcement functions (escape) in a multiele-
Lindgren et al. (2016) showed that we can conduct ment design for each child. We targeted negative-
behavioral assessment and treatment procedures reinforcement functions during treatment, based
effectively via telehealth. Suess, Wacker, Schwartz, on the functional analysis results. We imple-
Lustig, and Detrick (2016) suggest that the next mented and evaluated treatment in a nonconcur-
generation of questions will focus on timing and rent multiple-baseline design across participants.
dose of services. Relative to in vivo clinical pro- Treatment was FCT in a two-step chain of task
cedures, when should we use telehealth, for how completion and manding described by Wacker et
long, and on what schedule in relation to clinic, al. (2013a). We encouraged caregivers to practice
home, and school-based procedures? the FCT procedure with their children outside of
We have incorporated telehealth evaluations scheduled telehealth visits (Suess et al., 2014), and
into our biobehavioral service to provide brief our applied behavior analysts provided additional
behavioral assessment and treatment services to brief consultation via telehealth.
caregivers of children with ASD, who previously The results from the brief behavioral diagnoses
often waited up to 6 months for an in-person eval- via telehealth showed social functions of problem
uation in our clinic (see Suess et al., 2016). Five behavior for four out of five children. Problem be-
caregiver–child dyads received brief behavioral havior decreased relative to baseline by a mean of
diagnosis, intervention, and consultation via tele- 65% across children by the end of the third FCT
health. Each child participated in four telehealth visit. A statistical analysis showed a statistically
visits, one 60-minute visit for a functional analysis, significant reduction in levels of problem behav-
and three 15-minute FCT visits. The total time for ior from the functional analysis to the end of FCT
the telehealth assessment, 105 minutes, compared (Hedge’s g = 1.3, z = 3.15, p < .001). Mean increas-
favorably to the length of our 90- to 120-minute es in independent manding and task completion
clinic evaluations. However, rather than imple- were 88% and 34%, respectively.
menting functional analysis and FCT in one visit These results extended previous telehealth
as we do during in-clinic evaluations, we imple- studies (Wacker et al., 2013a, 2013b) by showing
mented the procedures during four telehealth vis- that behavior analysts could incorporate brief
its. Implementing frequent but brief visits allowed behavioral assessment and treatment procedures
us to integrate telehealth services more efficiently via telehealth into an existing outpatient service.
into our existing clinic caseload. This integration Brief behavioral assessment and treatment via
was important, because we still had not received telehealth may provide a means to initiate services
reimbursement for our telehealth services. while families remain on the wait list for clinic ser-
One question that researchers should address in vices. If the treatment initiated via telehealth is
future studies is how best to incorporate telehealth sufficient to reduce problem behavior, then we can
into existing services. For example, we wonder cancel the clinic evaluation and decrease the size
whether more frequent but brief sessions will be of our clinic wait list. However, if problem behav-
more effective and efficient than longer sessions ior continues, we can conduct a more extensive
distributed over a longer period. behavioral assessment and treatment in the clinic,
538 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
and we can tailor the clinic services directly to the (1982/1994) described. The pragmatic group will
telehealth results. receive a 1-hour descriptive assessment that does
One concern with behavioral telehealth re- not manipulate conditions in a single-case design.
lates to their security and ability to protect pa- Our goal is to determine whether the potential
tient confidentiality. We have used Skype as clinical benefits of the standard functional analy-
the video-conferencing software in some of our sis are significantly better than a more pragmatic
previous telehealth studies, but we informed the version that we will evaluate in a systematic and
caregivers in those studies of the security limits of quantitative fashion.
telehealth and had them provide written consent We believe that behavior analysts can conduct
before starting telehealth visits. More recently, we clinical and translational research via telehealth
have collaborated with the e-health group at the (e.g., Suess, Schieltz, Wacker, Detrick, & Podle-
University of Iowa Hospitals and Clinics to pilot- snik, 2020). We hope that these and other studies
test video-conferencing software that is compliant support the merits of using telehealth as a practi-
with the Health Insurance Portability and Ac- cal means to study problem behavior and deliver
countability Act (HIPAA) and directly connected effective interventions to families affected by such
to the patients’ electronic medical records. This behavior.
software allows patients to log into their medical
records from their homes and initiate telehealth
visits with providers at the University of Iowa LINGERING QUESTIONS AND CONCERNS
Hospitals and Clinics. We used this new software
initially to provide behavioral consultation for Behavior analysts have studied telehealth only as
feeding-related concerns, but we hope to expand a means for directly delivering behavioral assess-
it soon to the myriad of telehealth-based services ment and treatment via single-case designs since
we have summarized in this chapter. Other behav- 2006, and Internet access has been available wide-
ior analysts who wish to deliver ABA services via ly for only the last several years. Numerous ques-
telehealth should be aware of the health informa- tions and concerns regarding the future delivery of
tion privacy issues that are specific to this service these services remain.
delivery mode. Questions that researchers need to address im-
The results of Lindgren et al. (2016) suggest that mediately relate to the timing and dose of servic-
behavior analysts can conduct research and clini- es. Although researchers like Suess et al. (2016)
cal services successfully via telehealth. Research showed what we can accomplish via telehealth, we
conducted via telehealth can potentially increase also need to identify the limitations that warrant
the access that families have to relevant research in vivo services (see Schieltz et al., 2018). Wacker
protocols and can increase the diversity of partici- et al. (2016) offered a flowchart of questions to
pants enrolling in projects at The University of ask to determine when behavior analysts should
Iowa. We recently completed a randomized clini- select either telehealth or in vivo service delivery.
cal trial of FCT via a project funded by the Health For example, the first three questions were these:
Resources and Services Administration (HRSA) (1) Can we assess and treat the target behavior
Maternal and Child Health Bureau (Lindgren remotely? (2) Does the family have the necessary
& Wacker, 2011). In this project, participants equipment? and (3) Will insurance reimburse the
received FCT either immediately after we com- service? Researchers need to determine, study,
pleted the functional analysis or 3 months later. evaluate, and resolve these and many other ques-
We successfully completed this study entirely via tions as they consider wider clinical applications of
telehealth and in a much shorter period than if telehealth. Similarly, we need to evaluate whether
we had conducted the project in the home or the telehealth components, such as the visual presence
clinic (Lindgren et al., 2020). of a consultant or the caregiver’s wearing earbuds,
We have received funding from the National functions as an inadvertent independent variable
Institute of Mental Health to conduct a random- and alters behavior. If so, then further analysis of
ized clinical trial of functional analysis (Lindgren its impact on various dependent variables seems
& Wacker, 2015). We will randomly assign families warranted.
of young children with ASD who display problem Professional organizations have outlined ethi-
behavior to a group receiving either standard or cal guidelines for providing services via telehealth
pragmatic functional analysis. The standard group (American Psychological Association, 2013; Be-
will receive the functional analysis Iwata et al. havior Analysis Certification Board, 2014), but
Telehealth and Applied Behavior Analysis 539
this remains an area for professional discussion et al. (2016) scheduled three 15-minute treatment
(Pollard, Karimi, & Ficcaglia, 2017; Romani & sessions rather than one 45-minute session. But are
Schieltz, 2017). For example, under what condi- more frequent but briefer sessions easier or more
tions can we use family-friendly systems such as effective for caregivers?
Skype or FaceTime™ with informed consent to Another issue that we have considered con-
assess and treat problem behavior? Are different cerns a child’s behavior and a family’s situation.
levels of security needed for talking consultation Children who scream loudly and live in an apart-
versus direct implementation of procedures when ment may not be good candidates for telehealth-
children appear on screen, for example? If we re- based functional analyses, for example. Children
cord such sessions for later viewing using store- with problem behavior that presents a danger to
and-forward methods, how should we store and themselves, to others, to property, or a combina-
protect those video files? How can we best use tele- tion may also be difficult to assess and treat via
health in settings such as schools, day care centers, telehealth. Caregivers who cannot follow a consul-
and group homes? tant’s vocal instructions or who do not have sup-
Behavior analysts may need a different skill set port at home may not be candidates for telehealth,
for telehealth services than for in-person services. either. Most of our telehealth-based treatments
Many applied behavior analysts model the skills have been for problem behavior maintained by
they are trying to teach a caregiver. We frequently social reinforcement. Reducing problem behavior
use gestures, such as pointing, and may have care- maintained by automatic reinforcement has been
givers watch us implement a procedure several more challenging with telehealth (Schieltz et al.,
times before asking the caregivers to try it. We 2018).
frequently work directly with the children during A final consideration is that of reinforcers for
meltdowns or extinction bursts. Telehealth service the applied behavior analysts who are conduct-
delivery requires a different set of strategies, relying ing telehealth. Over the past 10 years that we
more on vocal instructions or video models. We have been conducting telehealth, not a single staff
need to provide clear and specific vocal instruc- member or student has indicated that he or she
tions during telehealth services, and we often must would prefer to conduct sessions exclusively via
prompt desired caregiver behavior to prevent fidel- telehealth. Most of us find working directly with
ity errors. Caregivers are likely to implement pro- clients and caregivers reinforcing. As supervisors,
cedures differently than we do, and as consultants, we find watching how skilled our trainees become
we must quickly discriminate whether variations reinforcing. We cannot completely remove these
in procedures are acceptable or constitute fidelity sources of positive reinforcement without consid-
errors. Thus we need to consider the skills applied ering what will replace those sources. As we con-
behavior analysts need to have and the best ways sider telehealth options in the future, we will need
to teach those skills. to conduct research to address timing and dosage
In providing training to caregivers, we need to issues for the providers as well as the patients.
consider how best to use procedures such as video In general, the right mix of telehealth with
modeling and interactive webpages. Suess et al. clinic, in-home, center-based, and school-based
(2014) asked caregivers to practice the procedures programs warrants further consideration. The re-
between appointments, and an applied behavior sults to date certainly support telehealth, but we
analyst reviewed recordings of those sessions be- need further research to determine how best to
fore the next appointment with a caregiver. As we implement services via telehealth.
have done with the dosage and timing questions
raised earlier for delivering services via telehealth,
we need to address questions for what occurs dur- FUTURE DIRECTIONS
ing times when an applied behavior analyst is not
present. How much practice should caregivers The options available for future telehealth appli-
conduct, and is this a function of their skills? For cations and extensions appear to be limited only
example, should caregivers practice a procedure by the resourcefulness of applied behavior analysts
several times per day at the beginning of treat- who are implementing those services. Most medi-
ment, or should they wait until treatment effects cal subspecialties have embraced telehealth, and
have occurred in session before they practice pro- insurance providers have increasingly accepted
cedures in the home? Should we schedule addi- this approach. We are embarking on three sets of
tional but briefer sessions early in treatment? Suess future directions for our services in Iowa.
540 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
A doctoral dissertation (Lee, 2016) showed that that maximally benefit families. Service teams
a consultant in the United States could deliver across many disciplines have been studying how to
functional analysis and FCT to families in Korea. consult, diagnose, train, and treat via telehealth.
Five young children with ASD who displayed Many caregivers will need access to each of these
problem behavior participated in the study. The services, which means that we need to consider
children’s caregivers conducted assessment and how to blend services into an efficient and effec-
treatment procedures in their homes. The func- tive package. We need studies, for example, that
tional analysis identified social functions for the identify (1) the conditions under which caregiver-
children’s problem behavior, and FCT reduced training needs require in-home service delivery
problem behavior by 100% of baseline. We ob- of behavioral treatment, (2) the need for ongoing
tained these results by using smartphones for the consultation after treatment, and (3) the need for
telehealth consultations. functional analysis before treatment. We need re-
Similarly, Tsami, Lerman, and Toper-Korkmaz search to identify how to incorporate telehealth
(2019) conducted functional analyses and FCT services into existing service delivery programs.
via telehealth with 12 children with ASD who re- A final direction for the future is to determine
sided across eight countries. Behavior analysts in how to use telehealth to expand our research pro-
the United States served as the families’ coaches, grams. Telehealth offers opportunities for more di-
and when needed, used interpreters who had been verse groups of families to participate in research,
raised in the countries of the families. Across all including clinical trials. We have recently initi-
participants, results were positive. Specifically, ated a randomized clinical trial of preceding FCT
problem behavior was reduced by at least 80% with either a standard functional analysis or a brief
from baseline levels and approporiate communica- functional assessment. Three research centers at
tion increased to at least 90% across opportuni- Iowa, Atlanta, and Houston are working jointly
ties. Additionally, parent procedural fidelity was on the project; we will conduct assessment and
accurate (average of at least 90%), parent ratings treatment procedures via telehealth in the homes
of treatment acceptability were high (average of of the participating families. We can conduct this
6.6 on a 7 point Likert scale), and the telehealth large-N trial using single-case designs, because
connections between the United States and the telehealth permits us to work with far more fami-
families’ homes in other countries remained high lies than would be possible if we needed to travel
(at least 92% of the scheduled appointment times to the families’ homes. Families do not need to live
occurred without connection issues). near any of the centers conducting the study, po-
There are many implications of these studies, tentially increasing the overall diversity of partici-
and we mention two here. First, telehealth permits pating families.
applied behavior analysts to work efficiently and As we discussed in Schieltz and Wacker (2020),
immediately internationally. If future studies repli- the current COVID-19 crisis has dramatically
cate Lee (2016) and Tsami et al. (2019), we will be increased the relevance of telehealth-delivered
able to show the international appeal of ABA and services. The good news is researchers in behav-
to meet the worldwide need for the service. Sec- ior analysis have laid the foundation of how to
ond, we have most often used laptop computers for effectively deliver these services across consulta-
in-home telehealth services to date, but Lee and
tion, diagnosis, training, and intervention. The
Tsami et al. obtained equally positive results with
not so good news is there is still much to learn
smartphones and other electronic devices (e.g.,
relative to how best to utilize this service delivery
tablets). As part of our routine outpatient clinical
method. Although not systematically planned for,
services, we now use smartphones to conduct ses-
this health crisis has pushed more behavior ana-
sions via our electronic medical records. Within
lysts into using telehealth, and we hope that this
10 years, we have moved from a telehealth system
encourages these researchers and practictioners to
involving fiber-optic cables that connected only
contribute to our understanding of the use of tele-
a few locations to smartphones that can connect
health in behavior analysis.
almost everywhere. We need to consider carefully
how to maximize our use of telehealth to reach
as many families as possible with state-of-the-art REFERENCES
equipment and procedures.
A related future direction involves determining American Academy of Pediatrics. (2016). Epilepsy in
how to combine telehealth services into packages pediatrics: Project ECHO. Retrieved from www.
Telehealth and Applied Behavior Analysis 541
Izquierdo, R., Morin, P. C., Bratt, K., Moreau, Z., Meyer, A., Garwick, A. E., & Finkelstein, S. M. (2015). Ef-
S., Ploutz-Snyder, R., et al. (2009). School-centered fects of telehealth care coordination intervention on
telemedicine for children with Type 1 diabetes mel- perceptions of health care by caregivers of children
litus. Journal of Pediatrics, 155(3), 374–379. with medical complexity: A randomized controlled
Kienzle, M. (2000, March). Rural–academic integra- trial. Journal of Pediatric Health Care, 29(4), 352–363.
tion: Iowa’s national laboratory for the study of Machalicek, W., O’Reilly, M., Chan, J. M., Lang, R.,
rural telemedicine. Retrieved from http://collab.nlm. Rispoli, M., Davis, T., et al. (2009). Using video-
nih.gov/tutorialspublicationsandmaterials/telesympo- conferencing to conduct functional analysis of chal-
siumcd/UIowaFinalReport.pdf. lenging behavior and develop classroom behavioral
Kobak, K. A., Stone, W. L., Wallace, E., Warren, Z., support plans for students with autism. Education and
Swanson, A., & Robson, K. (2011). A web-based Training in Developmental Disabilities, 44, 207–217.
tutorial for parents of young children with autism: Machalicek, W., O’Reilly, M., Chan, J. M., Rispoli, M.,
Results from a pilot study. Telemedicine and e-Health, Lang, R., Davis, T., et al. (2009b). Using videocon-
17(10), 804–808. ferencing to support teachers to conduct preference
Langkamp, D., McManus, M. D., & Blakemore, S. D. assessments with students with autism and develop-
(2015). Telemedicine for children with developmen- mental disabilities. Research in Autism Spectrum Dis-
tal disabilities: A more effective clinical process than orders, 3, 32–41.
office-based care. Telemedicine and e-Health, 21(2), Machalicek, W., O’Reilly, M. F., Rispoli, M., Davis, T.,
110–114. Lang, R., Hetlinger Franco, J., et al. (2010). Train-
Lee, G. (2016). The application of telehealth procedures to ing teachers to assess the challenging behaviors of
provide behavior assessment and treatment to families students with autism using video tele-conferencing.
with young children with autism spectrum disorder in Education and Training in Autism and Developmental
Korea. Unpublished doctoral dissertation, The Uni- Disabilities, 45, 203–215.
versity of Iowa, Iowa City, IA. Mahnke, C. B., Mulreany, M. P., Inafuku, J., Abbas,
Lee, J. F., Schieltz, K. M., Suess, A. N., Wacker, D. P., M., Feingold, B., & Paolillo, J. A. (2008). Utility of
Romani, P. W., Lindgren, S. D., et al. (2015). Guide- store-and-forward pediatric telecardiology evaluation
lines for developing telehealth services and trouble- in distinguishing normal from pathologic pediatric
shooting problems with telehealth technology when heart sounds. Clinical Pediatrics, 47(9), 919–925.
coaching parents to conduct functional analyses and Mazurek, M. O., Brown, R., Curran, A., & Sohl, K.
functional communication training in their homes. (2017). ECHO autism: A new model for training pri-
Behavior Analysis in Practice, 8, 190–200. mary care providers in best-practice for children with
Lindgren, S. D., & Wacker, D. P. (2009). Behavioral autism. Clinical Pediatrics, 56, 247–256.
treatment for autism in community settings using a tele- McDuffie, A., Machalicek, W., Oakes, A., Haebig, E.,
health network. Washington, DC: U.S. Department Weismer, S. E., & Abbeduto, L. (2013). Distance
of Health and Human Services, National Institute of video-teleconferencing in early intervention: Pilot
Mental Health. study of a naturalistic parent-implemented language
Lindgren, S. D., & Wacker, D. P. (2011). Behavioral treat- intervention. Topics in Early Childhood Special Educa-
ment through in-home telehealth for young children with tion, 33(3), 172–185.
autism. Washington, DC: U.S. Department of Health Nelson, E., Citarelli, M., Cook, D., & Shaw, P. (2003).
and Human Services, Health Resources and Services Reshaping health care delivery for adolescent par-
Administration, Maternal and Child Health Bureau. ents: Healthy steps and telemedicine. Telemedicine
Lindgren, S. D., & Wacker, D. P. (2015). Comparing Journal and e-Health, 9(4), 387–392.
behavioral assessments using telehealth for children Northup, J., Wacker, D., Sasso, G., Steege, M., Cigrand,
with autism. Washington, DC: U.S. Department of K., Cook, J., et al. (1991). A brief functional analysis
Health and Human Services, National Institute of of aggressive and alternative behavior in an outclinic
Mental Health. setting. Journal of Applied Behavior Analysis, 24,
Lindgren, S., Wacker, D., Schieltz, K., Suess, A., Pelzel, 509–522.
K., Kopelman, T., et al. (2020). A randomized con- Pollard, J. S., Karimi, K. A., & Ficcaglia, M. B. (2017).
trolled trial of functional communication training Ethical considerations in the design and implemen-
via telehealth for young children with autism spec- tation of a telehealth service delivery model. Behavior
trum disorder. Journal of Autism and Developmental Analysis: Research and Practice, 17, 298–311.
Disorders. [Epub ahead of print] Reese, R. J., Slone, N. C., Soares, N., & Sprang, R.
Lindgren, S., Wacker, D., Suess, A., Schieltz, K., Pelzel, (2015). Using telepsychology to provide a group par-
K., Kopelman, T., et al. (2016). Telehealth expands enting program: A preliminary evaluation on effec-
access and reduces costs for treating challenging tiveness. Psychological Service, 12(3), 274–282.
behavior in young children with autism spectrum Reimers, T., Wacker, D., Cooper, L., & DeRaad, A.
disorders using applied behavior analysis. Pediatrics, (1992). Clinical evaluation of the variables associ-
137(Suppl. 2), S168–S175. ated with treatment acceptability and their relation
Looman, W. S., Antolick, M., Cady, R. G., Lunos, S. to compliance. Behavioral Disorders, 18, 67–76.
Telehealth and Applied Behavior Analysis 543
Romani, P. W., & Schieltz, K. M. (2017). Ethical consid- Tsami, L., Lerman, D., & Toper-Korkmaz, O. (2019). Ef-
erations when delivering behavior analytic services fectiveness and acceptability of parent training via
for problem behavior via telehealth. Behavior Analy- telehealth among families around the world. Journal
sis: Research and Practice, 17, 312–324. of Applied Behavior Analysis, 52, 1113–1129.
Sable, C. (2003). Telemedicine applications in pediatric Volkert, V. M., Lerman, D. C., Call, N. A., & Trosclair-
cardiology. Minerva Pediatrica, 55(1), 1–13. Lasserre, N. (2009). Evaluation of resurgence during
Santiago Lipana, L., Bindal, D., Nettiksimmons, J., & treatment with functional communication training.
Shaikh, U. (2013). Telemedicine and face-to-face Journal of Applied Behavior Analysis, 42, 145–160.
care for pediatric obesity. Telemedicine and e-Health, Wacker, D. P., Berg, W. K., Harding, J. W., Derby, K.
19(10), 806–808. M., Asmus, J. M., & Healy, A. (1998). Evaluation and
Schieltz, K. M., Romani, P. W., Wacker, D. P., Suess, A. long-term treatment of aberrant behavior displayed
N., Huang, P., Berg, W. K., et al. (2018). Single-case by young children with disabilities. Journal of Devel-
analysis to determine reasons for failure of behavioral opmental and Behavioral Pediatrics, 19, 260–266.
treatment via telehealth. Remedial and Special Educa- Wacker, D. P., Harding, J. W., Berg, W. K., Lee, J. F.,
tion, 39, 95–105. Schieltz, K. M., Padilla, Y. C., et al. (2011). An evalu-
Schieltz, K. M., & Wacker, D. P. (2020). Functional as- ation of persistence of treatment effects during long-
sessment and function-based treatment delivered via term treatment of destructive behavior. Journal of the
telehealth: A brief summary. Journal of Applied Be- Experimental Analysis of Behavior, 96, 261–282.
havior Analysis, 53, 1242–1258. Wacker, D. P., Lee, J. F., Padilla Dalmau, Y. C., Kopel-
Shaikh, U., Cole, S. L., Marcin, J. P., & Nesbitt, T. S. man, T. G., Lindgren, S. D., Kuhle, J., et al. (2013a).
(2008). Clinical management and patient outcomes Conducting functional communication training via
among children and adolescents receiving telemedi- telehealth to reduce the problem behavior of young
cine consultations for obesity. Telemedicine and e- children with autism. Journal of Developmental and
Health, 14(5), 434–440. Physical Disabilities, 25, 35–48.
Spooner, S. A., Gotleib, E. M., the Steering Committee Wacker, D. P., Lee, J. F., Padilla Dalmau, Y. C., Kopel-
on Clinical Information Technology, & the Com- man, T. G., Lindgren, S. D., Kuhle, J., et al. (2013b).
mittee on Medical Liability. (2004). Telemedicine: Conducting functional analyses of problem behavior
Pediatric applications. Pediatrics, 113(6), e639–e643. via telehealth. Journal of Applied Behavior Analysis,
St. Peter Pipkin, C., Vollmer, T., & Sloman, K. (2010). 46, 31–46.
Effects of treatment integrity failures during differen- Wacker, D. P., Schieltz, K. M., & Romani, P. (2015).
tial reinforcement of alternative behavior: A trans- Brief experimental analyses of problem behavior
lational model. Journal of Applied Behavior Analysis, in a pediatric outpatient clinic. In H. S. Roane, J.
43, 47–70. E. Ringdahl, & T. S. Falcomata (Eds.), Clinical and
Suess, A. N., Romani, P. W., Wacker, D. P., Dyson, S. organizational applications of applied behavior analysis
M., Kuhle, J. L., Lee, J. F., et al. (2014). Evaluating the (pp. 151–177). Amsterdam: Elsevier.
treatment fidelity of parents who conduct in-home Wacker, D. P., Schieltz, K. M., Suess, A. N., Romani,
functional communication training with coaching P. W., Padilla Dalmau, Y. C., Kopelman, T. G., et
via telehealth. Journal of Behavioral Education, 23, al. (2016). Telehealth. In N. N. Singh (Ed.), Clini-
34–59. cal handbook of evidence-based practices for individuals
Suess, A. N., Schieltz, K. M., Wacker, D. P., Detrick, J., with intellectual disabilities (pp. 585–614). New York:
& Podlesnik, C. A. (2020). An evaluation of resur- Springer.
gence following functional communication training Wacker, D. P., Steege, M. W., Northup, J., Sasso, G.,
conducted in alternative antecedent contexts via Berg, W., Reimers, T., et al. (1990). A component
telehealth. Journal of the Experimental Analysis of Be- analysis of functional communication training across
havior, 113, 278–301. three topographies of severe behavior problems. Jour-
Suess, A. N., Wacker, D., Schwartz, J. E., Lustig, N., & nal of Applied Behavior Analysis, 23, 417–429.
Detrick, J. (2016). Preliminary evidence in the use of Xie, Y., Dixon, F., Yee, O. M., Zhang, J., Chen, A.,
telehealth in an outpatient behavior clinic. Journal of DeAngelo, S., et al. (2013). A study on the effective-
Applied Behavior Analysis, 49(3), 686–692. ness of videoconferencing on teaching parent train-
Telehealth Resource Centers. (2016). Types of telemedi- ing skills to parents of children with ADHD. Tele-
cine specialty consultation services. Retrieved from medicine and e-Health, 19(3), 192–199.
www.telehealthresourcecenter.org/toolbox-module/ Yang, N. H., Dharmar, M., Yoo, B. K., Leigh, J. P., Kup-
types-telemedicine-specialty-consultation-services. permann, N., Romano, P. S., et al. (2015). Economic
Tiger, J. H., Hanley, G. P., & Bruzek, J. (2008). Func- evaluation of pediatric telemedicine consultation to
tional communication training: A review and practi- rural emergency departments. Medical Decision Mak-
cal guide. Behavior Analysis in Practice, 1, 16–23. ing, 35(6), 773–783.
CHAP TER 32
David A. Wilder and Nicole E. Gravina
Behavior analysts have used behavioral principles pothetical/deductive model of research, in which
successfully in a range of settings and with a vari- an investigator tests a hypothesis to determine the
ety of populations. We refer to the application of likelihood of its validity, is generally not the focus
these principles to individuals and groups in busi- in behavior analysis. Instead, behavior-analytic
ness, industry, government, and human-service researchers focus on manipulation of individual
settings as organizational behavior management environmental variables of interest. Subsequent
(OBM). In this chapter, we first provide a descrip- studies are based on the results of previous re-
tion of OBM and its relation to applied behavior search with no formal tested hypothesis (Wilder,
analysis and industrial–organizational psychology. Austin, & Casella, 2009).
Next, we briefly review the history of OBM. Final- OBM is also related to the field of industrial–
ly, we examine the various subdisciplines of OBM. organizational psychology, in that the focus of
In doing so, we focus on common assessment and both disciplines is the application of psychologi-
intervention strategies in each. cal or behavioral knowledge to work settings (see
Bucklin, Alvero, Dickinson, Austin, & Jackson,
2000, for a comparison). The Hawthorne studies,
OBM AND RELATED DISCIPLINES conducted in the 1920s at an electric power plant
in Illinois, influenced the two fields. One of the
OBM is a branch of applied behavior analysis, main findings of the Hawthorne studies was that a
which is a branch of the discipline of behav- variety of environmental changes in work settings,
ior analysis or the science of behavior. Behavior including simply observing employees, can affect
analysis also includes the experimental analysis worker performance and productivity.
of behavior (the basic science branch of the dis- Beyond these similarities, however, the two
cipline) and behaviorism (the conceptual or theo- fields differ in several ways. First, industrial–orga-
retical branch of the discipline). Applied behavior nizational psychology has an eclectic conceptual
analysis consists of the application of operant and, background. It derives the concepts on which it
to a lesser extent, respondent principles to behav- is based from several theoretical orientations. By
iors of social significance. Unlike mainstream psy- contrast, OBM has a strictly behavior-analytic
chology, behavior analysis has a unified theoreti- orientation. A second difference between the two
cal approach, behaviorism, and uses an inductive disciplines is the techniques in which the prac-
model of investigation. The more traditional hy- titioners of the two fields engage. Industrial–or-
544
Organizational Behavior Management 545
ganizational psychologists spend much of their sales by over $2 million in a single year (O’Brien,
professional time on selection and placement of Dickinson, & Rosow, 1982, p. 459). Fortune maga-
employees in organizations. By contrast, OBM zine featured a story about this success, which did
practitioners spend most of their time assessing the much to expand familiarity with OBM in the busi-
variables contributing to employee performance ness world. The discipline expanded rapidly in the
deficits and developing programs to improve per- late 1960s and 1970s, with professionals founding
formance for individual employees and systems. A consulting firms such as Praxis and Behavioral
third difference between the two disciplines is the Systems, Inc., and researchers founding the flag-
topics that each discipline studies. Industrial–or- ship journal in the field, the Journal of Organiza-
ganizational psychologists typically study person- tional Behavior Management. The first editor of the
nel selection and placement, organizational cul- Journal of Organizational Behavior Management,
ture, and leadership/decision making, and often Aubrey Daniels, was also the founder of one of the
use correlational or between-participants designs first OBM consulting firms, Behavioral Systems,
in their research. OBM researchers typically study Inc. Interestingly, Fran Tarkenton, a famed quar-
performance management (Griffin, Gravina, terback for the Minnesota Vikings and New York
Matey, Pritchard, & Wine, 2019), safety (Gravina, Giants, cofounded this company. Daniels later left
King, & Austin, 2019), and organizational systems Behavioral Systems and founded Aubrey Daniels
(Kelley & Gravina, 2018), and often use within- and Associates, which eventually became Aubrey
participants designs in their research (Bucklin et Daniels International. Aubrey Daniels Interna-
al., 2000; Vergason & Gravina, 2020). The disci- tional remains one of the top OBM consulting
plines also differ in size. The Society for Industrial firms (Dickinson, 2001).
and Organizational Psychology represents indus- The mission of the Journal of Organizational
trial–organizational psychology and has more Behavior Management was and still is to publish
than 9,600 members (www.siop.org/benefits). The articles on “scientific principles to improve orga-
Organizational Behavior Management Network nizational performance through behavior change”
represents OBM, is a special-interest group of the (http://obmnetwork.com/publications/journal-of-
Association for Behavior Analysis International, organizational-behavior-management-jobm). The
and has fewer than 500 members (H. McGee, per- journal publishes both conceptual and applied ar-
sonal communication, May 11, 2016). ticles, highlights reports from the field, and invites
both academics and practitioners to contribute.
Journal of Organizational Behavior Management is
A BRIEF HISTORY OF OBM the official journal of the Organizational Behavior
Management Network.
The precursors to OBM as a discipline date as A few specific topic areas have developed
far back as the 1950s (see Dickinson, 2001). B. F. throughout OBM’s history, to the point that we
Skinner’s 1953 text Science and Human Behavior now consider them subdisciplines. These include
included a chapter on “Economic Control” and in- performance management (Daniels & Bailey,
troduced ideas about wage schedules and differen- 2014), behavioral safety (McSween, 2003), and
tially reinforcing high-quality work performance. behavioral-systems analysis (Rummler & Brache,
Many magazines and journals in the1960s, includ- 2012).
ing the Harvard Business Review and the Journal of
Advertising Research, published articles on manag-
ing performance. In addition, a formal organiza- PERFORMANCE MANAGEMENT
tion devoted to performance improvement, the
International Society for Performance Improve- Although some use the terms performance manage-
ment, was established in 1962. Many founding ment and OBM synonymously, we consider perfor-
International Society for Performance Improve- mance management a subdiscipline of OBM. The
ment members were behavior analysts (Dickinson, focus in performance management is on assessing
2001). and changing the performance of individuals or
Perhaps the most widely publicized application groups of employees to increase productivity and
of OBM took place at Emery Air Freight in the late efficiency. Performance management is often
1960s. Edward Feeney, a sales manager at Emery, conducted in a step-by-step fashion. The major
used a behavioral-systems package that he learned steps include pinpointing or operationally defin-
about in an OBM workshop to increase Emery ing a target performance; developing a system to
546 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
measure the target performance; assessing the en- sessing a performance problem. He arranged their
vironmental variables that contribute to the oc- responses into four categories: antecedents and
currence of the target performance; intervening information; equipment and processes; knowledge
on the target performance; assessing the social and skills; and consequences. Austin created a list
validity, costs, and benefits of the intervention; of three to six questions for each category and a
and evaluating the maintenance of performance dichotomous response system. Austin intended
change. This step-by-step process has produced that consultants, managers, or supervisors would
substantial performance improvements in organi- use the tool to identify targeted interventions to
zations large and small, in the public and private improve employee performance. The first empiri-
sector, and in many industries (Daniels & Bailey, cal demonstration of the Performance Diagnostic
2014). Checklist took place in a store in a shopping mall.
The store was experiencing chronic cash short-
ages for each cash register at the end of shifts.
Assessment
The researchers used the Performance Diagnostic
Assessment in performance management consists Checklist to examine potential variables that ex-
of evaluating the reinforcers supporting both the plained the cash shortages. Eight customer-service
performance targeted for change and the alterna- representatives at the store served as participants.
tive or desired performance. Consultants com- The Performance Diagnostic Checklist identified
monly use the PIC/NIC analysis (Daniels & Bai- the equipment and processes and the consequence
ley, 2014). In a PIC/NIC analysis, the consultant categories as problematic. Using these results, the
analyzes whether the target performance and an researchers changed employee assignments during
alternative performance produce positive, imme- shifts and delivered verbal and posted feedback.
diate, and certain consequences versus negative, The intervention reduced the cash shortage at the
future, and uncertain consequences. The PIC in store dramatically (Rohn, Austin, & Lutrey, 2003).
the PIC/NIC analysis represents positive, immedi- Researchers have used the Performance Diag-
ate, and certain; the NIC represents negative, im- nostic Checklist since then to identify the vari-
mediate, and certain. When conducting a PIC/NIC ables responsible for poor employee performance
analysis, the manager or consultant categorizes in many settings, including a coffee shop (Pampi-
the possible performance consequences as either no, Heering, Wilder, Barton, & Burson, 2004), a
positive, immediate, and certain, or negative, fu- department store (Eikenhout & Austin, 2005), a
ture, and uncertain. Those consequences that are retail framing and art store (Pampino, MacDon-
positive, immediate, and certain are highly likely ald, Mullin, & Wilder, 2004), restaurants (Amigo,
to influence performance. The consultant’s job Smith, & Ludwig, 2008; Austin, Weatherly, &
is to adjust the positive, immediate, and certain Gravina, 2005; Rodriguez et al., 2006), and a
consequences so that they occur after the desired health clinic (Gravina, VanWagner, & Austin,
performance. 2008). The Performance Diagnostic Checklist
A study by Doll, Livesey, McHaffie, and Ludwig has become a common assessment tool in perfor-
(2007) nicely illustrates this tool. The researchers mance management and has recently been adapt-
analyzed cleaning performance by employees at a ed for specialized settings, topics, and populations
ski shop. Their PIC/NIC analysis found that the (e.g., Smith & Wilder, 2018).
consequences for cleaning were often negative, in Researchers have recently adapted the Perfor-
that cleaning the store took time away from inter- mance Diagnostic Checklist for human-service
acting with customers and required physical labor. settings (Carr, Wilder, Majdalany, Mathisen, &
The researchers implemented an intervention that Strain, 2013). Although performance manage-
reversed the PIC/NIC analysis results. After the ment in particular and OBM in general can be
intervention, cleaning produced positive conse- applied in any setting, Carr et al. (2013) argued
quences, such as comments from a supervisor and that human-service settings are sufficiently differ-
a graph depicting cleaning performance. ent from for-profit settings to warrant a specialized
Another common form of assessment in per- version of the Performance Diagnostic Checklist.
formance management is an informant-based The Performance Diagnostic Checklist—Human
tool called the Performance Diagnostic Check- Services includes questions designed specifically
list (Austin, 2000). Austin developed the Per- for employee performance problems likely to be
formance Diagnostic Checklist by asking expert encountered in schools, clinics, group homes,
OBM consultants what they do when initially as- and hospitals. Unlike the original Performance
Organizational Behavior Management 547
Diagnostic Checklist, it also includes a direct-ob- task clarification meeting and found that the em-
servation component and a list of suggested inter- ployees who received the intervention performed
ventions matched to the results. Carr et al. used well, but those who did not performed poorly.
the Performance Diagnostic Checklist—Human Researchers have used task clarification in many
Services to evaluate poor preparation of therapy other studies, with similar results (e.g., Austin et
rooms by therapists at a university-affiliated clinic al., 2005; Choi, Johnson, Moon, & Oah, 2018;
for early, intensive behavior intervention. The re- Durgin, Mahoney, Cox, Weetjens, & Poling, 2014;
sults suggested that a lack of proper training and Palmer & Johnson, 2013).
insufficient feedback were responsible for the per- Another common antecedent-based interven-
formance deficits. During the intervention, the re- tion in performance management is training.
searchers trained the therapists to prepare therapy Sasson and Austin (2005) evaluated the effects
rooms adequately. They also provided graphed of training to increase correct ergonomic per-
feedback on the therapists’ performance. The in- formance among office workers. The training
tervention was effective; mean correct preparation consisted of a one-on-one meeting in which the
of therapy rooms greatly improved during the in- instructor first described correct wrist, neck, and
tervention. Interestingly, the researchers also im- shoulder positions when typing on a keyboard.
plemented an intervention that was not based on Next, the instructor modeled the correct posi-
Performance Diagnostic Checklist—Human Ser- tions. Finally, each participant demonstrated the
vices results, and it was ineffective, suggesting that correct performance, and the instructor provided
the Performance Diagnostic Checklist—Human feedback. Training in performance management
Services correctly identified an appropriate inter- often follows this three-step model of description,
vention. Ditzian, Wilder, King, and Tanz (2015) modeling, and feedback. In addition, training in
replicated this study with a different dependent performance management is often criterion based,
variable and obtained similar results. which means that the learner or employee must
meet a mastery criterion before completing train-
ing (Howard & DiGennaro Reed, 2015). Notably,
Intervention
although researchers commonly use training, its
We can divide performance management inter- effectiveness without other performance manage-
ventions into two broad categories: antecedent- ment interventions is often modest. For example,
based and consequence-based procedures. Anteced- Sasson and Austin (2005) added feedback and em-
ent-based interventions include task clarification, ployee involvement in conducting observations of
training, manipulation of the effort required to peer performance to increase performance to high
perform a task, and goal setting. Consequence- levels. Other performance management studies
based interventions include feedback and incen- (Nordstrom, Lorenzi, & Hall, 1991) and reports
tives or contingent access to money or work perks. from the field (Haberlin, Beauchamp, Agnew, &
Researchers commonly use antecedent-based O’Brien, 2012) have also used training.
interventions to address performance problems Reducing the effort required to perform a task,
in performance management. Task clarification often called reduced response effort, is another ante-
consists of reviewing employee responsibilities re- cedent-based performance management interven-
lated to the deficient target performance. Some- tion. Abellon and Wilder (2014) reduced the dis-
times the tasks an employee is responsible for are tance that employees had to travel to use protective
posted in a salient location. Rice, Austin, and eyewear in a manufacturing facility, thereby reduc-
Gravina (2009) used task clarification to improve ing the effort of accessing the eyewear. They found
the customer-service behaviors of employees at a that employees used protective eyewear much
grocery store. The researchers taught the manager more often when it was close to their workstations.
to clarify how employees were supposed to greet Brothers, Klantz, and McClannahan (1994) ma-
customers and what to say to them at the end of an nipulated the proximity of recycling containers to
interaction. Researchers taught the manager to de- office workers, thereby reducing the effort required
liver social praise contingent on correct employee to recycle documents. They found that recycling
performance. The intervention was effective; cor- increased substantially when recycling containers
rect employee performance improved from a mean were close to employees. Other studies have also
of fewer than 15% of opportunities to more than demonstrated using reduced response effort to im-
60% of opportunities. In addition, the researchers prove employee performance (Casella et al., 2010;
collected follow-up data 48 weeks after the initial Ludwig, Gray, & Rowell, 1998).
548 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
Goal setting is another strategy researchers have Marshall, Bellizzi, & Wilder, 2013), and to improve
used to improve employee performance. Goals for employee performance in an aluminum smelt
improving performance should be challenging but manufacturing facility (Jessup & Stahelski, 1999).
achievable and should be set in collaboration with Pampino, Wilder, and Binder (2005) evaluated
the employees. Several studies have evaluated goal feedback as a component of an intervention to
setting. For example, Downing and Geller (2012) increase accurate record keeping and submission
used goal setting with feedback to increase the of time sheets by construction foremen building a
frequency with which cashiers in a large grocery neighborhood. The feedback was part of a train-
store checked customer identification to prevent ing procedure that required foremen to practice
identity theft. Identification checks increased the correct performance repeatedly; they received
from under 1% to nearly 10% due to the interven- immediate feedback on several aspects of their
tion. The researchers used participative goal set- performance. These and other applications of
ting, which means that they asked each cashier feedback illustrate the robust nature of this inter-
for assistance in setting individual goals. Other vention.
performance management studies also have used One common and popular form of feedback is
goal setting effectively (Amigo et al., 2008; Loewy the sandwich method, in which a supervisor pro-
& Bailey, 2007). vides positive comments about an employee’s
By far the most common consequence-based performance, followed by a comment describing
intervention in performance management is feed- what the employee is doing incorrectly, followed
back (Alvero, Bucklin, & Austin, 2001). Although by another positive comment. Research does not
feedback can be an antecedent, such as when it support the effectiveness of sandwich feedback.
is presented immediately before a performance Henley and DiGennaro Reed (2015) compared
opportunity, we review it here as a consequence- the feedback sandwich (positive–corrective–posi-
based procedure. Researchers have evaluated many tive) to other sequences of feedback (positive–pos-
characteristics of feedback, such as the person who itive–corrective and corrective–positive–positive),
delivers feedback (peer, supervisor, consultant), and found that the corrective–positive–positive
the frequency of feedback delivery (immediate, sequence was most effective, although the differ-
daily, weekly, monthly), and the format in which ences among the sequences were small. Interest-
feedback is delivered (oral, written, graphic). In ingly, these authors also looked at the timing of
general, research has found that feedback is most feedback delivery. They delivered feedback either
effective when someone with authority over an immediately after performance or just before the
employee, such as a supervisor, delivers feedback; next opportunity participants had to perform the
when the feedback provider uses it frequently, ide- task. They found no differences in performance
ally immediately, daily, or weekly; and when the based on the timing of the feedback.
feedback is in an easy-to-understand graphic for- Another consequence-based performance man-
mat (Alvero et al., 2001). Ludwig, Biggs, Wagner, agement intervention consists of incentives or
and Geller (2002) used publicly posted feedback access to money or other work perks contingent
to increase the correct driving performance of 82 on improved performance. Several studies have
pizza delivery drivers. Specifically, the research- found that incentive or pay-for-performance sys-
ers targeted using turn signals, using safety belts, tems result in more productive performance than
and coming to a complete stop at intersections. salary-based or hourly pay systems (Long, Wilder,
Researchers posted employees’ driving scores on Betz, & Dutta, 2012; Oah & Lee, 2011; Slowiak,
a public graph that included each driver’s name. Dickinson, & Huitema, 2011). Thurkow, Bailey,
Correct driving increased by 17–22% above base- and Stamper (2000) compared the effects of three
line levels. Other performance management stud- types of incentive pay systems on the performance
ies have used feedback to improve performance of telephone company employees. The researchers
(Palmer & Johnson, 2013; So, Lee, & Oah, 2013). found that both individual and group incentives
Researchers have used feedback to improve per- produced better performance than that during
formance in many settings with a variety of em- baseline. Individual incentives produced the high-
ployees. For example, researchers have used feed- est rate of employee performance.
back to improve the performance of supervisors Although we have been describing each per-
and animal trainers in a nongovernmental orga- formance management intervention individually,
nization in east Africa (Durgin et al., 2014), to in- most performance management interventions in-
crease credit card use in a retail setting (Loughrey, clude more than one component. So-called pack-
Organizational Behavior Management 549
age interventions increase the likelihood of the 2000). One meta-analysis of behavioral-safety ap-
interventions’ effectiveness. Unfortunately, the plications reported a reduction in injuries in 32
specific component or components most respon- of 33 studies reviewed (Sulzer-Azaroff & Austin,
sible for the effectiveness of the package interven- 2000). The settings included construction sites,
tion are often difficult to identify. For that reason, utility companies, manufacturing plants, mines,
we encourage performance management research- and shipyards; most were in the United States,
ers and practitioners to introduce intervention but some were in other countries. A more recent
components systematically, one at a time. behavioral-safety meta-analysis with more strin-
gent inclusion criteria found a reduction in in-
juries in 12 of 13 studies evaluated (Tuncel, Lot-
BEHAVIORAL SAFETY likar, Salem, & Daraiseh, 2006). The techniques
researchers use in behavioral-safety applications
Behavioral safety is the use of behavior-analytic can vary, depending on the unique needs of a site
principles and techniques to improve safe per- and assessment findings. Therefore, Tuncel et al.
formance (Krause, 1997). The Journal of Organi- (2006) have recommended that researchers clearly
zational Behavior Management published the first report assessment methods and results in behav-
OBM study addressing safety in 1978 (Komaki, ioral safety, so that they can more closely link as-
Barwick, & Scott, 1978). Shortly afterward, Fox, sessment to intervention selection.
Hopkins, and Anger (1987) began working on
one of the longest-running evaluations of a behav-
Assessment
ioral approach to improving safety in a coal mine
and in a uranium ore mine. Many other behavior One of the first activities that usually occurs in a
analysts have contributed to the development and behavioral-safety implementation is a safety assess-
evolution of behavioral safety since then, includ- ment. A safety assessment has several objectives,
ing Sulzer-Azaroff (1980), McSween (1995), Geller including (1) identifying behaviors and conditions
(1996), and Daniels and Agnew (2010). Today, to target that are likely to have an impact on safe-
behavioral safety is one of the most popular ap- ty and injury reduction, (2) identifying variables
proaches to increasing safe performance in many that influence the targets, (3) understanding the
industries (e.g., Hagge, McGee, Matthews, & Ab- context in which those behaviors and conditions
erle, 2017). occur, (4) identifying existing safety programs,
The U.S. Bureau of Labor Statistics (BLS) re- and (5) building employee and management sup-
ported that nearly 4 million workplace injuries port (Agnew & Snyder, 2008; McSween, 2003).
and illnesses and over 4,600 workplace fatalities in A consultant reviews workplace practices and
the United States occurred in 2014 (BLS, 2014a). policies and current safety initiatives and programs
Safety-related incidents and injuries have been on in a behavioral-safety assessment (McSween,
the decline for many years, due to a greater under- 2003). Initiatives and programs often emphasize
standing of ways to prevent them and an increased areas the organization views as deficient and may
focus on reducing them. Although workplace in- provide clues for behavioral targets. In addition,
juries are declining, fatalities have flat-lined (BLS, the consultant gathers information about the site,
2014b), which has produced a new focus on process including number of employees, supervisors, man-
safety (Occupational Health and Safety Adminis- agers, nature of the work, union affiliation, worker
tration, 2000). Process safety includes identifying age, rate of turnover, and shift scheduling.
strategies to influence safe behaviors and noticing The consultant reviews data on injuries; inci-
and reporting safety concerns—things that be- dents; close calls; and environmental issues, when
havioral safety is well suited to address (Bogard, applicable (e.g., spills at a chemical manufacturing
Ludwig, Staats, & Kretschmer, 2015). site) for the previous 2–5 years (McSween, 2003).
The popularity of behavioral safety is likely The goal is to uncover trends in how, when, where,
due to its effectiveness for increasing safe behav- and why injuries and incidents occur. This may
iors and reducing injuries (Krause, Seymour, & require grouping the data by area, job, body part
Sloat, 1999). Many studies have demonstrated a injured, task, shift, and other variables, to deter-
reduction in injuries, and even more have demon- mine behaviors that may be valuable safety targets
strated an increase in safe behaviors after the use and situations in which injuries are more likely to
of behavioral-safety techniques (Grindle, Dickin- occur. In addition to the site-specific data, safety
son, & Boettcher, 2000; Sulzer-Azaroff & Austin, information on the specific industry or job may
550 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
provide more ideas for safety targets. For example, consequences discourage behaviors. Often, safety-
certified nursing assistants often have back inju- related behaviors do not have positive, immediate,
ries while transferring patients; therefore, patient certain consequences, or may even have negative,
lifting may be an important target for reducing immediate, uncertain consequences. The reverse
these employees’ back injuries (BLS, 2011). is often true for at-risk behavior. For example,
The assessment should also include interviews, wearing a hard hat may feel hot and uncomfort-
a survey, and direct observation to gather more able, which are negative, immediate, and certain
information on the workplace’s context, potential consequences. Working without it may produce
targets, and safety culture. The consultant should social support for being “tough” and completing
conduct interviews with front-line workers, super- work faster, which are positive, immediate con-
visors, managers, and leaders, to learn about safety sequences. Changing this behavior involves at-
challenges at the site from multiple perspectives tempting to adjust the consequences it produces.
(McSween, 2003). Using an anonymous survey This level of analysis for safety targets allows for a
with behaviorally oriented questions is a low-effort more customized behavioral-safety process.
way to identify safety concerns that may warrant The safety assessment should produce a report
further investigation. For example, most workers that lists behaviors and conditions that a behav-
disagreeing with the statement “All injuries on ior-based safety intervention can target and rec-
this site are reported” may indicate that someone ommendations for intervention components. In
has punished injury reporting in the past, and that addition, the assessment may indicate the start-
gathering accurate data in this area may require ing point for intervention, who to involve in the
concerted effort. Or employees reporting that process, the cultural challenges, and when to start.
their supervisor talks to them about safety less We describe the steps in behavioral-safety inter-
than once per month may indicate that engaging ventions next.
supervisors in the behavioral-safety process will be
important. Surveys are often followed by direct ob-
Intervention
servation to verify accuracy.
The consultant may include several OBM tools Although there is some variation, practitioners
in an assessment to facilitate a better understand- and consultants use the following steps in the
ing of why at-risk behaviors are occurring. For behavioral-safety process: (1) Form a safety com-
example, the Performance Diagnostic Checklist— mittee; (2) identify behaviors and conditions to
Safety provides a list of questions aimed at identi- target; (3) develop a measurement system; (4) cre-
fying potential factors impeding safe performance ate a feedback, reinforcement, and problem-solv-
and directions for intervention selection (Marti- ing plan; and (5) continually improve the process
nez-Onstott, Wilder, & Sigurdsson, 2016). Like (Austin, 2006; Agnew & Snyder, 2008; McSween,
similar assessments in OBM, the questionnaire 2003). The safety committee usually consists of
covers four areas: antecedents and information; representatives from the workforce, supervisors,
knowledge and skills; equipment and processes; safety department (if one exists), and leadership
and consequences. Research has shown that in- (McSween, 2003). The safety committee manages
terventions selected from the results of a perfor- the behavioral-safety process, which is often de-
mance diagnostic checklist are effective (Johnson, scribed as employee-led (Krause, 1995). Research
Casella, McGee, & Lee, 2014). in OBM suggests that participation by employees
An antecedent–behavior–consequence (A-B- may increase the impact of the intervention (Sig-
C) analysis can also provide valuable insight for urdsson & Austin, 2006), and behavioral-safety
understanding antecedents and consequences that practitioners suggest that employee participation
may influence safe and at-risk behaviors (Health increases acceptance and maintenance of behav-
and Safety Executive, 2012). A typical A-B-C ioral-safety programs (Geller, 2002). Research has
analysis examines both the desired and undesired also demonstrated that participation in setting
behaviors and lists antecedents and consequences safety goals may have a positive impact on perfor-
for each. The consultant scores each behavior as mance (Fellner & Sulzer-Azaroff, 1985). The safe-
positive or negative, immediate or future, and cer- ty committee is one way that the behavioral-safety
tain or uncertain (Daniels & Bailey, 2014). Posi- process increases employee engagement.
tive, immediate, and certain consequences encour- The consultant uses the safe behaviors and
age behaviors, and negative, immediate, uncertain conditions the safety assessment has identified to
Organizational Behavior Management 551
create an observations checklist. Such checklists intervention, plus goal setting and supervisor
often include a space for describing barriers to per- praise, produced a dramatic improvement in safe
formance (Austin, 2006). The observations can performance. The researchers also trained an em-
improve worker safety, produce better hazard rec- ployee to continue the intervention.
ognition, produce improved feedback on safe per- An intervention package can also include in-
formance, provide information for more focused centives for improving safety. McSween (2003)
safety intervention targets, and create social con- suggests offering a menu of awards and incentives
sequences for performing safely (McSween, 2003). in a tiered fashion based on effort. Hickman and
Additionally, the observation process may lever- Geller (2005) provided a small incentive ($1 per
age the observer effect, which occurs when a per- self-observation form) to short-haul truck drivers
son conducting an observation improves his or her for completing self-observations before or after
own safe behaviors (Alvero & Austin, 2004). Peer each shift. Participation in the prebehavior group
observations are the most common, but research was 42%, and participation in the postbehavior
has shown that self-observations increase safe be- group was 75%. Both groups demonstrated an im-
haviors and are a good alternative for employees provement in overspeed and extreme braking. At
who work alone (e.g., Hickman & Geller, 2005; the other end of the spectrum, Fox et al. (1987)
Olson & Austin, 2001). Employees only (e.g., Coo- created an elaborate token-economy reward sys-
per, Phillips, Sutherland, & Makin, 1994), super- tem for behaving safely and avoiding injuries at
visors only (e.g., Chhokar & Wallin, 1984), both, two mines. Workers could exchange the tokens
everyone in the organization, or outside observers in the local community to purchase products and
only (e.g., Ludwig & Geller, 1997) can conduct the services. The intervention produced a substantial
observations. The observation checklist serves as reduction in injuries. Although it cost between
the basis for the measurement system. $9,000 and $22,000 per year per site, the return
The observation and feedback process can also on investment was substantial, ranging from 13:1
focus on supervisory and leader behavior. For ex- to 28:1 depending on the year. The intervention
ample, Cooper (2006) evaluated the impact of had been in place for 10 years when the study was
management support in a behavioral-safety pro- published, demonstrating long-term maintenance
cess in a paper mill. Cooper asked employees to of a behavioral-safety program.
complete a checklist that indicated whether lead- There are several other considerations in de-
ers with whom they interacted that week provided signing a behavioral-safety intervention. For ex-
visible ongoing support. An exploratory analysis ample, conducting observations can be mandatory
showed statistically significant correlations be- or voluntary and can be scheduled daily, weekly,
tween visible ongoing support from leaders and monthly, or even less frequently. Observation
safety performance of employees, ranging from .47 checklists can be lengthy or brief. Observers can
to .72 in the three areas included in the study, al- announce their presence and give feedback, or can
though the effects seemed to diminish over time. observe discreetly and not provide feedback in the
Zohar and Luria (2003) examined an intervention moment. Unfortunately, the research available to
to increase line supervisors’ safety-related inter- inform these decisions is limited. One study sug-
actions and found that safe behaviors and safety gested that mandatory behavioral-safety processes
climate scores increased as safety interactions in- produced higher participation and satisfaction
creased. Although research on the importance of with behavioral-safety interventions than volun-
manager engagement in safety observations and tary ones (DePasquale & Geller, 2000). Behav-
interactions is still limited, this is clearly an av- ioral-safety interventions could benefit from more
enue worthy of further investigation and is poten- research on these and related topics.
tially an important part of the observation system. Although behavioral safety has had a signifi-
After a consultant establishes an observation cant impact on improving safe practices and re-
system, he or she uses the data to provide feedback ducing injuries in the workplace, opportunities to
to employees and to reinforce progress. Feedback strengthen and refine the practice exist. For exam-
delivery often includes graphic, publicly posted ple, the field needs more research to determine the
feedback on safe behaviors. For example, Komaki best strategies for implementing a peer observation
et al. (1978) used publicly posted graphic feedback system. Whether overt or discrete observations,
updated after each observation period to improve long or short checklists, or supervisor-to-employee
the safe practices of employees in a bakery. This or peer-to-peer observations yield the most accu-
552 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
rate data and greatest behavior change is unclear. systems lens may not improve intended results. For
In addition, we need to focus on leadership behav- example, Johnson and Frederiksen (1984) con-
ior and design, to evaluate interventions that have ducted a study in a mental health institution in
a positive impact on leader behavior, and to evalu- which they compared the provision of feedback on
ate the impact of changes in leader behavior on process (behavioral) versus outcome (results) mea-
worker behavior and safe performance. Also, al- sures. The process intervention increased process
though research has demonstrated the efficacy of behaviors but did not affect outcome; the outcome
behavioral safety, some behavioral-safety processes intervention did not improve outcome or process
are short-lived or meet resistance and never fully behaviors. This suggests that the researchers did
take hold in practice. Further investigation into not identify the correct behaviors that would have
the factors that produce active, well-received, and improved outcomes. Redmon (1991) suggests that
sustained behavioral-safety processes will allow OBM interventions should clearly link to key or-
OBM to have an even more substantial impact on ganizational goals. Behavioral-systems analysis
workplace safety in the years to come. Finally, we helps OBM to place its effective interventions in
need additional research on safety belt use and safe the context of the whole business and align behav-
driving in general (see Wilder & Sigurdsson, 2015, ior targets with important goals and results.
for a review).
Assessment
BEHAVIORAL‑SYSTEMS ANALYSIS Many tools exist in behavioral-systems analysis
to assess organizational systems and processes,
Another subdiscipline of OBM is behavioral-sys- although research has empirically evaluated few
tems analysis, which involves understanding an of them (Johnson et al., 2014). Rummler and
organization by outlining the system’s parts and Brache (2012) examined organizations at three
processes and determining how they interact. levels: the organization, the process, and the job
Behavioral-systems analysis developed through and performer. The organizational level describes
the work of many pioneers in the fields of OBM the organization in the greater context in which
and human performance improvement (e.g., Bre- it exists, highlighting inputs, primary processes,
thower, 1982; Gilbert, 1996; Glenn, 1988; R. W. outputs, and outside influences and sources of
Malott, 1974; M. E. Malott, 2003). OBM tends to feedback (such as competitors, regulations, and
focus on specific behaviors. By contrast, behavior- stakeholders). The process level describes how the
al-systems analysis emphasizes the broader context work gets done from first input to final output in a
in which the behaviors occur and is more likely step-by-step fashion; it includes information about
to focus on results such as sales, expenses, and tools, quality checks, and decision points. The job
customer service (Abernathy, 2014; Hyten, 2009). and performer level focuses on factors affecting job
Some researchers have suggested that OBM should completion that could cause and reduce human
adopt a more systemic approach to increase the error and increase worker efficiency.
durability of its interventions and its attractive- Assessments at each of these levels are designed
ness and relevance to business; behavioral-systems to describe the current or is state of a system or
analysis incorporates this approach (Abernathy, process, and to identify opportunities for improv-
2014; Hyten, 2009). ing efficiency, reducing costs, reducing cycle time,
Rummler and Brache (2012) argue that the and increasing quality (Sasson, Alvero, & Austin,
greatest opportunities for improving overall per- 2006). An assessment often involves using a tool
formance and results arise in the handoffs between to create a visual depiction of the level, such as
employees or departments, which are often not the the Total Performance System Relationship Map
focus of OBM consultations or research. Unless the described by Brethower (1982) or the Process Map
entire system is considered, an intervention that described by Rummler and Brache (2012), and a
strengthens performance in one area could be det- list of performance deficiencies or opportunities
rimental to another (Abernathy, 2014; Rummler for improving the system. In addition, the consul-
& Brache, 2012). To illustrate, if manufacturing tant gathers data on current performance levels
triples production and sales are unable to keep up, and performance potentials (Gilbert, 1996). The
the organization will have spent more money than upper panel of Figure 32.1 depicts an example of
necessary on products that it is now forced to store. an is process map for an employee’s trip approval
Additionally, behavior targets identified without a process.
Creates travel
budget and Makes travel Submits
Travels
sends for arrangements receipts
Employee
approval
Reviews Reviews
budget, signs, signature and
and sends for sends back to
approval employee
Supervisor
Reviews budget,
signs, and sends
back to
Manager
supervisor
Creates travel
budget and Makes travel Submits
Travels
sends for arrangements receipts
Employee
approval
Reviews Reviews
budget, signs, signature and
and sends for sends back to
approval employee
Supervisor
Manager
FIGURE 32.1. The upper panel depicts an is process map for an employee’s trip approval process. The lower panel depicts a should process map for the same process. The consultant
553
determined that manager approval was unnecessary.
554 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
setting: A process-focused evaluation. International cies: Toward a synthesis of behavior analysis and cul-
Journal of Medical Informatics, 77, 546–554. tural materialism. Behavior Analyst, 11(2), 161–179.
Daniels, A. C., & Agnew, J. (2010). Safe by accident? Gravina, N., King, A., & Austin, J. (2019). Training
Atlanta, GA: Performance Management. leaders to apply behavioral concepts to improve safe-
Daniels, A. C., & Bailey, J. (2014). Performance man- ty. Safety Science, 112, 66–70.
agement: Changing behavior that drives organizational Gravina, N., VanWagner, M., & Austin, J. (2008). In-
effectiveness (5th ed.). Tucker, GA: Performance creasing physical therapy equipment preparation
Management. using task clarification, feedback, and environmen-
DePasquale, J. P., & Geller, E. S. (1999). Critical success tal manipulations. Journal of Organizational Behavior
factors for behavior-based safety: A study of twenty Management, 28(2), 110–122.
industry-wide applications. Journal of Safety Research, Griffin, M., Gravina, N. E., Matey, J., Pritchard, J., &
30(4), 237–249. Wine, B. (2019). Using scorecards and a lottery to
Dickinson, A. M. (2001). The historical roots of orga- improve the performance of behavior technicians in
nizational behavior management in the private sec- two autism treatment clinics. Journal of Organization-
tor. Journal of Organizational Behavior Management, al Behavior Management, 39(3–4), 280–292.
20(3), 9–58. Grindle, A. C., Dickinson, A. M., & Boettcher, W.
Ditzian, K., Wilder, D. A., King, A., & Tanz, J. (2015). (2000). Behavioral safety research in manufacturing
An evaluation of the Performance Diagnostic settings: A review of the literature. Journal of Organi-
Checklist—Human Services to assess an employee zational Behavior Management, 20(1), 29–68.
performance problem in a center-based autism treat- Haberlin, A. T., Beauchamp, K., Agnew, J., & O’Brien,
ment facility. Journal of Applied Behavior Analysis, 48, F. (2012). A comparison of pyramidal staff training
199–203. and direct staff training in community-based day
Doll, J., Livesey, J., McHaffie, E., & Ludwig, T. D. (2007). programs. Journal of Organizational Behavior Manage-
Keeping an uphill edge: Managing cleaning behav- ment, 32(1), 65–74.
iors at a ski shop. Journal of Organizational Behavior Hagge, M., McGee, H., Matthews, G., & Aberle, S.
Management, 27(3), 41–60. (2017). Behavior-based safety in a coal mine: The re-
Downing, C. O., & Geller, E. S. (2012). A goal-setting lationship between observations, participation, and
and feedback intervention to increase ID-checking injuries over a 14-year period. Journal of Organiza-
behavior: An assessment of social validity and be- tional Behavior Management, 37, 107–118.
havioral impact. Journal of Organizational Behavior Health and Safety Executive. (2012). Leadership and
Management, 32(4), 297–306. worker involvement toolkit: The ABC analysis. Re-
Durgin, A., Mahoney, A., Cox, C., Weetjens, B. J., & trieved from www.hse.gov.uk/construction/lwit/assets/
Poling, A. (2014). Using task clarification and feed- downloads/abc-analysis.pdf.
back training to improve staff performance in an Henley, A. J., & DiGennaro Reed, F. (2015). Should you
east African nongovernmental organization. Jour- order the feedback sandwich?: Efficacy of feedback
nal of Organizational Behavior Management, 34(2), sequence and timing. Journal of Organizational Be-
122–143. havior Management, 35(3), 321–355.
Eikenhout, N., & Austin, J. (2005). Using goals, feed- Hickman, J. S., & Geller, E. S. (2003). Self-management
back, reinforcement, and a performance matrix to increase safe driving among short-haul truck driv-
to improve customer service in a large department ers. Journal of Organizational Behavior Management,
store. Journal of Organizational Behavior Management, 23(4), 1–20.
24(3), 27–62. Howard, V. J., & DiGennaro Reed, F. (2015). An evalu-
Fellner, D. J., & Sulzer-Azaroff, B. (1985). Occupational ation of training procedures for animal shelter volun-
safety: Assessment the impact of adding assigned or teers. Journal of Organizational Behavior Management,
participative goal setting. Journal of Organizational 35(3), 296–320.
Behavior Management, 7, 3–23. Hyten, C. (2009). Strengthening the focus on business
Fox, D. K., Hopkins, B. L., & Anger, W. K. (1987). The results: The need for systems approaches in OBM.
long-term effects of a token economy on safety per- Journal of Organizational Behavior Management,
formance in open-pit mining. Journal of Applied Be- 29(2), 87–107.
havior Analysis, 20, 215–224. Jessup, P. A., & Stahelski, A. J. (1999). The effects of
Geller, E. S. (1996). Working safe: How to help people ac- a combined goal setting, feedback, and incentive
tively care for health and safety. Boca Raton, FL: CRC intervention on job performance in a manufactur-
Press. ing environment. Journal of Organizational Behavior
Geller, E. S. (2002). The participation factor: How to Management, 19(3), 5–26.
increase involvement in occupational safety. Chicago: Johnson, D. A., Casella, S. E., McGee, H., & Lee, S.
ASSE. C. (2014). The use and validation of preintervention
Gilbert, T. F. (1996). Human competence: Engineering diagnostic tools in organizational behavior manage-
worthy performance. Amherst, MA: HRD Press. ment. Journal of Organizational Behavior Manage-
Glenn, S. S. (1988). Contingencies and metacontingen- ment, 34(2), 104–121.
Organizational Behavior Management 557
Johnson, R. P., & Frederiksen, L. W. (1984). Process vs. F. Maley (Eds.), Behavior analysis and systems analy-
outcome feedback and goal setting in a human ser- sis: An integrative approach to mental health programs
vice organization. Journal of Organizational Behavior (pp. 318–343). Kalamazoo, MI: Behaviordelia.
Management, 5(3–4), 37–56. Martinez-Onstott, B., Wilder, D., & Sigurdsson, S.
Kelley, D. P., & Gravina, N. (2018). Every minute (2016). Identifying the variables contributing to at-
counts: Using process improvement and performance risk performance: Initial evaluation of the Perfor-
feedback to improve patient flow in an emergency mance Diagnostic Checklist—Safety (PDC-Safety).
department. Journal of Organizational Behavior Man- Journal of Organizational Behavior Management,
agement, 38, 234–243. 36(1), 80–93.
Komaki, J., Barwick, K. D., & Scott, L. R. (1978). A McSween, T. E. (1995). The values-based safety process:
behavioral approach to occupational safety: Pin- Improving your safety culture with a behavioral ap-
pointing and reinforcing safe performance in a food proach. New York: Van Nostrand Reinhold.
manufacturing plant. Journal of Applied Psychology, McSween, T. E. (2003). The values-based safety process:
63, 424–445. Improving your safety culture with behavior-based safety
Krause, T. R. (1995). Employee-driven systems for safe be- (2nd ed.). Hoboken, NJ: Wiley.
havior: Integrating behavioral and statistical methodolo- Nordstrom, R., Lorenzi, P., & Hall, R. V. (1991). A be-
gies. New York: Van Nostrand Reinhold. havioral training program for managers in city gov-
Krause, T. R. (1997). The behavior-based safety process: ernment. Journal of Organizational Behavior Manage-
Managing involvement for an injury-free culture. New ment, 11(2), 189–212.
York: Van Nostrand Reinhold. Oah, S., & Lee, J. (2011). Effects of hourly, low-in-
Krause, T. R., Seymour, K. J., & Sloat, K. C. M. (1999). centive, and high-incentive pay on simulated work
Long-term evaluation of a behavior-based method for productivity: Initial findings with a new laboratory
improving safety performance: A meta-analysis of 73 method. Journal of Organizational Behavior Manage-
interrupted time-series replications. Safety Science, ment, 31(1), 21–42.
32, 1–18. O’Brien, R. M., Dickinson, A. M., & Rosow, R. P. (Eds.).
LaFleur, T., & Hyten, C. (1995). Improving the quality (1982). Industrial behavior modification. New York:
of hotel banquet staff performance. Journal of Organi- Pergamon Press.
zational Behavior Management, 15(1–2), 69–93. Occupational Health and Safety Administration. (2000,
Loewy, S., & Bailey, J. (2007). The effects of graphic reprinted). Process safety management. Retrieved
feedback, goal setting, and manager praise on cus- from www.osha.gov/Publications/osha3132.html.
tomer service behaviors. Journal of Organizational Olson, R., & Austin, J. (2001). Behavior-based safety
Behavior Management, 27(3), 15–26. and working alone: The effects of a self-monitoring
Long, R. D., Wilder, D. A., Betz, A., & Dutta, A. (2012). package on the safe performance of bus operators.
Effects of and preference for pay for performance: An Journal of Organizational Behavior Management,
analogue analysis. Journal of Applied Behavior Analy- 21(3), 5–43.
sis, 45, 821–826. Palmer, M. G., & Johnson, M. (2013). The effects of task
Loughrey, T., Marshall, G., Bellizzi, A., & Wilder, D. A. clarification and group graphic feedback on early
(2013). The use of video modeling, prompting, and punch-in times. Journal of Organizational Behavior
feedback to increase credit card promotion in a retail Management, 33(4), 265–275.
setting. Journal of Organizational Behavior Manage- Pampino, R. N., Heering, P. W., Wilder, D. A., Barton,
ment, 33(3), 200–208. C. G., & Burson, L. M. (2004). The use of the Per-
Ludwig, T. D., Biggs, J., Wagner, S., & Geller, E. S. formance Diagnostic Checklist to guide interven-
(2002). Using public feedback and competitive re- tion selection in an independently owned coffee
wards to increase the safe driving of pizza deliver- shop. Journal of Organizational Behavior Management,
ers. Journal of Organizational Behavior Management, 23(2), 5–19.
21(4), 75–104. Pampino, R. N., MacDonald, J. E., Mullin, J. E., &
Ludwig, T. D., & Geller, E. S. (1997). Assigned versus Wilder, D. A. (2004). Weekly feedback versus daily
participatory goal setting and response generaliza- feedback: An application in retail. Journal of Organi-
tion: Managing injury control among professional zational Behavior Management, 23(2), 21–43.
pizza deliverers. Journal of Applied Psychology, 82(2), Pampino, R. N., Wilder, D. A., & Binder, C. (2005). The
253–261. use of functional assessment and frequency building
Ludwig, T. D., Gray, T. W., & Rowell, A. (1998). In- procedures to increase product knowledge and data
creasing recycling in academic buildings: A system- entry skills among foreman in a construction orga-
atic replication. Journal of Applied Behavior Analysis, nization. Journal of Organizational Behavior Manage-
31, 683–686. ment, 25(2), 1–36.
Malott, M. E. (2003). Paradox of organizational change. Redmon, B. (1991). Pinpointing the technological fault
Reno, NV: Context Press. in applied behavior analysis. Journal of Applied Be-
Malott, R. W. (1974). A behavioral systems approach to havior Analysis, 24(3), 441–444.
the design of human services. In D. Harshbarger & R. Rice, A., Austin, J., & Gravina, N. (2009). Increasing
558 S u b s p e c ia lt i e s i n A pp l i e d B e h av i o r A n a ly s i s
customer service behaviors using manager-delivered als with intellectual disabilities. Behavior Analysis in
task clarification and social praise. Journal of Applied Practice, 11(2), 148–153.
Behavior Analysis, 42, 665–669. So, Y., Lee, K., & Oah, S. (2013). Relative effects of daily
Rodriguez, M., Wilder, D. A., Therrien, K., Wine, B., feedback and weekly feedback on customer service
Miranti, R., Daratany, K., et al. (2006). Use of the behavior at a gas station. Journal of Organizational
Performance Diagnostic Checklist to select an inter- Behavior Management, 33(2), 137–151.
vention designed to increase the offering of promo- Sulzer-Azaroff, B. (1980). Behavioral ecology and ac-
tional stamps at two sites of a restaurant franchise. cident prevention. Journal of Organizational Behavior
Journal of Organizational Behavior Management, Management, 2(1), 11–44.
25(3), 17–35. Sulzer-Azaroff, B., & Austin, J. (2000). Does BBS work?:
Rohn, D., Austin, J., & Lutrey, S. M. (2003). Using feed- Behavior-based safety and injury reduction: A survey
back and performance accountability to decrease of the evidence. Professional Safety, 45(7), 19–24.
cash register shortages. Journal of Organizational Be- Thurkow, N. M., Bailey, J. S., & Stamper, M. R. (2000).
havior Management, 22(1), 33–46. The effects of group and individual monetary incen-
Rummler, G. A., & Brache, A. P. (2012). Improving per- tives on productivity of telephone interviewers. Jour-
formance: How to manage the white space in the orga- nal of Organizational Behavior Management, 20(2),
nizational chart (3rd ed.). San Francisco: Jossey-Bass. 3–25.
Sasson, J. R., Alvero, A. M., & Austin, J. (2006). Effects Tuncel, S., Lotlikar, H., Salem, S., & Daraiseh, N.
of process and human performance improvement (2006). Effectiveness of behaviour based safety in-
strategies. Journal of Organizational Behavior Manage- terventions to reduce accidents and injuries in work-
ment, 26(3), 43–78. places: Critical appraisal and meta-analysis. Theoreti-
Sasson, J. R., & Austin, J. (2003). Performer-level sys- cal Issues in Ergonomics Science, 7(3), 191–209.
tems analysis. Journal of Organizational Behavior Vergason, C. M., & Gravina, N. E. (2020). Using a
Management, 22(4), 27–58. guest- and confederate-delivered token economy to
Sasson, J. R., & Austin, J. (2005). The effects of train- increase employee–guest interactions at a zoo. Jour-
ing, feedback, and participant involvement in behav- nal of Applied Behavior Analysis, 53(1), 422–430.
ioral safety observations on office ergonomic behav- Wilder, D. A., Austin, J., & Casella, S. (2009). Applying
ior. Journal of Organizational Behavior Management, behavior analysis in organizations: Organizational
24(4), 1–30. behavior management. Psychological Services, 6(3),
Sigurdsson, S. O., & Austin, J. (2006). Institutionaliza- 202–211.
tion and response maintenance in organizational Wilder, D., & Sigurdsson, S. (2015). Applications of
behavior management. Journal of Organizational Be- behavior analysis to improve safety in organizations
havior Management, 26(4), 41–77. and community settings. In H. S. Roane, J. E. Ring-
Slowiak, J. M., Dickinson, A. M., & Huitema, B. E. dahl, & T. S. Falcomata (Eds.), Clinical and organiza-
(2011). Self-solicited feedback: Effects of hourly pay tional applications of behavior analysis (pp. 583–604).
and individual monetary incentive pay. Journal of Or- San Diego, CA: Academic Press/Elsevier.
ganizational Behavior Management, 31(1), 3–20. Zohar, D., & Luria, G. (2003). The use of supervisory
Smith, M., & Wilder, D. A. (2018). The use of the Per- practices as leverage to improve safety behavior: A
formance Diagnostic Checklist—Human Services to cross level intervention model. Journal of Safety Re-
assess and improve the job performance of individu- search, 34(5), 567–577.
PAR T VIII
PROFESSIONAL ISSUES
IN APPLIED BEHAVIOR ANALYSIS
B ehavior analysis began as a subfield of psychology, but over time it has devel-
oped its own philosophy, scientific methods, applications, journals, organiza-
tions, credentialing board, and even a set of Current Procedural Terminology
(CPT) codes approved and published by the American Medical Association so
that applied behavior analysts can bill for the services they provide. Because
behavior analysis has evolved into a discipline separate from psychology, it has
become increasingly important for applied behavior analysts to develop ethical
and professional standards that address the unique features of their subfield; those
standards are topics addressed in this section of the handbook.
In Chapter 33, O’Donohue and Ferguson examine the relation between
the fields of behavior analysis and ethics at the levels of meta-ethics, norma-
tive ethics, and descriptive ethics. Their thoughtful discussion of current ethi-
cal guidelines relative to Skinner’s contingency analysis of moral and ethical
behavior should be of interest to applied behavior analysts. Carr, Nosik, Ratcliff,
and Johnston discuss the history and development of professional certification
in applied behavior analysis in Chapter 34. They describe the amazing growth in
the number of behavior analysts certified by the Behavior Analyst Certification
Board® to over 35,000 certificants, and they explain how the standards for certifi-
cation have changed in accordance with the progress of the profession of applied
behavior analysis.
559
CHAP TER 33
William O’Donohue and Kyle E. Ferguson
Behavior analysis has a complex relationship with of what ethical claims or beliefs some individu-
the field of ethics. In this chapter, we describe the als hold. This is an interesting empirical ques-
general structure of this relationship, and some of tion, and we might conduct research to see
its major complexities and unsettled issues. We what ethical claims behavior analysts hold. Be-
typically construe ethical discourse as occurring cause we could find no research on this issue,
at three distinct levels: we do not examine this dimension of ethics in
this chapter.
1. Meta-ethics addresses questions including
these: What kind of thing is ethical discourse?
Is it the same or different from empirical dis- BEHAVIOR ANALYSIS AND CONTROVERSIES
course? If it is the same, what kinds of dis- AT THE META‑ETHICAL LEVEL
course does ethical discourse belong to? If dif-
ferent, what kind of thing is ethical discourse, Behavior analysis has paid most attention to two
and how do we evaluate the truth value of ethi- key meta-ethical problems. First, behavior ana-
cal claims? lysts have followed Skinner and have responded
2. Normative ethics addresses questions of which to the meta-ethical question “Is ethical discourse
ethical claims, among many possible candi- even possible in a deterministic worldview?” in
dates for ethical principles, are correct. Is doing the affirmative, although it becomes a somewhat
X, for example, morally right, wrong, or indif- different kind of discourse than that in common
ferent, where X can range from being cultur- folk discourse. Folk perspectives assume points in
ally insensitive (Benuto, Casas, & O’Donohue, which an individual is free to choose morally bad
2018), to prescribing facilitated communica- or morally good alternatives. Skinnerians do not.
tion for a child with autism spectrum disorder, Second, behavior analysts have answered the ques-
to eating an orange? Normative ethics tries to tion “What kind of thing is ethical discourse?” with
address the questions of which moral prohibi- the response that it is naturalistic discourse, such
tions, permissions, or mandates are correct and as talk of observable properties of the world (e.g.,
how we justify them. color, smell). We first turn our attention to the issue
3. Descriptive ethics addresses empirical questions of morality from a deterministic standpoint.
561
562 P r o f e s s i o n a l I s s u e s i n A pp l i e d B e h av i o r A n a ly s i s
Meta‑Ethical Question 1: Determinism and Kant’s the existence of moral choices. Thus meta-ethical
Dictum “Ought Implies Can” morality is fundamentally different for the deter-
minist than for any libertarian.
The moral philosopher Immanuel Kant (1997) as-
Ringen (1996) addresses this question as fol-
serted, “Ought implies can”; that is, asserting that
lows:
someone ought to have done something makes no
sense unless this person could have performed the The scientific account of human action requires that
action. For example, asserting that Jane ought to every event is determined causally, so there are no
have jumped 6 meters in the air makes no sense acts that are free of determining causal influences.
when Jane cannot physically do so. Therefore, The conclusion is that the causal determinism in-
moral discourse presupposes choice: A person can volved in the scientific account of human action is
do something because he or she is free to choose to incompatible with the account of autonomy and self-
do this. Thus morality presupposes free will. determinism that legal, political, and ethical argu-
Skinner, particularly in Beyond Freedom and ments require. . . . (p. 356)
Dignity (1971/2002), discussed the conflict be-
tween a scientific worldview and a view based on This view means that we cannot accept scientif-
choice and free will. Science assumes some form ic determinism and the traditional concept that
of determinism because it assumes order, and it as- human beings act in a self-determining way. The
sumes that there are causal relations to discover reason is that if determinism is true, no human
(Fischborn, 2018). Science assumes that the world act is free of controlling influences. Accordingly,
is orderly, and we can characterize this order as if we presume determinism in the science of be-
lawful relationships. A scientific law, however, havior, does an ethics code for behavior analysts
describes an impossibility. According to the law even make sense? If “ought implies can,” then
of gravity, for example, for two objects to behave how can we hold an individual morally culpable
other than according to the law is simply impos- when he or she could not have behaved otherwise
sible. Objects are neither free to choose other ways (O’Donohue & Ferguson, 2003, p. 7)? As we shall
to behave, nor free to choose different relation- see later, if we want behavior analysts to behave
ships with other objects. The law describes the ethically, we need to understand the controlling
only way that the objects can behave. Skinner influences or contingencies (i.e., antecedents and
points out that if people are free to choose their consequences) of which ethical and unethical be-
behavior, then a science of behavior is not pos- havior is a function.
sible. Thus Skinner aptly titled his book Beyond
Freedom and Dignity to argue that the scientific Meta‑Ethical Issue 2: Skinner’s Naturalistic
enterprise, when applied to human behavior, must Approach to Ethics and Moral Reasoning
move beyond the notion of free choice. Therefore,
if they take Kant seriously, behavior analysts must Philosophers have made many claims about what
assume that ethical discourse does not make sense. kind of entity ethical discourse is. A brief list fol-
How can behavior analysts come to terms with lows:
this issue? Behavior analysts, of course, can take
the stance that adopting an ethical code for be- 1. Ethical discourse derives from some divine
havior analysts is just more determined behavior. source, usually resulting in some divine com-
However, behavior analysts must see that to re- mandments, and is a kind of religious discourse.
main consistent, they must accept that a determin- This view has held wide sway in the history of
istic view of ethical codes and discourse diverges humankind and still has a significant number
greatly from normal ethical discourse. Behavior of proponents. We can argue that much of the
analysts do not choose to behave in one way or founding discourse of Western civilization is
another; thus we should not praise or punish them based on this view; witness the theological–
for their behavior, because they could have done ethical connection in the U.S. Declaration of
nothing else. That is, to remain consistent, they Independence and the Constitution. However,
must be seen as compelled to emit the behavior behavior analysis clearly does not ascribe to
in question. Normal ethical discourse takes choice this view, and we may regard its emphasis upon
seriously and is a kind of argument for individuals naturalistic, evolutionary accounts as critical
to make good choices when they are at a moral of the view that ethical discourse is based fun-
crossroad. Behavior analysts do not countenance damentally on theological precepts.
Behavior Analysis and Ethics 563
2. Ethical discourse is meaningless. It is “lan- p. 328), in his later writings he adopted a position
guage gone on holiday,” to use Wittgenstein’s akin to naturalism (e.g., Skinner, 1953). Accord-
(1953) phrase. The logical positivists held this ing to Skinner (1971/2002),
view. They offered an emotivist meta-ethical
theory and claimed that “X is bad” has no di- Good things are positive reinforcers. . . . When we
rect observational basis; it violates their veri- say that a value judgment is a matter not of fact but
how someone feels about a fact, we are simply distin-
fiability principle and is meaningless. Their
guishing between a thing and its reinforcing effect.
positive claim, though, was that humans tend Physics and biology study things themselves, usually
to make these utterances because humans without reference to their value, but the reinforcing
are not only logical, rational animals but also effects of things are the province of behavioral sci-
emotional ones, and their emotional capacity ence, which is a science of values to the extent that it
causes them to make these utterances. “Mur- focuses on operant reinforcement. (pp. 103–104)
der is bad” is equivalent to “Uck!” or “I hate
murder,” and nothing more. Logical positivism Naturalism is the theory that we can derive
has not influenced behavior analysis to the moral values from facts about the world (Ma-
extent that many distant observers think (see cIntyre, 1998). From this perspective, is can
Smith, 1986). At the same time, it does not imply ought. Presumably, one ought to behave in
hold an emotivist view of ethics, either. ways that produce reinforcement from a Skin-
3. Ethical discourse refers to an entirely empiri- nerian point of view. Or, conversely, one ought
cal, natural phenomenon maintained by its not to behave in ways that produce punishment.
consequences, like other empirical, descriptive One should escape or avoid aversive stimulation,
discourse or even causal discourse. “X is bad” is as in the case of negative reinforcement. Skinner
equivalent to “I have observed that X has some (1971/2002) added:
empirical property.” Skinner (1971/2002) held
this view; thus this view has been most influ- Things are good (positively reinforcing) or bad
(negatively reinforcing) presumably because of the
ential in behavior analysis. We regard Skinner
contingencies of survival under which the species
as an authority on this and many other top- evolved. . . . It is part of the genetic endowment
ics in the science of behavior. In brief, Skinner called “human nature” to be reinforced in particular
held that “X is good” means that “X is reinforc- ways by particular things. (p. 104).
ing,” where reinforcing is an empirical matter
(i.e., a reinforcer increases the response rate Thus natural processes select human beings
of the behavior it is contingent on). Skinner to know right from wrong, good from bad, which
argued that ethics is a completely natural dis- is part of our genetic endowment. What follows,
course. This view, of course, is not above criti- therefore, is that “all reinforcers eventually derive
cism. their power from evolutionary selection” (Skin-
4. Ethical discourse is sui generis. Ethical utter- ner, 1971/2002, pp. 104–105). Simply, individuals
ances and claims are not like descriptive, em- whose behavior is maintained by certain types
pirical discourse. In fact, they are not like any of reinforcers live to reproduce; those insensitive
other kind of discourse: Ethical discourse is to comparable reinforcers do not. Thus morality
wholly unique. This is the view of many ethical evolved insofar as our species evolved.
theorists, such as Immanuel Kant (1997) and One inherent problem with Skinner’s naturalis-
G. E. Moore (1903/1988). There is an ethical tic position is that the concept of reinforcement is
realm, and ethical claims have a wholly differ- not clear. Early accounts were entirely functional.
ent status from that of empirical claims in the If animals were appropriately deprived (e.g., 80%
observable world. The exact status of ethical of ad libitum weight), certain events were highly
discourse depends on the specific ethical theo- likely to function as reinforcers for many behaviors
rist. Behavior analysis has generally disagreed (e.g., delivery of food pellets). Thus we also must
with this position in favor of naturalizing the understand reinforcing effects by contextualizing
normative. the behavior in a matrix of variables that affect
reinforcement. Ethical statements become some-
Although Skinner originally maintained that what complex in this view. For example, when X
ethics and moral reasoning are beyond the pur- is deprived of Y to extent Z and emits behavior B,
view of behavior analysis (e.g., Skinner, 1953, R is reinforcing—morally correct—to X. When
564 P r o f e s s i o n a l I s s u e s i n A pp l i e d B e h av i o r A n a ly s i s
the subject reaches satiation point S, R becomes able to discover their existence unless we had been
punishing—morally wrong (see Allison & Tim- struck first by the difference of quality between the
berlake’s [1975] response deprivation model of re- different colors. The most we can say of those vibra-
inforcement). How many parameters need to be a tions is that they are what corresponds in space to
the yellow that we perceive. (Section 10, Paragraphs
part of such sentences because the field’s under-
1 and 2, emphasis in original)
standing of the mechanisms of reinforcement is
incomplete or is unclear? Thus a problem with this
To the extent that yellow does not possess the
view is that the construct of reinforcement is by
property of yellowness, reinforcers do not possess
no means clear, and the controversies surrounding
the property of goodness. We must make such
this construct are germane to ethical pronounce-
value judgments regardless of their reinforcing
ments from a naturalistic standpoint.
properties; otherwise, we run into the same prob-
In addition, problematic counterexamples exist.
lem as in our pedophilia example. Labeling behav-
Take pedophilia as a case in point. Pedophilia is
ior good or bad, right or wrong, seems to concern
sexual attraction to prepubescent children. Few
would argue that pedophiles’ behavior is not mor- aspects of consequent stimuli unrelated to their
ally wrong, blameworthy, and so forth. However, reinforcing characteristics.
stimuli associated with prepubescent children,
such as child pornography or contact with chil-
dren, are highly reinforcing for certain behaviors NORMATIVE ETHICS: CONTROVERSIES
of such individuals. According to Skinner’s posi- CONCERNING AN ETHICAL CODE
tion, therefore, the pedophiles’ sexual behavior
is morally good, at least for the pedophiles. They The Florida Association for Behavior Analysis
might like pedophilic stimuli (i.e., those stimuli (1987) was the first chapter of what was then the
serve as reinforcers), but can we say nothing else Association for Behavior Analysis to develop an
here? Does Skinner’s naturalistic account force us ethical code. The Texas Association for Behavior
into a relativism in which we can say that child Analysis (1995) soon followed in the early 1990s,
pornography is good for X because it functions as a followed by the California Association for Behav-
reinforcer for X? Certainly, the problem is clear. X ior Analysis (1996). The Association for Behavior
can be a reinforcer, but still can be morally wrong Analysis International (as it now is) adopted the
in a broader sense. American Psychological Association’s (2002) Eth-
The problem with Skinner’s approach is due to ics Code, thus obviating those developed under in-
what Moore (1903/1988) called the naturalistic fal- dividual chapters. Likewise, the Behavior Analyst
lacy. Moore stated: Certification Board (BACB) by and large adopted
the American Psychological Association’s 2002
“Good,” then, if we mean by it that quality which Ethics Code in 2004, with some modification.
we assert to belong to a thing, when we say that the Accordingly, criticisms of the American Psycho-
thing is good, is incapable of any definition, in the logical Association’s Ethics Code are relevant with
most important sense of that word. The most impor- respect to the Association for Behavior Analysis
tant sense of definition is that in which a definition International’s code, because they are virtually the
states what are the parts which invariably compose a
same.
certain whole; and in this sense good has no defini-
tion because it is simple and has no parts. It is one
Ethical codes are not above criticism (O’Don
of those innumerable objects of thought which are ohue, 2016; O’Donohue & Ferguson, 2003). Ethi-
themselves incapable of definition because they are cal codes are the product of human behavior and
the ultimate terms of reference to which whatever is are fallible, like all human activity. The American
capable of definition must be defined. . . . There is, Psychological Association’s (2002) Ethics Code is
therefore, no intrinsic difficulty in the contention no exception.
that good denotes a simple and indefinable qual- The Ethics Code is a fallible document. How
ity. . . . fallible and in what ways constitute the crux of the
Consider yellow, for example. We may try to de-
matter. Of concern is that we pronounce individ-
fine it by describing its physical equivalent. We may
state what kind of light vibrations must stimulate the
ual behavior as unethical or ethical on the basis
normal eye so we may perceive it. But a moment’s of a flawed code. The code’s fallibility is easy to
reflection is sufficient to show that those light vibra- discern. First, why would it have undergone nine
tions are not what we mean by yellow. They are not revisions since its inception if it were infallible?
what we perceive. Indeed, we should never have been Of course, the American Psychological Associa-
Behavior Analysis and Ethics 565
tion can argue that the code has morphed in ac- does not explicitly state this. Namely, behavior
cordance with changing ethics, and that succes- analysts have special knowledge and skill sets that
sive iterations have accommodated these changes clients typically do not, and that place them at an
accurately. However, the American Psychological unfair advantage. Thus the relationship is unilat-
Association has not made this argument, and pro- eral; there is a power differential. This specialized
fessional morality generally remains largely static knowledge puts therapists in the role of authority.
over time. That is, the standards of beneficence Behavior analysts, therefore, should not engage
and nonmaleficence, fidelity, responsibility, integ- in sexual relationships with clients under any cir-
rity, justice, and respect for individual rights and cumstances or at any time, because “such relation-
dignity do not change appreciably (O’Donohue & ships easily impair judgment and become exploi-
Ferguson, 2003). tive (due to this power differential).”Now consider
Second, the American Psychological Associa- the BACB’s (2017) most recent version of its ethi-
tion used fallible procedures to develop the code. cal guidelines, Code 1.0.7.b: “Behavior analysts
Committee members voted for or against the refrain from any sexual relationships with clients,
codes, and no committee members were behavior students, or supervisees, for at least two years after
analysts. Knowledge by way of authority is fallible the date the professional relationship has formally
due to the committee members’ imperfect judg- ended.” (p. 6). This, of course, is vastly different
ment, heuristic biases (e.g., representative biases), from the BACB’s prior guidelines regarding such
and the like. We did not conduct experimental dual relationships. Ignoring the power differential
analyses; hence no safeguards were there to miti- mentioned earlier, Code 1.0.7.b states that such
gate against these biases, which probably were op- sexual relationships are not a problem if either
erating at some level during committee meetings. party holds out for a minimum of 2 years. “Why 2
Third, other professional organizations in the years?” is a reasonable question to ask. The BACB
behavioral sciences have their own ethics codes, has not provided any empirical evidence why 2
some of which differ on fundamental points in years is better than 1 year, 6 months, or 10 years,
relation to the American Psychological Associa- for example. Nor has it provided any evidence to
tion (2002) Ethics Code. For example, what was support that even having sexual relationships with
then the Association for Advancement of Behav- former clients, irrespective of the time interval, is
ior Therapy (1977; now Association for Behav- beneficial to either party or does no harm at the
ioral and Cognitive Therapies) long ago adopted very least (i.e., primum non nocere).
a higher standard of evidence clinicians must use Reason would suggest that the BACB’s (2017)
before they select interventions. Ethical codes of more recent standard is patently wrong. Once
individual chapters, such as the California Asso- behavior analysts begin having sex with former
ciation for Behavior Analysis (1996) and the Flor- clients, they might start viewing current clients
ida Association for Behavior Analysis (1987), also differently. For example, they might view current
have inconsistencies relative to the Ethics Code. clients as potential conquests. This, of course, can
The question then becomes “Which one is right?” obfuscate professional boundaries and compro-
Although no hard and fast rules exist in determin- mise the therapeutic alliance.
ing this experimentally, we can evaluate these A fourth problem concerns the relation between
standards by following their lines of reasoning and the American Psychological Association (2002)
turning to the empirical literature for support or Ethics Code and justification for these standards.
counterevidence. Simply making the claim that one action is ethi-
Let us take the BACB’s (2004) Guidelines for cal and another is unethical, without alluding to
Responsible Conduct for Behavior Analysts as a some ethical theory or general ethical standard,
case in point. Consider Code 1.4.2 (Exploitative is insufficient for several reasons (O’Donohue &
Relationships): “Behavior analysts do not engage in Ferguson, 2003). First, how can we know if we are
sexual relationships with clients, students, or super- interpreting a given standard correctly? For exam-
visees in training over whom the behavior analyst ple, deontological theory judges the rightness or
has evaluative or direct authority, because such rela- wrongness of an action based solely on the nature
tionships easily impair judgment or become exploit- of the act itself or its structural characteristics, ir-
ative” (emphasis added). A likely interpretation of respective of its consequences. We deem some ac-
this guideline is that behavior analysts should not tions unethical, even though they bear functional
engage in sexual relationships with clients (ever), similarities to actions that are not. For example, we
because authority never ends, although the code might view stealing from a client’s purse as unethi-
566 P r o f e s s i o n a l I s s u e s i n A pp l i e d B e h av i o r A n a ly s i s
cal, but not view charging excessive professional foster ethical behavior and avoid conditions that
fees as unethical. Utilitarian ethics, by contrast, promote unethical behavior.
judge the rightness or wrongness of an act based According to Skinner (1972), “We cannot
solely on its consequences. If no one got hurt, we choose a way of life in which there is no con-
would not view the act as unethical, even if ma- trol. . . . We can only change the controlling con-
licious intent was involved. Accordingly, deontic ditions” (pp. 97, 99, 194–195). What he meant by
and utilitarian ethical theories provide different this is that our behavior is always under the control
interpretations of the same code. Given that the of genetic and environmental determinants (i.e.,
BACB has not stated its position on this matter antecedent and consequential stimuli). Skinner’s
explicitly, either interpretation is defensible. Fur- views on promoting ethical behavior, therefore,
thermore, the decision to take disciplinary steps entail changing these controlling conditions. The
would differ considerably, depending on the theo- most important aspect of these controlling condi-
retical orientation of individual committee mem- tions is the presence of countercontrol (Skinner,
bers, should the BACB investigate an alleged ethi- 1972).
cal violation. There has been much controversy Countercontrol is the “emotional reaction of
in the American Psychological Association over anger . . . including operant behavior,” on the part
the ethical propriety of psychologists involved in of the controllee (i.e., the person whose behavior
interrogation, but whether these arguments are is under the control of another), that “injures or
based on deontic ethics, utilitarian ethics, virtue is otherwise aversive to the controller” (Skinner,
ethics, or some other normative theory is unclear 1953, p. 321). In other words, countercontrol is
(see O’Donohue et al., 2014). the controllee’s attempt at changing the aversive
A second problem is determining whether a be- contingencies of which his or her behavior is a
havior analyst has violated a standard (O’Donohue function. Several examples of countercontrol are
& Ferguson, 2003). For example, does the motive as follows: An abused woman murders her abus-
to do harm, as opposed to actual harm, constitute er; an incarcerated criminal escapes from prison;
an ethical violation? The American Psychological citizens protest against their government; and the
Association (2002) and the BACB (2017) are both volunteers who served as inmates in the infamous
silent on this matter. Technically speaking, these Stanford prison rebelled against the volunteers
contrasting views concern motivist ethical theory who were randomly assigned to be guards.
and utilitarian ethics, respectively. The intent The following passage summarizes how coun-
of wrongdoing warrants disciplinary action from tercontrol relates to ethics:
a motivist perspective, but not from a utilitarian
perspective. The utilitarian perspective requires The consequences responsible for benevolent, de-
evidence that a client was harmed. Accordingly, voted, compassionate, or public-spirited behavior are
because both the American Psychological Asso- forms of countercontrol, and when they are lacking,
these much-admired features of behavior are lack-
ciation and the BACB do not explain the moral-
ing. . . . The point is illustrated by five fields in which
ity upon which they based the Ethics Code, how control is not offset by countercontrol and which
can they ensure due process of ethical inquiries have become classical examples of mistreatment.
(O’Donohue & Ferguson, 2003, p. 7)? The answer They are the care of the very young, of the aged, of
is that they cannot, based on the arguments dis- prisoners, of individuals diagnosed with psychosis,
cussed above. and of individuals with intellectual disabilities. Crit-
ics often say that those who have these susceptible
people in their charge lack compassion or a sense of
SKINNER’S CONTINGENCY ANALYSIS ethics, but the conspicuous fact is that vulnerable
OF MORAL AND ETHICAL BEHAVIOR individuals are not subject to strong countercontrol.
The young and the aged are too weak to protest, pris-
oners are controlled by police power, and individuals
Although Skinner’s science of behavior is inad- diagnosed with psychosis or intellectual disabilities
equate in helping us determine what is moral or cannot organize or act successfully. Little or nothing
ethical behavior (i.e., normative ethics), his analy- is done about mistreatment unless external agencies
sis of the contingencies by which people behave introduce countercontrol from outside. (Skinner,
ethically is useful in helping us understand why 1971/2002, pp. 196–197)
people behave as they do (i.e., meta-ethics). By
identifying the contingencies of which ethical be- Here Skinner was very clear about what profes-
havior is a function, we can create conditions that sionals need to do to ensure ethical behavior. In
Behavior Analysis and Ethics 567
circumstances where individuals are so disadvan- countercontrol mechanisms are in place. Report-
taged that they have no effective mechanisms of ing and documentation have become increasingly
countercontrol, external agencies need to step in stringent. Moreover, skilled nursing facilities are
to offset this imbalance. In recent times, no other subject to random checks by government agencies.
area in health care illustrates the role of counter-
control and ethics better than the care of elderly
persons with neurocognitive disorders (dementia) CLOSING REMARKS
and other intellectual impairments.
Behavior analysis has a complex relation with
ethical discourse. This chapter has described the
Countercontrol and Elder Abuse
complexities of this relation and some unresolved
Persons with cognitive and behavioral disorders issues. One of the major unsettled issues is this: Is
who receive elder care are at particularly high risk it even possible to reconcile a scientific view of be-
of abuse, due to the burden their caregivers experi- havior with an indeterministic view or the notion
ence (Lachs, Williams, O’Brien, Hurst, & Horow- of free will, which presupposes the most recent
itz, 1997; Pillemer, Burnes, Riffin, & Lachs, 2016). versions of the codes of ethics for the American
Elder abuse is physical, psychological, sexual, and/ Psychological Association (2002) and the BACB
or financial exploitation; neglect; or a combina- (2017)? Does it even make sense to have an eth-
tion of these that may result from the actions of ics code for behavior analysts who espouse a sci-
others, such as caregivers (Dyer, Pavlik, Murphy, entific view of behavior? If indeed “ought implies
& Hyman, 2000). The prevalence of elder abuse can,” then how can we hold a professional morally
ranges from 1 to 12%, although these figures likely and ethically responsible when he or she could not
underestimate the problem, due to methodological have behaved otherwise (O’Donohue & Ferguson,
weaknesses and to both recipients’ and caregivers’ 2003, p. 7)? The professional’s unethical behavior
reluctance to report abuse (Burnes et al., 2015; was beyond his or her personal control. That is
Tueth, 2000). After all, elder abuse is a felony and to say, the unethical behavior was due to genetic
can result in criminal charges. Moreover, care re- variables and deficiencies in the prevailing contin-
cipients may be reluctant to report abuse for fear gencies of which such behavior and other, incom-
of being moved out of their homes to alternative patible ethical behavior are a function.
living arrangements (e.g., assisted living or skilled In Beyond Freedom and Dignity, Skinner
nursing facilities). (1971/2002) discussed this conflict between a
Characteristics that place care recipients with scientific worldview and a view based on choice
dementia at risk for abuse include challenging be- and free will. Science presupposes determinism
haviors, such as physical aggression, and their need because it assumes order in the universe. Based
for increasingly intense hands-on care with activi- on the presupposition that the world is orderly,
ties of daily living, such as bathing, toileting, and science then can elucidate those lawful relations
dressing, as they deteriorate (Dong, 2015; Lachs et (e.g., contingencies of survival and reinforcement)
al., 1997; Wangmo, Nordström, & Kressig, 2017). that enable our species to better predict and con-
Physical abuse is most likely to occur in the con- trol or influence natural phenomena. Insofar as
text of hands-on caregiving. When a caregiver has matter does not get to choose what natural forces
direct contact with aggressive behavior, noncom- operate on it, organisms do not get to choose what
pliant behavior, or both from an elderly person behavior gets emitted. Accordingly, the answer to
with dementia, the caregiver is likely to be emo- the earlier question is no. The codes of ethics for
tionally distressed, possibly angry, and in physical the American Psychological Association (2002)
pain if injured by the elderly person (Paveza et al., and the BACB (2017) are incompatible with the
1992; Teri et al., 1992). Most importantly, care science of behavior. Therefore, we should abandon
recipients with dementia are unable to exert ef- them entirely.
fective countercontrol, given their overall impair- Skinner’s contingency analysis of moral and
ment and often frail condition. ethical behavior should supplant those codes of
Although elder abuse has received widespread ethics. Although Skinner’s science of behavior is
attention in the media lately (Roberto, 2016), the inadequate in helping us determine what is moral
treatment of those who receive elder care has im- or ethical behavior, his analysis of the contingen-
proved considerably over the last several decades cies by which people behave ethically could be in-
as states have stepped in to ensure that better strumental in promoting ethical behavior among
568 P r o f e s s i o n a l I s s u e s i n A pp l i e d B e h av i o r A n a ly s i s
professionals. Although individuals might disagree Association for Advancement of Behavior Therapy.
about particulars, we can all agree about what is (1977). Ethical issues for human services. Behavior
and is not ethical in most instances. For example, Therapy, 8, v–vi.
striking patients under any circumstances is un- Behavior Analyst Certification Board (BACB). (2004).
Behavior Analyst Certification Board guidelines for
ethical. Likewise, an ethical practice might be
responsible conduct for behavior analysts. Retrieved
to maximize a patient’s independence, while de- from www.bacb.com.
creasing his or her dependence on other people. Behavior Analyst Certification Board (BACB). (2017).
By identifying the contingencies of which ethical Professional and ethical compliance code for behav-
behavior is a function, psychology and behav- ior analysts. Retrieved from www.bacb.com.
ior analysis can create the conditions that foster Benuto, L. T., Casas, J., & O’Donohue, W. T. (2018).
ethical behavior and obviate those conditions that Training culturally competent psychologists: A sys-
evoke unethical behavior. tematic review of the training outcome literature.
In closing, the wholesale adoption of the Amer- Training and Education in Professional Psychology, 12,
ican Psychological Association’s (2002) Ethics 125–134.
Burnes, D., Pillemer, K., Caccamise, P. L., Mason, A.,
Code should come as a surprise to the behavior-
Henderson, C. R., Berman, J., et al. (2015). Preva-
analytic community, given that it is antithetical to lence of and risk factors for elder abuse and neglect in
the tenets of the science of behavior. Surprisingly, the community: A population-based study. Journal of
the community has remained largely silent about the American Geriatrics Society, 63, 1906–1912.
this change. California Association for Behavior Analysis. (1996).
We might assume two things from the behavior- Cal-ABA’s code of ethics. Morro Bay, CA: Author.
analytic community’s lack of response. First, be- Dong, X. Q. (2015). Elder abuse: Systematic review and
havior analysts are silent on this issue because the implications for practice. Journal of the American Ge-
American Psychological Association’s (2002) Eth- riatrics Society, 63, 1214–1238.
ics Code is acceptable to them; hence they have Dyer, C. B., Pavlik, V. N., Murphy, K. P., & Hyman, D.
J. (2000). The high prevalence of depression and de-
no criticisms. This, of course, does not seem to be
mentia in elder abuse or neglect. Journal of the Ameri-
the case. Most behavior analysts would have been can Geriatrics Society, 48, 205–208.
outraged by the Association of Behavior Analysis Fischborn, M. (2018). Questions for a science of moral
International’s adoption of the American Psycho- responsibility. Review of Philosophy and Psychology, 9,
logical Association’s Ethics Code, had they con- 381–394.
sidered this action carefully. Second, they have Florida Association for Behavior Analysis. (1987). Code
not read the Ethics Code; they consider ethics of ethics of the Florida Association for Behavior Analy-
only superficially at best. Although this remains sis. Tallahassee, FL: Author.
an empirical question, we might safely assume that Kant, I. (1997). Lectures on ethics (P. Heath & J. B.
this is indeed the case. This attitude notwith- Schneewind, Trans.). Cambridge, UK: Cambridge
University Press.
standing, members of the behavior-analytic com-
Lachs, M. S., Williams, C., O’Brien, S., Hurst, L., &
munity should note that accepting the American Horowitz, R. (1997). Risk factors for reported elder
Psychological Association’s Ethics Code has se- abuse and neglect: A nine-year observational cohort
rious implications or consequences. The Ethics study. Gerontologist, 37, 469–474.
Code has no safeguards to ensure due process if MacIntyre, A. (1998). A short history of ethics (2nd ed.).
a behavior analyst should ever be the subject of Notre Dame, IN: University of Notre Dame Press.
an ethical inquiry (O’Donohue & Ferguson, 2003, Moore, G. E. (1988). Principia ethica. New York: Pro-
p. 7). Simply, there are no contingencies or coun- metheus Books. (Original work published 1903)
tercontrol mechanisms to ensure a fair trial. O’Donohue, W. T. (2016). Oppression, privilege, bias,
prejudice, and stereotyping: Problems in the APA
code of ethics. Ethics and Behavior, 26(7), 527–544.
REFERENCES O’Donohue, W. T., & Ferguson, K. E. (2003). Hand-
book of professional ethics for psychologists. New York:
Allison, J., & Timberlake, W. (1975). Response depri- SAGE.
vation and instrumental performance in the con- O’Donohue, W., Snipes, C., Dalto, G., Soto, C., Maraga-
trolled-amount paradigm. Learning and Motivation, kis, A., & Im, S. (2014). The ethics of enhanced in-
6, 112–142. terrogations and torture: A reappraisal of the argu-
American Psychological Association. (2002). Ethical ment. Ethics and Behavior, 24(2), 109–125.
principles of psychologists and code of conduct. Re- Paveza, G. J., Cohen, D., Eisdorfer, C., Freels, S., Semla,
trieved from www.apa.org/ethics/code. T., Ashford, J. W., et al. (1992). Severe family vio-
Behavior Analysis and Ethics 569
lence and Alzheimer’s disease: Prevalence and risk Smith, L. (1986). Behaviorism and logical positivism.
factors. Gerontologist, 32, 493–497. Stanford, CA: Stanford University Press.
Pillemer, K., Burnes, D., Riffin, C., & Lachs, M. S. Teri, L., Rabins, P., Whitehouse, P. V., Berg, L., Reis-
(2016). Elder abuse: Global situation, risk factors, berg, B., Sunderland, T., et al. (1992). Management of
and prevention strategies. Gerontologist, 56(Suppl. behavior disturbance in Alzheimer disease: Current
2), S194–S205. knowledge and future directions. Alzheimer Disease
Ringen, J. (1996). The behavior therapists’ dilemma: and Associated Disorders, 6, 77–88.
Reflections on autonomy, informed consent and sci- Texas Association for Behavior Analysis. (1995). Texas
entific psychology. In W. O’Donohue & R. F. Kitch- Association for Behavior Analysis code of ethics. Den-
ener (Eds.), The philosophy of psychology (pp. 352– ton, TX: Author.
360). Thousand Oaks, CA: SAGE. Tueth, M. J. (2000). Exposing financial exploitation of
Roberto, K. A. (2016). The complexities of elder abuse. impaired elderly persons. American Journal of Geriat-
American Psychologist, 71, 302–311. ric Psychiatry, 8, 104–111.
Skinner, B. F. (1953). Science and human behavior. New Wangmo, T., Nordström, K., & Kressig, R. W. (2017).
York: Macmillan. Preventing elder abuse and neglect in geriatric insti-
Skinner, B. F. (1972). Cumulative record (3rd ed.). New tutions: Solutions from nursing care providers. Geri-
York: Appleton-Century-Crofts. atric Nursing, 38, 385–392.
Skinner, B. F. (2002). Beyond freedom and dignity. New Wittgenstein, L. (1953). Philosophical investigations. New
York: Hackett. (Original work published 1971) York: Macmillan.
CHAP TER 34
Professional Certification
for Practicing Behavior Analysts
Holding a professional credential is the tradition- professional credential has become the primary
ally recognized way for an individual to demon- means of identifying qualified professionals.
strate to the public that he or she has met specific Behavior analysts had no specific and widely
and recognized standards of competence in a pro- recognized professional credential until the late
fession. Consumers generally use a profession’s cre- 1990s. Consumers seeking a qualified behavior
dential as a necessary qualification when selecting analyst before that did not have a credential on
a professional. For example, most of us would not which they could rely to aid their search: rather,
obtain the services of a surgeon, attorney, or den- they were forced to use less reliable means of as-
tist unless that person was properly credentialed. sessing practitioner qualifications, such as word of
Likewise, if we moved to a new city and needed mouth, a review of the practitioner’s curriculum
the services of an attorney, we might begin our vitae, and practitioner advertisements. Employ-
search by obtaining a list of attorneys who have ers and government agencies found themselves in
the proper professional credential. If we needed a a similar predicament and were forced either to
specialized service, such as a real estate lawyer, we develop their own criteria for a qualified behavior
might seek an attorney who also held a specialty analyst or to specify that individuals possess cre-
credential in that area. Of course, we might do dentials from other professions (e.g., a psychology
many other things to be sure of hiring a competent license, teacher certification). Neither of these
attorney (e.g., obtain references, talk with commu- options was particularly beneficial for consumers,
nity members, review pubic records about the at- due to their lack of specificity in relation to behav-
torney’s practice), but verifying that the attorney ior analysis.
was properly credentialed would be a reasonable
and customary first step in the selection process.
Employers and government agencies use a similar TYPES OF PROFESSIONAL CREDENTIALS
approach because it is an efficient and effective
means of hiring professionals or designating the Professional credentials sometimes take the form
required qualifications for a position. Although of a license issued by the state, provincial, or na-
merely having a credential may not be sufficient tional government. These licenses are usually es-
for an individual to obtain a given professional tablished through a law that limits either the use
position, it is most often a prerequisite. Thus, a of a professional title or engagement in specified
570
Professional Certification 571
professional practice to those holding the appro- written examination, continuing education for cre-
priate license (Green & Johnston, 2009; Redbird, dentialed individuals, and disciplinary oversight.
2017). Under a title law, for example, individuals Applicants must first meet specific eligibility stan-
may identify themselves as licensed behavior ana- dards to qualify for examination. These standards
lysts only if they are licensed by the government. usually require the applicant to have a minimum
Similarly, under a practice law, individuals may educational degree, specific university course-
practice behavior analysis only if they are licensed work, and some form of supervised experience.
as a behavior analyst. If the law combines title and The cornerstone of any professional credentialing
practice features, individuals may not use the title program, however, is the professionally developed
or engage in the practice of behavior analysis un- written examination. To be viable, the written ex-
less they are appropriately licensed or explicitly amination must meet both psychometric and legal
exempted from the law. Usually the licensure law standards; that is, the examination must separate
also establishes a professional board to oversee the “those who know” from “those who do not know”
licensing process and review disciplinary matters and must also be defensible in a court of law. To
involving licensed professionals. these ends, examinations must be constructed fol-
A second form of credential is professional cer- lowing established psychometric and legal stan-
tification. Although licensure and professional dards, which are often promulgated by standards-
certification take fundamentally the same ap- setting bodies (American Educational Research
proach to the credentialing process, professional Association, American Psychological Association,
certification differs from licensure in these ways: & National Council on Measurement in Educa-
A private organization rather than a government tion, 2014; ANSI, n.d.; Institute for Credentialing
usually issues it; most often it is not limited to a Excellence, n.d.; U.S. Equal Employment Oppor-
single governmental jurisdiction; and it can be vol- tunity Commission, 1978). Next, individuals who
untary for practice. The profession rather than the become credentialed must usually obtain continu-
government usually drives certification; thus, it is ing education (CE) to maintain their credential’s
not as subject to the idiosyncrasies of a politically currency. The purpose of a CE requirement is to
influenced governmental process, because a pri- help ensure that a credentialed professional con-
vate organization develops and issues professional tinues to grow professionally and keeps abreast of
certification. In addition, because professional cer- new developments and techniques in the field. Fi-
tification can cross state borders, it is more portable nally, most credentialing organizations include a
for certificants than local government licensure fourth step of disciplinary or ethical compliance
is (Hall & Lunt, 2005). Although licensure and based on a published code of ethics.
professional certification differ in many respects,
one parallel requirement is that both must adhere
to the same legal, psychometric, and professional THE HISTORY OF BEHAVIOR
standards in the development of their credentials ANALYST CREDENTIALING
and assessment instruments, and in the operation
of their credentialing programs. One of the ways There have been several attempts to establish
that professional certification programs demon- professional credentials for practicing behavior
strate their adherence to established credential- analysts (Johnston, Carr, & Mellichamp, 2017).
ing standards is to obtain accreditation from a For example, Minnesota initiated a certification
national standard-setting organization such as the program in the 1970s for practitioners working in
American National Standards Institute (ANSI, the state developmental disabilities system, but it
n.d.) or the National Commission for Certifying eventually fell victim to changes in philosophy
Agencies (NCCA; Institute for Credentialing Ex- and priorities in the state (Thomas, 1979). The
cellence, n.d.). Association for Behavior Analysis1 sponsored a
certification program in the mid-1980s, but later
dissolved it for legal and practical reasons (Shook
CREDENTIALING PROGRAM COMPONENTS et al., 1988). In addition, Florida implemented a
AND REQUIREMENTS certification program based on a professionally de-
veloped examination in the mid-1980s (Johnston of behavior analysis. Individuals who hold BCaBA
& Shook, 1993; Starin, Hemingway, & Hartsfield, certification may practice behavior analysis under
1993) because of misuse of behavioral procedures the supervision of a BCBA. In addition, behavior
(Bailey & Burch, 2016; Johnston & Shook, 1987). analysts who hold BCBA certification and meet
Oklahoma, Texas, California, Pennsylvania, and certain doctoral-degree requirements may qualify
New York followed suit in the 1990s with programs for a doctoral designation: Board Certified Be-
based on Florida’s successful model and using Flor- havior Analyst–DoctoralTM (BCBA-DTM). Since
ida’s examination. One state’s providing an exami- 2007, both the BCBA and BCaBA certification
nation to another state was highly unusual, but in programs have been accredited by NCCA contin-
the case of applied behavior analysis (ABA), this is uously, which indicates that the BACB adheres to
exactly what happened. Florida paved the way for established testing and legal standards for boards
other states to follow, and they did. As additional that grant professional credentials.
states and governments became interested in Flor- Both BCBA and BCaBA certification have
ida’s program, the need for a broader credentialing specific requirements for degree, coursework, and
program for the profession became clear. supervised experience (see Figure 34.2) that ap-
BACB Certifications
The BACB offers two professional certification
programs for behavior analysts:2 Board Certified
Behavior Analyst® (BCBA®) and Board Certified
Assistant Behavior Analyst® (BCaBA®) (see Fig-
ure 34.1). Individuals who hold BCBA certifica-
tion are graduate-level, independent practitioners
Certification Maintenance
BCBA or BCaBA certification is granted only after
an individual has met the eligibility requirements
mentioned above and passed the appropriate ex-
FIGURE 34.2. A visual depiction of the examination eli-
amination. After obtaining a BACB credential, a
gibility process for BCBA and BCaBA certification.
certificant must meet maintenance requirements
to help maintain an acceptable level of profession-
al currency and competence in behavior analysis.
plicants must meet to qualify for the appropriate Professionals with BCBA and BCaBA certifica-
written examination. Applicants for the BCBA tions must obtain CE, which can be earned in sev-
examination need at least a master’s degree in an eral ways. These include passing university courses
acceptable academic domain, graduate coursework in behavior analysis, teaching such courses, at-
in specific behavior-analytic content, and super- tending state and national conferences, taking
vised professional behavior-analytic experience3 online tutorials, presenting approved sessions at
to qualify. Applicants for the BCaBA examination conferences, and engaging in certain scholarly ac-
need at least a bachelor’s degree, undergraduate or tivities (such as editorial reviews and publication
graduate coursework in specific behavior-analytic of journal articles). Certificants must dedicate a
content, and supervised professional behavior-an- portion of the required CE to ethics content and
alytic experience to qualify. to supervision content for those providing supervi-
sion to ensure continued development in particu-
BACB Examinations larly important areas.
The BACB’s examinations are developed and re-
viewed routinely through processes that adhere to Ethics and Discipline
best-practice guidelines for credentialing, nation- In addition to the maintenance requirements
ally accepted examination development standards, described above, the BACB employs a disciplin-
and NCCA requirements (Johnston, Mellichamp, ary system for ethical and professional behav-
Shook, & Carr, 2014). The BACB uses multiple ior for certification applicants and certificants.
workgroups of carefully selected subject-matter ex- The BACB’s disciplinary system is based on the
perts to write and revise examination items and BACB’s ethics code, the Professional and Ethical
make recommendations for ongoing changes to Compliance Code for Behavior Analysts (hereaf-
the BCBA and BCaBA task lists: this input serves ter referred to as the Code; BACB, 2014a). The
as the basis of examination content. Task list revi- Code covers the wide range of ethical and profes-
sions are content-validated through a large-scale sional situations faced by practicing behavior ana-
lysts. It also serves as a reference for consumers and
3 There are currently two other eligibility options for BCBA employers when they are trying to determine what
certification for faculty members and senior, doctoral-level manner of behavior is appropriate and should be
practitioners. expected from a behavior analyst faced with ethi-
574 P r o f e s s i o n a l I s s u e s i n A pp l i e d B e h av i o r A n a ly s i s
cal and professional issues. The BACB enforces rized Continuing Education providers worldwide.
the Code through a process designed to provide A marked increase in the availability of CE con-
consumer protection, appropriate due process for tent on ethics and professional behavior has oc-
the certificant involved, mentorship where ap- curred after the 2008 implementation of the
propriate, and consequences to the certificant for BACB’s requirement that every certificant obtain
violations. Disciplinary sanctions, including the a proportion of CE in that area. A similar phe-
suspension or revocation of an individual’s certi- nomenon occurred in response to the BACB’s
fication, are posted on the BACB’s website (www. training and CE requirements for certificants who
bacb.com). wish to provide supervision (BACB, 2012). These
data suggest that the BACB’s CE requirements
have influenced the landscape of available train-
Growth of BACB Certification
ing for behavior analysts.
The profession of behavior analysis has grown sub-
stantially in recent years (Carr & Nosik, 2017). As
Standards Changes
one example, the total number of individuals who
hold BCBA or BCaBA certification worldwide has Regulatory standards change in any profession ac-
increased more than sevenfold since 2008 when cording to the progress of the profession, such as
there were 6,636 certificants (i.e., 4,747 BCBAs available training programs, consumer needs, and
and 1,889 BCaBAs; Shook & Favell, 2008). As published research findings. The BACB addresses
of August 2020, the total number of certificants is this need for behavior analysts by reviewing and
48,219 (i.e., 43,476 BCBAs and 4,743 BCaBAs). As revising its standards periodically using processes
demand for the BACB’s certification programs has mandated by the NCCA. These processes involve
increased, there has also been substantial growth workgroups of subject-matter experts who make
in the number of educational programs providing recommendations regarding standards and their
behavior-analytic training. For example, there are revision to the BACB’s Board of Directors. For il-
currently over 300 institutions worldwide offer- lustrative purposes, the following paragraphs de-
ing coursework identified as meeting the require- scribe changes to three standards areas.
ments for BCBA or BCaBA eligibility (i.e., verified Eligibility standards for the BCBA and BCaBA
course sequences). As the number of institutions certification include specific requirements for
with verified course sequences has grown, so has qualifying degree, coursework, and supervised ex-
the need for feedback regarding students’ per- perience. Unlike the coursework and experience
formance on BACB examinations. The BACB requirements, which have always been behavior-
provided data in the early years for each verified analytic, the original degree requirement was not
course sequence on the percentage of its graduates limited to any specific field of study. The BACB
who passed a BACB examination on the first at- restricted the BCBA-level requirement in 2011 to
tempt. Since 2014, the BACB has publicly posted degrees from certain academic domains. It further
pass rates online. These changes are efforts to as- restricted the BCBA degree requirement in 2016
sist prospective students in making informed deci- to degrees in behavior analysis, education, psy-
sions about their training before enrollment, and chology, or degree programs in which an applicant
to provide feedback for faculty about one outcome completed a verified course sequence. The prima-
of their training. Beginning in 2019, ABAI as- ry pathway toward BCBA certification will require
sumed responsibility for administering the verified a graduate degree from a program accredited by
course sequence system. ABAI beginning in 2022, although a secondary
The increase in the number of BACB cer- pathway will allow degrees from other academic
tificants has created a corresponding increase in domains.
demand for CE opportunities in behavior analy- The BACB convened a workgroup in 2017 to
sis. The BACB operates a process for identifying review the supervised experience standards that
providers of BACB-approved CE content at the apply to individuals pursuing BCBA or BCaBA
organizational and individual levels (Authorized certification (BACB, 2017). The recommenda-
Continuing Education providers). Although CE tions of that workgroup produced changes to sev-
opportunities were previously difficult to find out- eral features of the experience standards, which
side of a few annual conventions, they are now will take effect in 2022. First, new BCBAs will
plentiful. There are currently over 1,000 Autho- need to wait 1 year before supervising individu-
Professional Certification 575
als pursuing BCBA or BCaBA certification. The Autism Spectrum Disorder Practice Guidelines
BACB revised acceptable activities to limit direct
In response to the growing demand for ABA ser-
implementation of behavioral programs to no more
vices for autism spectrum disorder, the BACB pub-
than 40% and 60% of the total supervised experi-
lished practice guidelines in 2012 for healthcare
ence for BCBAs and BCaBAs, respectively. It in- funders and managers, consumers, and service
creased the amount of supervised experience from providers. It later revised the practice guidelines in
1,500 to 2,000 and from 1,000 to 1,300 hours for 2014 (BACB, 2014c). The purpose of the guidelines
BCBA and BCaBA trainees, respectively. Finally, is to support ABA treatment for individuals diag-
it clarified the standards regarding the nature and nosed with autism spectrum disorder that is con-
structure of supervision, appropriate activities, and sistent with the best available scientific evidence
supervision documentation. and expert clinical opinion. Some of the unique
The BACB convened a workgroup in 2014 to features of ABA treatment that the guidelines ad-
review several standards, including the standards dress include: training and credentialing of behav-
for the ongoing supervision of BCaBAs. Origi- ior analysts; assessment, formulation of treatment
nally developed in 2007 and implemented in 2009, goals, and measurement of client progress; service
the policy at that time required every BCaBA to authorization and dosage; tiered service delivery
obtain 1 hour of supervision per month from a models; behavior technicians; and case supervi-
BCBA. The new policy, which went into effect in sion. In addition, the document provides general
2017, requires BCaBAs to be supervised for 2% of guidance and basic descriptions of typical ABA
the time they spend providing ABA services. The service delivery that can be tailored to fit individu-
policy also addresses many issues that the previ- al, local, and regional requirements and needs. As
ous version did not, such as supervisory caseload, of 2020, these practice guidelines have been trans-
supervisor qualifications, the supervisor’s responsi- ferred to the Council of Autism Service Providers
bility for the BCaBA’s clinical work, contracts, and (CASP, 2020), who will be responsible for manag-
group supervision (BACB, 2014b). ing and updating the document in the future.
FIGURE 34.3. Cumulative number of U.S. licensure laws enacted through 2020 (N = 31). The states are indicated below
the x-axis.
576 P r o f e s s i o n a l I s s u e s in App l ie d B e h av i o r A n a ly s i s
U.S. STATE LICENSURE OF BEHAVIOR ANALYSTS existing ones and the other resources the BACB
has been able to provide will continue to serve as
Since 2009, 31 U.S. states have enacted laws to reg- an important resource to the profession as it con-
ulate the practice of ABA through licensure (As- tinues to mature.
sociation of Professional Behavior Analysts, n.d.;
see Figure 34.3 on p. 575). Thirty-one states cur-
rently provide a pathway to licensure for behavior ACKNOWLEDGMENT
analysts with a graduate degree (Alabama, Alas-
Gerald L. Shook contributed to the version of this chap-
ka, Arizona, Connecticut, Hawaii, Iowa, Kansas,
ter in the first edition of the Handbook of Applied Behav-
Kentucky, Louisiana, Maryland, Massachusetts, ior Analysis.
Michigan, Mississippi, Missouri, Montana, Ne-
vada, New Jersey, New York, North Dakota, Ohio,
Oklahoma, Oregon, Rhode Island, South Dakota, REFERENCES
Tennessee, Texas, Utah, Vermont, Virginia, Wash-
ington, Wisconsin), and 24 of those states provide American Educational Research Association, Ameri-
a pathway to licensure for assistant behavior ana- can Psychological Association, & National Council
lysts with an undergraduate degree (Alabama, on Measurement in Education. (2014). Standards for
Alaska, Iowa, Kansas, Kentucky, Louisiana, Mas- educational and psychological testing (2014 ed.). Wash-
sachusetts, Michigan, Mississippi, Missouri, Mon- ington, DC: American Educational Research Asso-
tana, Nevada, New Jersey, New York, North Dako- ciation.
ta, Oklahoma, Oregon, Rhode Island, Tennessee, American National Standards Institute. (n.d.). Accred-
itation program for personnel certification bodies
Texas, Utah, Vermont, Virginia, Washington)4.
under ANSI/ISO/IEC 17024. Retrieved from https://
The BACB’s credentials or standards constitute a anab.ansi.org/credentialing/personnel-certification.
pathway to licensure in these states, which assists Association of Professional Behavior Analysts. (n.d.).
behavior analysts who hold BACB credentials to Licensure and other regulation of ABA practitioners.
move to different states and still qualify for licen- Retrieved from www.apbahome.net/general/recom-
sure. We have yet to determine the impact of licen- mended_links.asp.
sure on the practice of behavior analysis, because Bailey, J. S., & Burch, M. R. (2016). Ethics for behavior
these legislative activities have occurred in the last analysts (3rd ed.). New York: Routledge.
decade, and several other states are still pursuing Behavior Analyst Certification Board (BACB). (2012,
licensure. However, the existence and growth of September). Special edition on supervision. BACB
licensure constitutes another indicator of the pro- Newsletter. Retrieved from https://www.bacb.com/wp-
content/Sep2012_Newsletter.
fession’s increasing maturity.
Behavior Analyst Certification Board (BACB). (2014a).
Professional and ethical compliance code for behav-
ior analysts. Retrieved from www.bacb.com/ethics/
CONCLUSION ethics-code.
Behavior Analyst Certification Board (BACB). (2014b,
As a profession, ABA has undergone substantial December). Special edition on BCaBA supervision
growth in recent years. The growth is evident in standards. Retrieved from https://www.bacb.com/wp-
the increasing number of credentialed behavior content/Dec2014_Newsletter.
analysts, behavior-analytic training programs, Behavior Analyst Certification Board (BACB). (2014c).
and laws for the practice of behavior analysis; the Applied behavior analysis treatment of autism spectrum
increasing demand for behavior analysts (BACB, disorder: Practice guidelines for healthcare funders and
managers (2nd ed.). Littleton, CO: Author.
2019); and the development of behavior-analytic
Behavior Analyst Certification Board (BACB). (2017,
infrastructure around the world. The BACB’s pro- October). Revised BCBA and BCaBA experience
fessional certification programs, which provide a and supervision requirements. BACB Newsletter.
foundation for U.S. state licensure, represent cur- Retrieved from https://www.bacb.com/wp-content/
rent standards for behavior analysts developed by Oct2017_Newsletter.
behavior analysts around the world. The BACB’s Behavior Analyst Certification Board (BACB). (2019).
systems for developing new standards and revising U.S. employment demand for behavior analysts: 2010–
2018. Littleton, CO: Author.
4 Some states use the titles certified or registered instead of Carr, J. E., & Nosik, M. R. (2017). Professional creden-
licensed for their behavior analyst or assistant behavior ana- tialing of practicing behavior analysts. Policy Insights
lyst credentials. from the Behavioral and Brain Sciences, 4, 3–8.
Professional Certification 577
Council of Autism Service Providers. (2020). Applied Martin, N. T., & Carr, J. E. (2020). Training and cer-
behavior analysis treatment of autism spectrum dis- tification of behaviour analysts in Europe. European
order: Practice guidelines for healthcare funders and Journal of Behaviour Analysis, 21, 9–19.
managers (2nd ed.). Retrieved from casproviders.org/ Redbird, B. (2017). The new closed shop?: The econom-
asd-guidelines. ic and structural effects of occupational licensure.
Green, G., & Johnston, J. M. (2009). A primer on pro- American Sociological Review, 82, 600–624.
fessional credentialing: Introduction to invited com- Shook, G. L. (2005). An examination of the integrity
mentaries on licensing behavior analysts. Behavior and future of Behavior Analyst Certification Board
Analysis in Practice, 2(1), 51–52. credentials. Behavior Modification, 29, 562–574.
Hall, J. E., & Lunt, I. (2005). Global mobility for psy- Shook, G. L., & Favell, J. E. (2008). The Behavior
chologists: The role of psychology organizations in Analyst Certification Board and the profession of
the United States, Canada, Europe, and other re- behavior analysis. Behavior Analysis in Practice, 1(1),
gions. American Psychologist, 60, 712–726. 44–48.
Hughes, J. C., & Shook, G. L. (2007). Training and Shook, G. L., Johnston, J. M., Cone, J., Thomas, D.,
certification in behaviour analysis in Europe: Past, Greer, D., Beard, J., et al. (1988). Credentialing, qual-
present, and future challenges. European Journal of ity assurance and right to practice. Kalamazoo, MI: As-
Behavior Analysis, 8, 239–249. sociation for Behavior Analysis.
Institute for Credentialing Excellence. (n.d.). National Shook, G. L., Johnston, J. M., & Mellichamp, F. (2004).
Commission for Certifying Agencies (NCCA) stan- Determining essential content for applied behavior
dards. Retrieved from https://www.credentialingexcel- analyst practitioners. Behavior Analyst, 27, 67–94.
lence.org/page/ncca. Starin, S., Hemingway, M., & Hartsfield, F. (1993). Cre-
Johnston, J. M., Carr, J. E., & Mellichamp, F. H. (2017). dentialing behavior analysts and the Florida behav-
A history of the professional credentialing of applied ior analysis certification program. Behavior Analyst,
behavior analysts. Behavior Analyst, 40, 523–538. 16, 153–166.
Johnston, J. M., Mellichamp, F. H., Shook, G. L., & Thomas, D. R. (1979). Certification of behavior analysts
Carr, J. E. (2014). Determining BACB examination in Minnesota. Behavior Analyst, 2(1), 1–13.
content and standards. Behavior Analysis in Practice, U.S. Equal Employment Opportunity Commission.
7, 3–9. (1978). United States Equal Employment Opportunity
Johnston, J. M., & Shook, G. L. (1987). Developing be- Commission guidelines on employment testing proce-
havior analysis at the state level. Behavior Analyst, dures. Washington, DC: Author.
10, 199–233. Virués-Ortega, J., Shook, G. L., Arntzen, E., Martin,
Johnston, J. M., & Shook, G. L. (1993). A model for the N., Rodríguez-García, V., & Rebollar-Bernardo, M.
statewide delivery of programming services. Mental A. (2009). Campo profesional y procedimientos de
Retardation, 31, 127–139. certificación en análisisaplicado del comportamiento
Johnston, J. M., & Shook, G. L. (2001). A national cer- en España y Europa [Professional field and certifica-
tification program for behavior analysts. Behavioral tion in applied behavior analysis: Spain and Europe].
Interventions, 16, 77–85. Papeles del Psicologo, 30, 1–10.
Author Index
Abellon, O. E., 547 Akmanoglu, N., 259 Altman, K., 355, 365
Aberle, S., 549 Alavosius, M. P., 287 Altmann, J., 143
Abernathy, B., 552 Alber-Morgan, S. R., 259, 261 Altomari, T. S., 328
Abramowitz, A. J., 362 Alberto, P. A., 256, 259, 262, 357, 376 Altschaefl, M. R., 464
Abt, K. A., 62, 224 Albin, R. W., 38, 197 Altus, D. E., 512, 519
Acar, G., 259 Alderman, N., 208 Alvarez, B., 20
Acheson, A., 125 Aldrich, N. J., 262 Alvero, A. M., 287, 288, 544, 548, 551,
Acierno, R. E., 26 Alessandri, S. M., 327 548, 552
Acker, J., 126 Alessi, C., 518 Alverson, L. G., 146
Ackerlund Brandt, J. A., 183, 244 Alessi, G., 450, 452 Alwis, D. S., 304
Acquisto, J., 311, 341, 353 Alessi, S. M., 491, 494, 495, 498, 500, Alzrayer, N., 275
Adams, C. M., 30 501 Aman, M. G., 72
Adams, G., 262 Alevizos, P., 142 Amaral, D. G., 271
Adams, G. L., 259, 454 Alexander, J. L., 83 Amari, A., 141, 179, 199, 435
Adams, L. A., 410 Alexander, P. A., 261 Amigo, S., 546, 548
Adams, M., 453 Alfieri, L., 262 Amlung, M. T., 126
Adams, R. M., 144 Algozzine, B., 468 Anand, N. K., 433
Addison, L. R., 118, 182, 203, 242, 274, Algozzine, K., 468 Andelman, M., 327
397, 400, 435 Allard, A. M., 390 Anders, B. M., 122, 304, 343
Adelinis, J. D., 216, 218, 225, 304, 345, Allen, E. M., 377 Anderson, C. A., 355
356, 366 Allen, K. D., 307, 356, 422 Anderson, C. M., 204, 208, 209, 433,
Adkins, V. K., 286, 289, 512 Allen, K. E., 142, 206 438
Adolf, K. J., 126, 240, 252 Allen, S., 408 Anderson, J., 354
Agar, S., 521 Allison, D. B., 302, 314 Anderson, T., 453
Agnew, J. L., 210, 546, 547, 549, 550 Allison, J., 20, 361, 439, 564 Andrade, I., 378
Ahearn, W. H., 73, 102, 103, 104, 106, Allman, M. J., 118 Andree, P. J., 312, 343
109, 122, 177, 185, 186, 203, 208, Almason, S. M., 276, 398 Andrews, G., 238
258, 277, 329, 396, 433, 434, 448, Almeida, D., 291, 468 Andrews, V. H., 416
471 Alnemary, F. M., 468 Ang, A., 503
Ahlers, A. A., 184 Alonso-Perez, F., 494 Anger, W. K., 382, 549
Ainsleigh, S. A., 313 Alpern, M., 293 Anhalt, K., 375, 380
Akers, J. S., 88, 203, 228, 247 Alsop, B., 100, 101, 515 Annest, J. L., 477
Aklin, W. M., 504, 505 Alterson, C. J., 136, 225, 228 Antonitis, J. J., 327
579
580 Author Index
Aparicio, C. F., 100 Bachmeyer, M. H., 322, 324, 432, 434, Barry, L. M., 468
Applegate, H., 141 435, 437, 439 Bartholomew, C. C., 471
Aradóttir, B., 278 Badger, G. J., 125, 491 Bartlett, B. L., 276
Arah, O. A., 431 Baer, A. M., 357 Barton, C. G., 546
Archer, A. L., 453 Baer, D. M., 10, 11, 31, 78, 82, 85, 88, Barton, E., 512
Ardoin, S. P., 235, 246, 252, 261, 263, 135, 147, 155, 202, 238, 239, 246, Barton, L. E., 357, 363
264, 445, 449, 455, 457, 458 254, 260, 275, 276, 323, 357, 375, Barton, R. P., 362
Areán, P. A., 517, 519 379, 380, 393, 401, 402, 449, 485 Barwick, K. D., 549
Arevalo, A. R., 324 Baeyens, D., 125 Basile, V. C., 302
Arfken, C. L., 500 Baez, C. N., 402 Bass, R. F., 138
Ariely, D., 127 Bagne, A., 481 Bassingthwaite, B. J., 531, 534, 536
Armstrong, A., 228 Baham, M., 326 Bastian, A. M., 477
Armstrong, F. D., 477 Bahl, A. B., 375 Battaglia, A. A., 180
Armstrong, K. J., 361 Bai, J. Y. H., 108 Bauer, J., 279
Arndorfer, R. E., 205, 207 Bailey, A., 390 Bauld, L., 494
Arntzen, E., 37, 365, 516 Bailey, D. B., 447 Baum, W. M., 8, 58, 70, 71, 97, 98, 99,
Arora, S., 532 Bailey, J. A., 95 100, 101
Arrington, R. E., 147 Bailey, J. S., 138, 144, 146, 169, 287, Bauman, K. E., 9, 95, 198, 204, 215,
Arvedson, J. C., 428 289, 293, 465, 471, 482, 545, 548, 312, 321, 337, 358, 391, 534
Ary, D. V., 144, 492 550, 572 Bauman, R., 146
Ascione, F. R., 519 Baires, N. A., 194 Bauman, R. A., 123
Ash, D. W., 260 Bakeman, R., 208 Baumann, A. A., 124
Ashbaugh, R., 205 Baker, A. G., 366 Baumeister, A. A., 302, 327, 365, 366
Asmus, J. M., 222, 228, 327 Baker, C. I., 31 Baxley, N., 85
Atak, J. R., 60 Baker, J. C., 388, 512, 513, 516, 519, Beale, I. L., 146
Athens, E. S., 96, 166, 167, 168, 237, 520, 521, 522, 523 Beaton, S., 522
245, 256, 321, 332, 342, 391 Baker, S. A., 477 Beattie, S., 520
Atwell, A., 303 Baker, V., 287 Beauchamp, K. L., 122, 547
Atwell, J., 166, 307, 433 Bakken, J., 289 Beaudry-Bellefeuille, I., 411
Augustine, M., 243 Balcazar, F., 288 Beaulieu, L., 199, 464
Auguston, E. M., 87 Balsam, P. D., 17 Beaver, B. R., 141
Ault, M. H., 138, 202 Baltes, M., 512 Beavers, G. A., 216, 222, 311, 321,
Ault, M. J., 82, 258, 401 Balu, R., 446 338, 342
Ausenhus, J. A., 291, 292 Bamburg, J. W., 196 Becirevic, A., 117
Austin, J., 210, 287, 294, 544, 545, 546, Bamon, M. J., 374 Beck, A. M., 513
547, 548, 549, 550, 551, 552, 554 Banco, L. I., 477 Beck, K., 482, 485
Austin, J. E., 246, 344 Bancroft, S., 258, 329 Becker, J., 365
Austin, J. L., 226, 239, 247 Banda, D. R., 275 Becker, W. C., 382, 453, 454
Austin, M., 491 Bandini, L. G., 430 Becker, W. S., 262
Avsar, K. B., 125 Bandura, A., 44, 257 Becraft, J. L., 238, 239
Axe, J. B., 274 Bangs, K. E., 455 Bedell, M. A., 105
Axelrod, S., 262, 279, 352, 455 Banko, K. M., 52, 186 Begeer, S., 390
Ayllon, T., 242, 358, 374, 375, 376, Banks, R. K., 362 Begeny, J. C., 387, 445, 446, 453, 467
377, 381 Barbera, M. L., 271 Belanger, S., 82
Ayres, J. J. B., 18, 21 Barker, R., 206 Belfiore, P. J., 313, 471
Ayres, K. M., 83, 257 Barlow, D. H., 26, 140, 142 Belisle, J., 399
Azrael, D. R., 477 Barnes, M., 363 Bell, M. C., 103
Azrin, N. H., 42, 44, 45, 46, 242, 243, Barnes-Holmes, D., 5, 85, 87, 253 Bellizzi, A., 548
302, 327, 354, 355, 358, 362, 363, Barnett, D. W., 263, 446, 457 Belsky, J., 512, 516
364, 366, 374, 375, 376, 377, 378, Barnett, L., 436 Benedict, J. O., 21
381, 382, 390 Barnett, N. P., 497 Benishek, L. A., 375, 493
Baroff, G. S., 305, 365 Bennett, C. M., 390
Baron, A., 241, 341, 352, 362, 363, Bennett, G. M., 185
B 513, 523 Bennett, K. D., 256
Barone, V. J., 422 Bensemann, J., 102
Babbitt, R. L., 324, 433, 434, 435, 436, Barretto, A., 530, 534 Benuto, L. T., 561
438 Barrish, H. H., 247, 467 Bereiter, C., 262
Babcock, R. A., 285, 292 Barry, C. L., 477 Berenson, G. S., 438
Bachman, J. E., 302, 303 Barry, D., 501 Berg, I., 416
Author Index 581
Berg, T., 512 Bloom, S. E., 52, 118, 136, 140, 183, Bradfield, G., 477
Berg, W. K., 246, 323, 530 184, 203, 222, 225, 226, 227, 338, Bradley, T. A., 457
Bergan, J. R., 464, 465, 466, 470 468 Bradley-Johnson, S., 82
Bergmann, S. C. J., 8, 30, 387, 389, Bloomsmith, M. A., 120, 229 Bradshaw, B., 293
536 Blough, D. S., 80 Brady, J. V., 366, 490, 507
Bergstrom, M. K., 455 Blum, N. J., 408, 409, 422 Brady, K. T., 124, 496
Berk, L. B., 414 Blum-Kulka, S., 206 Brady, M. P., 476
Berkler, M. S., 142 Blyler, E., 433 Braga-Kenyon, P., 471
Berkson, G., 215, 303 Bobo, H. E., 302 Branch, M. N., 29
Bernstein, I. L., 26, 430 Boerke, K., 244, 374 Brand, D., 519
Berotti, D., 348 Boettcher, W., 549 Brandauer, C., 62
Berridge, K. C., 32 Bogard, K., 549 Brandt, C. W., 160
Bersh, P. J., 49, 331 Bogoev, B. K., 183 Braun, L. M., 453
Berth, D. P., 323, 324, 432 Boiron, M., 433 Braunling-McMorrow, D., 86
Bessellieu, F. B., 262 Bojak, S. L., 183 Bravo, A., 277
Bethune, K. S., 263 Boles, S. M., 492 Bray, M. A., 376, 380
Bettin, B., 287 Bolles, R. C., 366 Brekstad, A., 37
Betz, A. M., 214, 247, 260, 346, 347, Bolton, D., 293 Bremer, B. A., 361
391, 548 Bondy, A., 88, 235, 270, 271, 273, 275, Brendlinger, J., 354
Beutler, L. E., 302 276, 278 Brethower, D. M., 552
Bevan, D., 239 Bonfiglio, C. M., 449, 451 Bricker, W. A., 288
Bible, G. H., 287 Bonner, A. C., 324 Briesch, A. M., 264
Bicard, D. F., 72 Bonner, R. T., 100 Briggs, A. M., 194, 228, 311
Bice-Urbach, B., 464 Booth, D., 411 Bristow, A., 365
Bickel, W. K., 122, 124, 125, 126, 127, Bootzin, R. R., 244, 374, 409, 410 Brobst, B., 446
184, 382 Borgealt, A. J., 28 Brodhead, M. T., 247
Bigelow, G. E., 492, 502, 504 Borrero, C. S. W., 95, 96, 97, 98, 99, Brodsky, L., 428
Biggs, J., 548 100, 101, 203, 204, 206, 209, 326, Bronfenbrenner, U., 466
Bihm, E. M., 194, 196 329, 330, 431, 438 Broocks, A., 302
Bijou, S. W., 138, 139, 142, 202, 206, Borrero, J. C., 65, 118, 122, 125, 173, Brooks, A., 264
207, 237, 241, 242 177, 182, 185, 202, 203, 206, 209, Brooks, P. J., 262
Billingsley, F. F., 258 228, 238, 239, 240, 246, 329, 330, Brookshire, R. H., 146, 147
Bilyew, D. D., 518 338, 374, 376, 383, 431, 438 Broster, B. S., 308
Bindal, D., 531 Borsari, A., 427 Brothers, K. J., 256, 547
Binder, C. V., 263, 448, 449, 450, 455, Bosch, A., 486 Broussard, C. D., 179, 226
548, 554 Boscoe, J. H., 382, 433, 436 Brousseau, D. C., 532
Binder, K. S., 264 Bosman, A. M., 290 Brouwer, R., 482
Binder, L. M., 72 Bosworth, K., 467 Brower-Breitwieser, C., 486
Binkoff, J. A., 47, 214, 312, 314 Bourgeois, M., 522 Brown, D., 512
Biondi, D. R., 123 Bourland, G., 100 Brown, J., 86
Biran, S., 194 Bourret, J. C., 20, 203, 208, 244, 245, Brown, K. M., 293
Bird, F. L., 291, 313 329, 400, 401 Brown, M. R., 277
Birnbrauer, J. S., 365 Bouton, M. E., 26, 105, 106, 107 Brown, P. L., 26
Birren, J. E., 513 Boutot, E. A., 275, 276 Brown, R., 271, 532
Bishop, J. R., 159, 363 Bouts, L., 205 Brownell, M. T., 452
Bishop-Fitzpatrick, L., 279 Bouxsein, K. J., 160, 219, 239, 276, Bruce, A., 430
Bitgood, S. C., 354, 366 337, 391 Brucker, D. L., 463
Bittle, R. G., 365 Bowen, C. N., 185, 260, 276, 278 Bruellman, J., 482
Blackhurst, A. E., 256 Bowman, L. G., 43, 68, 107, 136, 141, Brulle, A. R., 147, 357
Blackman, D. K., 522 179, 180, 184, 199, 218, 225, 229, Bruner, N. R., 125
Blakely, M. R., 471 240, 243, 304, 310, 353, 354, 356, Brusca, R., 137
Blakemore, S. D., 531 357, 359, 360, 361, 364, 376, 390, Bruzek, J., 339, 536
Blampied, N. M., 326, 328, 409, 410 391 Bryceland, J. A., 453
Bland, V. J., 108 Boyajian, A. E., 74 Bryson, S. E., 402
Blanton, W. E., 452 Boyer, J., 446 Buchanan, J. A., 512, 521
Blasé, K. A., 290 Boyle, M. A., 118, 183, 184 Buckley, J. L., 100
Blazer, D., 517 Brache, A. P., 545, 552, 554 Buckley, S. D., 433
Blevins, T., 305 Bracken, M., 279 Bucklin, A., 195
Blew, P. A., 302 Brackett, L., 289 Bucklin, B. R., 287, 544, 545, 548
582 Author Index
De Vries, T. J., 127 Diefenbach, G. J., 517 Doyle, P. M., 82, 258, 401
de Wit, H., 125 Diergaarde, L., 127 Doyne, E. J., 302
DeBar, R. M., 291 Dietz, A., 139 Dozier, C. L., 183, 194, 203, 222, 244,
DeBernardis, M., 342 Dietz, J., 433 343, 365, 376, 391
Deci, E. L., 186 Dietz, W. H., 438 Drabman, R. S., 357, 375, 376, 377,
DeFulio, A., 490, 504, 505, 507 DiGennaro Reed, F. D., 126, 127, 210, 380, 381, 382
Deitz, D. E. D., 288, 357 291, 292, 387, 402, 446, 463, 468, Dracobly, J. D., 194, 227
Deitz, S. M., 453 469, 470, 471, 547, 548 Draper, R., 344
Delano, M. E., 264 Dillon, C. M., 277 Drapkin, M., 505
Delany, D. D., 382 DiLorenzo, T. M., 356 Drasgow, E., 277
Delbrück, N., 98 Dinnebeil, L. A., 465 Dresang, L. T., 477
DeLeon, I. G., 14, 43, 104, 107, 109, Dinsmoor, J. A., 18, 22, 79, 81 Driskell, J. E., 263
115, 118, 120, 121, 122, 123, 125, DiPipi-Hoy, C., 259 Drossel, C., 523
140, 160, 177, 178, 181, 182, 183, Diquelou, M. C., 427 Droungas, A., 26
184, 185, 194, 195, 204, 209, 216, DiScala, C., 477 Drummond, M. B., 500
223, 228, 240, 241, 278, 304, 305, Dissanayake, C., 390 Du, L., 277
306, 308, 309, 310, 314, 340, 343, Disterhoft, J. F., 514 Dube, W. V., 54, 73, 83, 87, 102, 109,
344, 353, 354, 374, 376, 378, 383, Ditzian, K., 294, 547 203, 204, 207, 225, 329
398, 519 Dixon, L. S., 84 Dubey, D. R., 465
Delia, M. D., 304, 345 Dixon, M. J., 364 Duchaine, E. L., 470
Delmendo, X., 122 Dixon, M. R., 72, 126, 139, 399, 521, Duffy, G. G., 261
DeLuca, R. B., 61, 62 522 Dugas, M., 517
Demchak, M. A., 256, 286 Djurhuus, J. C., 415 Duggan, A. K., 478
DeMeyer, A. M., 430 Dobson, R. J., 447 Dugosh, K. L., 499
DeMeyer, M. K., 390 Doerksen-Klopp, B., 471 Duhon, G. J., 256, 468, 471
Dempsey, C. M., 137 Doerner, M., 289 Duker, P. C., 195, 196, 228, 277, 365,
Dempsey, J. R., 436, 437 Doggett, R. A., 209, 375 366
Demurie, E., 125 Doke, L. A., 365 Dukes, C., 476
Deno, S. L., 456 Doll, B., 142 Dulcan, M. K., 408, 409
DePasquale, J. P., 551 Doll, J., 546 Duncan, B. A., 223
DePhilippis, D., 505 Dominguez, A., 353, 363 Duncan, R., 194
DeRaad, A., 536 Domire, S. C., 257 Dunham, P. J., 366
Derby, K. M., 187, 222, 229, 308, 323 Domjan, M., 104 Dunlap, G., 182, 244, 302, 314, 343,
Derenne, A., 241, 513, 523 Donahoe, J. W., 4, 13, 15, 16, 17, 18, 366, 398
DeRisi, W., 142 19, 20, 21, 23, 24, 25, 27, 28, 30, 31, Dunn, E. K., 65, 468
Derkay, C. S., 428 32, 40, 51 Dunn, J., 144
DeRosa, N. M., 11, 133, 155, 159, 304, Donaldson, J. M., 82, 177, 247, 260, Dunn, K. E., 494
339, 340, 345, 363 356, 378, 382, 383, 402, 512 Dunn, T., 366
DeShaw, J. M., 179 Donelson, F., 446 DuPaul, G. J., 72, 254, 313, 380, 464
Desmond, J. E., 514 Doney, J. K., 291, 435, 468 DuPont, R. L., 507
DeSouza, A. A., 88, 274, 344 Dong, X. Q., 567 Dupree, L., 513
Detrick, J., 108, 468, 537, 538 Donley, C. R., 256 Dupuis, D. L., 288, 356, 446
Detweiler, D. D., 390 Donlin, W. D., 504 Durand, V. M., 106, 107, 194, 195, 196,
Devany, J., 325 Donn, L. K., 222 199, 276, 330, 337, 348
Devine, B., 143 Donnelly, D. R., 240 Durgin, A., 547, 548
Devine, S. L., 144, 147 Donovan, M., 477 Dutcher, D. L., 83
DeVries, H. M., 517 Dool, E. J., 449, 457 Dutra, L., 491
Diament, C., 139, 463 Dorey, N. R., 229 Dutta, A., 548
DiCarlo, C. F., 256 Dorsey, M. F., 9, 95, 194, 198, 204, 215, Dworkin, B. R., 25
Dichter, G. S., 30 229, 312, 321, 322, 325, 337, 338, Dworkin, S. I., 29
Dick, L. P., 517 342, 353, 358, 364, 391, 534 Dwyer-Moore, K. J., 521
Dickerson, D. S., 83 Doucette, M. L., 477 Dyer, C. B., 567
Dickerson, F. B., 374 Dougher, M. J., 87, 361 Dyer, E. J., 120, 182
Dickes, N., 41, 100, 226, 400 Dougherty, D. M., 497, 498 Dyer, K., 182, 244, 292, 302, 314
Dickinson, A., 25 Douglas, J. E., 430
Dickinson, A. M., 544, 545, 548, 549 Dowdy, A., 323
Dickson, C. A., 83 Downing, C. O., 548 E
Dickson, M. J., 446 Downing, R. H., 374
Didden, R., 205, 228, 229 Downs, A., 290 Eack, S. M., 279
Dieckhaus, K., 498 Downs, R. C., 290 Earls, F., 409
Author Index 585
Flessner, C., 480, 482, 483, 484 Fritz, J. N., 52, 136, 137, 222, 225, 226, Gaylord-Ross, R. J., 343,357
Flood, W. A., 66 227, 229, 314, 332, 338 Gazdag, G. E., 277
Florentino, S., 486 Frost, L., 271, 273, 275, 276, 278 Geier, L., 28
Floress, M. T., 468 Fruzetti, A. E., 411 Gelfand, S., 288
Flory, J. D., 125 Fry, A. F., 124 Geller, E. S., 382, 548, 549, 550, 551,
Fombonne, E., 390 Fryling, M. J., 194, 313, 402, 433 554
Fong, G. W., 30 Fuchs, D., 445, 446, 456 Gendreau, P., 382
Foody, C., 352 Fuchs, L. S., 445, 446, 456 George, T., 179
Ford, J. E., 288 Fucile, S., 433 Gerada, C., 491
Ford, W. B., 180 Fuhrman, A. M., 79, 136, 246, 311, Gerber, M., 395
Forehand, R., 327, 363 330, 347 Gerencser, K. R., 247
Foreman, A. P., 366 Fullen, J. A., 5 Gersten, R., 447, 453
Foreman-Yates, K., 451 Fullerton, J. A., 108 Ghaemmaghami, M., 202, 227, 241,
Forman, D., 209 Fulton, R. T., 80, 89 246
Forman, G., 482 Fuqua, R. W., 115, 302, 303, 411, 522 Ghezzi, P. M., 72, 435
Forman, S. G., 463 Furniss, F., 194 Ghio, L., 532
Forsyth, B., 408 Fusilier, I., 177 Giacin, T. A., 206, 312
Foshee, T. J., 293 Futterman, A., 517 Giacobbe-Grieco, T., 314
Foster, N., 417 Fyffe, C. E., 323, 331, 345 Giacomelli, A. M., 123
Foulkes, C., 147 Gianoumis, S., 394
Fowler, S. A., 239 Gibbons, B., 439
Fox, B., 179 G Gibbons, K. A., 455, 456, 457
Fox, D. K., 382, 549, 551 Gibbs, C. M., 22, 27
Fox, J. J., 195 Gabel, S., 412 Gibson, A. B., 180, 354
Fox, R. A., 179 Gadaire, D. M., 173, 193, 207 Gibson, J. L., 534, 535
Foxx, R. M., 302, 353, 354, 356, 357, Gaffaney, T., 205 Gilbert, T. F., 552
361, 365, 366, 390 Galbicka, G., 245, 256, 500 Gilbertson, D., 468
Frades, B., 183, 512 Galensky, T. L., 331, 342 Giles, A. F., 180, 354, 362, 468
Fraley, L. E., 3 Galiatsatos, G. T., 207 Gillat, A., 293
Frampton, S. E., 88 Galizio, M., 84, 85, 87, 341 Gilles, C. N. R., 465
France, K. G., 326, 328 Gallagher-Thompson, D., 516, 517 Gillett, J. N., 394, 395
Francis, L. L., 226 Gallegos, L., 5 Gilligan, K., 435
Francisco, M. T., 115, 118, 122, 182 Gallup, G. G., 328 Gillis, J. M., 66
Frank-Crawford, M. A., 118, 123, 209 Gansle, K. A., 142, 255, 378, 402 Gilroy, S. P., 117, 127
Franke, T. M., 477 Garcia, D., 476 Gingiss, P. M., 467
Franklin, A. N., 120 Garcia, E., 22 Girardeau, F. L., 83
Franklin, L., 439 Garcia, E. E., 275 Girolami, P. A., 324, 432, 436, 439
Frantz, S. E., 355 Garcia, J., 26 Gisel, E., 433
Franzese, J. C., 222 Garcia, S., 519 Gitlin, L. N., 523
Frazer, T. J., 366 Garcia-Albea, E., 256 Gladstone, B. W., 287
Frederiksen, L. W., 327, 552 Garcia-Rodriguez, O., 494 Glasgow, R. E., 492
Fredrick, L. D., 338, 453, 455, 470 Gardella, L., 435 Glass, R., 433
Fredrikson, M., 26 Gardenier, N. C., 73, 102, 205 Glasscock, S. G., 420
Freed, D. E., 119 Gardiner, K., 366 Gleason, D. L., 435
Freed, T., 140 Gardner, J. M., 285, 286, 288 Gleeson, S., 240
Freedman, D. S., 438 Gardner, K. B., 447 Glenn, I. M., 491, 494
Freeland, J. T., 255, 402 Gardner, S. M., 293 Glenn, S. S., 552
Freeman, K. A., 411, 422, 435 Garfinkle, A. N., 279 Glodowski, K. R., 125, 203
Freeston, M. H., 517 Garner, M., 534 Glover, A. C., 120, 155, 182, 291
Freitag, G., 214 Garon, N., 402 Glover, M., 494
French, J. R. P., 466 Garro, J., 468 Goddard, K. M., 381
Frese, J., 430, 436 Gassaway, L. J., 259 Godley, M. D., 499
Fridley, D., 256 Gast, D. L., 82, 86, 87, 242, 245, 253, Goel, N., 467
Frieder, J. E., 534, 535 256, 258, 259, 400, 401 Goetz, E. M., 246, 260
Friman, P. C., 313, 375, 377, 387, 408, Gatchalian, K. M., 126 Goff, G. A., 123, 185
409, 410, 411, 412, 413, 414, 415, Gatheridge, B. J., 480, 481, 482, 483, Goh, H. L., 121, 182, 216, 304, 306,
416, 417, 418, 420, 422, 450 484, 486 309, 310, 326, 343, 345
Frittro, E., 345 Gatti, S. L., 203, 208 Gold, M. S., 507
Fritts, K. M., 471 Gatz, M., 517 Gold, V. J., 214
Author Index 587
Goldberg, I. D., 408 Greene, D. J., 453 Hagopian, L. P., 43, 107, 138, 160, 178,
Goldberg, M. C., 118, 119 Green-Paden, L. D., 374 180, 181, 184, 204, 206, 208, 218,
Goldiamond, I., 255 Greenway, D. E., 87 219, 223, 224, 240, 243, 246, 306,
Goldman, K. J., 115 Greer, B. D., 45, 79, 108, 115, 136, 310, 311, 324, 331, 341, 342, 344,
Goldsmith, T. R., 252 228, 229, 246, 276, 311, 312, 330, 346, 347, 352, 353, 354, 356, 358,
Goldstein, H., 88, 138, 273 332, 337, 344, 345, 347, 352, 356, 359, 360, 362, 364, 366, 376, 391,
Goldwater, B., 255 391, 536 435, 519, 520
Golkar, A., 26 Greer, R. D., 273, 277, 279 Hahn, E. C., 430
Gomes, L. A., 366 Gregerson, L., 468 Hake, D. F., 42, 327, 363, 364, 366
Gomes, R., 532 Gregory, M. K., 118 Hale, L. M., 465
Gongola, L., 292 Gresham, F. M., 142, 378, 402, 456, Haley, R., 63
Good, W., 195 457, 469 Halfon, N., 477
Goodall, E., 144, 146, 147 Griffin, A. K., 259 Hall, G., 276
Goodall, G., 362 Griffin, A. S., 427 Hall, H. R., 416
Goodkin, K., 260 Griffin, C. C., 452 Hall, J. E., 571
Gordon, B. D., 356 Griffin, J. C., 242 Hall, L. J., 470
Gormezano, I., 17, 22, 27 Griffin, M., 545 Hall, R. V., 455, 547
Gormley, M. J., 464 Griffin, P., 253, 450 Hall, R. W., 531
Gortmaker, V. J., 254, 451, 452 Grimes, J. A., 104, 109 Hall, S. S., 203, 209, 256, 342
Gotjen, D. L., 180, 216, 244, 306, 344 Grindle, A. C., 549 Hall, W. G., 309
Gotleib, E. M., 531 Gritter, A. K., 59 Hallam, S. C., 28
Goto, A., 431 Groden, G., 206 Halldorsdottir, L. Y., 435
Goto, M., 431 Groen, A., 392 Halle, J. W., 82, 245, 276, 277
Gottlieb, H., 382 Groff, R. A., 1, 3, 429, 435, 436, 437 Halliday, S., 416
Gottman, J. M., 208 Gross, A. C., 388, 476, 486, 522 Halpern, S. D., 503, 505, 506
Goudie, A., 127 Gross, A. M., 357 Hamler, K. R., 449
Goyal, M., 476 Gross, T., 471 Hammer, D. G., 30
Goza, A. B., 365 Grossen, B., 261, 262, 264 Hammond, J. L., 52, 136, 203, 222,
Grabinski, M. J., 494, 502 Grossman, D. C., 477, 479 227, 338
Grabman, J. H., 415 Groves, E. A., 226, 247 Hammond Natof, T., 66
Grabowski, J. G., 288 Grow, L. L., 8, 30, 84, 120, 182, 276, Hamo, M., 206
Grace, N. C., 140, 141, 362 326, 339, 387, 389 Handen, B. J., 82
Grace, R. C., 59, 73, 74, 100, 101, 103, Gruber, D. J., 258 Handleman, J. S., 302
105, 109, 126 Gudleski, G. D., 124 Handler, M. W., 463
Graff, R. B., 177, 181, 183, 207, 288, Guenther, S. L., 136 Handwerk, M. L., 377, 415
291, 398 Guevremont, D., 413 Haney, J. I., 482
Graham, S., 261 Gulotta, C. S., 307, 324, 434, 435, Hanley, G. P., 54, 65, 66, 67, 68, 74,
Grahame, N. J., 28 436, 437 79, 105, 138, 147, 160, 178, 179, 180,
Granmo, S., 522 Gupta, A., 433 183, 185, 186, 187, 193, 194, 199,
Grant, D. A., 26 Gureasko-Moore, S., 254, 264 202, 204, 218, 222, 226, 227, 235,
Gratz, O. H., 122 Gutierrez, M., 531 237, 238, 240, 241, 242, 243, 244,
Grauerholz-Fisher, E., 139 Gutowski, S. J., 142 246, 247, 304, 306, 307, 308, 309,
Graupner, T. D., 392 Gutshall, K., 306 311, 330, 331, 338, 339, 342, 344,
Graves, C. A., 513, 514, 515 Guttman, N., 366 345, 346, 353, 362, 364, 394, 398,
Gravina, N. E., 286, 388, 544, 545, Gynther, M. D., 28 434, 464, 486, 516, 523, 536
546, 547 Hanna, E. S., 84
Gray, K. M., 501 Hannigan, K. F., 246
Gray, T. W., 547 H Hanratty, L. A., 227, 482, 486
Green, C., 408 Hansen, B., 88, 276
Green, C. E., 502 Haberlin, A. T., 122, 547 Hansen, B. D., 143
Green, C. W., 40, 179, 285, 286, 289, Hackbert, L., 324, 434, 436 Hansen, C. L., 446
290, 293, 354 Hackenberg, T. D., 25, 65, 125, 240, Hansen, S. G., 352
Green, G., 84, 85, 86, 87, 199, 205, 244, 374 Hantula, D. A., 63, 69, 276, 286
271, 392, 397, 571 Hacohen, G., 206 Harbers, H. M., 277
Green, L., 71, 72, 119, 124, 125 Hadley, A. J., 430 Harchik, A. E., 288, 290, 292
Green, S. B., 146, 147 Hadley, E. C., 518 Hardin, J. M., 521
Green, V. A., 392 Hagaman, J. L., 254 Harding, J. W., 61, 204, 207, 246, 253,
Greene, B. F., 287 Haggar, J. L., 25, 118, 374 530
Greene, D., 52, 186 Hagge, M., 549 Hardy, M. S., 477, 478, 479, 481, 482
588 Author Index
Haremza, J. L., 288 Heflin, L. J., 470 Hineline, P. N., 49, 323
Haring, N. G., 446 Hegedus, A. M., 412 Hirsch, S. E., 247
Harker, S., 437 Heiberg, A., 430 Hirst, J. M., 402
Harkins, S. W., 518 Heidorn, S. D., 323 Hixson, M., 276
Harley, C. M., 431 Heil, S. H., 421, 491, 494 Ho, E. E., 517
Harley, D. A., 259 Heineke, M., 523 Hoagwood, K. E., 463
Harlow, H. F., 78, 81 Heisenberg, W., 420 Hobbs, S. A., 363
Harniss, M., 262 Heitzman-Powell, L. S., 534, 535 Hoch, H., 136, 216, 256, 313
Harper, J. M., 216, 218 Heldt, J., 261 Hoch, T. A., 324, 434, 436, 439
Harrell, R. T., 356, 379 Helsel, W. J., 289, 364 Hodos, W., 118, 181
Harris, A. C., 122 Hemenway, D., 477 Hoerger, M. L., 67,68, 72, 105, 107
Harris, B., 67, 107 Hemingway, M., 572 Hoffman, H. S., 63, 366
Harris, C., 229 Henderson, D., 208 Hoffman, K., 341
Harris, F. R., 142, 206 Hendricks, C. M., 477 Hoffmann, A. N., 118, 184
Harris, J., 312 Henley, A. J., 127, 548 Hogan, A., 287, 446
Harris, K. R., 261 Henrickson, M. L., 144 Holburn, C. S., 361
Harris, M. B., 477, 478 Henrot, A., 433 Holburn, S. W., 61, 62
Harris, S. L., 40, 302 Henry, L. M., 522 Holding, D. H., 260
Harrison, C. D., 180 Herbert-Jackson, E., 356 Hollahan, M. S., 143, 313
Harrison, R. H., 78 Hergenrader, M., 452 Holland, J. G., 274
Harrop, A., 144, 146, 147 Herman, B. C., 471 Holland, K. E., 532
Hart, B., 206, 393 Herman, K. C., 467 Holland, P. C., 28, 32
Hart, J. M., 446 Herman, K. M., 122 Hollander, D. L., 278
Hart, K. J., 413 Hernandez, E., 74, 240, 244, 398 Hollander, M., 285
Hart, S., 438 Heron, T. E., 8, 136, 164, 175, 205, 253, Hollenbeck, K., 262
Hartel, F., 365 326, 357 Hollis, J. F., 492
Hartman, B., 448 Herring, M., 203 Hollis, J. H., 86
Hartmann, D. P., 135, 137, 141, 142, Herrnstein, R. J., 5, 49, 59, 70, 96, 97, Hollobaugh, T., 276
143 98, 99, 101, 103, 123, 330, 341 Holman, E., 517
Hartmann-Boyce, J., 493 Herscovitch, B., 208 Holmberg, M., 147
Hartsfield, F., 572 Hersen, M., 26, 140, 142 Holmes, N. M., 19
Harvey, M. T., 446 Hessler, T., 261 Holmes, P., 307
Harvey, T., 522 Hetrick, W. P., 312, 324 Holt, D. D., 124, 125
Hatch, T. F., 412 Heward, W. L., 8, 136, 164, 175, 205, Holth, P., 37
Hatti, M., 415 253, 259, 326, 357, 446 Holtyn, A. F., 388, 490, 499, 507
Hatton, G. I., 328 Heyman, G. M., 100 Holtz, R., 366
Haught, K., 477 Hickey, D. T., 253 Holz, W. C., 44, 46, 358, 363, 364,
Hausherr, L., 289 Hickman, C., 324 378, 382
Hawkins, A. M., 289 Hickman, J. S., 551 Homlitas, C., 292
Hawkins, L., 446 Hicks, S. C., 263 Honaker, S. M., 411
Hawkins, R. P., 378 Hier, B. O., 449 Honig, W. K., 366
Hawkins, S. M., 470 Higbee, T. S., 180, 181, 185, 247, 257, Honsberger, T., 256
Haydon, T., 263 260, 344, 398 Hook, J., 277
Hayes, S. C., 5, 85, 88, 383, 421, 422 Higgins, J. W., 375 Hooper, S., 259
Hay-Hansson, A. W., 534, 535 Higgins, S. T., 126, 382, 421, 490, 491, Hopkins, B. L., 144, 288, 382, 549
Haymes, L. K., 242 494, 500, 507 Hopko, D. R., 517
Hays, R., 433 Higgins, W. J., 291, 292 Hopper, J. S., 534
Haywood, H. C., 252 Hightower, A. D., 179 Horn, K., 124
He, A. B., 26 Hilgard, E. R., 16 Horn, S., 438
Headley, A., 287 Hilker, K., 187, 308 Horne, P. J., 85, 88, 277, 522
Heal, N. A., 105, 180, 331 Hill, A. P., 390 Horner, R. D., 64, 255, 303, 304, 344,
Healy, O., 352 Hill, B. P., 439 401
Healy, S., 247 Hill, W., 366 Horner, R. H., 159, 160, 161, 197, 198,
Heard, K., 62, 521 Hillhouse, M., 503 240, 264, 314, 337, 339, 340, 343,
Heathfield, L. T., 240, 343 Hillman, C., 446 392, 456, 468
Heckaman, K. A., 259 Hilton, L. C., 512, 521 Horner, V., 285
Hedges, L. V., 161 Himle, M. B., 477, 479, 480, 481, 482, Horowitz, R., 567
Heering, P. W., 546 483, 484, 486 Horwitz, S. M., 408
Heffner, V. A., 532 Hine, R. A., 202 Hosp, J. L., 194
Author Index 589
Kowalewicz, E. A., 375, 380, 382 Lamb, R. J., 245, 500, 501 Leaf, P. J., 408
Kozak, C., 482 Lambert, J. M., 226 Leaf, R., 274, 400
Kozisek, J. M., 436 Lambert, J. V., 49 LeBlanc, B. A., 400, 536
Kozloff, M. A., 262, 263 Lancioni, G., 277 LeBlanc, J. M., 238, 256, 257, 323
Kraaykamp, M., 277 Landa, R., 79, 202 LeBlanc, L. A., 82, 252, 277, 278, 292,
Kramme, K. W., 366 Landa, R. K., 276 303, 306, 307, 311, 341, 353, 356,
Krantz, P. J., 256, 260, 275, 393, 547 Landon, J., 101 388, 390, 392, 393, 394, 395, 447,
Kranzler, H. R., 491 Lane, J. D., 253 512, 519, 521, 522, 523
Krapfl, J. E., 285 Lane, K. L., 166 LeBlanc, M., 375
Kratochwill, T. R., 160, 161, 162, 464, Lane, L. L., 62 Lechago, S. A., 276, 277
465, 466, 468, 469, 470 Lane, S. J., 411 Ledbetter-Cho, K., 253, 264
Kratzer, D. A., 276 Lang, R., 157, 158, 228, 229, 534 Ledgerwood, D. M., 500
Kraus, J. F., 477 Langdon, N. A., 359, 360, 364 Lee, D. L., 313
Krause, T. R., 549, 550 Langer, S. N., 205 Lee, F. C., 517
Krell, D. M., 324, 434, 436 Langford, A., 289 Lee, G., 540
Krentz, H., 244 Langkamp, D., 531 Lee, H., 26
Kressig, R. W., 567 Langone, J., 257 Lee, J., 530, 548
Kretlow, A. G., 471 Langston, J. L., 100 Lee, J. F., 246, 533
Kretschmer, D., 549 Lan-Hsiang, R. L., 449 Lee, K., 380, 548
Krishnan-Sarin, S., 503 Lannie, A. L., 138, 450 Lee, M. S. H., 181
Krivacek, D., 361 Lanovaz, M. J., 339 Lee, R., 184, 246, 260, 261
Krival, K. R., 430 Lantz, S., 206 Lee, S. C., 550
Kroesbergen, E. H., 253 LaNuiziata, L., 262 Legacy, S. M., 306, 341
Kronfli, F. R., 332 Lapidus, G. D., 477 Leger, E., 517
Krueger, K., 146 Laraway, S., 58, 215, 243, 301, 308, Legierski, C. M., 123
Krug, A., 240 338, 381, 524 Lehardy, R. K., 229
Kryger, M., 419 Larrabee, D., 521, 522 Leirvick, S. E. S., 522
Kuhn, B. R., 411, 415, 422 Larsen, J., 181 Leitenberg, H., 106, 107, 358, 363
Kuhn, D. E., 225, 228, 311, 345 Larsen, K. M., 331 Lemanek, K. L., 411, 418
Kuhns, J. B., 382 Larsen, S., 430 Lennox, D. B., 199, 240, 361
Kulp, S., 144, 146 Larson, P. J., 468 Lentz, F. E., 446, 457
Kunnavatana, S. S., 226, 468 Larson, T. A., 229 Lentz, R. J., 376, 379, 383
Kuroda, T., 102 LaRue, R. H., Jr., 173, 193, 207, 274, Leon, Y., 125, 240, 374
Kurt, O., 259 306, 329, 342, 397, 428, 435, 439 Leontovich, M., 262
Kurtz, P. F., 42, 216, 228, 242, 306, 344, Laski, K. E., 394, 398 LePage, J. P., 380
353, 354, 357, 359, 360, 361 Latham, S. B., 27 Lepper, M. R., 52, 186
Kymissis, E., 277 Laties, V. G., 48 Lepper, T. L., 23, 143, 186, 278
Kyse, E. N., 468 Latkowski, M. E., 374 Lerer, T. J., 477
Lattal, A. D., 341 Lerman, D. C., 42, 64, 65, 105, 106,
Lattal, K. A., 37, 43, 107, 118, 240, 109, 118, 119, 136,139, 161, 181, 182,
L 308, 341 185, 203, 204, 207, 210, 216, 223,
Lattal, K. M., 104 224, 229, 241, 242, 256, 285, 288,
La Londe, K. B., 80 Lattimore, L. P., 288 292, 305, 309, 311, 312, 325, 326,
Lachowicz, J., 374 Lau, C., 433 327, 328, 329, 331, 338, 339, 343,
Lachs, M. S., 567 Laud, R. B., 209 347, 352, 354, 356, 362, 363, 364,
Lackgren, G., 415 Lauderdale-Littin, S., 374 366, 391, 397, 400, 401, 446, 468,
Lacy, B. E., 428 Laudont, C. L., 183, 244 536, 540
Ladouceur, R., 517 Lauer, E. A., 463 LeSage, D. L., 229
LaFleur, K., 449 Laugesen, M., 126 LeSage, M. G., 122, 135
LaFleur, T., 554 Lavie, T., 287, 291 Leslie, J., 259
LaFrance, D. L., 271, 274 Lawyer, S. R., 125 Lesser, A. D., 203, 431
Lagorio, C. H., 124 Layer, S. A., 180, 183, 185, 307, 324, Lester, A. C., 125
Lahey, B. B., 376 394, 434, 436, 437 Levensky, E. R., 421
Lake, C. M., 160 Layng, T. V. J., 278, 448, 449 Leventhal, B. L., 271
Lake, S. L., 498 Lazarchick, W., 314 Leventhal, J. M., 408
Lalli, J. S., 52, 53, 61, 64, 65, 203, 206, Le, D. D., 325, 354 LeversLandis, C. E., 418
209, 229, 240, 242, 246, 306, 330, Lea, S. E. G., 119 Levesque, M. A., 277
332, 337, 338, 343, 354 Leadbetter, S., 477 Levin, D. S., 436
LaLonde, K. B., 278 Leader, G., 117 Levin, J. R., 261, 264
LaMarre, J., 274 Leaf, J. B., 82, 256 Levin, L., 279
592 Author Index
Levine, M. D., 412, 413 Loewy, S., 548 Lyketsos, C. G., 517, 523
Levy, A., 186 Logan, P., 468 Lyon, D. O., 60
Levy, R., 287 Logsdon, R. G., 519 Lyon, V. B., 532
Levy, S., 391 Løkke, J. A., 522
Lewallen, J. D., 288 LoLordo, V. M., 26
Lewin, A. B., 178, 352 Loman, S. L., 468 M
Lewis, A. V., 413 Lomas, J. E., 11, 160, 198, 338, 343
Lewis, C., 471 Long, A. C. J., 463, 467, 468, 469, 470 Maag, J. W., 468
Lewis, D. C., 492 Long, E. S., 107, 138, 139, 178, 204, Maas, A. P. H. M., 205
Lewis, D. J., 328, 329 205, 208, 209, 246, 344, 356, 482, Maas, C. J. M., 253
Lewis, M., 327 519 MacDonald, J. E., 546
Lewis, M. I., 516 Long, J., 396 MacDonald, J. M., 329
Lewis, M. W., 498 Long, R. D., 548 MacDonald, R., 205
Lewis, T. J., 194, 195 Looman, W. S., 531 MacDonald, R. B., 302
Leyland, C. M., 26 Lopez-Nunez, C., 494 MacDonald, R. F., 205
Li, X., 105 Lorah, E. R., 275, 276 Mace, F. C., 6, 13, 52, 58, 59, 61, 64,
Li, Y., 514 Lord, C., 271, 390 65, 67, 68, 71, 72, 73, 74, 101, 102,
Libby, M. E., 183, 208, 258 Lorenz, R. A., 410, 418 103, 104, 105, 106, 107, 108, 109,
Liberman, R., 142 Lorenzi, P., 547 141, 184, 198, 203, 209, 228, 229,
Liberman, R. P., 287, 290 Lorincz, E. N., 31 241, 243, 306, 313, 356, 410, 417
Liem, G. D., 253 Lotlikar, H., 549 Mach, E., 115
Lienemann, T. O., 254 Loughrey, T., 548 Machalicek, W., 352, 464, 534, 535
Lieving, G. A., 107 Lovaas, E. E., 275 Macht, J., 81
Lillenstein, L., 457 Lovaas, N. W., 275 MacIntyre, A., 563
Lin, H., 126 Lovaas, O. I., 31, 214, 271, 274, 287, Mackay, H. A., 54, 86, 87
Lindauer, S. E., 136, 225, 229, 353 322, 324, 357, 364, 365, 392, 396, MacKenzie-Keating, S. E., 353, 354
Lindberg, J. S., 138, 178, 185, 186, 400, 401 MacKillop, J., 126, 127
242, 352 Love, J. R., 260, 276, 398, 400 Mackintosh, N. J., 26, 109, 329
Lindblad, T. L., 366 Love, S. R., 256 MacLean, W, E., Jr., 302
Lindenberg, A. M., 144 Lovelace, B., 229 MacNeill, M., 388, 512
Lindgren, S. D., 388, 530, 534, 536, Loveland, D. H., 70 MacQuarrie, L. L., 448
537, 538 Loven, M. D., 142 Macrae, M., 18
Lindsay, J. A., 502 Lovitt, T. C., 446, 450 Macropoulis, G., 63
Lindsley, O. R., 446, 458, 512 Lowe, C. F., 85, 88, 277 Macurik, K. M., 236, 285, 292
Lindström, B., 26 Lowery, S., 413 Macvean, M. L., 470
Ling, W., 503 Lowry, M. W., 84 Madden, G. J., 63, 115, 124, 125, 126,
Linnerooth, P. J. N., 420 Lozoff, B., 409 127, 184
Linscheid, T. R., 242, 243, 356, 365, Lubetsky, M. J., 364 Maffei, J., 273
366, 433 Lucas, G. A., 21 Magee, S. K., 61, 64, 359
Lionello-DeNolf, K. M., 109 Luce, S. C., 271, 292, 397 Maggin, D. M., 381, 382, 470
Lipinski, D. P., 357 Lucena, S. L., 433 Magito McLaughlin, D., 314
Lipner, C., 357 Luck, K. M., 288, 292, 446 Maglieri, K. A., 140, 180, 238, 310,
Listwan, S. J., 382 Lucker, K. D., 169 331, 353, 356, 362, 364, 366, 394
Litow, L., 247 Luczynski, K. C., 180, 238, 246, 463 Maguire, H., 291, 402, 468
Litt, M. D., 496 Ludwig, C., 285 Mahmoud, E., 378
Liu, S. S., 28 Ludwig, T. D., 382, 546, 547, 548, 549, Mahnke, C. B., 532
Liu-Constant, B., 291, 468 551 Mahoney, A., 218, 547
Livanis, A., 468 Luiselli, J. K., 65, 208, 229, 253, 435, Mahoney, M. J., 44
Livesey, J., 365, 546 463 Mahy, C. V., 254
Livezey, K., 52, 64, 354 Lukens, C. T., 430 Maier, S. F., 38
Lloveras, L. A., 332 Lumley, V. A., 482, 483 Majdalany, L., 290, 546
Lloyd, B. P., 125, 193, 208 Lund, P. A., 522 Makin, P. J., 551
Lloyd, J. W., 247 Lunt, I., 571 Makrygianni, M. K., 392
Lochbaum, M. R., 302, 303 Luria, G., 551 Malanga, P., 292
Locher, J., 518 Lussier, J. P., 491 Malley, J. L., 59
Lochner, D. G., 61, 63 Lustig, N., 468, 537 Malone, F., 531
Locke, B. J., 365 Lutrey, S. M., 546 Malott, M. E., 552
Locke, P. A., 177 Lutz, M. N., 198 Malott, R. W., 89, 254, 552
Lockshin, S. B., 66 Lutzker, J. R., 247 Mancil, G., 276
Loeb, J., 292 Lydon, S., 352, 359 Mandell, C., 60
Author Index 593
Mang, K., 479 Matheny, A. P., 414 McGhan, A. C., 242, 400
Manning, P. J., 363 Matherne, P., 363 McGimsey, J. F., 41, 302, 304
Manoli, C., 86 Mathews, J. R., 378, 412 McGinnis, J. C., 375, 415, 450
Mansell, J., 147 Mathews, R. M., 512, 516, 519 McGinty, E. E., 477
Manset-Williamson, G., 261 Mathisen, D., 290, 546 McGlynn, A. P., 365
Manuck, S. B., 125 Matson, J. L., 141, 194, 196, 256, 356, McGoey, K. E., 380
Maple, T. L., 229 362, 391 McGrath, A. M., 430
Marakovic, N. N., 124 Matter, A. L., 260 McHaffie, E., 546
Marano, C. M., 517 Matthews, G., 549 McHenry, M. M., 160
Maras, M. A., 467 Mattos, M., 456 McHenry, W. C., 361
Marchand-Martella, N., 482 Mattson, T., 451 McIlvane, W. J., 54, 73, 83, 84, 86, 87
Marchant, N. J., 105 Mauk, J. E., 228 McIntyre, L. L., 352, 446, 468
Marchese, D., 471 Maupin, A. N., 264, 289, 471 McIntyre, T., 312, 343
Marchese, N. V., 82 Maurice, C., 271, 274, 397, 401 McKay, J. R., 503, 505
Marcin, J. P., 532 Mauro, B. C., 71, 74, 102, 184, 198, McKee, J., 179
Marckel, J., 72 243 McKenzie, S. D., 311, 364, 366
Marcus, B. A., 65, 121, 177, 187, 198, Mauzy, C. R., 323, 432 McKerchar, P. M., 179, 203, 244, 288,
223, 303, 307, 309, 310, 314, 332, Maxfield, T. C., 476 331
337, 342, 520 May, M. E., 303 McKerchar, T. L., 353
Marder, T. J., 228, 433, 434 Mayer, G., 225 McLane, M., 519
Marechal, J. M., 412 Mayer, K. L., 210 McLaughlin, K. J., 482
Marholin, D., 365 Mayhew, G. L., 289 McLaughlin, R., 45
Marion, C., 446, 447 Maynes, N. P., 256 McLaughlin, T. F., 356, 375, 382
Marion, S. D., 209 Mays, M. N., 303 McLay, L. K., 31, 88, 275
Markham, M. R., 87 Mazaleski, J. L., 64, 176, 177, 224, 306, McLean, A. P., 101, 109
Markus, K. A., 468 312, 313, 322, 325, 327, 331, 339, 340 McLellan A. T., 492, 507
Marquis, J. G., 516 Mazur, J. E., 8, 58, 110, 123, 124, 181, McMahon, W. M., 374
Marr, M. J., 229 184, 198 McManus, M. D., 531
Marshall, A. M., 82, 276 Mazurek, M. O., 532 McMillan, K., 438
Marshall, G., 548 McAdam, D. B., 228, 260 McNabney, M. K., 512
Marsteller, T. M., 65, 106 McAuley, K., 32 McNamara, B., 471
Martel. L., 203 McCall, M., 263, 449 McNamara, J. R., 382
Martens, B. K., 61, 63, 71, 125, 138, McCarn, J. E., 290 McNeil, C. B., 375
203, 252, 263, 330, 377, 382, 383, McCarthy, D., 70, 95, 97, 98, 99, 100, McPherson, S., 504
387, 445, 446, 447, 448, 449, 450, 101 McRae-Clark, A. L., 496
451, 456, 457, 458, 466, 467, 468, McCarty, J., 463 McReynolds, W. T., 247
470 McClannahan, L. E., 256, 260, 275, McStay, R. L., 390
Martin, A. J., 253 393, 521, 547 McSween, T. E., 353, 545, 549, 550,
Martin, A. L., 120, 229 McCloskey, M., 125 551
Martin, B., 491 McClure, E. A., 127 McSweeney, F. K., 301
Martin, B. L., 477 McComas, J. J., 136, 141, 203, 216, McWilliam, R. A., 288
Martin, C. A., 277 246, 260, 301, 313, 400, 446, 447 Meadow, S. R., 416
Martin, G. L., 181, 446 McConnell, O. L., 245 Medin, D. L., 84
Martin, N. T., 139, 572 McCord, B. E., 160, 204, 222, 331, 338, Mees, H., 50, 322
Martin, T. L., 181, 239 342, 523 Mehta, K. M., 517
Martindale, A., 144, 146 McCoy, J. F., 147 Meichenbaum, D., 517
Martindale, B., 146 McCuller, G., 63 Meier, A. E., 313, 433
Martinez, C. J., 324, 433, 436, 439 McCurdy, M., 254, 451, 452 Meindl, J. N., 375
Martinez-Onstott, B., 550 McDonald, L., 354 Mellichamp, F. H., 571, 573
Marx, J., 491 McDonell, M. G., 495, 498 Meltzer, L. J., 411
Marya, V., 276 McDowell, C., 258, 449 Mendres-Smith, A. E., 239
Masler, E. A., 203, 431 McDowell, J. J., 70, 98, 309, 330 Menza, T. W., 495
Mason, D., 393 McDowell, L., 354 Merckelbach, H., 26
Mason, S. A., 185 McDuffie, A., 535 Mercugliano, M., 344
Mason, W. A., 215, 303 McEachin, J. J., 274, 392, 400 Mercy, J. A., 477
Massie, D., 448 McElwee, J., 88 Meredith, S. E., 502, 503
Masuda, A., 161, 326 McFadden, E., 124 Merges, E., 276
Mataras, T. K., 83 McGee, G. G., 185, 393 Merlo, L. J., 507
Matese, M., 365 McGee, H., 545, 549, 550 Merrell, K. W., 141
Matey, J., 545 McGee, K. S., 480 Merritt, T. A., 522
594 Author Index
Mesmer, E. M., 449, 468 Mohan, M., 433 Murphy, M. A., 377, 379
Meszaros, L., 279 Mojtabai, R., 517 Murphy, M. J., 147
Methot, L. L., 135, 293 Molinari, V., 512 Murphy, P. K., 261
Meunier, P., 412 Monaghen, B. D., 445 Murphy, T. B., 433
Mevers, J. E. L., 160, 219, 338 Mongeon, J. A., 491 Murray, R. G., 363
Meyer, K. A., 198, 312 Mongrella, J., 356 Muskat, L. R., 278
Meyskens, F. L., 517 Moniz-Cook, E., 521 Musser, E. H., 376
Mezhoudi, N., 256 Montgomery, R. J. V., 523 Myers, D. M., 470
Michael, J. L., 23, 38, 58, 185, 186, Montgomery, R. W., 363 Myerson, J., 124
198, 215, 242, 243, 272, 273, 274, Moon, K., 547
275, 276, 301, 310, 324, 332, 338, Moore, A. L., 253
341, 343, 381, 396, 524 Moore, B., 504 N
Michalski, D. N., 144 Moore, B. A., 411
Mick, B. R., 183, 244 Moore, G. E., 563, 564 Nabeyama, B., 287
Midgley, B. D., 3 Moore, J., 5, 293 Najdowski, A. C., 5, 432, 435, 438, 439
Miguel, C. F., 88, 186, 235, 270, 271, Moore, J. W., 28, 209, 291, 292, 375, Namasivayam, A., 513
272, 275, 277, 278, 395, 396 400, 468 Napolitano, D. A., 260
Mikami, A. Y., 382 Moran, L., 352 Nash, H. M., 377
Milan, M. A., 363, 515, 517 Moreira, B. C. D. M., 127 Nau, P. A., 246, 353
Milby, J. B., 493 Moreira, M. E. L., 433 Naudé, G. P., 127
Miles, N. I., 288 Moreno-Sierra, V., 516 Navarro, A. D., 127
Millard, T., 253 Morgan, D. G., 288 Nay, R. N., 141
Millard, W. J., 20 Morgan, P. L., 453 Neef, N. A., 71, 72, 184, 198, 243, 246,
Miller, B. G., 229 Morizio, L. C., 208 285, 292, 342, 446
Miller, D. L., 140, 239, 433 Morley, A. J., 229 Nefdt, N., 395
Miller, J. R., 107 Morral, A. R., 245, 500 Nehs, R., 355
Miller, J. T., 100 Morren, J., 352 Neidert, P. L., 122, 226, 304, 343, 391
Miller, L. J., 436 Morris, E. K., 3, 327 Nelson, B., 182
Miller, M., 477 Morris, S. L., 183 Nelson, E., 531
Miller, M. S., 184, 198 Morrison, D. C., 206 Nelson, J. M., 261
Miller, N. E., 328 Morrison, H., 303 Nelson, J. R., 468
Miller, S. A., 65, 89, 276, 345 Morrison, R., 507 Nelson, R. O., 141, 420
Miller, S. J., 235, 246, 252 Morse, P. S., 353 Nemier, C., 264
Miller, T. M., 534 Morse, T. E., 259 Nepo, K., 279
Milnes, S. M., 436 Morse, W. H., 29, 103 Nesbitt, T. S., 532
Miltenberger, R. G., 139, 199, 205, 215, Mortweet, S. M., 409 Ness, M. E., 194
240, 244, 256, 257, 289, 307, 312, Moser, H., 390 Nettiksimmons, J., 531
321, 337, 356, 361, 374, 375, 378, Moses, L. J., 254 Neugebauer, S. R., 446
379, 382, 388, 445, 446, 476, 477, Moss, J., 204, 209 Neuringer, A., 8, 38, 260, 308
479, 480, 481, 482, 483, 484, 485, Mostofsky, S. H., 125 Neveus, T., 415
486, 521 Mousa, H., 430 Nevin, J. A., 42, 59, 60, 72, 73, 74, 94,
Mims, P. J., 263 Mousetis, L., 88 99, 100, 102, 103, 104, 105, 106, 107,
Mineka, S., 26 Mouzakitis, A., 468, 471 108, 109, 313, 329
Ming, S., 88 Mowrer, O. H.16 Newchok, D. K., 433
Minkin, N., 210 Mozingo, D. B., 287 Newcomer, L., 471
Minshew, N. J., 279 Mudford, O. C., 137, 138, 139, 146, 519 Newland, M. C., 88, 97, 100
Mintz, J. C., 346 Mudric, T., 492, 504 Newsom, C. D., 47, 214, 312, 314, 324
Mintz-Resudek, C. M., 291, 468 Mueller, E. T., 126 Newton, I., 72
Miranda-Dukoski, L., 102, 108 Mueller, M. M., 86, 8, 8361, 435, 438 Newton, J. T., 194
Mirenda, P., 275 Mulick, J. A., 106 Newton, S. J., 468
Mitchell, C., 453 Mullin, J. E., 546 Nguyen, K. V., 264
Mitchell, R. C., 453 Mundy, P., 275 Nicholas, L. S., 146
Mitchell, W. S., 31 Munk, D. D., 436 Nicholson, J., 194
Mitteer, D. R., 108, 228, 332, 347, 356 Murdoch, A., 457 Nicolson, A. C., 185
Miyashita, Y., 31 Muris, P., 26 Niemeier, J. J., 277
Mizunami, M., 20 Murphy, C., 126 Nieminen, G. S., 137
Modi, A. C., 418 Murphy, E. S., 301, 304, 308, 312 Nigam, M., 262
Moeller, F. G., 502 Murphy, G., 144, 146, 147 Niileksela, C. R., 119
Moes, D. R., 331, 345 Murphy, J. G., 126, 127, 497 Ninness, H. A. C., 63, 215, 306, 330
Moffatt, M. E., 415 Murphy, K. P., 567 Nipe, T., 323
Author Index 595
Nisbett, R. E., 52 O’Leary, S. G., 362, 375, 379, 465 Paden, A. R., 41, 100, 226, 400
Nissen, M. A., 338, 446 Oleson, C. R., 512, 520, 522, 523 Page, S., 260
Nitecki, L. A., 123 Olfson, M., 517 Page, T. J., 120, 136, 138, 139, 175, 198,
Nixon, J., 203 Olin, S. S., 463 229, 246, 288, 356, 359, 376
Nobrega, L. D., 433 Olinger, E., 137 Pakula, L. C., 478
Noda, M., 431 Oliver, A. C., 203, 210 Palinkas, L. A., 392
Noell, G. H., 142, 208, 209, 235, 246, Oliver, C., 203, 208, 209 Palmer, D. C., 16, 17, 18, 22, 23, 27, 28,
252, 253, 255, 256, 378, 379, 383, Oliver, J., 433 32, 51, 273, 396
402, 468, 469, 470, 471 Oliver, J. M., 390 Palmer, M. G., 547, 548
Nogales-González, C., 183, 512 Oliver, J. R., 285, 293 Palmer, S., 433, 438
Noguchi, D., 512, 513 Oliver, R. M., 468 Pampino, R. N., 546, 548
Noll, J., 515 Ollendick, T. H., 362, 482 Pancari, J., 206, 312
Nordström, K., 567 Olmi, J., 86 Pangborn, M. M., 438
Nordstrom, R., 547 Olness, K., 416 Panos, R., 140
Norgaard, J. P., 415 Olsen-Woods, L. A., 482 Pantalon, M. V., 504
Normand, M. P., 64, 203, 229, 274, Olson, R., 551 Paone, D., 216, 313
306, 307, 340, 433 Olson, S. C., 263, 446 Papka, M., 514
North, S. T., 138, 243 Olsson, A., 26 Paramore, N. W., 181
Northup, J., 141, 177, 179, 180, 216, Olweus, D., 467 Parikh, K., 476
226, 228, 375, 376, 378, 379, 534, O’Neill, D., 521 Parker, T., 287
536 O’Neill, R., 197 Parnell, A., 276
Norton, G. E., 496 O’Neill, S. J., 258 Parrish, J. M., 246, 342
Nosik, M. R., 559, 570, 574 Ong, P., 353 Parrot Hayes, L., 69
Notterman, J. M., 331 Ono, K., 308 Parrott, K. A., 259
Novak, M. D., 304 Oppenheimer, M., 82 Parry-Cruwys, D., 329
Novotny, M. A., 476 Oppenheim-Leaf, M. L., 256 Parsons, M. B., 179, 285, 286, 288, 289,
Nuernberger, J. E., 183 Oram, M. W., 31 290, 291, 292, 293, 354, 469
Nunn, R. G., 355 Ordog, G. J., 477 Partington, J. W., 23, 271, 275, 278,
Nuzzo, P. A., 502 O’Reilly, M. F., 198, 277, 312, 343, 396, 398, 400
534, 535 Passante, S. C., 290
Orlando, R., 241, 242 Patel, M. R., 307, 313, 324, 433, 434,
O Orme, S. F., 463 435, 436, 437, 439
Ortega, J. V., 183 Patel, S. J., 476
Oah, S., 547, 548 Ortman, J. M., 512 Pattij, T., 127
O’Brien, C. P., 492 Osborne, L. A., 402 Paul, C. M., 433
O’Brien, F., 242, 547 Osborne, S., 260 Paul, G. L., 376, 379, 383
O’Brien, M., 240, 343 Oscar-Berman, M., 100 Paveza, G. J., 567
O’Brien, R. M., 545 O’Steen, L. E., 243 Pavlik, V. N., 567
O’Brien, S., 356, 361, 420, 567 Ostrosky, M. M., 277 Pavlov, I. P., 15, 16, 32, 39
O’Callaghan, P. M., 307, 375, 377, 379, Ott, S. M., 356 Pawich, T. L., 41, 100
380, 381 Ottinger, E., 437 Payne, S. W., 343, 376
Ochsner, C. A., 437 Ottoni, E. B., 427 Pearce, L., 518
O’Connor, A., 229 Overley, E., 375 Peck, S. M., 205
O’Connor, C., 326 Owen, T. M., 42, 45, 137, 143, 228, Pedersen, E. B., 415
O’Connor, J. T., 216, 306, 344 276, 312, 344, 345, 346, 536 Peeler, C. M., 522
O’Dell, M. C., 394 OwenDeSchryver, J. S., 215, 306, 330 Peirce, J. M., 492
Odom, A., 437 Owens-Stively, J. A., 413 Pejeau, C., 356
Odom, S. L., 142 Ozenne, L., 63 Pelham, W. E., 377
O’Donnell, J., 88, 366 Pelios, L., 352
O’Donohue, W. T., 421, 559, 561, 562, Pellegrin, L. D., 517
564, 565, 566, 567, 568 P Pellegrino, J. W., 253
Odum, A. L., 122, 124, 125 Pelletier, K., 471
O’Dwyer, J. K., 477 Pabico, R. R., 345 Pence, S. T., 180, 203, 207, 209, 354,
Ogawa, I., 142 Pace, G. M., 40, 65, 120, 175, 176, 179, 366, 468
Ogle, L., 257 181, 184, 198, 215, 229, 312, 321, Penney, H., 13, 58, 67, 107, 241
O’Grady, A. C., 160 322, 323, 337, 338, 342, 343, 346, Pennington, R., 273
Ohman, A., 26 359, 376, 468, 521 Pennington, R. C., 534
O’Kane, N. P., 236, 285, 292 Pachis, J. A., 512 Pennypacker, H. S., 6, 8, 9, 10, 60, 135,
Ólafsdóttir, A. R., 278 Packard, D., 325 136, 138, 140, 142, 418
O’Leary, K. D., 137, 139, 378, 382 Paclawskyj, T. R., 176, 177, 194, 196, 208 Penrod, B., 120, 182, 380, 435
596 Author Index
Peoples, A., 325 Pisman, M. D., 106 Quigg, A. M., 135, 156
Perera, R., 493 Pitchert, J. W., 418 Quilitch, H. R., 285
Perez, A., 520 Plaud, D. M., 515 Quittner, A. L., 418
Pérez-González, L. A., 279, 516 Plaud, J. J., 515
Perkins, L. I., 293 Plavnick, J. B., 264, 289, 471
Perlman, J. E., 120 Pliskoff, S. S., 99 R
Perone, M., 45 Plutchik, R., 285
Perrett, D. I., 31 Poche, C., 482, 485 Rabinowitz, S., 431
Perrott, M. C., 79 Podlesnik, C. A., 14, 94, 102, 103, 104, Rachlin, H. C., 5, 71, 72, 100, 124, 127
Perry, A. C., 122 105, 107, 108, 260, 538 Rackliffe, G., 261
Persampieri, M. J., 254, 449, 451, 452 Poindexter, A. R., 194 Radiker, J. E., 285
Person, D. A., 531 Polaha, J., 408 Radley, K. C., 180
Peters, K. P., 139, 332 Poler, J. E., 415 Raetz, P. B., 292, 388, 447, 512, 519,
Peters. L. C., 354 Poling, A., 58, 64, 80, 135, 138, 144, 521, 522
Peters, M. E., 517 182, 215, 218, 243, 301, 306, 338, Rahman, B., 208
Peters, R., 468 340, 361, 381, 524, 547 Raiff, B. R., 494
Peters, R. D., 354 Politylo, B. C., 469 Rainaud, C. P., 124, 125
Peterson, C. R., 42, 98, 338 Pollack, M. J., 290 Raineri, A., 124
Peterson, J. L., 411 Pollard, J. S., 260, 539 Rainey, S. K., 413
Peterson, K. M., 390, 429, 430, 433, Poole, A. D., 142, 144 Rajan, R., 304
434, 435, 436, 439 Pope, D. A., 122 Ramasamy, R., 82, 256
Peterson, L., 414 Porterfield, J. K., 356 Ramos, J. R. M., 433
Peterson, N., 39 Post, T. M., 303 Rand, M. S., 341
Peterson, R. F., 88, 138, 142, 202, 206 Potter, J. N., 243 Randich, L., 126
Peterson, R. L., 357 Potthoff, S., 467 Rankin, B. E., 433
Peterson, S. M., 534 Poulson, C. L., 257, 258, 275, 277 Rankin, C. H., 308
Petitjean, S. A., 495 Powell, J., 144, 146, 147, 361 Rape, R. N., 433
Petry, N. M., 122, 124, 125, 247, 491, Powell, S., 182, 307 Rapoff, M. A., 417, 418, 419, 420, 421
492, 494, 495, 496, 498, 500, 501, Powers, M. A., 362 Rapp, J. T., 139, 144, 183, 203, 244,
505 Powers, S., 302 292, 303, 305, 339, 366, 400, 482
Petscher, E. S., 95, 97, 289, 471 Powers, S. W., 433 Rasey, A., 229
Petursdottir, A. I., 23, 143, 273, 274, Prager, K. L., 67 Rash, C. J., 495, 505
277, 278, 395, 397, 398 Prakash, O., 433 Rasmussen, E. B., 100
Pfadt, A., 205, 391 Pratt, J. L., 67 Rasmussen, T., 271
Phelps, M. C., 243 Pratt, L. A., 203 Ratcliff, C. L., 559, 570
Phillips, C. L., 82, 402 Premack, D., 20, 31, 41, 361, 376 Rau, K., 290
Phillips, J. F., 287, 288, 290, 354 Presbie, R. J., 356 Raulston, T. J., 352
Phillips, K. M., 203, 210 Pressley, M., 261, 264 Raven, B. H., 466
Phillips, R. A., 551 Preston, K. L., 492 Rawson, R. A., 106, 358, 363
Piaget, J., 271 Price, J. N., 380 Ray, K. P., 468
Piazza, C. C., 108, 136, 141, 168, 169, Price, R., 433 Raymond, G. A., 45
173, 174, 175, 179, 180, 181, 184, Prihodova, T., 125 Rea, J., 255
185, 186, 187, 199, 203, 207, 214, Prinz, R. J., 142 Reade, K. K., 288
215, 216, 217, 218, 225, 228, 229, Pritchard, D., 13, 58, 67, 73, 105, 107, 241 Reagan, R., 229
238, 242, 243, 244, 276, 304, 306, Pritchard, J., 545 Reagon, K. A., 257
307, 308, 311, 324, 330, 331, 337, Protopopova, A., 238 Realon, R. E., 288, 361
344, 345, 347, 353, 354, 356, 357, Pruett, A., 435 Reaney, J. B., 414
359, 360, 361, 362, 364, 366, 376, Pumroy, D. K., 247 Reavis, A. R., 118, 182
387, 390, 391, 394, 427, 429, 430, Puri, R., 467 Reay, D. T., 477
431, 432, 433, 434, 435, 436, 437, Pustejovsky, J. E., 161 Redbird, B., 571
438, 439 Putnam, R. F., 463 Redmon, B., 552
Piché, L. M., 177 Pyles, D. A., 289 Reed, D. D., 98, 110, 115, 117, 119, 120,
Pierce, K. L., 256 121, 125, 126, 127, 208, 377, 382,
Pierce, L. E., 276 383, 402
Pierce, W. D., 52, 71, 186, 330 Q Reed, G. K., 307, 330, 434, 435, 439
Pilgrim, C., 84, 85, 87 Reed, P., 392, 402
Pillemer, K., 567 Quattrochi-Tubin, S., 522 Reese, H. W., 87
Pilling, S., 491, 492 Querim, A. C., 79, 216, 217, 223, 311, Reese, M. R., 376
Pimenta, H. P., 433 345, 347 Reese, N. M., 238, 323
Pinkston, E. M., 238, 323, 522 Quick, M. J., 520 Reese, R. J., 535
Author Index 597
Reeve, K. F., 84, 160, 256 Rimmo, P. A., 26 Rollins, J. D., 125
Reeve, S. A., 160, 256 Rincover, A., 42, 217, 286, 325, 344 Rollyson, J. H., 285, 289, 290, 469
Reeves, D., 208 Ringdahl, J. E., 65, 120, 155, 177, 187, Roma, P. G., 115, 116, 117, 126, 127
Regier, D. A., 408 198, 223, 226, 303, 307, 308, 310, Romanczyk, R. G., 66, 363
Rehfeldt, R. A., 88, 126, 331 332, 342, 520, 521 Romani, P. W., 228, 356, 533, 539
Reichert, A., 477 Ringen, J., 562 Romaniuk, C., 307, 343, 398
Reichle, J., 162 Riordan, M. M., 433 Romano, S., 514
Reichow, B., 392 Riordan, M. R., 287, 289 Romanowich, P., 500, 501
Reid, D. H., 179, 236, 256, 285, 286, Risley, T. R., 10, 11, 50, 135, 155, 185, Romer, L. T., 258
287, 288, 289, 290, 291, 292, 293, 206, 238, 322, 356, 383, 393, 518, Rommel, N., 430, 431, 439
325, 354, 469 521 Roncati, A. L., 278
Reid, N., 501 Rispoli, M., 468, 534, 535 Rooker, G. W., 160, 204, 209, 218, 222,
Reid, R., 254, 261, 264 Ritchie, H., 520, 523 223, 225, 324, 338, 344, 345, 347,
Reile, P. A., 100 Ritchie, K. E., 88 352, 353
Reimers, T. M., 222, 410, 536 Ritschl, C., 356 Roper, T. J., 119
Reinke, W. M., 467, 468, 471 Rittig, S., 415 Rortvedt, A. K., 205
Reiss, A. L., 256, 342 Rivara, F. P., 479 Rosales, R., 292
Reiss, M. L., 287, 289 Rivas, K. M., 168, 169, 429, 435, 437 Rosales-Ruiz, J., 229, 254
Reitman, D., 244, 253, 374, 375, 377, Roane, H. S., 1, 3, 11, 65, 118, 119, Rosati, T. C., 82
379, 380, 381, 382, 383 120, 122, 133, 136, 155, 156, 159, Rosch, E. H., 84
Reitzel, J., 254 160, 164, 173, 175, 177, 178, 181, Rosch, K. S., 125
Repp, A. C., 137, 142, 144, 146, 147, 182, 187, 198, 219, 223, 224, 229, Roscoe, E. M., 54, 65, 181, 183, 185,
206, 228, 229, 288, 357, 436 241, 243, 276, 277, 303, 304, 306, 186, 194, 203, 204, 207, 208, 209,
Repp, C. F., 142 310, 326, 327, 339, 345, 346, 356, 210, 225, 226, 229, 243, 291, 303,
Reschly, D. J., 445, 455 363, 391, 520 309, 314, 326, 341, 353, 360
Rescorla, R. A., 18, 19, 20, 22, 28, 32, Roath, C. T., 118 Roscoe, N., 436
104 Robbins, F. R., 314, 343 Rose, K., 302
Retzlaff, B. J., 173, 175, 219, 228, 276, Roberson, A. A., 464 Rose, T. L., 142
301, 398 Roberto, K. A., 567 Rosica, L. A., 436
Reuter-Yuill, L., 520 Roberts, D. M., 142 Rosow, R. P., 545
Rey, C., 95, 353 Roberts, J. A., 482 Ross, D., 257
Rey, J., 463 Roberts, M. L., 58, 71 Ross, D. E., 82
Reynish, L. C., 226 Roberts-Gray, C., 467 Ross, L. V., 420
Reynolds, B., 124 Robertson, J., 208 Ross, N. A., 122
Reynolds, G. S., 64, 69 Roberts-Pennell, D., 194, 277 Ross, S. A., 257
Reynolds, L. K., 375, 378 Robinson, N., 400, 446 Ross, S. G., 453
Ricciardi, J. N., 435 Robinson, S. E., 395 Rossito, A. L., 86
Rice, A., 547 Robles, E., 492, 504 Rotatori, A. F., 179
Richard, E., 256 Robson, K. M., 375 Roth, M. P., 433
Richards, J. B., 124 Rocha, A. D., 433 Rotholz, D. A., 287
Richards, K., 521 Roche, B., 5, 85 Rourke, A. J., 89
Richardson, A. R., 332 Rodger, S., 439 Rourke, D. A., 194
Richardson, J., 62 Rodgers, T. A., 194, 242, 325 Roux, S., 433
Richardson, V., 252 Rodriguez, M., 546 Rovee-Collier, C. K., 327
Richardson, W. K., 28 Rodriguez, N. M., 42, 89, 143, 203, Rowbury, T., 357
Richman, D. M., 206, 327, 353, 354, 204, 206, 274, 277 Rowell, A., 547
363, 365 Rodriguez-Catter, V., 122, 140, 304, Rowsey, K. E., 399
Richman, G. S., 9, 95, 198, 204, 215, 306, 310, 340, 343, 353 Rubin, H., 242
289, 290, 312, 321, 337, 358, 391, 534 Roehler, L. R., 261 Rubio, E. K., 437
Rick, G., 246 Roeyers, H., 125 Ruckstuhl, L. E., 416
Rickert, V. I., 410 Rohn, D., 546 Rudolph, C., 430
Ricketts, R. W., 242, 365 Rohrbeck, C. A., 179 Rudrud, E. H., 183
Ridgel, A. L., 430 Rohrer, N., 279 Rule, S., 465
Riegel, K., 439 Rohsenow, D. J., 501, 504 Rummler, G. A., 545, 552, 554
Ries, R. K., 493 Rojahn, J., 194, 364 Rush, D., 312
Riffin, C., 567 Rolider, A., 354, 362, 365, 366 Rush, K. S., 138, 178, 194, 196, 246,
Riis, G., 430 Rolider, N. U., 137, 138, 243, 365 344, 356, 519
Riley, A. R., 422 Roll, D., 242 Rusinko, L. C., 534
Riley, D. S., 109 Roll, J. M., 504, 505 Russell, D., 25, 118, 323, 345, 374
Rimington, D. D., 124 Rollings, J. P., 365, 366 Russo, D. C., 217, 286, 342
598 Author Index
Russo, L., 325, 354 Sauttfr, R. A., 252 Scott, L. M., 255
Rutter, M., 391 Saville, B. K., 446 Scott, L. R., 549
Ryan, G. W., 477 Scanlon, D. M., 452 Scott, T. M., 194
Ryan, J. B., 357 Schaefer, H. H., 310 Scotti, J. R., 324, 432, 439
Ryan, S. R., 496 Schafe, G. E., 430 Sears, L., 390
Ryder, J., 100 Schaie, K. W., 513 Seaver, J. L., 400, 401
Ryle, G., 4 Schechter, G. L., 428 Seaver, J. P., 20
Scheeren, A., 390 Secades-Villa, R., 494
Schell, R. M., 293, 304 Seckinger, K. E., 135, 156
S Schepers, A., 517 Seefeldt, D., 513
Schepis, M. M., 290 Sege, R., 477
Sable, C., 532 Scherer, E. A., 496 Sehgal, S. K., 433
Sabo, B., 63 Scheres, A., 125 Seiverling, L., 394, 438
Sacks, J., 477 Scheuermann, B., 357 Seligman, M. E. P., 43
Sadek, S., 517 Schieber, E., 115 Sellers, J. A., 226
Sadowski, K. A., 522 Schieltz, K. M., 108, 388, 530, 533, Sellers, T. P., 244, 260
Saggers, E. A., 225, 226 537, 538, 539, 540 Sembi, S., 125
Saini, V., 45, 65, 106, 173, 174, 175, Schienebeck, C. J., 465 Senesac, B. V., 456
194, 214, 219, 229, 276, 301, 303, Schinke, S. P., 142 Senturia, Y. D., 477
312, 325, 329, 330, 345, 352, 391, Schleien, S. J., 256 Serna, R. W., 83
398, 536 Schlichenmeyer, K. J., 54, 203, 204, Severtson, J. M., 292
Saka, N., 463 206, 207, 225 Sevin, B. M., 140, 307, 310, 324, 353,
Sakagami, T., 102 Schlinger, H. D., 237, 241, 261 434, 436, 437
Sakai, K., 31 Schmid, T. L., 109 Sevin, J. A., 256
Saladin, M. E., 124 Schmidt, A. C., 183, 394 Seymour, B. L., 499
Salama, F., 307 Schmidt, J. D., 256 Seymour, K. J., 549
Salem, S., 549 Schmitt, B. D., 409 Seys, D. M., 365, 366
Salend, S. J., 356, 376, 377 Schmitz, J. M., 502 Sgro, G. M., 345
Salhi, C., 477 Schneider, S. M., 3 Shabani, D. B., 312, 314
Saliba, E., 433 Schneiderman, N., 19 Shade, D., 71, 184, 198, 243
Sallows, G. O., 392 Schnell, L. K., 291 Shadish, W. R., 161
Salvy, S. J., 123, 364, 365 Schnelle, J. F., 518 Shafer, K., 273
Samaha, A. L., 118, 183, 184, 208, Schnerch, G., 144 Shaham, Y., 105
226, 468 Schnoes, C. J., 409, 410, 411 Shahan, T. A., 25, 94, 102, 104, 106,
Sameoto, D., 353 Schoen, S. F., 256, 258 107, 122
Sampson, H. A., 431 Schoenbach, V. J., 480 Shaikh, U., 531, 532
Samuels, F. H., 453 Schoenfeld, W. N., 6, 22, 331 Shalev, R. A., 436
Sanchez, S., 388, 476, 481 Schoffelmeer, A. N., 127 Shantz, J., 433
Sanders, R. Q., 198 Schonbachler, H. E., 62, 224 Shapiro, E. S., 362
Sandman, C. A., 209, 312, 324 Schonfeld, L., 513 Shapiro, M. M., 21, 28
Sands, D., 517 Schottenfeld, R. S., 504 Shapiro, S. T., 356
Sanetti, L. M. H., 387, 463, 468, 469, 470 Schreck, K. A., 390 Sharp, W. G., 390, 437
Sanservo, A. C., 517 Schreibman, L., 82, 256, 394, 395, 401 Shattuck, J., 453
Sanson-Fisher, R. W., 142, 144 Schultz, R. T., 390 Shattuck, P. T., 279
Santana, C. M., 324, 433, 436, 437 Schultz, W., 24, 25 Shaw, P., 531
Santarcangelo, M., 127 Schulze, K. A., 183 Shea, M. C., 184
Santiago Lipana, L., 531 Schuster, J. W., 259 Sheldon, J. B., 288, 290, 291, 305, 376
Sarokoff, R. A., 287, 469 Schuster, M. A., 477 Shepard, J. A., 415
Sasso, G. M., 222 Schuster, M. M., 411, 413 Shepley, S. B., 83
Sasson, J. R., 547, 552, 554 Schusterman, R. J., 54 Sheridan, S. M., 463, 464, 466, 467
Saudargas, R. A., 144, 147, 382 Schutz, R. E., 382 Sherman, F. T., 516
Saulsgiver, K. A., 494 Schwartz, I. S., 279, 292 Sherman, J. A., 83, 88, 245, 288, 290,
Saunders, K. J., 83, 84, 85, 86, 87, 88, Schwartz, J. E., 468, 537 291, 305, 376
366 Schwarz, S. M., 431 Shernoff, E. S., 453
Saunders, M., 247, 467 Schweitzer, J. B., 30, 72, 246 Shields, F., 376
Saunders, R. R., 82, 83, 84, 85, 86, 87, Schwendiman, J. W., 103 Shiffman, S. S., 183
89, 516 Scilley, K., 521 Shillingsburg, M. A., 88, 118, 182,
Saunders, S. S., 100 Scott, D., 255 185, 260, 274, 276, 277, 278, 326,
Sautter, R. A., 277 Scott, J., 292, 476 339, 397
Author Index 599
Shimamura, J. W., 247 Skinner, B. F., 3, 4, 6, 7, 8, 9, 15, 16, 17, Snycerski, S., 58, 215, 243, 301, 338,
Shimoff, E., 52 18, 21, 22, 25, 28, 29, 32, 37, 44, 50, 381
Shippen, M. E., 262 52, 54, 60, 78, 97, 103, 214, 240, 241, Snyder, A., 433
Shireman, M. L., 356, 446 255, 256, 259, 270, 271, 272, 273, Snyder, G., 549, 550
Shirley, M. J., 218, 339 275, 276, 278, 279, 326, 328, 330, So, Y., 548
Shlesinger, A., 291 343, 374, 379, 395, 396, 397, 398, Soares, N., 535
Shoemaker, J., 288, 289 400, 411, 421, 445, 452, 512, 522, Sodawasser, A. J., 228
Shoemaker, M., 391 559, 561, 562, 563, 564, 566, 567 Soderlund, M. J., 311, 364
Shogren, K. A., 194 Skinner, C. H., 313, 446, 457 Sohl, K., 532
Shook, G. L., 293, 571, 572, 573, 574, Skinner, K. A., 108 Sokolowska, M., 26
576 Slack, D. J., 142 Sol, G. D., 427
Shore, B. A., 109, 123, 185, 216, 223, Slaton, D. B., 256 Solberg, K. M., 185, 186
305, 308, 338, 343, 354, 433, 435 Slaton, J., 241, 243, 246 Solnick, J. V., 42
Showers, B., 468, 469, 471 Slaven, J., 144 Solnick, S. J., 477
Shugart, M. A., 408 Slider, N. J., 468, 469 Solomon, B. G., 469, 470
Shull, R. L., 73, 99, 102, 104, 109, 329 Slifer, K. J., 9, 95, 198, 204, 215, 312, Solomon, P. R., 513, 514, 515
Sicherer, S. H., 431 321, 337, 358, 391, 534 Sonuga-Barke, E. J. S., 125
Sidener, T. M., 243, 291, 392, 519 Slivka, R. M., 366 Sor, S., 477
Sideridis, G. D., 453 Sloan, W., 83 Soraci, S. A., Jr., 83
Sidman, E. A., 479 Sloat, K. C. M., 549 Soriano, H. L., 446
Sidman, M., 5, 31, 47, 49, 81, 83, 84, Slocum, S. K., 139, 260, 332, 343 Sorocco, K., 516
85, 88, 341, 382 Slocum, T. A., 160, 262, 263, 264 Soto, P. L., 99
Siegel, G. M., 89 Sloman, K. N., 96, 203, 326, 536 Souza, A. C., 278
Siegel, S., 26 Slone, N. C., 535 Spangler, P. F., 518
Sierp, B. J., 436 Slowiak, J. M., 548 Sparkman, C. R., 392
Sigafoos, J., 194, 195, 196, 225, 226, Small, G. A., 142 Spector, A., 517
277 Smalls, Y., 194, 196 Spector, V., 395
Sigmon, S. C., 421, 494 Smeets, P. M., 83, 84 Speece, D. L., 446
Sigurådóttir, Z. G., 87, 435 Smethells, J. R., 122 Speechley, K. N., 408
Sigurdsson, S. O., 140, 507, 550, 552 Smith, A., 546 Speltz, M. L., 247
Silbaugh, B. C., 260 Smith, A. C., 147 Spence, B. H., 286
Silberberg, A., 117 Smith, A. D., 513 Spencer, C. J., 287
Silkowski, E. L., 159, 304, 363 Smith, A. F., 390 Spencer, T. J., 87
Silla-Zaleski, V. A., 264 Smith, C. A., 183 Spencer, W. A., 308
Silver, A., 476 Smith, C. E., 206, 312, 314 Spiga, R., 115
Silverman, A. H., 430 Smith, C. J., 182 Spirito, A., 413
Silverman, K., 388, 490, 491, 492, 493, Smith, E. E., 84 Spitalnik, K., 380
502, 503, 504, 505, 506, 507 Smith, G. J., 260 Spitalnik, R., 380
Simmelhag, V. L., 21 Smith, I. M., 402 Splett, J. W., 467
Simmons, C. A., 203, 330 Smith, J., 125 Spooner, D., 259, 260
Simmons, J. N., 309, 310, 311, 364 Smith, K. A., 83 Spooner, F., 259, 260, 276
Simmons, J. Q., 214, 322, 364, 365, Smith, L., 563 Spooner, S. A., 531
396 Smith, M., 435, 546 Spradlin, J. E., 14, 78, 80, 82, 84, 85,
Simmons, S. F., 518 Smith, M. C., 17 86, 89, 260, 276
Simon, H. K., 477, 478 Smith, R. G., 64, 123, 176, 216, 224, Sprague, J. R., 197, 240, 343
Simon, J. L., 14, 78, 260 227, 228, 229, 301, 303, 305, 306, Sprang, R., 535
Simonsen, B., 470 308, 309, 311, 312, 313, 314, 322, Spreat, S., 357
Simo-Pinatella, D., 196 323, 325, 327, 331, 340, 343, 354, Sprinkle, E. C., 277
Sing, L. Y., 439 364 Squires, J. L., 274
Singer, G., 395 Smith, S. J., 452 Squires, N., 98
Singer, L. T., 439 Smith, T., 260, 287, 391, 392, 401, Sran, S. K., 122, 376, 383
Singer-Dudek, S., 273 403, 447 Srinivasan, S. R., 438
Singh, N. N., 146, 194, 196, 363 Smith, T. L., 10 Srour, J., 413
Singletary, L. L., 468 Smits, M. G., 205 St. Peter, C. C., 65, 96, 105, 106, 180,
Siqueland, E. R., 329 Smits-Seemann, R. R., 125 203, 209, 245, 256, 354, 366, 400,
Sistrunk, F., 260 Smyth, C. A., 468, 470 446, 468, 536, 537
Sivan, E., 261 Sneed, T. J., 287 Staats, A. W., 382
Sizemore, O. J., 43 Snell, M. E., 259 Staats, C., 549
Skidmore-Roth, L., 415 Snow, C. E., 253, 450, 458 Staats, C. K., 382
600 Author Index
Staddon, J. E. R., 21, 98, 103 Strobel, M. J., 207, 468, 519 Tait, R., 366
Stage, S. A., 453 Strohmeier, C. W., 136 Tanney, J., 4
Stahelski, A. J., 548 Stromer, J. B., 83, 87 Tanz, J., 294, 547
Stahmer, A. C., 392 Stromer, R., 83, 86, 87, 142 Tanzman, V. S., 452
Stamper, M. R., 548 Strouse, M. C., 288, 290, 291 Tapp, J., 139
Stanger, C., 496 Strunck, P. G., 433 Tarasenko, M. A., 486
Stanislaw, H., 392 Stuart, R. B., 422 Taravella, C. C., 136, 181
Stanley, A. E., 433 Stubbs, K. H., 436, 437 Tarbox, J., 69, 356, 433
Stanley, M. A., 517 Studer, L., 291 Tarbox, R. S. F., 291, 468
Starin, S., 572 Sturmey, P., 118, 164, 182, 194, 287, Tardif, M., 491
Stark, L. J., 413, 414, 433 288, 289, 291, 294, 394, 438, 469 Tarkenton, F., 545
Starnes, M., 479 Suarez, Y., 288, 354 Tasky, K. K., 183
Stasolla, F., 238 Subramaniam, S., 388, 490 Tate, B. G., 305, 365
Staubitz, J. L., 125 Suda, K. T., 289, 482 Taub, E., 245
Steege, M. W., 59, 197, 323, 339 Suen, H. K., 144, 146 Taube, C. A., 408
Steers, M. E., 519 Suess, A. N., 108, 388, 468, 530, 534, Taubman, M., 287
Stein, B., 428 536, 537, 538, 539 Taylor, B. A., 256, 279
Stein, M. B., 517 Sugai, G. M., 194, 447, 456, 470 Taylor, D. V., 312
Steinberg, M., 517 Sullivan, K. J., 161 Taylor, E., 125
Steingrimsdottir, H. S., 516 Sullivan, M. A., 375, 379 Taylor, M. A., 144
Steinhilber, J., 184 Sullivan, M. T., 311, 341 Taylor, R. L., 82
Stenberg, A., 415 Sullivan, M. W., 327 Taylor, S. A., 139
Stenhoff, D. M., 534 Sullivan, W. E., 11, 133, 155, 260, 387, Tees, R. C., 275
Stennies, G., 477 445 Teichman, H. M., 135
Stephenson, K., 182 Sulzer-Azaroff, B., 30, 72, 225, 246, Tekin-Iftar, E., 259, 531
Sterling-Turner, H., 468 287, 290, 293, 549, 550 Temple, J. L., 123
Sternberg, L., 259 Sumnall, H., 127 Tenenbaum, H. R., 262
Sterner, K., 292 Sundberg, C. A., 23, 274 Tenhula, W. N., 374
Stevens, K. B., 256 Sundberg, M. L., 23, 271, 273, 274, 275, Terao, K., 20
Stevenson, S. S., 328 276, 278, 396, 398, 399, 400, 522 Terauchi, Y., 431
Stewart, I., 88 Sunderland, T., 516 Teret, S. P., 477
Stewart, J., 26 Sunderman, P., 82 Terhune, J. G., 361
Stewart, K. K., 160 Sunstein, C. R., 127 Teri, L., 519, 567
Stewart, K. L., 84 Sutherland, D., 88 Terrace, H. S., 80
Stewart, V., 437 Sutherland, V. J., 551 Terris, W., 363
Stickney, K. A., 20 Sutton, N. P., 101 Tesch, D., 352
Stiefler, N. A., 332 Swain, J. C., 356, 382 Test, D. W., 142, 276
Stitzer, M. L., 127, 492, 504 Swain, R., 253, 259 Testa, J. R., 144
Stocco, C. S., 203, 206, 446 Swanson, V., 177 Testa, T. J., 38
Stock, L. Z., 517 Swartzwelder, D. M., 437 Tetreault, A. S., 207, 468
Stoddard, L. T., 54, 81, 83, 84, 86, 87 Swearer, S. M., 415 Tevyaw, T. O., 504
Stoffelmayr, B. E., 31 Swearingen, M., 482 Thaler, R. H., 123, 127
Stojek, M. K., 126 Swearingin, W., 433 Theios, J., 109
Stokes, G., 521 Sweeney, M. M., 106, 107 Theodore, L. A., 380
Stokes, T. F., 11, 31, 78, 239, 260, 375, Swift, R. M., 497, 501 Therrien, W. J., 453
379, 380, 393, 401, 402, 449, 485 Swisher, M. J., 54 Thibadeau, S., 302
Storey, K., 197 Swithers, S. E., 309 Thibault-Sullivan, M., 353
Stormont, M., 467, 471 Switzky, H., 179 Thibodeau, M. G., 82, 401
Stotts, A. L., 502 Sy, J. R., 118, 126, 182, 185 Thiemann, K. S., 138
Strain, E. C., 502 Symon, J. B. G., 468 Thiesse-Duffy, E., 482
Strain, L. A., 290, 546 Symons, F. J., 162, 206, 208 Thoeny, A. L., 125
Strand, P. S., 203 Syncerski, S., 524 Thomas, D. R., 571
Strang, J., 491 Szadokierski, I., 447 Thomas, J., 255
Strauss, D., 5 Thomas, J. R., 363
Strawn, K. N., 477 Thomason-Sassi, J. L., 226, 229, 356
Street, E. M., 454 T Thommessen, M., 430
Stricker, J. M., 363 Thompson, L. W., 517
Stricklin, S. B., 256 Tai, S. S., 446 Thompson, R., 514
Striefel, S., 83, 84, 88, 89, 199 Tailby, W., 85 Thompson, R. F., 28, 308
Author Index 601
Thompson, R. H., 64, 65, 66, 136, 140, Trosclair-Lasserre, N. M., 105, 118, Van Luit, J. E. H., 253
173, 179, 180, 183, 185, 202, 203, 122, 182, 242, 343, 347, 536 van Mourik, Y., 127
206, 209, 210, 216, 225, 228, 240, Troutman, A. C., 256, 357, 376 Van Norman, E. R., 445
242, 243, 244, 245, 288, 309, 311, Truchlicka, M., 356, 382 Van Norman, R., 446
314, 330, 331, 345, 353, 358, 360, Tryon, G. S., 468 van Oorsouw, W. M., 290
363, 364, 366, 446 Tsami, L., 356, 540 van Schooten, E., 252
Thompson, S., 208 Tucker, D., 62 van Steensel, R., 252
Thompson, T., 206 Tucker, J. A., 419, 448 van Stone, M., 43, 310
Thompson, T. M., 276 Tucker, R. D., 375, 376, 380, 381 Van Willigen, F., 18
Thomson, C., 147 Tueth, M. J., 567 VanBiervliet, A., 86
Thomson, R. J., 331, 342 Tullis, C. A., 303 VanDerHeyden, A. M., 203, 208, 468
Thorndike, E. L., 15, 16, 17 Tuncel, S., 549 Vandrey, R. G., 127
Thurkow, N. M., 548 Tung, S. B., 177 Vanselow, N. R., 227, 486
Tidey, J. W., 497, 501 Tunstall, K. R., 453 VanWormer, J. J., 141
Tiger, J. H., 66, 67, 68, 74, 105, 118, Turk, D., 517 Vargas, A. U., 313
147, 180, 183, 185, 235, 237, 239, Turner, J. A., 302 Vargas, E. A., 3
240, 244, 246, 260, 311, 330, 331, Tustin, R. D., 121 Vargo, K. K., 446
339, 340, 344, 398, 434, 536 Tversky, A., 123 Vassilopoulos, A., 244, 374
Tighe, T., 492 Twyman, J., 274 Vaughan, M. E., 23, 242, 324, 343
Tilly, W. D., 455, 456, 457 Tyler, B. J., 453 Vaughn, B. J., 314
Timberlake, W., 20, 21, 361, 564 Tyler, D. J., 430 Vaughn, S., 453
Tinbergen, N., 202 Tyler, S., 144, 145, 146, 147 Vaz, P. C. M., 429, 430, 436, 437
Tincani, M. J., 88, 235, 270, 273, 275, Tyler-Merrick, G., 88 Veazey, S. E., 447
276, 279, 323 Tzu, L., 13 Veenstra, R. A., 229
Tingstrom, D. H., 147, 209, 375 Veereman-Wauters, G., 430
Tiry, A. M., 125 Vegas, R., 13, 15, 19, 40
Todd, A. W., 264, 468 U Vellutino, F. R., 452
Todd, J. T., 3, 327 Vener, S. M., 257
Todd, T. P., 105 Ubdegrove, K., 194 Vergason, C. M., 545
Tofflemoyer, S., 264 Uden, D., 416 Verhave, T., 85
Toien, O., 427 Ulicny, G., 259 Vernick, J. S., 477
Tolson, D., 523 Umbreit, J., 199, 226 Verriden, A. L., 185, 360
Tomassone, S., 229 Umbricht, A., 492, 502 Vesloski, M. J., 264
Tombari, M. L., 465 Upadhyaya, H., 496 Vincent, S. A., 414, 415
Tompkins, B. F., 83 Upright, J. J., 453 Violano, P., 476
Tonga, J. B., 517 Urcuioli, P. J., 54, 84 Virués-Ortega, J., 239, 392, 512, 520,
Tontsch, C., 125 Usher, S., 354 521, 572
Toole, L. M., 107, 180, 347, 352, 356, Uttal, W. R., 4 Visser, E., 277
365, 366, 391 Uy, M., 306 Vittorio, T. D., 289
Toper-Korkmaz, O., 540 Uzzell, R., 286 Vladescu, J. C., 160, 183, 243, 244,
Torquato, R. D., 73, 102, 329 291, 398
Torres-Viso, M., 136 Vogel-Sprott, M., 362
Tota, M. E., 73, 102, 329 V Vogt, K., 279
Touchette, P. E., 81, 82, 84, 205, 209, Volkert, V. M., 105, 107, 291, 324, 343,
245, 276 Vaca, L., 468 347, 390, 429, 430, 433, 435, 436,
Tourigny Dewhurst, D. L., 179 Vaccarello, C. A., 465 436–437, 438, 536
Toussaint, K. A., 41, 100, 118, 183, Vadasy, P. F., 453 Volkmar, F. R., 390, 391, 392
244, 398 Vaidya, M., 88 Vollmer, D., 416
Towery, D. L., 288 Valbuena, D., 244, 388, 476, 481 Vollmer, T. R., 61, 62, 64, 65, 68, 71,
Townsend, N. M., 365 Valdimarsdottir, H., 435 74, 82, 96, 98, 120, 126, 139, 143,
Trachtenberg, S., 292 Valdovinos, M. G., 524 166, 167, 168, 176, 177, 179, 183, 187,
Trahan, M. A., 512, 522, 523 Valentino, A. L., 260, 276, 447 194, 196, 198, 203, 206, 207, 208,
Trahant, D., 361 Valerius, K. S., 430 209, 223, 224, 228, 229, 237, 240,
Trapman, M., 252 Van Camp, C. M., 65, 106, 206, 224, 243, 244, 246, 256, 301, 303, 305,
Traub, M. R., 143 309, 327, 338, 366 306, 307, 309, 310, 314, 321, 322,
Trenholme, I. A., 362, 363 van Gelderen, A., 252 323, 325, 327, 329, 330, 332, 337,
Trevino, N., 378 Van Hasselt, V. B., 26 338, 340, 342, 343, 344, 356, 357,
Tripp, G., 515 Van Houten, R., 246, 353, 354, 356, 391, 400, 402, 411, 520, 536
Troia, G. A., 263, 264 357, 362, 365 Vom Saal, W., 19
602 Author Index
von Duvillard, S., 515 Weeden, M., 218 Williams, B. A., 22, 30, 97
Vorndran, C. M., 65, 118, 181, 241, 305, Weeks, M., 343, 357 Williams, C., 567
309, 327, 329, 362, 364, 366, 400 Weetjens, B. J., 80, 547 Williams, C. D., 322
Vowles, K. E., 161 Wehby, J. H., 143, 313, 468 Williams, C. E., 482
Vriniotis, M., 477 Wehman, P., 179, 256, 291 Williams, D., 255
Vuchinich, R. E., 419 Weigle, P. A., 109 Williams, D. C., 83, 84, 88, 366
Weil, T. M., 169 Williams, D. E., 242, 363
Weiner, H., 241, 331 Williams, G., 256, 279, 390, 409
W Weiner, J., 348 Williams, J. A., 70
Weinstein, A., 28 Williams, J. C., 263
Wachter, J., 85, 86, 89 Weinstein, K. S., 447 Williams, K. E., 390, 433, 435, 438, 439
Wacker, D. P., 102, 103, 104, 107, 108, Weinstock, J., 498 Williams, K. L., 468
222, 229, 246, 253, 313, 323, 327, Weinstock, L. B., 411 Williams, L., 365
337, 339, 353, 362, 388, 433, 468, Weise, M. J., 323 Williams, R. L., 375
530, 533, 534, 536, 537, 538, 540 Weiss, B., 48 Williams, S. A., 125
Waddington, H., 277 Weiss, J. S., 258 Williams, W. L., 289, 293, 344
Waelti, P., 25 Weisz, J. R., 383 Williamson, P. N., 519
Wagner, A. R., 19, 20 Welch, M., 463 Willis, B. S., 287, 293
Wagner, L., 455 Weldy, C. R., 292 Willis, R. P., 263
Wagner, R. K., 261 Welker, R. L., 32 Willson-Morris, M., 85
Wagner, S., 520, 548 Weller, R. E., 125 Wilson, A. G., 125
Wahler, R. G., 203, 206 Welsch, W. V., 285 Wilson, A. N., 122
Wald, A., 412 Welsh, F., 5 Wilson, C. L., 476
Walden, T., 139 Welter, K. M., 307 Wilson, C. W., 144
Waldrop, A. E., 496 Werder, C. S., 449 Wilson, D. M., 140, 331
Walker, B., 262 Werker, J. F., 275 Wilson, J. L., 304
Walker, G., 394, 401 Werle, M. A., 433, 438 Wilson, K., 422
Walker, N., 494 Werner, S. B., 365 Wilson, K. M., 363, 515
Wallace, C. J., 287 West, M. J., 51 Wilson, M. E. H., 478
Wallace, M. D., 42, 64, 65, 120, 178, Westbrook, R. F., 19 Wimberly, B. S., 332
182, 222, 229, 291, 309, 313, 326, Wetherby, B., 85, 88 Windsor, J., 177, 179
402, 433, 435, 468 Wheeler, A. J., 288 Wine, B., 166, 184, 545
Wallin, J. A., 551 Wheeler, D. J., 205 Winick, M., 438
Walsh, M. E., 82 Wheeler, E. E., 204, 225 Winkel, G. H., 206
Wang, Y., 26 Whelan, B. M., 206, 312 Winkler, R. C., 115
Wangmo, T., 567 Whitby, P. S., 276 Winstanley, E. L., 502
Wankoff, L. S., 275 White, G. P., 254 Wintemute, G. J., 477
Ward, P., 446 White, W. A. T., 447 Winterbauer, N. E., 105, 106, 107
Ward, R. D., 104 White, W. L., 507 Winterling, V., 182, 244, 314
Ward-Horner, J. C., 164, 184 White-Blackburn, G., 247 Wirth, O., 144, 146, 147
Warren, J. L., 517 Whitehead, W. E., 411, 413, 518 Wiskow, K. M., 247, 260
Warren, S. F., 277 Whitehouse, M. H., 453 Witt, J. C., 140, 203, 255, 263, 378,
Warzak, W. J., 415 Whiting, S. W., 139 446, 457, 466, 468
Wasik, B. H., 142 Whitlow, J. W., 309 Wittgenstein, L., 563
Waters, M., 481 Whitman, T. L., 286, 288, 289, 293 Wohl, M. K., 433
Wathen, S. N., 105 Whitmarsh, E. L., 208 Wohn, T., 52, 64
Watkins, C. L., 262, 263, 264, 454 Wichnick-Gillis, A. M., 257 Wolery, M., 82, 242, 245, 256, 258,
Watson, J. B., 4 Wierzbicki, J. J., 391 259, 392, 400, 401, 447, 448
Watson, L. S., 286 Wightman, D. C., 260 Wolf, A. W., 409
Watson, T. S., 62, 197, 355, 417, 457, Wilczynski, S. M., 209 Wolf, M. M., 10, 11, 50, 135, 155, 247,
468, 521 Wild, B., 517 322, 323, 365, 374, 382, 467
Watters, J. K., 260 Wilder, D. A., 66, 166, 184, 229, 274, Wolfe, K., 160
Wearden, J. H., 97 286, 288, 290, 294, 307, 326, 331, Wolfe, P., 257
Weatherly, N. L., 546 344, 353, 356, 388, 433, 544, 546, Wong, C. J., 504
Weaver, L. A., 447 547, 548, 550, 552 Wong, S. E., 142
Webber, C., 456 Wiley, C. C., 477 Wood, C. L., 263
Webber, L., 457 Wilke, A. E., 228, 356, 433 Wood, D. D., 135, 137, 142, 143
Weber, L. H., 260 Wilkins, J. W., 429, 430, 436 Wood, K. D., 452
Webster, D. W., 477, 478, 479 Wilkinson, K. M., 84 Woodruff-Pak, D., 514, 515
Author Index 603
ABAB design, 162. See also Single- cigarette smoking, 494 elder abuse and, 567
case experimental designs marijuana, 496–497 overview, 512–513, 523–524
A-B-C checklist, 207 opioids, 495 Alcohol, 497–498
A-B-C model. See Antecedent– overview, 388, 490, 507 Allocation of behavior, 94, 95–96. See
behavior–consequence (ABC) pharmacotherapy and, 501–502 also Response allocation
A-B-C recording, 206–207 polydrug use, 498–499 Alternative reinforcement, 344–345
Abolishing operations, 272, 309 respondent conditioning and, Alternative schedules of
Abstinence reinforcement procedures. 32–33 reinforcement, 68t, 74. See
See also Addiction stimulants, 495–496 also Combined schedules of
cigarette smoking, 494 Add-in analysis, 164 reinforcement; Schedules of
dissemination of, 505–506 Adherence to treatment. See also reinforcement
history of the research regarding, Treatment integrity Alzheimer’s disease, 304
490–493 assessing, 417–420 Amphetamines, 495–496
improving outcomes from, 499–505 behavioral pediatrics and, 416–422 Anorexia nervosa, 428. See also
overview, 490, 506–507 improving, 420–422 Feeding disorders
recent advances in, 493–506 Adjusting-amount procedure, 124 Antecedent exercise, 302–303
utility of, 493–499 Agency, 288 Antecedent interventions. See also
Acceptability of treatment, 380, 536 Aggression Interventions; Noncontingent
Acceptance escape and, 47 reinforcement (NCR)
feeding disorders and, 428–429, extinction and, 327 default interventions, 301–305
433–435 as an indirect effect of punishment, escape behavior and, 311–314
organization behavior management 366 feeding disorders and, 433–434,
(OBM) and, 550 noncontingent reinforcement 439
Accuracy, 135, 140. See also (NCR) and, 306–307 function-based interventions,
Measurement response restriction and, 305 305–314
Acquisition-level training, 447–448 schedules of reinforcement and, 74 noncontingent reinforcement
Active-learning approach, 481 using punishment effectively and, (NCR) and, 306–310
Activity assessment, 361. See also 355–356, 365 organization behavior management
Assessment Aging. See also Behavioral gerontology (OBM) and, 547
Addiction. See also Abstinence behavior-analytic research on, overview, 301–302, 314–315
reinforcement procedures 513–516 stimulus control strategies, 310–311
alcohol, 497–498 clinical behavior-analytic research Antecedent Variables Assessment
behavioral economics and, 122–123 on, 516–523 Form, 197
604
Subject Index 605
Antecedent–behavior–consequence behavioral gerontology and, 513 Auditory stimuli, 25, 79, 355. See also
(ABC) behavioral pediatrics and, 417–420 Exteroceptive stimuli
consultation model and, 465 clinical behavior-analytic research Auditory–visual conditional
learning/instructional hierarchy on aging and, 518–523 discriminations, 396–397
model and, 447 complexity and, 254–255 Autism Curriculum Encyclopedia
organization behavior management direct instruction and, 262 (ACE), 399
(OBM) and, 550 encopresis and, 412–413 Autism spectrum disorder (ASD)
overview, 4 enuresis and, 415 clinical considerations and, 397–402
staff training and management evaluating reinforcement effects conditioned reinforcement and,
and, 287–289, 290–291 and, 180–182 30–31
token-economy systems and, 374 factors that may influence the early intensive behavioral
Antecedent–behavior–consequence effectiveness of reinforcement interventions (EIBI) and,
(ABC) chain, 253 and, 184–185 391–397
Antecedent–behavior–consequence feeding disorders and, 430–432, 433 enriched environment and, 304
recording, 206–207 indirect behavioral assessments, features of, 390–391
Antecedents 193–199 future research, 397–402
antecedent control and, 364 issues related to stimuli as issues related to stimuli as
applied-verbal behavior approach reinforcement, 182–184 reinforcement and, 183–184
with ASD and, 395–396 methods to identify preferred stimuli overview, 387, 389–390, 403
feeding disorders and, 433–434 and, 175–180, 176f, 186–187 professional certification and
Functional Analysis Interview noncontingent reinforcement credentials and, 575
and, 198 (NCR) and, 308 reinforcement and, 41–42
generalized matching law and, 100 organization behavior management staff training and management
overview, 299 (OBM) and, 546–547, 549–550, and, 291–292
private events and, 5–6 551, 552–554, 553f structural versus functional
staff training and management safety skills and, 480–481, 483, classification of behavior and, 8
and, 287–288 485, 485t telehealth services and, 535, 537
token-economy systems and, 374 selecting procedures for Autoclitic verbal operants, 395–396
Anxiety, 516–517 punishment and, 357–362, 358f, Automatic reinforcement. See also
Aphasia, 522. See also Verbal behavior 359f, 360t Reinforcement
Appetitive unconditioned responses, stimulus preference and, 186–187 differential reinforcement
26–27. See also Unconditioned teacher consultation and, 463, 464, procedures and, 242–243
response (UR) 467–471 functional analyses and, 216–217
Applied behavior analysis (ABA) teaching verbal behavior and, 278 function-based extinction
behavioral pediatrics and, 409, 422 telehealth services and, 530f procedures and, 324–326, 332
complexity and, 254 treatment adherence and, 417–420 function-based reinforcement
early intervention models for Assessment of Basic Language and procedures and, 343–344
autism spectrum disorder based Learning Skills—Revised noncontingent reinforcement
on, 392–396 (ABLLS-R), 278, 398–399 (NCR) and, 307–308
overview, 3, 10–12, 408 Asymmetrical choices, 71, 115. See overview, 23
professional certification and also Choice Autoshaping procedure, 26. See also
credentials and, 570–576, 572f, Attention seeking behaviors Shaping
573f, 575f feeding disorders and, 432 Aversive stimuli, 312–314
respondent conditioning and, function-based extinction Aversive unconditioned responses,
29–33 procedures and, 321, 322 26–27. See also Unconditioned
response-to-intervention models noncontingent reinforcement response (UR)
and, 457–458 (NCR) and, 306–307 Avoidance, 46–48, 305
telehealth services and, 533–537 Attention-deficit/hyperactivity Avoidant/restrictive food intake
Applied-verbal behavior approach, disorder (ADHD) disorder, 428. See also Feeding
395–396 autism spectrum disorder (ASD) disorders
Approximations, 255 and, 400
Assessment. See also Data collection; conditioned reinforcement and,
Functional analyses; 29–30 B
Measurement; Observational delay discounting and, 125
procedures schedules of reinforcement and, 72 Background stimuli, 79
addiction and, 507 telehealth services and, 535 Backup reinforcers, 376. See also
applied behavior analysis and, 11 token-economy systems and, 377 Reinforcers
autism spectrum disorder (ASD) Audience functional relations, Backward chaining, 245–246,
and, 394, 397–399, 400 271–272 259–260. See also Chaining
606 Subject Index
Backward chaining with leap ahead, Behavioral consultation, 387–388, Behavioral skills training. See also
259. See also Chaining 463–467. See also Consultation; Skills training
Baseline conditions Teacher consultation education and, 445–446
function-based extinction Behavioral contrast, 69–70, 69f future research, 486
procedures and, 328–330, 332 Behavioral cusps, 254 safety skills and, 482–484, 486
function-based reinforcement Behavioral definitions, 136. See also staff training and management
procedures and, 337 Operational definitions and, 286–287
single-case experimental design Behavioral economics. See also Choice Behavioral Stream Interview, 197
and, 157 applied relevance of assessing Behavioral velocity
visual inspection of graphed data stimulus value and, 120–123, behavioral momentum theory and,
and, 159 121f 103–104
Bedtime pass program, 410–411 delay and probability as cost, schedules of reinforcement and, 73
Bedtime problems, 409–411. See also 123–126 Behavioral-systems analysis, 552–554,
Sleep problems effort as cost and, 116–120, 117f, 553f. See also Organizational
Behavior. See also Decreasing problem 119f behavior management (OBM)
behaviors; Disruptive behavior; emerging themes in, 126–127 Behavior-analytic intervention, 435–436
Increasing desirable behaviors factors that influence demand and, Behaviorism, 3, 11–12
applied behavior analysis and, 11 118–120 Bias
autism spectrum disorder (ASD) overview, 115–116, 127 generalized matching law and, 99,
and, 391 Behavioral gerontology. See also Aging 100–101
behavioral discrepancy and, 17, behavior-analytic research on aging observational procedures and, 137
19–20, 20t and, 513–516 Biofeedback, 414, 415–416
behavioral dysfunction, 29–31 clinical behavior-analytic research Blending, 435
behavioral pediatrics and, 409–411 on aging and, 516–523 Block-by-block agreement, 138–139.
change and, 25–27 elder abuse and, 567 See also Interobserver agreement
disruptive behavior, 306–307, 391 overview, 512–513, 523–524 Blocking design, 19–20, 20t
Functional Analysis Interview Behavioral management, 289–290, Brain systems and functioning, 23–25,
and, 198 535 23f, 24f, 31, 514–516
maintenance of, 32–33 Behavioral mass Brief punisher assessment, 360–361.
during mealtimes, 429, 434–436 behavioral momentum theory and, See also Assessment
natural environment and, 31–33 103–104 Bulimia nervosa, 428. See also Feeding
persistence of, 94, 101–109 schedules of reinforcement and, 59, disorders
respondent conditioning and, 59f, 73
15–17, 16f Behavioral measurement. See
schedules of reinforcement and, Measurement C
59–60 Behavioral momentum. See also
as subject matter for study, 3–5 Matching law; Momentum Caregiver reports, 179, 419
token-economy systems and, 374, escape behavior and, 312–313 Certification, professional. See
375 mitigating the persistence and Professional certification and
Behavior analysis. See also Applied relapse of problem behavior and, credentials for behavior analysts
behavior analysis (ABA) 107–109 Chained schedules of reinforcement,
as a natural science, 8 overview, 14, 94, 102–110 68t, 71–74, 72f, 246. See
overview, 3, 11–12 schedules of reinforcement and, 59, also Combined schedules of
professional certification and 59f, 72–74 reinforcement; Schedules of
credentials and, 570–576, 572f, Behavioral pediatrics reinforcement
573f, 575f applied behavior analysis and, 409 Chaining
Behavior Analyst Certification Board influence of behavioral variables complexity and, 253
(BACB) on medical problems, 414–416 differential reinforcement
ethical codes and, 564–566, overview, 408, 422 procedures and, 245–246
567–568 physiological variables in overview, 259–260
overview, 572–575, 572f, 573f, 575f behavioral problems, 411–414 Change
Behavior modification procedures, 381 routine behavior problems and, equivalence classes and, 87
Behavior penalty. See Response cost 409–411 persistence and, 103–105
Behavior support plans, 471 telehealth services and, 531–532 schedules of reinforcement and, 60
Behavioral assessment, 173–174, 468. treatment adherence and, 416–422 Change-over delay
See also Assessment; Functional Behavioral procedures, 286–290, 293 establishing control by a single
behavioral assessment Behavioral safety, 549–552. See stimulus and, 79
Behavioral chains, 253. See also also Organizational behavior generalized matching law and, 99–100
Chaining management (OBM) overview, 70
Subject Index 607
Changing-criterion design, 163. See response modality selection and, Conditioned response (CR). See also
also Single-case experimental 275–276 Response
designs teaching the mand, 276–277 behavioral momentum theory and,
Charts, behavioral, 377–378 teaching verbal behavior, 276–280 104
Chaser, 437 Communication-training procedures, behavior-analytic research on aging
Chewing, 430, 437–438. See also 396 and, 513–516
Feeding disorders Competency-based training, 287 conditioned reinforcement and, 22
Child Access Prevention Law, 479 Complexity. See also Behavior; overview, 16–17, 16f
Chin prompt, 436. See also Prompting Response phenomena associated with
Choice. See also Behavioral assessment and, 254–255 respondent conditioning and,
economics; Decision making; behavior and, 21–25, 24f 27–29, 27f
Preference chaining and, 259–260 unified conception of the
assessment of, 361–362 direct instruction and, 261–263 conditioning process and, 17–21,
autism spectrum disorder (ASD) generalization of complex skills 17f, 18f, 20t
and, 397, 398 and, 263–264 Conditioned stimulus (CS). See also
delay-discounting tasks to assess overview, 252–254, 264 Stimulus; Stimulus control
preference and, 123–124 promoting response generalization behavioral momentum theory and,
differential reinforcement and variety and, 260–261 104
procedures and, 244 prompt fading and, 257–259 characteristics of, 25–26
generalized matching law and, response and, 181 conditional stimulus control and,
98–101, 98f self-management and, 264 82–84
matching law and, 94, 95–96, 101–102 shaping and, 255–256 equivalence classes and, 87
as reinforcement, 182–183, 187 strategic instruction and, 261 overview, 16–17, 16f, 22
response allocation and, 95–96 Compliance with treatment. See also respondent conditioning and,
schedules of reinforcement and, Treatment adherence 27–29, 27f
58–59, 59f, 71, 72, 73–74 behavioral pediatrics and, 417 unified conception of the
selecting reinforcers and, 120–121 clinical behavior-analytic research conditioning process and, 17–21,
stimulus preference and, 123–124, on aging and, 517–518 17f, 18f, 20t
180, 187 Component analyses, 164–166, 166f, Conditioning. See also Operant
Cigarette smoking, 494, 503, 504, 506 167f contingencies; Respondent
Clarity, 136 Component-teaching procedure, 84 conditioning
Classical conditioning. See Comprehensive behavioral treatment behavior-analytic research on aging
Respondent conditioning for young children, 392. See and, 513–516
Closed economy, 118–119. See also also Early intensive behavioral overview, 13, 15–17, 16f
Behavioral economics intervention (EIBI) unified conception of, 17–21, 17f,
Coaching, 471 Concurrent chains in preference 18f, 20t
Cocaine, 495–496, 504–505 assessment, 180, 187 Confidentiality, 538–539
Cognitive functioning, 513–516 Concurrent schedules of Conjoint behavioral consultation,
Cognitive-constructivist approaches, reinforcement, 68t, 70–71, 181. 466–467. See also Behavioral
252–253 See also Combined schedules consultation; Consultation
Combined designs, 164–169, 165f, of reinforcement; Schedules of Conjoint schedules of reinforcement,
166f, 167f, 168f, 169f. See also reinforcement 68t, 70–71. See also Combined
Single-case experimental designs Concurrent validity, 194. See also schedules of reinforcement;
Combined schedules of reinforcement, Validity Schedules of reinforcement
66–74, 68t, 69f, 72f, 108 Concurrent-operant schedules, 181, Conjunctive schedules of reinforce
Commodities, 116. See also Behavioral 187 ment, 68t, 74. See also Combined
economics Conditional discrimination, 83–88 schedules of reinforcement;
Common Core State Standards Conditional probabilities, 208 Schedules of reinforcement
Initiative, 467 Conditioned punishers, 364 Consequence Variables Assessment
Communication. See also Functional- Conditioned reinforcement. See also Form, 197
communication training (FCT); Reinforcement Consequences. See also Outcomes of
Language; Speech; Verbal complex behavior and, 21–25, 24f a response
behavior; Vocalizations neural systems of, 23–25, 23f, 24f establishing control by a single
clinical behavior-analytic research overview, 22 stimulus and, 79–80
on aging and, 518 response–consequence feeding disorders and, 436–437
Functional Analysis Interview contingencies, 38 Functional Analysis Interview
and, 198 role of in dysfunctional behavior, and, 198
learning/instructional hierarchy 29–31 function-based extinction
model and, 447–448 token-economy systems and, 376 procedures and, 331
608 Subject Index
Demand functions, 116–120, 117f. See education and, 445 function-based reinforcement
also Behavioral economics equivalence classes and, 86–87 procedures and, 342
Dependent-group contingencies, 247 features of behavior to target with, mitigating the persistence and
Depression, 516–517 238–240 relapse of problem behavior and,
Deprivation, 185 feeding disorders and, 434–435, 436 107–109
Descriptive analysis. See also Direct group contingencies and, 247 overview, 61t–62t, 65, 66
observation overview, 237–238, 247 quantitative law of effect and,
A-B-C recording and, 206–207 procedures to complement, 97–98
choosing a procedure for, 203 244–246 weakening undesirable behavior
data analysis and, 208–209 prompting and, 256 and, 123
developing hypotheses about response maintenance and, 246 Differential stimulus control, 81–82.
behavioral function and, schedule thinning and, 246 See also Stimulus control
203–204 schedules of reinforcement and, 60 Direct instruction, 261–264, 453–454,
future research, 210 shaping of operant classes and, 455. See also Instruction
overview, 202–203, 209–210 50–51 Direct observation. See also
scatterplots and, 205–206 Differential reinforcement of Observational procedures
Descriptive ethics, 561, 566–567. See alternative behavior (DRA) applied behavior analysis and, 11
also Ethical factors behavioral momentum theory and, choosing a procedure for, 203
Desirable behaviors, increasing. See 104–105, 106–107 data analysis and, 208–209
Increasing desirable behaviors choice and persistence and, 102 developing hypotheses about
Detailed Behavior Report prompts, feeding disorders and, 433–434 behavioral function and,
206 mitigating the persistence and 203–204
Determinism, 8–9, 562, 567. See relapse of problem behavior and, future research, 210
also Experimental analysis of 107–109 overview, 140, 142–147, 145f, 146f,
behavior overview, 61t, 65 202–203, 209–210
Developmental disabilities, 523–524 quantitative law of effect and, variations in measurement during,
Diagnosis 97–98 204–207
autism spectrum disorder (ASD) selecting procedures for Discontinuous measurements,
and, 389 punishment and, 358 142, 144–147, 145f, 146f.
clinical behavior-analytic research topography and, 238 See also Interval recording;
on aging and, 517 Differential reinforcement of high- or Measurement; Momentary time
encopresis and, 411 low-rate behavior (DRH–DRL) sampling
feeding disorders, 428, 439 behavioral momentum theory and, Discordant validity, 194. See also
nocturnal enuresis, 414 103 Validity
telehealth services and, 530f, 532, choice and persistence and, Discounting rate, 124
534 101–102 Discrete-trial training
Diagnostic and Statistical Manual of overview, 61t autism spectrum disorder (ASD)
Mental Disorders, fifth edition rate and, 238–239 and, 396–397, 402
(DSM-5) Differential reinforcement of other education and, 446
autism spectrum disorder (ASD) behavior (DRO) learning/instructional hierarchy
and, 389 combined schedules of model and, 447–448
encopresis and, 411 reinforcement and, 74 teacher consultation and, 471
feeding disorders, 428 duration and, 239 Discriminated avoidance, 47. See also
nocturnal enuresis, 414 function-based extinction Avoidance
Diarrhea, 412 procedures and, 322 Discriminated operant, 58, 59, 59f
Didactic training, 535 overview, 61t Discrimination training, 364
Dietary interventions, 517–518 rate and, 238 Discriminative control, 364, 366
Differential associability, 26 response allocation and, 95 Discriminative effects
Differential consequences, 79. See also weakening undesirable behavior conditional stimulus control and,
Consequences and, 123 82–84
Differential contingencies, 79. See also Differential reinforcement differential stimulus control and,
Contingencies schedules. See also Differential 81–82
Differential preferences, 179. See also reinforcement; Schedules of equivalence classes and, 84–88
Preference reinforcement function-based extinction
Differential reinforcement. See also behavioral momentum theory and, procedures and, 332
Differential reinforcement 103, 104–105, 106–107 overview, 89
schedules; Reinforcement choice and persistence and, punishment and, 45–46
considerations for, 240–243 101–102 successive discrimination and,
diverse responding and, 246–247 feeding disorders and, 433–434 80–81
610 Subject Index
Discriminative stimulus. See also Eating problems. See Feeding disorders effectiveness of reinforcement and,
Stimulus Echoic functional relations 184–185
autism spectrum disorder (ASD) applied-verbal behavior approach enriched environment, 303–304
and, 393 with ASD and, 395–396 experimentation and, 9
behavioral momentum and, 105 overview, 271–272 functional analysis in the natural
conditional stimulus control and, teaching verbal behavior and, environment, 226–227, 227f
82–84 277–278 Functional Analysis Interview
differential stimulus control and, Ecological events, 198 and, 198
81–82 Ecological validity, 186 function-based extinction
function-based reinforcement Economics, behavioral. See Behavioral procedures and, 332
procedures and, 345–347 economics mitigating the persistence and
function-based treatments and, Eddie Eagle GunSafe Program, relapse of problem behavior and,
337–338 481–482, 483–485, 486–487 108–109
mitigating the persistence and Education, behavioral approaches to. observational procedures and,
relapse of problem behavior See also Instruction; Teacher 157–158, 158f
and, 108 consultation overview, 6–7, 18–19
overview, 18–19, 89 generative-response repertoires punishment and, 45
procedures for clarifying results from and, 450–452 respondent conditioning and,
a functional analysis, 223–224 health education, 420 31–33
schedules of reinforcement and, 58 instructional programs and, response–consequence
Disruptive behavior. See also Behavior 452–455 contingencies, 37–39
autism spectrum disorder (ASD) learning/instructional hierarchy staff training and management
and, 391 model and, 387, 446–450 and, 292
noncontingent reinforcement overview, 445–446, 458 Equivalence, 84–88
(NCR) and, 306–307 response-to-intervention models, Error correction, 242
Distribution of reinforcement, 184. See 455–458 Errorless teaching, 276
also Reinforcement teaching strategies, 446–450 Escape
Diurnal enuresis, 414. See also Enuresis telehealth services and, 535 antecedent interventions and,
Divided attention, 216 Elastic demand, 117, 117f 311–314
Doing, 445 Elder abuse, 567 feeding disorders and, 432, 433,
Domain effect, 124–125 Electric shock. See Shock as a 434, 435
Dopamine system, 24, 25–26 punisher functional analyses and, 216
Drop-out analysis, 164 Electronic Children’s Hospital of the function-based extinction
Drug addiction. See Addiction Pacific, 531–532 procedures and, 323–324
Drug assays, 418 Electronic monitors, 418–419 function-based reinforcement
Dual-discrepancy approach, 456 Eliciting effects, 15–16, 16f, 45–46, 47. procedures and, 343
Dualist view of intelligent behavior, 4 See also Stimulus as an indirect effect of punishment,
Duration, 142–143, 156, 178–179, 187, Elopement, 47 366
239 Embedded reinforcement, 186. See also negative reinforcement and,
Dynamic fading, 471. See also Fading Reinforcement 46–47
Dynamic Indicators of Basic Early Emotional behavior, 327–328, 366 response restriction and, 305
Literacy Skills (DIBELS), 457–458 Empiricism, 8, 9. See also using punishment effectively and,
Dynamic-probability analyses, 208–209 Experimental analysis of 365
behavior Establishing event, 38, 49. See also
Employment-based services. See Motivational factors
E Organizational behavior Establishing operations, 272, 331
management (OBM) Ethical factors
Early intensive behavioral Encopresis, 411–414 descriptive ethics, 561, 566–567
intervention (EIBI) Engagement, 378–379, 380 discourse and, 562–564
autism spectrum disorder (ASD) Engagement-based reinforcement ethical codes, 564–566, 568
and, 391–397, 403 assessment, 519–520 meta-ethics, 561–564
features of, 396–397 Enriched environment, 303–304 normative ethics, 561, 564–566
Early-intervention models. See also Enuresis, 414–416. See also Urinary overview, 561, 567–568
Interventions incontinence professional certification and
autism spectrum disorder (ASD) Environmental factors credentials and, 573–574
and, 391–397 assessment and, 255 punishment and, 45
combining, 397 behavioral momentum theory and, telehealth services and, 538–539
curricular manuals based on ABA 102–103 Ethics Code (APA, 2002), 564–566,
principles for ASD, 400 direct observation and, 142 568
Subject Index 611
Evaluation. See also Assessment reinforcement and, 42–43 Fixed-ratio (FR) schedule of
encopresis and, 412–413 response blocking and, 325–326 reinforcement. See also
feeding disorders and, 431–432 response patterns associated with, Schedules of reinforcement
nocturnal enuresis and, 415 326–328 behavioral economics and, 116–117,
teacher consultation and, 463 schedules of reinforcement and, 61t, 118
Event recording, 156, 204 64, 246 behavioral momentum and, 105
Evolutionary processes, 50–51, 54–55 selecting procedures for clinical behavior-analytic research
Exact agreement, 138. See also punishment and, 358 on aging and, 519
Interobserver agreement shaping and, 255–256 differential reinforcement
Exchange rate, 376–377 side effects of, 42–43 procedures and, 241
Exclusion, 84, 356. See also Time out using punishment effectively and, function-based reinforcement
Executive control, 254 362, 364 procedures and, 338
Exercise, antecedent, 302–303 Extinction schedules of reinforcement, law of demand and, 116–117
Expectancy bias, 137 61t, 64. See also Schedules of overview, 60, 61t, 62
Experimental analysis of behavior. See reinforcement schedule thinning and, 246
also Determinism; Empiricism; Extinction-induced emotional selecting reinforcers and, 120–122,
Experimentation; Parsimony; behavior, 327–328 121f
Philosophical doubt; Reliability Extrinsic reinforcers, 52–53 using punishment effectively and,
compared to applied behavior 363
analysis, 10 Fixed-time (FT) schedules of
dimensions of, 8–10 F reinforcement. See also
overview, 3, 11–12 Schedules of reinforcement
Experimental design, 19–20, 20t, Fading combined schedules of
133, 222. See also Single-case differential reinforcement schedules reinforcement and, 73
experimental designs and, 66 differential reinforcement schedules
Experimentation, 8, 9. See also education and, 445 and, 65–66
Experimental analysis of escape behavior and, 312 noncontingent reinforcement
behavior feeding disorders and, 433, 435 (NCR) and, 308–309, 340–341
Expressive language, 397. See also instructional fading, 346 overview, 61t, 64–66
Language promoting response generalization Flipped spoon, 437
Expulsion, 429, 436. See also Feeding and variety and, 260 Fluency, 448–452
disorders prompt fading, 257–259, 401 Follow-up
Extension for Community Healthcare successive discrimination and, 80–81 safety skills and, 483
Outcomes project (Project teacher consultation and, 471 teacher consultation and, 470–471
ECHO), 532 token-economy systems and, 379–380 token-economy systems and,
External validity, 161–162. See also using punishment effectively and, 380–381
Validity 365 Food refusal
Exteroceptive stimuli, 25. See also Fecal retention. See Encopresis function-based extinction
Auditory stimuli; Conditioned Feeding disorders procedures and, 324
stimulus (CS); Tactile stimuli; autism spectrum disorder (ASD) noncontingent reinforcement
Visual stimuli and, 390 (NCR) and, 306
Extinction. See also Function-based dependent variables, 428–430 Forward chaining, 245–246, 259, 260.
extinction procedures etiology of, 430–431 See also Chaining
bedtime problems and, 410 evaluation and, 431–432 Free-operant preference assessment
behavioral momentum and, 105 overview, 387, 427–428, 438–439 clinical behavior-analytic research
contingencies and, 49 treatment and, 432–438 on aging and, 520
escape behavior and, 312 Firearm safety. See also Safety skills evaluating reinforcement effects
extinction burst and, 326, 409–410 firearm injuries and deaths, and, 181
factors that influence the efficacy 476–479 overview, 178, 187
of, 328–331 future research, 486–487 Frequency recording, 142, 207. See
feeding disorders and, 433, 435 overview, 388, 476, 484–485, 485t also Continuous measurements;
functional variations of, 322–326 Fixed-interval (FI) schedule of Direct observation
noncontingent reinforcement reinforcement. See also Function. See also Functional analyses
(NCR) and, 309 Schedules of reinforcement developing hypotheses about
overview, 28 differential reinforcement behavioral function and,
practical considerations in the procedures and, 241 203–204
implementation of, 331–332 overview, 60, 61t, 63 overview, 15, 214–215
promoting response generalization using punishment effectively and, private events and, 5–6
and variety and, 260 363 Functional alternative behavior, 198
612 Subject Index
Good Productivity Game, 247 Immediacy of effects, 159, 240–241 Instruction. See also Education,
Graduated extinction, 410. See also Implementation Planning for Logistics behavioral approaches to;
Extinction and Barriers approach, 469 Learning; Teacher consultation
Graduated guidance, 258. See also Imprinting, 39 autism spectrum disorder (ASD)
Fading Impulsivity and, 400–402
Graphing data choice and, 72 complexity and, 264
scatterplots and, 205–206 delay discounting and, 125 direct instruction and, 261–263
token-economy systems and, schedules of reinforcement and, 72 generalization of complex skills
377–378 In situ assessments, 481, 483, 485t. See and, 263–264
visual inspection of graphed data also Assessment generative-response repertoires
and, 158–161, 159f, 160f In situ training, 482–484, 485, 485t and, 450–452
Group arrangement in preference In vivo coaching, 535 instructional programs based on
assessment, 180 In vivo didactic training, 535 behavioral analysis principles,
Group comparison methods in Incentives. See also Reinforcers; 452–455
research, 6 Reward learning/instructional hierarchy
Group contingencies improving treatment adherence model and, 387, 446–450
abstinence reinforcement with, 421 overview, 445–446
procedures and, 498, 502–503 organization behavior management response-to-intervention models,
differential reinforcement (OBM) and, 548, 551 455–458
procedures and, 247 Incidental teaching, 393–394 safety skills and, 481–484, 485t
organization behavior management Increasing desirable behaviors. See self-management and, 264
(OBM) and, 548 also Behavior; Interventions; strategic instruction and, 261
token-economy systems and, 380 Reinforcement teaching the mand, 276–277
Gun safety. See Firearm safety assessment and, 254–255 verbal behavior and, 276–280
Gustatory stimuli, 79. See also Taste chaining and, 259–260 Instructional fading, 346. See also
stimuli direct instruction and, 261–263 Fading
overview, 235–236 Instructional planning, 254–255
prompting and, 256–259 Instruction-following task, 89
H shaping and, 255–256 Instrumentation effects, 162
strategic instruction and, 261 Integrated model of school
Habituation, 365 teaching verbal behavior, 276–280 consultation, 465–466. See
Health education, 420 Indifference point, 124 also Consultation; Teacher
Health Insurance Portability and Indirect behavioral assessments. See consultation
Accountability Act (HIPAA), also Assessment Integrity, treatment. See Treatment
538 benefits and limitations of, 193–194 integrity
Health maintenance, 516–518. See interviews, 197–199 Intellectual developmental disorder,
also Medical conditions; Physical overview, 193, 199 183–184
health issues rating scales, 194–196 Intensity, 143–144, 239
Helping Early Literacy with Practice Indirect measures, 11, 140–142. See Interdependent-group contingencies,
Strategies (HELPS), 453, 455 also Measurement 247
Higher-order classes, 51–54. See also Indiscriminable contingencies. See Interim link, 71
Response classes also Contingencies Intermittent-reinforcement
High-probability (high-p) sequences autism spectrum disorder (ASD) schedule. See also Schedules of
escape behavior and, 312–313 and, 402 reinforcement
feeding disorders and, 433–434 token-economy systems and, 379 function-based extinction
History effects, 162, 198–199 Individual differences, 6, 278–279 procedures and, 329
Hydration Interview, 513 Individuals with Disability Education token-economy systems and, 379
Hypothetical purchase tasks, 126 Improvement Act of 2004, using punishment effectively and,
455–456, 463 363
Industrial–organizational psychology. Internal validity, 161, 162. See also
I See Organizational behavior Validity
management (OBM) Internet-based services. See Telehealth
Identity-matching experiments, 83 Information technologies, 292. services
Ignore condition See also Technology as Interobserver agreement, 138–139,
functional analyses and, 216–217 reinforcement 160. See also Observational
procedures for clarifying results Informational-training approach, procedures
from a functional analysis, 223, 481–482 Interoceptive stimuli, 25, 28. See also
224f Informed consent, 539 Conditioned stimulus (CS)
Imitation, 89, 257 Initial link, 71, 73–74 Interrater agreement, 160
614 Subject Index
indirect versus direct measurement, Momentary differential reinforcement Multiple-baseline design, 162–163. See
140–147, 145f, 146f of other behavior, 79. See also also Single-case experimental
measurement error, 135 Change-over delay designs
overview, 133, 135–136, 148, Momentary time sampling, 142, Multiple-exemplar training, 260
149f–150f 145f, 146–147, 146f, 180. Multiple-stimulus assessment, 177, 187
variations in during direct See also Direct observation; Multiple-stimulus-without-
observation, 204–207 Discontinuous measurements; replacement assessment,
visual inspection of graphed data Time-sampling conventions 177–178, 187, 433
and, 158–161, 159f, 160f Momentum. See also Behavioral Multiple-treatment interference, 162
Mechanical recording, 139–140 momentum Multiply controlled problem behavior,
Mechanical restraint, 305 escape behavior and, 312–313 220, 221f
Media, 127 persistence and, 103–105
Medical conditions. See also Health versus response rate, 42
maintenance; Pediatric schedules of reinforcement and, N
medicine; Physical health issues; 72–74
Physiological variables Monitoring, 420 Natural environment training,
clinical behavior-analytic research Morality, 561–564 392–393, 396–397
on aging and, 516–518 Morningside Model of Generative Natural language approaches,
feeding disorders and, 430–431, Instruction, 454, 455 392–393, 394–395
438–439 Most-to-least fading, 258–259. See also Natural selection, 6–7
Medications Fading Naturalism, 563–564
behavioral gerontology and, 524 Motivating operations Naturalistic observations, 157. See also
combining with abstinence function-based treatments and, Observational procedures
reinforcement procedures, 332, 338 Naturalistic-teaching strategies,
501–502 overview, 301–302 392–393
nocturnal enuresis and, 415 primary verbal operants and, 272 Negative punishment, 50, 352,
token-economy systems and, 381 schedules of reinforcement and, 58 356–357. See also Punishment
Memory, 518 token-economy systems and, 381 Negative reinforcement. See also
Mental health, 408, 465–466, Motivation Assessment Scale, Punishment; Reinforcement
516–518. See also Behavioral 195–196 differential reinforcement
pediatrics Motivational factors procedures and, 242
Metabolism, 520–521 considerations in stimulus distinguishing from positive
Meta-ethics, 561–564. See also Ethical selection, 186 reinforcement, 48
factors private events and, 5–6 establishing events in, 49
Metaphorical extensions, 272 procedures for clarifying results extinction and, 49
Methamphetamines, 495–496 from a functional analysis, 224 feeding disorders and, 435
Methodological features and issues, response–consequence functional analyses and, 216
133 contingencies, 38 function-based extinction
Metonymic extensions, 272 self-injurious behavior (SIB) and, procedures and, 322–324
Microeconomic theory, 115–116, 127. 215 function-based reinforcement
See also Behavioral economics Mouth clean, 429–430, 435, 436–437. procedures and, 341–343
Minimal-response repertoire, 452 See also Feeding disorders meta-ethics and, 563
Mixed nocturnal and diurnal enuresis, Multicomponent interventions, 289 negative punishment, 50
414. See also Enuresis Multielement design, 163–164. See overview, 46–49
Mixed schedules of reinforcement, also Single-case experimental reinforcers in, 48–49
67, 68t, 69–72, 69f, 346–347. designs Neural systems
See also Combined schedules Multiple control, 273–274 behavior-analytic research on aging
of reinforcement; Schedules of Multiple incident-teaching sessions, and, 514–516
reinforcement 394 unconditioned and conditioned
Modeling. See also Prompting Multiple schedules of reinforcement. reinforcement and, 23–25, 23f,
generalization of complex skills See also Combined schedules 24f, 31
and, 263–264 of reinforcement; Schedules of Neurocognitive disorders, 516–518,
overview, 257 reinforcement 519, 520, 522–523
safety skills and, 485t function-based extinction Neurodevelopmental disorders, 121–
staff training and management procedures and, 330–331 122. See also Autism spectrum
and, 286 mitigating the persistence and disorder (ASD)
Model–lead–test–delayed-test relapse of problem behavior Neurological disorders, 304
procedure, 263 and, 108 Neutralizing routines, 314
Modified chin prompt, 436. See also overview, 67, 68t, 69–70, 69f No Child Left Behind Act of 2001,
Prompting schedule thinning and, 246 455–456, 458
616 Subject Index
No-choice condition, 182–183. See Occurrence agreement, 139. See also history of, 545
also Choice Interobserver agreement overview, 285–286, 388, 544–545,
Nocturnal enuresis, 414–416 Occurrence of behavior, 204, 207 555
No-cure criticism, 381 Olfactory stimuli, 79. See also Smell performance management and,
Noise as a punisher, 355 stimuli 545–549
Nonadherence to treatment, 417. See Omission errors, 105–106 related disciplines, 544–545
also Treatment adherence Online services. See Telehealth Outcomes of a response, 5–6. See also
Noncontingent reinforcement services Consequences
(NCR). See also Function-based Ontogenetic variables Overcorrection
reinforcement procedures; determinism and, 8–9 antecedent exercise and, 302
Reinforcement overview, 7 overview, 354–355
behavioral momentum theory and, reinforcement and cultural selecting procedures for
104–105, 106–107 selection and, 54–55 punishment and, 360t
choice and persistence and, 102 Open Court Reading program, 453 Overjustification effect, 186
differential reinforcement Open economy, 118–119. See also Overlapping data points, 159–160,
procedures and, 243 Behavioral economics 160f
feeding disorders and, 434 Operant behavior, 4, 94. See also Overlearning, 263
mitigating the persistence and Behavior Overmatching, 99
relapse of problem behavior and, Operant behavioral economics. See
107–109 Behavioral economics
overview, 301–302, 306–310, 314, Operant conditioning P
340–341 abstinence reinforcement
quantitative law of effect and, 98 procedures and, 493 Package interventions, 548–549
single-case experimental design behavior-analytic research on aging Packing, 436–437
and, 169 and, 515–516 Pair-coded neurons, 31
social negative reinforcement and, early intensive behavioral interven Paired-choice preference assessment,
342–343 tions (EIBI) and, 391–392 176–177, 176f, 187
social positive reinforcement and, 339 education and, 445 Pairing procedures, 185–186
time schedules of reinforcement feeding disorders and, 432–433 Panic disorder, 26
and, 64 overview, 16–17 Paradoxical diarrhea, 412
Nonexclusionary time out, 356. See Operant contingencies. See also Parametric analyses, 164, 166–169,
also Time out Conditioning; Contingencies; 168f, 169f
Nonoccurrence agreement, 139. See Punishment; Reinforcement Parent training, 530f, 535
also Interobserver agreement determinism and, 8–9 Parental behaviors, 477–480
Nonoccurrence of behavior, 204 higher-order classes and, 51–53 Parsimony, 8, 10. See also
Nonsocial reinforcement, 242 negative punishment, 50 Experimental analysis of
Normative ethics, 561, 564–566. See negative reinforcement, 46–49 behavior
also Ethical factors overview, 13, 37 Partial-interval recording, 143,
Norm-referenced approaches, 449 reinforcement and cultural 144–146, 145f, 146f, 204–205.
Nucleus accumbens, 24–25 selection and, 54–55 See also Direct observation;
reinforcer classes and reinforcer- Discontinuous measurements;
specific effects and, 53–54 Interval recording
O response–consequence Partial-reinforcement extinction
contingencies, 37–39 effect, 328–329
Observational procedures. See shaping of operant classes and, 50–51 Passive avoidance, 46. See also
also Assessment; Data Operant mechanisms, 337–338 Avoidance
collection; Direct observation; Operant psychology, 115, 127. See also Patient reports, 419
Measurement Behavioral economics Pavlov, Ivan, 15–16
behavioral pediatrics and, 418 Operant response, 87, 103 Pavlovian conditioning. See also
humans and machines as observers Operational definitions, 133, 136, 156 Respondent conditioning
and, 136–140 Opioids, 495 behavioral momentum theory and,
organization behavior management Oral reading fluency. See Fluency 103–105
(OBM) and, 551 Oral support, 433 matching law and behavioral
overview, 11, 133, 148, 149f–150f Oral–motor dysfunction, 431 momentum theory and, 94
setting effects, 157–158, 158f Organizational behavior management mitigating the persistence and
Observer bias, 137. See also Bias; (OBM). See also Staff training relapse of problem behavior
Observational procedures and management and, 108
Observer drift, 137. See also behavioral safety, 549–552 Paying attention, 21
Observational procedures behavioral-systems analysis and, Pediatric feeding disorders. See
Occam’s razor, 10 552–554, 553f Feeding disorders
Subject Index 617
Pediatric medicine, 387, 408. See also Pictorial representations in preference delay-discounting tasks to assess,
Behavioral pediatrics; Medical assessment, 179–180, 187 123–124
conditions Picture Exchange Communication evaluating reinforcement effects
Percentile schedules, 245 System (PECS), 275 and, 180–181
Performance Diagnostic Checklist, Pill counts, 419 factors that may influence the
546–547, 554 Pivotal-response training effectiveness of reinforcement
Performance Diagnostic Checklist— autism spectrum disorder (ASD) and, 184–185
Human Services, 546–547 and, 393, 395 feeding disorders and, 433
Performance Diagnostic Checklist— teaching verbal behavior and, 278 methods to identify preferred
Safety, 550 Pleasant Events Schedule— stimuli and, 175–180, 176f
Performance feedback, 470, 471 Alzheimer’s Disease, 519 overview, 186–187
Performance level, 456 Policy preference reversal, 123
Performance management, 545–549. behavioral economics and, 127 selecting procedures for
See also Organizational behavior firearm injuries and deaths and, punishment and, 361–362
management (OBM) 479 selecting reinforcers and, 120–122,
Performance-based training, 287. See response-to-intervention models 121f
also Training and, 455–456 stability of, 184–185
Permanent-product recording, telehealth services and, 531 teacher consultation and, 468
140–141, 156. See also Indirect Polydrug use, 498–499 teaching new preferences, 185–186
measures; Measurement Positive psychology, 43 Prefrontal cortex, 23–24, 23f, 24f
Persistence of behavior. See also Positive punishment, 353–356. See also Presentation, 49
Behavior Punishment Pretest sensitization, 162
behavioral momentum theory and, Positive reinforcement. See also Prevention
102–109 Reinforcement abstinence reinforcement
matching law and, 94, 101–102 differential reinforcement procedures and, 492, 503–504
mitigating, 107–109 procedures and, 242 firearm injuries and deaths and,
Personal restraint, 305 distinguishing from negative 479–484
Pharmacotherapy, 501–502. See also reinforcement, 48 response-to-intervention models
Medications feeding disorders and, 432–433, 434 and, 456–457
Phasing out, 379–380. See also Fading functional analyses and, 215–216 teachers and, 463
Philosophical doubt, 8, 9, 10. See function-based reinforcement Price elasticity of demand, 117
also Experimental analysis of procedures and, 342 Primary care. See Behavioral
behavior methods to identify preferred pediatrics; Pediatric medicine
Phonemes, 452 stimuli and, 175–180, 176f Private events, 5–6
Phonics instruction. See Education, overview, 43, 46, 175, 322 Probability, 123–126
behavioral approaches to; time out from, 354 Probability analyses, 208–209. See
Instruction Positive routines, 410 also Data analysis; Statistical
Phonological awareness, 452. See Positive-practice overcorrection, 354. analyses
also Education, behavioral See also Overcorrection Probability discounting, 124
approaches to Postponement, 47 Problem analysis
Phylogenetic variables Postreinforcement pauses, 60. See also consultation model and, 465
determinism and, 8–9 Fixed-ratio (FR) schedule of response-to-intervention models
overview, 6–7 reinforcement and, 457–458
reinforcement and cultural Praise Problem behavior. See Behavior
selection and, 54–55 clinical behavior-analytic research Problem Behavior Questionnaire,
Physical dependence, 32–33 on aging and, 518 194–195
Physical health issues, 517–518. See maintenance of behavior and, 33 Problem evaluation
also Health maintenance; safety skills and, 481, 485t consultation model and, 465
Medical conditions token-economy systems and, 376 response-to-intervention models
Physical restraint. See Restraint Precursor functional analysis, 227. See and, 458
Physiological variables. See also also Functional analyses Problem identification, 465
Medical conditions Preference. See also Choice; Stimulus Problem-solving consultation model,
behavioral problems and, 411–414 preference 465. See also Consultation
feeding disorders and, 430–431 autism spectrum disorder (ASD) Problem-solving model, 457–458, 465
influence of behavioral variables and, 398 Professional certification and
on medical problems and, behavioral gerontology and, 513 credentials for behavior analysts
414–416 clinical behavior-analytic research Behavior Analyst Certification
nocturnal enuresis and, 414–415 on aging and, 519–521 Board (BACB) and, 572–575,
Pica, 306, 308, 355 considerations in stimulus 572f, 573f, 575f
PIC/NIC analysis, 546 selection, 185–186 credentialing programs, 571
618 Subject Index
methods to identify preferred feeding disorders and, 432 implications for applied behavior
stimuli and, 175–180, 176f Functional Analysis Interview and, analysis and, 29–33
mitigating the persistence and 197–199 phenomena associated with, 27–29,
relapse of problem behavior and, function-based extinction 27f
107–109 procedures and, 329 Response. See also Conditioned
natural environment and, 31–33 in negative reinforcement, 48–49 response (CR); Unconditioned
negative reinforcement, 46–49 reinforcer classes and reinforcer- response (UR)
overview, 5, 13, 15, 39–44, 237, 299 specific effects and, 53–54 behavior-analytic research on aging
procedures for clarifying results selecting, 120–122, 121f, 197–199 and, 513–516
from a functional analysis, 224 specificity of, 40 differential reinforcement
quantitative law of effect and, token-economy systems and, procedures and, 246–247
96–98, 97f 375–376 evaluating reinforcement effects
reinforcement and cultural types of, 242–243 and, 181
selection and, 54–55 Reinforcing consequence, 338 generalized matching law and,
response allocation and, 95–96 Reinforcing stimulus. See also 99–100
response–consequence Reinforcement; Reinforcers generative-response repertoires
contingencies, 37–39 autism spectrum disorder (ASD) and, 450–452
shaping and, 50–51, 255–256 and, 397–398 patterns of, 224, 326–328
side effects of, 42–43 behavioral discrepancy and, 19–20, promoting generalization and
strict matching law, 96, 96f 20t variety and, 260–261
teaching verbal behavior and, Rejecting behaviors, 276–277. See also reinforcer classes and reinforcer-
278–279 Communication specific effects and, 53–54
token-economy systems and, 376 Relapse respondent conditioning and,
unified principle of, 20–21 abstinence reinforcement 15–17, 16f
using punishment effectively and, procedures and, 492, 503–504 response allocation and, 95–96
362, 363–364, 365 behavioral momentum and, response restriction, 304–305
verbal behavior and, 52–53 105–107 response variation, 326–327
weakening undesirable behavior mitigating, 107–109 response–consequence
and, 122–123 overview, 507 contingencies, 37–39
Reinforcement-based treatments. See relapse prevention, 492, 503–504 selecting reinforcers and, 120–121
Decreasing problem behaviors; using punishment effectively and, shaping of operant classes and,
Function-based reinforcement 365 50–51
procedures; Interventions Relational-frame theory, 5 teaching verbal behavior and, 278
Reinforcer Assessment for Individuals Reliability, 8, 9–10, 135, 194. See Response allocation, 95–96
with Disabilities (RAISD), 179 also Experimental analysis of Response blocking
Reinforcer choice, 244. See also behavior; Measurement automatically reinforced behaviors
Choice Reminders, 420–421 and, 325–326
Reinforcer immediacy or contiguity, Remote services. See Telehealth overview, 354
240–241 services selecting procedures for
Reinforcer magnitude, 242 Reprimands, 358, 360t. See also Verbal punishment and, 360t
Reinforcer pathology model, 126–127 reprimands Response chaining, 245–246. See also
Reinforcer quality, 184, 242 Requesting behavior, 276–277. See also Chaining
Reinforcer relations, 103 Communication Response classes, 51–53
Reinforcer variation, 243–244 Resistance to change. See also Change Response cost. See also Negative
Reinforcers. See also Reinforcement; behavioral momentum theory and, punishment
Reinforcing stimulus; Reward 103–104 overview, 352, 356–357
abstinence reinforcement equivalence classes and, 87 selecting procedures for
procedures and, 492–493 schedules of reinforcement and, 60 punishment and, 357, 360t
assessing, 40–41, 116–118, 119f, Resistance to extinction, 328–329. See time out and, 356
120–123, 121f also Extinction token-economy systems and, 376,
versus bribes, 53 Resititutional overcorrection, 354. See 378, 382
choice as reinforcement, 182–183 also Overcorrection Response effort, 240
delay discounting and, 125–126 Respondent (classical) conditioning, 13 Response generalization, 253. See also
delivery of in a token economy, Respondent conditioning. See also Generalization
376–377 Conditioning Response interruption and
evaluating reinforcement effects behavioral change and, 25–27 redirection, 355
and, 180–182 behavior-analytic research on aging Response periodicity, 59–60
factors that may influence the and, 513–516 Response prompts, 256, 258–259. See
effectiveness of, 184–185 historical context of, 15–17, 16f also Prompting
620 Subject Index
Self-nomination preference 83. See also Discriminative Social positive reinforcement. See
assessment, 179 effects also Positive reinforcement;
Self-recording, 261 Simultaneous presentation, 433–434 Reinforcement; Social
Self-Regulated Strategy Development Simultaneous prompting, 259. See also reinforcement
(SRSD), 261 Prompting functional analyses and, 215–216,
Self-regulation, 44, 261, 264 Single-case experimental designs. See 217–218
Self-reinforcement, 43–44. See also also Experimental design functional variations of extinction
Reinforcement data collection and, 156 and, 322
Self-report interpreting functional-analysis function-based reinforcement
applied behavior analysis and, 11 data and, 222 procedures and, 338–341
behavioral pediatrics and, 419 overview, 155–156, 170 Social reinforcement. See also
private events and, 6 response-to-intervention models Reinforcement
safety skills and, 480, 483 and, 457–458 abstinence reinforcement
Self-stimulation behaviors setting effects, 157–158, 158f procedures and, 502–503
contingent delivery of aversive types of, 161–169, 165f, 166f, 167f, differential reinforcement
tastes and smells and, 355 168f, 169f procedures and, 240–241, 242
function-based extinction visual inspection of graphed data functional variations of extinction
procedures and, 324–325 and, 158–161, 159f, 160f and, 322–326
Sensitivity Singled-reinforcement, 330–331 function-based reinforcement
behavioral momentum and, 105 Single-function problem behavior, procedures and, 337, 341–343
generalized matching law and, 219, 220f noncontingent reinforcement
99–100 Single-stimulus preference assessment, (NCR) and, 306–307
Sensory extinction approach, 325 175–176, 187. See also teaching verbal behavior and, 279
Sensory-integration intervention, Assessment; Preference Social skills, 447–448
435–436 Skills training. See also Behavioral Social stimuli, 183–184. See also
Sensory-reinforced behavior, skills training; Training Reinforcers; Stimulus
324–325 educational applications of, Socially mediated reinforcement,
Sequence effects, 163 445–446 322–326
Sequential presentation, 434 generative-response repertoires Sound Partners program, 453
Sequential-sensory approach, 436 and, 450–452 Speaker’s behavior, 271. See also
Setting events, 198, 314 safety skills and, 482–484 Verbal behavior
Severity levels of ASD, 389. See staff training and management Speech, 275–276. See also
also Autism spectrum disorder and, 286–287, 468–469 Communication; Verbal
(ASD) teacher consultation and, 468–469 behavior
Sexual abuse, 412 telehealth services and, 530f, 535 Spontaneous recovery, 32, 328
Shaping Sleep Assessment and Treatment Stability, 158–159, 184–185. See also
autism spectrum disorder (ASD) Tool, 199 Preference
and, 394 Sleep problems Staff training and management. See
autoshaping procedures, 26 autism spectrum disorder (ASD) also Professional certification
differential reinforcement and, 390 and credentials for behavior
procedures and, 66, 245 behavioral pediatrics and, analysts; Training
of operant classes, 50–51 409–411 Behavior Analyst Certification
overview, 255–256 nocturnal enuresis, 414–416 Board (BACB) and, 572–575,
prompt fading, 257–258 Small-group instruction, 453–454. See 572f, 573f, 575f
Shock as a punisher also Instruction continuing education (CE)
overview, 355–356 Smell stimuli requirements and, 571, 573, 574
selecting procedures for classical conditioning and, 26 current technology of, 286–290
punishment and, 360t clinical behavior-analytic research future research, 290–294
using punishment effectively and, on aging and, 520–521 gaps in technology for, 290–294
365 as a punisher, 355, 360t overview, 285–286
Sidman avoidance, 47. See also Smoking cessation, 494, 503, 504, professional certification and
Avoidance 506 credentials and, 570–576, 572f,
Sign language, 275–276, 278 Social disapproval, 215–216 573f, 575f
Signaled avoidance, 47. See also Social negative reinforcement. See teachers and, 455, 468–469
Avoidance also Negative reinforcement; token-economy systems and,
Signaled-extinction schedules, Reinforcement; Social 378–379, 380–381
330–331 reinforcement Standards, 574–576
Simple responses, 181. See also functional analyses and, 216 Static-probability analyses, 208
Response function-based reinforcement Statistical analyses, 161, 170. See also
Simultaneous discrimination, 81–82, procedures and, 341–343 Data analysis
622 Subject Index
Variable-interval (VI) schedule primary verbal operants, 271–273 autism spectrum disorder (ASD)
of reinforcement. See also response modality selection and, and, 393
Schedules of reinforcement 275–276 feeding disorders and, 434
behavioral momentum theory and, role of the listener in, 273 noncontingent reinforcement
103 teaching, 276–280 (NCR) and, 306–307
differential reinforcement Verbal Behavior Milestones teaching new preferences and,
procedures and, 241–242 Assessment and Placement 185–186
generalized matching law and, Program (VB-MAPP), 278, teaching verbal behavior and,
99–100 398–399 277–278
overview, 60, 61t, 63–64 Verbal discriminative stimuli, 272 vocal skills, 447–448
response allocation and, 95 Verbal operants, 271–273, 276–280 Voucher-based reinforcement. See also
Variable-ratio (VR) schedule Verbal reprimands. See Punishment Reinforcement; Token-economy
of reinforcement. See also Video self-modeling, 257 systems
Schedules of reinforcement Video-conferencing services. See abstinence reinforcement
behavioral momentum theory and, Telehealth services procedures and, 491, 492–493,
103 Videos 499–505
overview, 60, 61t, 62–63 observational procedures and, 139 cigarette smoking and, 494
schedule thinning and, 246 prompting and, 257 stimulants and, 495–496
using punishment effectively and, staff training and management
363 and, 292
Variable-time (VT) schedules Visual inspection method of data W
of reinforcement. See also analysis, 218–222, 220f, 221f. See
Schedules of reinforcement also Data analysis Wait-to-fail model, 456
differential reinforcement schedules Visual media, 139 Water mist procedure, 355, 360t
and, 65–66 Visual representation of data. See also Whole-interval recording, 144, 204–
establishing control by a single Graphing data 205. See also Direct observation;
stimulus and, 79 interpreting functional-analysis Discontinuous measurements;
noncontingent reinforcement data and, 218–222, 220f, 221f Interval recording;
(NCR) and, 308–309, 340–341 scatterplots and, 205–206 Measurement
overview, 61t, 64–66 token-economy systems and, Within-stimulus fading, 401. See also
Ventral tegmental area, 23–25, 23f, 377–378 Fading
24f Visual screen in time out, 356. See also Work functions, 116–118, 117f
Verbal behavior. See also Time out Workplace services. See
Communication; Language; Visual stimuli, 25, 79. See also Organizational behavior
Speech; Vocalizations Exteroceptive stimuli management (OBM)
applied-verbal behavior approach Vocal report preference assessment, Writing skills, 275
with ASD and, 395–396 179, 187
clinical behavior-analytic research Vocalizations. See also
on aging and, 522–523 Communication; Language; Y
multiple control and, 273–274 Verbal behavior
operant contingencies and, 52–53 applied-verbal behavior approach Young Autism Project. See UCLA
overview, 270–271, 279–280 with ASD and, 395–396 Young Autism Project