Child Psychiatry & Human Development (2022) 53:1349–1358

https://doi.org/10.1007/s10578-021-01201-x

ORIGINAL ARTICLE

Effectiveness of Positive Discipline Parenting Program on Parenting

Style, and Child Adaptive Behavior
Paul Carroll1

Accepted: 4 June 2021 / Published online: 3 July 2021

© The Author(s) 2021

Abstract
In this study, a community sample of parents attending free 7-week Positive Discipline parenting workshops were recruited,
as well as a non-randomized community control. Both samples consisted of primarily Hispanic parents with similar demo-
graphic information and attrition rates (initial N = 91), as well as children of similar age (mean age 6.89 and 6.95 years) and
gender. Parenting stress, parenting style, and parent-reported child adaptive behavior were assessed at baseline and after three
months. Longitudinal analysis was performed using mixed-effects regression modeling. Results indicate that attendance in
Positive Discipline parenting workshops was related to a decrease in authoritarian parenting style, a decrease in permissive
parenting style, and a decrease in parental stress. It was also related to an increase in child academic competence, and a
decrease in externalizing-hyperactive behavior (both parent-report). These results suggest that positive discipline parenting
workshops may alter parenting style and may positively impact children of parents who attend.

Keywords  Positive discipline · Parenting style · Authoritative · Intervention · Effectiveness

Introduction on parenting style remains correlational. Nor is there very

much to demonstrate a method by which it may be modified.
The importance of parenting style in child development has If such a method were available though, it might enable a
been acknowledged for a long time, ever since prototypi- true experimental test of the effects of parenting style, via a
cal concepts were developed to distinguish between certain direct manipulation, and a follow-up observation.
styles, such as authoritative, authoritarian, and permissive For now, the strongest evidence of a causal connection
parenting styles [1]. The authoritative parenting style for between parenting styles and the outcomes they are linked to
example, which is the combination of high warmth with high comes from longitudinal studies that do their best to predict
limit-setting, has been linked to children with improved psy- and control for other possible confounds. One such study
chosocial maturity, more academic competence, less inter- examined parenting style in parents of preschool aged chil-
nalized distress, and less externalizing problems [2] includ- dren, who were followed all the way to early adolescence
ing in samples of serious juvenile offenders [3]. An overly [9]. After controlling for initial child differences, the study
punitive style on the other hand has been linked to increases found that certain authoritarian-distinctive power-assertive
in both internalizing and externalizing behavior [4], a greater practices, such as verbal hostility and psychological con-
risk of juvenile delinquency [5], and even a greater risk of trol, were the most detrimental to children, while the most
obesity [6]. Permissive or indulgent parenting style has been competent and well-adjusted children tended to have more
similarly linked to a range of undesirable outcomes, from democratic or authoritative parents [9]. Another longitudinal
sleep problems in preschool aged children [7], to entitle- study of adolescents looked at parenting style and school
ment, narcissism, and lower work ethic in young adulthood dropout rates, finding that students with authoritative parents
[8]. Yet despite all of this and more, most of the research were the most likely to have completed secondary school,
and the least likely to have dropped out, after controlling for
* Paul Carroll numerous other variables [10]. This relationship was found
[email protected] to be partly mediated through adolescent’s school engage-
ment, emphasizing the importance of quality parent–child
1
University of California Merced, 5200 North Lake Rd, relationships for student’s school engagement [11]. There is
Merced, CA 95343, USA

13
Vol.:(0123456789)

1350 Child Psychiatry & Human Development (2022) 53:1349–1358

therefore reasonable evidence to suggest that a more authori- [15]. What makes the Positive Discipline model unique
tative and less authoritarian style may be causally related to though is the teaching of core concepts through the use of
better outcomes in children—and at least some evidence of experiential techniques, which help a participant to not only
the specific mechanisms or mediating factors that drive this practice a specific skill, but to “feel” what it is like to be on
relationship. There is less evidence, however, on what inter- both sides of the approach, versus another approach [22].
ventions may be successful at modifying parenting style, The goal of the approach is to achieve effective discipline,
or any of the identified components of it which seem to be which is defined as that which helps children feel a sense of
harmful or helpful. connection, is mutually respectful, is effective long-term,
Among the very few intervention studies in which parent- teaches important social skills, and builds a sense of per-
ing style itself was a direct outcome variable and not just a sonal capability in children [15, 22].
predictor variable, two somewhat similar and yet distinct While the focus of Positive Discipline is mainly to help
parenting training programs stand out. One is the widely parents to achieve effective discipline, it is very possible that
known, family-centered Triple P program [12, 13], which is it may also be helping parents modify their parenting style
possibly the most researched parenting education program towards an authoritative style, and away from a permissive or
in existence [14]. The Triple P program appears to be an authoritarian style. Many of the specific tools and techniques
effective program at reducing dysfunctional parenting strate- that Positive Discipline advocates [15, 23] are very similar to
gies and increasing confidence in parents [12], making it a or overlap completely with the dimensions of parenting style
worthy choice and a promising approach. Another program identified as authoritative by existing parenting dimensions
which is lesser-known but the subject of this study, is the scales, like the Parenting Dimensions Questionnaire [24].
Positive Discipline parenting program [15, 16]. Although Many items which were empirically derived to character-
much less researched, Positive Discipline is beginning to ize authoritativeness in four different sub-dimensions, are
show similar evidence that it may modify parenting attitudes directly representative of techniques and virtues extolled by
and behaviors [17, 18]. The first of these studies looked at Positive Discipline. For instance, in the warmth/involvement
101 parents attending Positive Discipline parenting classes factor, there are items such as “Gives praise when the child
and found that parents experienced a decrease in authoritar- is good,” “gives comfort and understanding when the child
ian and permissive parenting style along with an increase is upset,” and “expresses affection by hugging, kissing, and
in authoritative style, at 3-months after completion of the holding the child.” These items are analogous to whole sec-
program [17] A second study looked at 107 parents of more tions of techniques recommended in Positive Discipline, like
diverse ethnicity and lower SES and found a similar pattern “give encouragement freely,” and “give a hug,” [23]. Each of
of effects on parenting style, sustained to three months after the other sub-dimensions identified as belonging to authori-
attending the workshops [18]. Thus, the Positive Discipline tative parenting style also have clear analogous representa-
program seems like it may be a promising possibility in tions in Positive Discipline, including Reasoning Induction
modifying parenting behavior—and may even be effective (“take time for training”), Democratic Participation (“hold
with low SES populations who could possibly benefit from weekly family meetings”), and Good Natured / Easy Going
the intervention the most. So, what exactly is Positive Dis- (“look for improvement, not perfection”) [23]. Thus, Posi-
cipline, and how does it differ from the many other existing tive Discipline appears to be very well representative of the
parenting programs? authoritative style, at least as it is intended. The question
Positive Discipline is an approach to raising children remaining though, is how effective a typical applied work-
that is based on the teachings of Alfred Adler and Rudolph shop is at modifying parent’s behaviors, and if so whether
Dreikurs, which emphasizes the need for belonging as a fun- these modifications carry over into benefits in children.
damental motivator of human beings [16, 19, 20]. A basic So far, the existing research on Positive Discipline inter-
premise of the approach is the understanding that “misbe- ventions is limited. Jane Nelsen’s 1979 dissertation provided
having children are discouraged children who have mistaken the first trial of what would later form the foundation of
ideas (faulty private logic) about how to achieve their pri- Positive Discipline, although this 12-week intervention
mary goal—to belong” [21]. Thus, a major focus of Positive involved both parents and teachers, and many of the tech-
Discipline is to help parents and educators understand these niques that are taught today had not been developed yet [25].
mistaken ideas that children may hold, and to use a variety Other studies have focused on school-based applications of
of specific strategies to help children feel a sense of belong- the program, such as implementing class meetings [26],
ing, which is their root goal. For example, there are several or democratic problem-solving [27]. One study looked at
main concepts which are emphasized, such as the use of Adlerian-based parenting classes broadly, of which Positive
encouragement (not praise), and the use of family and class Discipline as a sub-type, but was not focused specifically on
meetings to solve problems in a democratic manner, which Positive Discipline [28]. Only three studies have focused on
helps children feel a sense of belonging and significance Positive Discipline parenting trainings in their current and

13
Child Psychiatry & Human Development (2022) 53:1349–1358 1351

widely delivered format [17, 18, 29], but the results so far a two group quasi-experiment is still an improvement over a
are promising. A dissertation by Holliday found that Posi- one-group quasi-experiment. This study therefore serves the
tive Discipline parenting workshops increased authoritative aim of a program of research, which is to seek to success-
parenting style (and decreased authoritarian and permissive fully modify parenting style to reduce negative influences
style), although in relatively affluent participants [17]. A on children, to observe the positive effects of this change in
second study in a much lower SES population also found parenting style on children, and to understand and optimize
the workshops altered parenting attitudes and behaviors, the maximum benefit possible to render via this approach.
although using different measures [18]. The effects in either
study did not appear to be limited or moderated by common
factors such as SES, ethnicity, or being a single parent [18, Method
29]. Thus, the parenting workshops have the potential to not
only improve the long-term prospects of many children, but Participants and Procedure
potentially shrink socio-developmental disparities as well,
if applied properly. In this study, parents invited to participate in a free, seven-
The purpose of this study was therefore to replicate and week workshop on Positive Discipline were recruited by a
extend previous findings, in a population of mostly low- community partner (a local non-profit service provider),
income parents and community controls, attending free while a comparison group of parents involved in local adult
Positive Discipline parenting training workshops (or com- schools where recruited directly by the research team. The
munity controls). One main addition of this study, relative to comparison group in this study was therefore a non-rand-
previous studies, was the recruitment of a non-randomized omized control group, but both samples consisted of pri-
control group—making it a two group quasi-experimental marily Hispanic, Spanish speaking parents, and had similar
design. This type of control cannot eliminate personal con- demographic information, attrition rates, and children of
founds, but it can protect against some threats to internal similar age and gender. Participants where recruited from
validity such as testing effects, maturation, and history a total of 18 sites, in two waves. Outreach efforts for the
threats. Another main addition was the use of new and workshops varied by site, but consisted of contact with com-
extensive measures, including both extensive measures on munity leaders, distributed flyers, direct outreach by facili-
parenting style and parent-report of their children’s adaptive tators, and through word of mouth. Outreach efforts for the
behavior, broken into several domains. Theoretically, several non-randomized control group consisted of showing up with
of these domains correspond directly to parenting style, such permission to classes being taught at the adult school, intro-
as academic competence, and types of both internalizing ducing the study and asking if there were any parents of
and externalizing behavior [2]. The study collected mail-in young children who would be interested in participating, and
questionnaires at baseline, and after three months. soliciting names and addresses of those interested.
The primary hypotheses of this study were that authorita- In the first wave of recruitment, after the names and
tive parenting style would increase in the intervention group, addresses of those interested were gathered in person,
but not in the control, while authoritarian style, permissive potential participants were mailed an initial packet with
style, and parenting stress would decrease in the intervention more information about the study, a consent form, and an
group but not in the control. These findings would serve to initial survey. Fifty-two participants chose to return the ini-
replicate and extend previous results. Secondary hypotheses tial survey and were enrolled in this manner, which was a
concerned outcomes in children—where it was hypothesized response rate of approximately 58 percent. In the second
that academic and social competence would increase in the wave, interested parties were simply given a self-addressed
intervention group but not in the control, while internal- stamped envelope with the initial packet of information,
izing and externalizing behavior would decrease in the consent form, and survey—and enrolled if they chose to fill
intervention group, but not the control. These hypotheses it out and return it. An additional thirty-nine participants
are theoretically implicated but have not been examined. were enrolled in this manner, for an initial total of N = 91
Beyond that, exploratory analyses would examine the many participants. Three months after initial contact, a follow-up
sub-domains within each dimension of parenting style, as survey was mailed to the address on file for each partici-
well as the sub-domains within each major dimension of pant, and a total of 54 were returned (37 were not returned),
children’s adaptive behavior—without any specific hypoth- which is an approximately 59% retention rate. However, par-
eses of effects. Of course, a major limitation that must be ticipants who returned surveys more than three weeks late
understood is that all the data are parent reported, as well as at either timepoint (n = 9) were further restricted from the
the comparison group being a non-randomized comparison. analysis of primary outcomes, although their demographics
Nonetheless, the measurement of these outcomes is a step are reported. Thus, the initial sample was 91 participants, the
forward in connecting theoretical desirable endpoints, and final sample after attrition was 45 total participants—while

13

1352 Child Psychiatry & Human Development (2022) 53:1349–1358

workshops were being offered primarily as a community ser-

vice, not as a research project, it was not ethical or feasible to
randomize any participants to not receive the workshops—
but a semi-equivalent comparison group was nonetheless
recruited. While not as ideal as a randomized control, the
non-randomized control could still act as a control against
several possible threats to validity, including a simple testing
effect, maturation in the sample, or natural attrition causing
those with deteriorating life circumstances to drop-out of the
study. Since it was known from previous research that attri-
tion was likely to be very high in this very low SES sample
[18], it was hoped that the non-randomized control would
experience equivalent rates of attrition, and thus somewhat
protect against this threat. Similarly, if a simple testing effect
or maturation over the three-month follow-up had any effect,
then the non-randomized control group should experience
this effect as well. And indeed, as can be seen in the results,
both groups were similar in demographic characteristics, had
similarly aged children, and had similar rates of attrition—
hopefully mitigated some of these threats to validity.
Fig. 1  Recruitment of study participants
Measures
the useable sample which included useable primary outcome
data from either timepoint, was 82 participants. A flowchart The survey measures included several demographic vari-
of the recruitment process is shown in Fig. 1. ables including age, gender, ethnicity, education, income,
A mail-in survey was chosen as the primary data collec- financial strain, number of children, and single parent sta-
tion tool in this study for two reasons. One—it limited the tus. Positive Discipline specific attitudes and behaviors were
potential for facilitators of parenting workshops to bias the assessed using the 7-item Positive Discipline Parenting Scale
results, as all responses would be filled out at home. And (PDPS), which has shown acceptable internal consistency
two—it made sure there was adequate time to complete each for community research (α = 0.775–0.772) [30]. Parental
survey, which were each 8 pages in length, plus a tear-off stress and reward were assessed using the 12-item Span-
page. To ensure confidentiality and to limit potential experi- ish adaptation of the Parental Stress Scale (PSS), which has
menter bias, participants did not put their name or address also shown acceptable internal consistency (α = 0.76–0.77)
anywhere on the bulk of the survey, but instead only wrote [31], which breaks down into the sub-domains of stress
that information on a final tear-off page of each survey— and reward from parenting. The main dimensions of of
which asked where they should have a gift card mailed to authoritative, authoritarian, and permissive parenting style
for their participation. The surveys were then self-addressed were assessed using the 62-item Parenting Dimensions
back to a neutral third party, who physically separated names Questionnaire-Revised (PDQ-R) [24], which also breaks
and addresses from each survey, writing a codename on each those main dimensions down into 11 sub domains. The sub-
one and delivering them in batches to the principal investi- domains for authoritative parenting style include “warmth/
gator. Thus, anonymity of the participants was maintained involvement,” “reasoning induction,” “democratic partici-
while allowing responses to be matched by their codename, pation,” and “good natured/easygoing.” The sub-domains
and any experimenter bias in the coding or judgement of for authoritarian parenting style include “verbal hostility,”
survey responses should have been limited. Each participant “corporal punishment,” “non-reasoning/punitive strate-
also read and signed a consent form, which was also sepa- gies,” and “directiveness.” The sub-domains for permissive
rated from their responses and stored separately. For their parenting style include “lack of follow through,” “ignoring
participation, each participant was mailed a generic $10 gift misbehavior,” and “self-confidence.” (reverse coded) [24].
card for each response, for a total of $20 in incentives if they The PDQ-R has been shown to have acceptable to excellent
completed both surveys. internal consistency for the main parenting style dimensions
The overall design of the study was thus a pre-test, post- (α = 0.75–α = 0.91), while the sub-dimensions are more
test design, with a non-randomized control. Most of both exploratory. These measures are the sum of what we asked
samples were Hispanic, and approximately half of those parents to describe about themselves, and their parenting
recruited were Spanish speaking. Because the parenting style.

13
Child Psychiatry & Human Development (2022) 53:1349–1358 1353

Next, we asked each parent to describe their youngest two additional information about a typical parenting workshop
children who were at least 3–18 years old, using a 54-item of this kind should consider reaching out to the Positive Dis-
short form of the Child Adaptive Behavior Inventory (CABI) cipline Association [22], or individual workshop facilitators.
[32]. The short form used measures 20 individual attributes,
such as task orientation, creativity, introversion, extrover-
sion, so on—which are then z-scored and combined into six Data Analysis
main factors of adaptive behavior. These six main factors
of adaptive behavior assess (1) Academic competence; (2) The primary hypotheses of this study were that authoritative
Social competence; (3) Externalizing-aggressive; (4) Exter- parenting style and positive discipline parenting style would
nalizing-hyperactive; (5) Internalizing—socially isolated; increase in the intervention group, but not in the control,
and (6) Internalizing—psychological symptoms (anxiety, while authoritarian style, permissive style, and parenting
depression, etc.). The CABI has been validated in samples stress would decrease in the intervention group but not in the
of children from age 4 in preschool to age 12 in 4­ th grade, control. A mixed effects regression model was used for each
with samples of up to n = 1723 [32, 33]. In this context, of these outcomes, with subject as a random effect and both
the six main domains are considered reliable for research group and group by time interaction as fixed effects, using
purposes, while the 20 smaller domains are somewhat less an unstructured covariance matrix. A mixed effects model
reliable, and considered only for exploratory purposes. The accounts for the intra-class correlation between repeated
only other information we asked parents to provide were the measures by subject, allows for them to vary as a random
age and gender of each child. effect, and controls for the imbalance of observations as a
result of attrition—without restricting the sample to only
Treatment participants with complete data [35]. This preserves statisti-
cal power as opposed to dropping all the observations that
Each seven-week workshop was carried out by a pair of did not complete both time points. Including a fixed effect
trained facilitators, all of whom had been trained and certi- for group models the overall differences between compari-
fied by the Positive Discipline Association as Certified Posi- son groups which exist at the outset and is appropriate for
tive Discipline Parenting Educators (a fourteen-hour certifi- a non-randomized control. The group by time interaction is
cation) and had at least some previous experience facilitating thus the primary outcome of importance—which estimates
workshops. All facilitators at Spanish-language sites were if the treatment group experienced any significant changes
bilingual and had access to Spanish language materials. over time, and if they varied significantly from changes
In addition, these facilitators were generally recruited and experienced by the comparison group. This significance of
trained from the same cultural background as the program this interaction was assessed using a one-tailed F-test, since
participants, and therefore had an enhanced cultural compe- there was a directional hypothesis regarding the interaction.
tency to be able to relate to and help program participants The secondary hypotheses of this study were that aca-
with their unique challenges. The structure of each workshop demic competence and social competence would increase
is flexible by design, but generally includes warm-up activi- in children of parents in the intervention group, but not the
ties, lessons, experiential activities, and parents-helping-par-
ents in every 2-h session. Sessions are typically focused on
Table 1  Demographic characteristics of participants
hands-on experiential learning, often with the use of props
and visual aids, but without any standardized audio-visual Demographic variable Treatment Comparison (n = 32)
group
presentation, such as DVDs or recorded lectures, in this for-
(n = 59)
mat. Many of the activities involve participants role-playing
both as children and adults, as they process different parent- Age (M, SD) 36.4 (11.25) 35.1 (7.99)
ing behaviors and their reactions to them from both perspec- Number of children (M, SD) 2.34 (1.36) 2.81 (1.18)
tives [34]. Facilitators tailor each session to the needs and Hispanic (%) 44 (75) 24 (75)
challenges of each group however, with different warm-ups, Female (%) 44 (75) 30 (94)
lessons, and more than 70 possible experiential activities Education: high school or 45 (78) 21 (68)
in a facilitators training manual to choose from [34]. The greater (%)
goal is to help parents overcome the particular challenges Education: 4-year degree or 6 (10) 8 (28)
higher (%)
they are facing to effective discipline, while becoming both
Income: < $39,999 47 (81) 20 (63)
kind and firm. To ensure confidentiality and allow the pro-
Financial worry (%) 30 (52) 13 (42)
cess to take place with potentially sensitive disciplinary
Single parent (%) 11 (19) 5 (16)
issues, members of the research team did not attend the ses-
sions beyond the initial recruitment visits. Those seeking Individual variables may have missing data, and lower individual n’s

13

1354 Child Psychiatry & Human Development (2022) 53:1349–1358

Table 2  Results for main Parameter Estimate 95% CI (Lower/Upper) Sig Interaction Sig
outcomes in parents
PDPS* .408 0.095/0.717 .011* .024*
Authoritative style .154  − 0.125/0.433 .274 .535
Authoritarian style*  − .493  − 0.729/ − 0.256 .000* .001*
Permissive style*  − .514  − 0.789/ − 0.239 .000* .002*
Parental stress*  − .284  − 0.515/ − 0.053 .017* .054†
*
 Significant at p < .05, †Significant at p < .05 given a directional hypothesis

control, while the domains of externalizing (aggressive), expected and statistically controlled for as a fixed effect
externalizing (hyperactive), internalizing (socially isolated), regardless.
and internalizing (psychological symptoms) would decrease
in the children of parents in the intervention group, but not Primary Outcomes
the control. A mixed-effects regression model was also used
for these outcomes, with the same group and group by time The primary hypotheses of this study were that authorita-
interaction modeled as fixed effects. Finally, several explora- tive parenting style and positive discipline parenting style
tory analyses were then conducted using the same mixed- (PDPS) would increase in the intervention group but not in
effects regression model, but for each of the 11 sub-domains the control, while authoritarian style, permissive style, and
of the PDQ-R, and each of the 20 sub-scales of the CABI. parenting stress would decrease in the intervention group
Since these subscales are likely to be of lesser reliability and but not in the control. These hypotheses were tested using a
since they were not the primary outcomes of the study, the mixed-effects regression model, and the results and param-
results of these analyses should be considered exploratory— eter estimates for each variable are displayed in Table 2.
and the significance of any findings should be considered The analysis revealed a significant group by time interaction
preliminary, given the large number of significance tests. for PDPS; F (2, 50.45) = 4.01, p = 0.024, for authoritarian
These analyses are nevertheless included though, to aid in a style; F (2, 53.45) = 8.74, p = 0.001, for permissive style;
more theoretical understanding of the results, and to guide F (2, 49.34) = 7.09, p = 0.002, and for parental stress; F (2,
future directions of research. 46.62) = 3.10, p = 0.054,1 but not for authoritative style—
which was estimated to increase, but not significantly; F (2,
49.89) = 0.63, p = 0.535. These variables were transformed
Results into z-scores, and z-scored parameter estimates for the treat-
ment group are provided in Table 2. None of these outcomes
Demographics changed significantly for the comparison group, indicating
that the changes do not appear to result from a simple test-
Demographic information for both groups at time one ing effect, maturation, or other methodological artifact that
is reported in Table  1. Of the 91 participants originally would equally influence the comparison condition.
recruited, approximately 81% were female, 75% were His- Next, the second set of hypotheses were that academic
panic, and 18% reported being a single parent. Highest level competence and social competence would increase in chil-
of education achieved was about average for the region, with dren in the intervention group but not the control, while the
26% reporting that they had not completed high school, 74% domains of externalizing (aggressive), externalizing (hyper-
having completed high school or greater, and 16% having active), internalizing (socially isolated), and internalizing
obtained a four-year degree or higher. Income was also very (psychological symptoms) would decrease in the children of
low on average, with 74% reporting a combined household parents in the intervention group but not the control. These
income of less than $39,999/year (near the median for the hypotheses were also tested using a mixed-effects regres-
region), and 47% reported that they “sometimes struggle sion model, and the results and parameter estimates for each
with finances” or that “financial worries are a serious prob- variable are displayed in Table 3. The analysis revealed a
lem” for them each month. The number of children par- significant group by time interaction for academic compe-
ticipants had ranged from 1 to 8, with a mean of 2.51. The tence; F (2, 77.81) = 5.36, p = 0.007, and for externalizing
average age of participants was 35.9. The only significant (hyperactive); F (2, 74.86) = 3.54, p = 0.034, but not for any
demographic difference between groups was that there were
more women in the treatment versus the comparison group;
t (79.4) = 2.67, p = 0.009. Otherwise the groups were fairly 1
  P-values reflect two-tailed significance test. Result is significant at
similar demographically, although group differences were p < .05 in light of a directional hypothesis.

13
Child Psychiatry & Human Development (2022) 53:1349–1358 1355

Table 3  Results for main Parameter Estimate 95% CI (Lower/Upper) Sig Interaction Sig
outcomes in children
Academic competence* .432 0.158/0.705 .002* .007*
Social competence .225  − 0.059/0.509 .119 .194
Externalizing—aggressive  − .165  − 0.387/0.057 .144 .320
Externalizing—hyperactive*  − .288  − 0.513/ − 0.063 .013* .034*
Internalizing—socially isolated  − .193  − 0.422/0.036 .098† .221
Internalizing—psychological symptoms)  − .094  − .360/0.172 .482 .363
*
 Significant at p < .05, †Significant at p < .05 given a directional hypothesis

Table 4  Results for exploratory Parameter Estimate 95% CI (Lower/Upper) Sig Interaction sig
outcomes in parents
Warmth and involvement .139  − .161/.440 .356 .648
Reasoning/induction .006  − .287/.298 .970 .799
Democratic participation .218  − .089/.525 .160 .347
Good natured/easy ­going† .256  − .028/.540 .077† .154
Verbal ­hostility†  − .284  − .568/.001 .051† .139
Corporal punishment*  − .427  − .674/ − .179 .001* .004*
Non-reasoning/punitive*  − .387  − .640/ − .135 .003* .012*
Directiveness*  − .396  − .720/ − .073 .017* .053†
Lack of follow-through†  − .255  − .520/.010 .059† .164
Ignoring misbehavior  − .208  − .550/.134 .229 .478
Low self confidence*  − .661  − .962/ − .359 .000* .000*
*
  Significant at p < .05, †Significant at p < .05 given a directional hypothesis

of the other four domains of adaptive behavior. Transformed a standard deviation in size, including reported decreases in
(z-scored) parameter estimates for the children in the treat- distractable and antisocial behavior, and increases in kind-
ment group are provided in Table 3. As expected, none of ness, intelligence, and creativity (all parent-report), as well
these outcomes changed significantly for the comparison as possible decreases in social isolation and increases in
group, indicating that constructs in question appear stable, social perception. These results should be considered pre-
lacking an intervention. liminary pending a replication—but it seems unlikely that
several of them are due to error. There were no significant
Exploratory Analysis changes in the children in the comparison group—despite
the numerous analyses and the increased odds of type I error.
Finally, several exploratory analyses were then conducted This would seem to suggest that spontaneous changes in
using the same mixed-effects regression model, but for each these traits are rare outside of some type of specific influ-
of the 11 sub-domains of the PDQ-R (Table 4), and each of ence, such as parenting class.
the 20 sub-scales of the CABI (Table 5). These results pre-
sented with z-scored parameter estimates for ease of inter-
pretability and comparison. On the parenting dimensions Discussion
questionnaire, the largest single effect was a large increase in
parenting self-confidence (reversed as “low self-confidence”, Taken together, the findings from this study indicate that
part of the Permissive Style), while there were several other parenting style did indeed change significantly over time in
decreases including in verbal hostility, corporal punishment, conjunction with Positive Discipline parenting workshops,
directiveness, and the use of non-reasoning and punitive including decreases authoritarianism and permissiveness—
strategies (all part of the Authoritarian Style). All observed as well as decreases in parental stress, and increases in posi-
changes in the treatment group were in the hypothesized and tive discipline style parenting. The changes do not appear to
positive direction, and there were no significant changes in be a result of a possible testing effect, maturation, or other
the comparison group, despite the increased odds of type I factors that would equally influence the comparison condi-
error. On the child’s adaptive behavior questionnaire, there tion. Furthermore, the findings appear to indicate a reduction
were several notable changes of roughly one quarter to a half in externalizing—hyperactive behavior, and an increase in

13

1356 Child Psychiatry & Human Development (2022) 53:1349–1358

Table 5  Results for exploratory Parameter Estimate 95% CI (Lower/Upper) Sig Interaction Sig
outcomes in children
Fair .126  − .188/.440 .427 .651
Kind* .428 .127/.728 .006* .018*
Anxious  − .219  − .505/.068 .132 .319
Hyperactive  − .127  − .366/.112 .294 .554
Antisocial*  − .297  − .530/ − .064 .013* .034*
Oppositional  − .200  − .477/.077 .154 .343
Hostile .018  − .189/.226 .859 .926
Intelligent* .432 .171/.692 .001* .004*
Creative* .436 .149/.723 .003* .007*
Task oriented .232  − .080/.544 .142 .291
Distractable*  − .376  − .631/ − .121 .004* .011*
Extroverted .140  − .105/.386 .257 .523
Introverted  − .110  − .356/.156 .374 .671
Depressed .121  − .107/.348 .294 .351
Somaticized  − .157  − .449/.135 .289 .307
Impaired development .067  − .161/.295 .559 .825
Socially ­isolated†  − .296  − .591/ − .001 .049* .092†
Socially rejected .032  − .298/.362 .845 .854
Socially ­perceptive† .314 .021/.607 .036* .098†
Socially skilled  − .100  − .409/.209 .522 .590
*
 Significant at p < .05, †Significant at p < .05 given a directional hypothesis

academic competence, in children of parents attending the this may be responsible for some or even all their improve-
Positive Discipline workshops. While some other hypotheses ment. However, the lack of any significant changes in any
were not supported, overall these are promising results that of the observed variables in the control group suggests that
may be of interest to both parents and educators. these traits are relatively stable, lacking a directed inter-
These findings are consistent with previous research in vention. Another limitation was the use of entirely parent-
their demonstration of an effect on parents [17, 18], but also reported measurements. Although the study used validated
extend those findings with the measurement of outcomes measures for all reported outcomes, they are nonetheless
in children, and the use of a control group. The finding of self-reported, and may be influenced by participant response
increased academic competence and decreased hyperactive bias. At the very least, the comparison group somewhat
behavior are consistent with research that shows authorita- protects against this threat—as the parents in this condition
tive parenting style (and lower permissive or authoritarian might also be motivated to report socially desirable informa-
style) predicts school readiness and achievement in the first tion about themselves or their children, and yet reported no
grade [36]. Although there was only a three-month follow up changes. Another limitation was the relatively low preva-
in this study, parenting style has also been shown to predict lence of single-parents taking the workshops. It is not known
dropout rates in secondary school [10, 11], making these how well these results would generalize to other demo-
early childhood interventions of potentially far-reaching graphic groups, such as single parents—although previous
importance. More broadly speaking, this research adds to research has not found outcomes to be moderated by single-
the base of support for implementing effective parenting pro- parent status [18, 29]. Some final notable limitations include
grams in general, [12, 13]. Hopefully future research will the relatively small sample, high attrition, and a relatively
continue to explore such programs with experimental stud- short 3-month follow-up period. All of these limitations rep-
ies, and extended follow-up observations. resent the challenges of doing field research with vulnerable,
Of course, the present study has some limitations. First, low-resource participants—and yet future research should
it used a nonrandomized control group rather than a rand- attempt to do better in these areas if possible.
omized control, which introduces the possibility of selec- Despite the limitations, this study demonstrated several
tion bias into the intervention group—and cannot rule out significant changes in parenting style for the participants
the possibility of personal confounds contributing to the who attended Positive Discipline parenting workshops—
observed effects. It is possible that parents attending the which is not something that many programs have been able
parenting workshops were simply motivated to improve, and to accomplish. It also appeared to show sizeable increases

13
Child Psychiatry & Human Development (2022) 53:1349–1358 1357

in children’s academic competence, as well as potential Summary

improvements in several types of adaptive behavior, which
are all desirable outcomes. It is possible that there are Positive Discipline is a type of parenting program based on
unique features to Positive Discipline that help parents Adlerian psychology, which emphasizes a style of parenting
make a philosophical switch, such as the central concept that is analogous to authoritative parenting, though some-
that a misbehaving child is a discouraged child, and that what unique. This study is the first to examine the effects
their behavior will improve when they feel their need for of Positive Discipline parenting workshops on outcomes in
belonging being met [15]. Or perhaps it is possible that it children of those attending, as well as the first to incorporate
is the way that these concepts are taught, through activi- a control group (albeit non-randomized). The results indicate
ties like role-playing and experiential exercises, which that attendance in Positive Discipline parenting workshops
have been invented and refined over the years by dozens was related to a decrease in authoritarian parenting style,
of individual facilitators [34]. Either way, there appears including all four subcomponents (verbal hostility, cor-
to be a high level of consumer satisfaction with parents poral punishment, non-reasoning punitive, and directive-
who have attended a Positive Discipline workshop, and ness), as well as a decrease in permissive style parenting,
a relatively low cost per participant to implement [18]. and a decrease in parental stress. It was also related to an
Hopefully future research can continue to explore this cost increase in child academic competence (parent-report), and
versus benefit, as well as the longer-term effectiveness, if a decrease in externalizing-hyperactive behavior (parent-
any, of conducting Positive Discipline training workshops. report). These results suggest that Positive Discipline parent-
A few of the questions that remain to investigated ing workshops appear to be effective at improving parenting
include whether there is an optimum dose of the train- style in the studied population and may further positively
ing, whether the effects are sustained, and whether there impact the children of parents who attend. It should continue
is any change in Authoritative parenting style specifically, to be implemented and explored.
which has seen inconsistent results in two different stud-
ies. It may be that the average effect on Authoritativeness
lies somewhere between what was observed in this study Declarations 
and one previously [17], or it’s possible that the greatest
part of the effect is delayed, as was observed previously. Conflict of interest  The author declares they have no conflict of inter-
Perhaps future research could explore if there is an opti- est.
mum number of Positive Discipline workshops to expose Ethical Approval  All procedures performed were in accordance with
participants to before they reach a point of diminishing the ethical standards of the University of California Institutional
interest or benefits, or if there is an optimum number of Review Board, and with the 1964 Helsinki declaration and its later
participants to have in each group setting? Information amendments or comparable ethical standards.
like this would greatly aid stakeholders who might wish Informed Consent  Informed consent was obtained from all individual
to offer these classes to the population they serve, such as participants included in this study.
school-districts or human services agencies. The great-
est need for future research is probably still for a well- Open Access  This article is licensed under a Creative Commons Attri-
conducted randomized controlled trial, with a longer-term bution 4.0 International License, which permits use, sharing, adapta-
follow-up, and multiple third-party observations. Such as tion, distribution and reproduction in any medium or format, as long
study would ultimately offer the strongest validity to its as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
conclusions and produce the best estimates of the pro- were made. The images or other third party material in this article are
gram’s effectiveness. included in the article’s Creative Commons licence, unless indicated
There may ultimately be several different parenting pro- otherwise in a credit line to the material. If material is not included in
grams that are effective, and each may be effective in dif- the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
ferent ways or with different populations. Hopefully this need to obtain permission directly from the copyright holder. To view a
research contributes to the overall goal of finding out what copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
programs exist that are effective at all—and isn’t seen as
placing Positive Discipline in competition with any other
particular program. Indeed, as some researchers have sug- References
gested, it is time to start treating investments in parenting
style as investments in human development [37] and when 1. Baumrind D (1966) Effects of authoritative parental control on
making such investments there is always room for different child behavior. Child Dev 37:887–907
2. Lamborn SD, Mounts NS, Steinberg L, Dornbusch SM (1991)
approaches.
Patterns of competence and adjustment among adolescents from

13

1358 Child Psychiatry & Human Development (2022) 53:1349–1358

authoritative, authoritarian, indulgent, and neglectful families. 22. “What is Positive Discipline?” (2020). https://w ​ ww.p​ ositi​ vedis​ cipl​
Child Dev 62:1049–1065 ine.​org/​about-​posit​ive-​disci​pline. Accessed 14 Sept 2020
3. Steinberg L, Blatt-Eisengart I, Cauffman E (2006) Patterns of 23. Nelsen J (2009) Parenting with love: an introduction to “positive
competence and adjustment among adolescents from authorita- discipline,” with 14 proven strategies for raising children without
tive, authoritarian, indulgent, and neglectful homes: a replication blame, shame or pain. Hinduisim Today. 1–16. https://w ​ ww.h​ imal​
in a sample of serious juvenile offenders. J Res Adolesc 16:47–58 ayanac​ ademy.c​ om/m​ edia/b​ ooks/p​ arent​ ing-w
​ ith-l​ ove_e​ i/p​ arent​ ing-​
4. Zubizarreta A, Calvete E, Hankin B (2019) Punitive parenting with-​love_​ei.​pdf
style and psychological problems in childhood: the moderating 24. Robinson C, Mandleco B, Olsen S, Hart C (1995) Authoritative,
role of warmth and temperament. J Child Fam Stud 28:233–244 authoritarian, and permissive parenting practices: development of
5. Mowen TJ, Schroeder RD (2018) Maternal parenting style and a new measure. Psychol Rep 77:819
delinquency by race and the moderating effect of structural dis- 25. Nelsen J (1979) The effectiveness of Adlerian parent and teacher
advantage. Youth Soc 50:139–159 study groups in changing child maladaptive behavior in a positive
6. Chen Y, Kawachi I, Berkman L, Trudel-Fitzgerald C, Kubzansky direction. A dissertation (Order No. 8017442). ProQuest Disserta-
L (2019) Does optimal parenting style help offspring maintain tions & Theses A&I (303035071)
healthy weight into mid-life? Prevent Med 123:84–90 26. Potter S (1999) Positive interaction among fifth graders: is it a
7. Tyler D, Donovan C, Scupham S, Shiels A, Weaver S (2019) possibility? The effects of classroom meetings on fifth-grade stu-
Young children’s sleep problems: the impact of parental distress dent behavior. Submitted as partial fulfillment of the requirements
and parenting style. J Child Fam Stud 28:2098–2106 for the degree of Master of Education. Southwest Texas State
8. Lonestein J (2019) A correlational analysis: permissive parent- University, San Marcos, Texas June 1999
ing, millennials, work ethic, narcissism, and entitlement. ProQuest 27. Browning L, Davis B, Resta V (2000) What do you mean “think
Dissertations & Theses A&I (10830440) before I act?”: conflict resolution with choices. J Res Child Educ
9. Baumrind D, Parselene RE, Owens EB (2010) Effects of preschool 14:232–238
parents’ power assertive patterns and practices on adolescent 28. McVittie J, Best AM (2009) The impact of Adlerian-based parent-
development. Parent Sci Pract 10:157–201 ing classes on self-reported parental behavior. J Individual Psy-
10. Blondal KS, Adalbjarnardottir S (2009) Parenting practices and chol 62:264–285
school dropout: a longitudinal study. Adolescence 44:729–749 29. Brown KA (2018) The examination of parental demographics as
11. Blondal KS, Adalbjarnardottir S (2014) Parenting in relation to moderators of the effects of a positive discipline parenting train-
school dropout through student engagement: a longitudinal study. ing model. Doctoral dissertation, Adler School of Professional
J Marriage Fam 76:778–795. https://​doi.​org/​10.​1111/​jomf.​12125 Psychology
12. Fujiwara T, Kato N, Sanders M (2011) Effectiveness of group 30. Carroll P, Hamilton WK (2016) Positive Discipline Parenting
positive parenting program (triple P) in changing child behavior, Scale: reliability and validity of a measure. J Individual Psychol
parenting style, and parental adjustment: an intervention study 72(1):60–74
in Japan. J Child Fam Stud 20:804–813. https://​doi.​org/​10.​1007/​ 31. Oronoz B, Alonso-Arbiol I, Balluerka N (2007) A Spanish adapta-
s10826-​011-​9448-1 tion of the Parental Stress Scale. Psicothema 19:687–692
13. Shapiro CJ, Kilburn J, Hardin JW (2014) Prevention of behav- 32. Cowan P, Cowan C, Heming G, Miller N (2004) Manual for the
ior problems in a selected population: stepping stones triple P child adaptive behavior inventory; a behavior checklist for teach-
for parents of young children with disabilities. Res Dev Disabil ers and parents. University of California, Berkeley
35:2958–2975 33. Cowan PA, Cohn DA, Cowan CP, Pearson JL (1996) Parents’
14. Sanders MR, Pickering JA, Kirby JN, Turner KMT, Morawska A, attachment histories and children’s externalizing and internalizing
Mazzucchelli T, Ralph A, Sofronoff K (2012) A commentary on behaviors: exploring family systems models of linkage. J Consult
evidenced-based parenting programs: redressing misconceptions Clin Psychol 64:53–63
of the empirical support for triple P. BMC Med 10:145–149 34. Lott L, Nelsen J (2008) Teaching parenting the positive discipline
1 5. Nelsen J (2006) Positive discipline. Ballantine Books, New York way; a step-by-step approach to starting and leading parenting
16. Gfroerer K, Nelsen J, Kern RM (2013) Positive discipline: helping classes, 6th edn. (n.p.) Authors
children develop belonging and coping resources using individual 35. Hedeker D, Gibbons R (2006) Mixed-effects regression models
psychology. J Individual Psychol 69:294–304 for continuous outcomes. In: Longitudinal data analysis. Wiley,
17. Holliday MJ (2014) Authoritative parenting and outcomes of Hoboken, NJ, pp 47–79
positive discipline parent training: Parenting style and perceived 36. Kim Y, Calzada EJ, Barajas-Gonzalez RG, Huang K-Y, Brotman
efficacy (Order No. 3662133). ProQuest Dissertations & Theses LM, Castro A, Pichardo C (2018) The role of authoritative and
A&I (1648390208). authoritarian parenting in the early academic achievement of
18. Carroll P, Brown P (2020) The effectiveness of positive discipline Latino students. J Educ Psychol 110:119–132
parenting workshops on parental attitude and behavior. J Indi- 37. Cobb-Clark DA, Salamanca N, Zhu A (2018) Parenting style as an
vidual Psychol 76(3):286–303 investment in human development. J Popul Econ 32:1315–1352
19. Adler A (1927) Understanding human nature. The World Publish-
ing Company, New York Publisher’s Note Springer Nature remains neutral with regard to
20. Dreikurs R, Soltz V (1964) Children: the challenge. Penguin Com- jurisdictional claims in published maps and institutional affiliations.
pany, New York
21. Nelsen J (1999) Positive time-out: and over 50 ways to avoid
power struggles in the home and the classroom. Three Rivers
Press, New York

13