Professional Documents
Culture Documents
(Auditing and Accounting Studies) Kristina Yankova (Auth.)-The Influence of Information Order Effects and Trait Professional Skepticism on Auditors’ Belief Revisions_ a Theoretical and Empirical Analy
(Auditing and Accounting Studies) Kristina Yankova (Auth.)-The Influence of Information Order Effects and Trait Professional Skepticism on Auditors’ Belief Revisions_ a Theoretical and Empirical Analy
Herausgegeben von
A. Köhler, Duisburg-Essen, Deutschland
K.-U. Marten, Ulm, Deutschland
R. Quick, Darmstadt, Deutschland
K. Ruhnke, Berlin, Deutschland
M. Wolz, Dortmund, Deutschland
Herausgegeben von
Prof. Dr. Annette Köhler Prof. Dr. Klaus Ruhnke
Universität Duisburg-Essen Freie Universität Berlin
Springer Gabler
© Springer Fachmedien Wiesbaden 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, speci¿cally the rights of translation, reprinting, reuse of illus-
trations, recitation, broadcasting, reproduction on micro¿lms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by
similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a speci¿c statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.
In the aftermath of the numerous accounting scandals and unfavorable audit inspection results
of the last years, professional skepticism has evolved to one of the most intensively and con-
troversially discussed topics in auditing. While auditing regulators and society tend to consid-
er professional skepticism as a universal panacea for problems related to insufficient audit
quality and increasingly demand from auditors to raise the bar with regard to the application
and maintenance of professional skepticism, auditing standards involve a rather neutral view,
recognizing that extremely high levels of professional skepticism can jeopardize audit effi-
ciency (cost and timeliness), while not necessarily improving audit effectiveness (quality).
Acknowledging the fundamental importance of profession skepticism to auditing, academic
research has exerted considerable effort in exploring the nature, determinants and implications
of professional skepticism in auditing. However, despite of the great academic and normative
interest, professional skepticism is a concept which due to its complexity and latency still in-
volves more questions than answers.
In the present work, Kristina Yankova addresses the question of whether professional skepti-
cism can prevent cognitive biases like information order effects from distorting the quality of
auditors’ judgments. She makes a significant contribution to the existing auditing literature by
providing a profound theoretical and empirical analysis on the role and behavioral conse-
quences of professional skepticism in generic audit settings. The present piece of research
impresses by its conceptual finesse, scientific rigor and high academic standard, but also by
its enormous relevance to current debates in auditing. The thesis of Kristina Yankova is a
must read for auditing researchers, standard setters and regulators as well as readers interested
in current topics in auditing.
I thank Kristina Yankova very much for her excellent support in teaching and research during
her research assistance activities at my chair and wish the work the deserved positive reso-
nance.
Prof. Dr. Annette G. Köhler
V
Acknowledgements
I relish this opportunity to acknowledge and thank all those who have generously provided
their time, efforts, knowledge and guidance to support me in this academic and personal expe-
rience of a lifetime. In particular, my deepest gratitude goes to my supervisor, Prof. Dr. An-
nette Köhler. Her generous academic and personal support and guidance throughout the last
years have been essential to the successful completion of this work. She is a great mentor, and
I cannot exaggerate my gratitude to her for encouraging me to develop and grow in my own
way and for changing the course of my life. I also want to express my gratitude to my co-
promoter, Prof. Dr. Christopher Koch, for teaching me the basics of experimental research
and for providing valuable comments on earlier versions of my case study. Furthermore, I
wish to thank Prof. Dr. Peter Anker and Prof. Dr. Joachim Prinz for their participation in my
dissertation committee and for the lively discussion.
A special word of thanks goes to Dirk Egbers for his ongoing tremendous support throughout
the process. He generously supplied his time, energy and expertise to make this project possi-
ble, and his contribution to the present work cannot be overstated.
I have benefited enormously from my personal and academic interactions with a number of
colleagues at the University of Duisburg-Essen. I especially want to thank Jun. Prof. Dr. Marc
Eulerich, Andrea Gantzhorn, Christian Hanke, Dr. Thilo Helpenstein, Meike Herbers, Dr.
Jens Jany, Dr. Katharina Köhler-Braun, Yu-Hui Liu, Dr. Mirko Mertenskötter, Jan Michael,
Dr. Johanna Souad Qandil, Monika Schmock, Dr. Jochen Theis, Markus Urban as well as the
student assistants at the Chair of Accounting and Auditing at the University of Duisburg-
Essen for their invaluable help and support. It has been a great pleasure working in such an
outstanding, cohesive, cheerful and friendly team.
Finally, I never could have accomplished this work without the unwavering encouragement,
support, patience and love of my family and my friends. I am so grateful to have you in my
life, and I thank you from the bottom of my heart for supporting me, for challenging me, for
teaching me, but most of all, for believing in me. I dedicate this work to my family.
Kristina Yankova
VII
Table of Contents
Foreword ................................................................................................................................... V
Acknowledgements .................................................................................................................VII
1 Introduction ........................................................................................................................ 1
X
4.2.3 Professional Skepticism as in the Auditing Literature ......................................... 95
4.2.3.1 Neutrality Perspective ................................................................................. 95
4.2.3.2 Presumptive Doubt Perspective .................................................................. 97
4.2.3.3 Other Perspectives .................................................................................... 100
4.3 The Nelson Model .................................................................................................. 102
4.3.1 Model Overview ................................................................................................. 102
4.3.2 Determinants ...................................................................................................... 104
4.3.2.1 Knowledge ................................................................................................ 104
4.3.2.2 Traits ......................................................................................................... 108
4.3.2.2.1 Problem-Solving Ability....................................................................... 108
4.3.2.2.2 Ethics and Moral Development ............................................................ 109
4.3.2.2.3 Dispositional Skepticism ...................................................................... 111
4.3.2.2.4 Dispositional (Dis)Trust ....................................................................... 112
4.3.2.2.5 Culture .................................................................................................. 113
4.3.2.2.6 Gender .................................................................................................. 115
4.3.2.3 Incentives .................................................................................................. 116
4.3.2.3.1 Skepticism-Promoting Incentives ......................................................... 117
4.3.2.3.1.1 Oversight and Enforcement ........................................................... 117
4.3.2.3.1.2 Litigation and Reputation Loss...................................................... 117
4.3.2.3.1.3 Audit Firm Culture ........................................................................ 118
4.3.2.3.1.4 Accountability ............................................................................... 119
4.3.2.3.2 Skepticism-Inhibiting Incentives .......................................................... 120
4.3.2.3.2.1 Client Retention Pressure and Fee Dependence ............................ 121
4.3.2.3.2.2 Tenure ............................................................................................ 123
4.3.2.3.2.3 Familiarity ..................................................................................... 124
4.3.2.4 Judgment ................................................................................................... 125
4.3.2.5 Action ........................................................................................................ 127
4.4 Hurtt’s Notion of Trait Professional Skepticism and the Hurtt Scale .................... 128
4.4.1 General Remarks ................................................................................................ 128
4.4.2 Constituent Elements of Trait Professional Skepticism ..................................... 130
4.4.2.1 Questioning Mind ..................................................................................... 130
4.4.2.2 Suspension of Judgment ........................................................................... 132
4.4.2.3 Search for Knowledge ............................................................................... 133
XI
4.4.2.4 Interpersonal Understanding ..................................................................... 135
4.4.2.5 Autonomy ................................................................................................. 136
4.4.2.6 Self-Esteem ............................................................................................... 137
4.4.3 Scale Development and Validation .................................................................... 138
4.4.4 Empirical Evidence ............................................................................................ 143
4.5 The Influence of Trait Skepticism on Auditors’ Belief Revisions ......................... 145
5 Empirical Analysis ......................................................................................................... 150
XII
Appendix 2 ............................................................................................................................. 207
XIII
List of Abbreviations
ca. circa
CAQ Center for Audit Quality
CICA Canadian Institute of Chartered Accountants
Co. Company
c.p. ceteris paribus
CPA Certified Public Accountant
CPAB Canadian Public Accountability Board
df degrees of freedom
Dipl.-Kff. Diplom-Kauffrau
XV
Dr. Doctor
f. following
fMRI functional magnetic resonance imaging
FRC Financial Reporting Council
GmbH & Co. KG Gesellschaft mit beschränkter Haftung & Compagnie Kommandit-
gesellschaft
GRE Graduate Record Examination
H1 Hypothesis 1
H2 Hypothesis 2
HGB Handelsgesetzbuch
http hypertext transfer protocol
No. Number
NZAuASB New Zealand Auditing and Assurance Standards Board
p. page
XVII
PCAOB Public Company Accounting Oversight Board
PhD philosophiae doctor
Prof. Professor
ProMES Productivity Measurement and Enhancement System
PS Prüfungsstandard
PSA Philosophy of Science Association
rd. rund
S. Seite
SAS Statement of Auditing Standard
SAT Scholastic Assessment Test
sbr Schmalenbach Business Review
SbS Step-by-Step
SD Standard Deviation
SE Standard Error
SEC Securities and Exchange Commission
sic sic erat scriptum
SOX Sarbanes-Oxley Act
SPSS Statistical Package for the Social Sciences
Sr. Senior
SSRN Social Science Research Network
StB Steuerberater
STPI State-Trait Personality Inventory
vs. versus
WiPrPrüfV Wirtschaftsprüferprüfungsverordnung
WP Wirtschaftsprüfer
WPK Wirtschaftsprüferkammer
WPO Wirtschaftsprüferordnung
www world wide web
XIX
List of Figures
XXI
List of Tables
XXIII
List of Symbols
% percent
& and
§ paragraph
* significance at a 5%-level
** significance at a 1%-level
*** significance at a 0.1%-level
῀ complement
│ conditional on
∩ conjunction
ൈ interaction
െെ information presentation order where two positive items are followed
by two negative items
െെ information presentation order where two negative items are followed
by two positive items
ߙ sensitivity to negative evidence
ߚ sensitivity to positive evidence
ߚ݅ regression coefficient of the parameter i
ߝ error term
ߟʹ eta-squared (effect size)
ܨ F-statistic
ܰ number of subjects / observations
probability / significance value
R reference point / coefficient of determinations
ݎ Pearson product-moment correlation coefficient
ݎ௦ Spearman rank correlation coefficient
ܵ degree of belief in a hypothesis after evaluating k pieces of evidence
ܵ initial strength of belief
ܵିଵ anchor (prior belief)
ݏሺݔ ሻ subjective evaluation of the kth piece of evidence
ݐ t-statistic
ݓ adjustment weight for the kth piece of evidence
ݖ standardized value
XXV
ACT auditor certification
AGE age
BELREV belief revision
CERTAIN certainty in the own assessment
CONC concentration
FINBEL final belief
GEN gender
GEXP general audit experience
IEXP industry-specific experience
IMPC1 assessments of the importance of the additional piece of information
concerning the rejection of state funding
IMPC2 assessments of the importance of the additional piece of information
concerning the insolvency of a major supplier
IMPC3 assessments of the importance of the additional piece of information
concerning the favorable external report and banks’ willingness to ne-
gotiate
IMPC4 assessments of the importance of the additional piece of information
concerning the announcement of a cash capital increase
INFORD information order
INTBEL initial belief
INTENS intensity of thought
POS position in the firm hierarchy
PS professional skepticism
SUFINF sufficiency of information
TEXP task-specific experience
XXVI
1 Introduction
The aim of an external audit is to express an independent professional opinion (i.e., a judg-
ment) on the compliance of financial statements1 with relevant generally accepted accounting
principles.2 In forming an overall opinion, auditors evaluate a wealth of facts and circum-
stances that relate to what is accurate, true, fair, material, and sufficient.3 Judgment pervades
practically every facet of auditing.4 The fundamental importance of judgment in auditing is
evident not only in the frequent use of the term in national and international auditing stand-
ards,5 but also in the substantial academic interest and research endeavor dedicated to this
topic in the last decades. Overall, this research has evolved around the notion that it is essen-
tial to profoundly understand the foundations, determinants, and complex processes that un-
derlie auditor judgment6 in order to improve judgment quality.7
An important feature of the judgment process in auditing is its sequential nature. In particular,
in arriving at a judgment, auditors generally proceed gradually by starting with an initial posi-
tion (i.e., belief) and successively adjusting it in light of newly incoming evidence rather than
by obtaining complete information en bloc and integrating it all within a single overall belief
update.8 Psychological theory, however, suggests that sequential information processing bears
the potential for systematic biases in judgment, the so-called “information order effects”,
which refer to the dependence of an individual’s response on the order in which new infor-
mation is presented and processed. Order effects imply that the very same informational stim-
uli can induce a fairly different response contingent on the order of presentation of the stimu-
li.9 Under a normative viewpoint, individuals’ responses shall be based on information
1
For a definition of the term “financial statements” and its constituent parts, see ISA 200.13(f). For a brief
historical overview of the origins of modern auditing dating back to the manorial households from the four-
teenth century, see FRC (2012): 9-11. For a very broad definition of auditing as a certification practice, see
Francis (2011b): 318.
2
See Anderson/Koonce/Marchant (1991): 46; Bazerman/Morgan/Loewenstein (1997): 91; Quick/Warming-
Rasmussen (2007): 1008.
3
See Boylan (2008): 230.
4
See KPMG (2011): 1; Solomon/Shields (1995): 139; Trotman (1998): 115.
5
See Trotman/Tan/Ang (2011): 279.
6
Note that a precise delimitation of the terms “auditor judgment” and “audit judgment” is missing in the rele-
vant auditing literature where both terms have frequently been used interchangeably. For an instance of the
parallel use of both terms, see the meta-analytical paper of Trotman (1998). The present work attempts to
use the terms “audit judgment” and “auditor judgment” as unambiguously and uniformly as possible. Never-
theless, in some situations, a clear-cut distinction between both terms appears not possible. In addition, it
should be noted that the terms “auditor judgment” and “auditors’ judgment” are used synonymously herein.
7
See Anderson/Koonce/Marchant (1991): 44; Libby (1981): 121; Libby/Lewis (1977): 245; Trotman (1998):
118. For a discussion of judgment quality, see Section 2.4.1.
8
See Gibbins (1984): 110.
9
See Anderson (1981): 144; Hogarth/Einhorn (1992): 1f.
1
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2_1, © Springer Fachmedien Wiesbaden 2015
content (e.g., diagnosticity, relevance, reliability), not on the order in which the information
happens to be received and evaluated.10
Although information order effects are not an audit-specific phenomenon, their importance
and relevance to auditing is fundamental for several reasons. First, auditing standards do not
provide any practical guidance on how individual pieces of evidence shall be aggregated into
an overall opinion. Second, order effects can have serious implications for audit efficiency
and effectiveness. Specifically, the order of information presentation can influence auditors’
search for additional evidence and the amount of audit testing considered necessary to cor-
roborate the own opinion, thereby potentially resulting in over- or underauditing.11 Third, as
indicated above, order effects can yield not substantively justified beliefs and judgments,
thereby jeopardizing the quality of auditor judgments.12 Fourth, auditors frequently obtain
information and data from the client who determines the order and sequence of information
presentation, and thus has the opportunity to unduly influence auditors’ views and judgments.
The existing auditing literature has broadly recognized information order effects as a serious
threat to the quality of auditor judgments and has dedicated considerable research endeavor to
exploring this cognitive phenomenon in auditing settings and identifying mechanisms for mit-
igating order effects. Overall, although prior research on information order effects has provid-
ed valuable insights into the roots and determinants of order effects, the current knowledge of
the influence of personal dispositions on the susceptibility to and magnitude of information
order effects in auditor judgments is still very limited.
Similarly, although the concept of professional skepticism is fundamental to auditing,13 the
knowledge of what skepticism essentially is, how it manifests itself, how it affects the way in
which auditors process information, and how it interacts with other relevant variables is still
fairly restricted.14 Until very recently, research on professional skepticism has been conducted
in an ad hoc manner, without the benefit of a unifying framework, a thorough conceptualiza-
tion, or an explicit measure. With the publication of the influential conceptual works of Nel-
son (2009) and Hurtt (2010), first important steps toward overcoming these deficiencies and
developing a deeper understanding of the nature of professional skepticism in auditing have
been made.
In his theoretical analysis of the antecedents of professional skepticism in auditing, Nelson
(2009) recognizes that cognitive biases like information order effects can systematically un-
dermine auditors’ professional skepticism.15 This contention is echoed by Glover and Prawitt
10
This aspect is further discussed in Section 2.4.2.1.4 as well as in Section 3.3.
11
See Ashton/Ashton (1988): 638; Krull/Reckers/Wong-On-Wing (1993): 144; Messier/Tubbs (1994): 58;
Tubbs/Messier/Knechel (1990): 459. This aspect is also addressed within Section 3.3.
12
See Schreiber (2000): 170; Schwind (2011): 73.
13
See Chekushina/Loth (2014): 87; Hurtt (2010): 149.
14
See Hammersley (2011): 108; Harding/Trotman (2011): 8.
15
See Nelson (2009): 13f. For a detailed discussion of the Nelson model of professional skepticism, see Sec-
tion 4.3.
2
(2013).16 In a similar vein, the American Institute of of Certified Public Accountants (AICPA)
acknowledges the potential for cognitive traps to compromise professional skepticism and
calls for future research on the cognitive and behavioral aspects underlying skepticism.17 The
present work responds to this call by investigating the interrelation between auditors’ profes-
sional skepticism and their susceptibility to information order effects in belief updating and
judgment formation.
Building on the seminar work of Hurtt (2010), herein professional skepticism is viewed as an
innate characteristic of an auditor. By conceptualizing professional skepticism as a personal
disposition (i.e., a trait), the present study offers some important insights into the role of indi-
vidual characteristics in the context of information order effects. In addition, the study pro-
vides valuable insights into the behavioral manifestations of trait professional skepticism in a
context that involves high complexity, uncertainty, limitedly available information, and no
skepticism-inducing conditions such as irregularities, fraud, or high environmental risk.
Overall, the objectives of the present work are two-fold: first, to provide additional experi-
mental evidence on the influence of information order effects on auditors’ judgments, and
second, to explore whether auditors exhibiting measurably different degrees of trait profes-
sional skepticism adjust their beliefs in a different manner and are to a different extent subject
to information order effects. In addressing these objectives, a positive scientific approach has
been adopted in the present study. This approach essentially aims at drawing causal infer-
ences.18 In investigating the interplay between information order effects and professional
skepticism, the present work builds on psychological and auditing models and theories to de-
rive the relevant research hypotheses and questions, which are then subjected to experimental
testing. Hence, the procedure adopted herein can be classified as deductive.19
As the ultimate aim of judgment and decision making research in auditing is to enhance the
quality of auditors’ judgments and decisions,20 the insights gained from the inquiry into the
relationship between information order effects and professional skepticism are used to derive
suggestions on how order effects can be reduced and the normative ideal of unbiased infor-
mation processing and judgment21 approached. Basically, this part of the study involves a
normative focus.22 However, it is important to note that in light of the current state of incom-
16
See Glover/Prawitt (2013): 8-11.
17
See AICPA (2010): 2.
18
See Kornmeier (2007): 28.
19
See Ghauri/Grønhaug (2005): 15; Wilson (2014): 13. The deductive approach has also been referred to as a
“theory then research” approach and has generally been affiliated with the quantitative type of research. An
alternative scientific approach relates to induction, or qualitative research, which proceeds from data collec-
tion and observation of findings toward the development of theory. This approach has also been referred to
as “research then theory”. See Ghauri/Grønhaug (2005): 15; Peecher/Solomon (2001): 194.
20
See Belkaoui (1989): xi; Kotchetova/Salterio (2007): 557; Peecher/Solomon (2001): 194; Trotman (1998):
118.
21
See Stahl (2012): 135.
22
See Chmielewicz (1994): 14; Fülbier (2004): 267f.; Kornmeier (2007): 28.
3
plete knowledge and understanding of the complex causal interrelations underlying auditors’
judgment and behavior, the derivation of prescriptions does not appear tenable.23
The reminder of this thesis proceeds as follows: In Section 2, the broad theoretical back-
ground is outlined. At the outset of this section, the need for an adoption of a psychological
lens in behavioral research is discussed (Section 2.1), followed by a consideration of the
structure of human cognition (Section 2.2) and the information processing approach (Section
2.3). Building on this broad psychological foundation, the central features and determinants of
auditor judgment are discussed (Section 2.4).
Section 3 sheds light on the first major pillar of the present work, the cognitive phenomenon
of information order effects in belief updating. At the beginning of this section, some general
remarks on the nature of beliefs are made (Section 3.1) as a prelude to the detailed discussion
of the psychological model predicting the emergence of information order effects, the belief-
adjustment model developed by Hogarth and Einhorn (1992) (Section 3.2). The section pro-
ceeds with a consideration of the fundamentality of the process of belief updating to auditing
(Section 3.3) and concludes with a review of the relevant auditing literature (Section 3.4).
The second major pillar of the present work, the fundamental concept of professional skepti-
cism in auditing, is discussed in Section 4. After some general introductory remarks are pro-
vided (Section 4.1), an inquiry into the meaning and nuances attached to the term “skepti-
cism” in the social sciences as well as in auditing is presented (Section 4.2). Building on this
broad terminological foundation, the conceptual model of determinants of professional skep-
ticism in auditing developed by Nelson (2009) is presented and underpinned with relevant
research findings (Section 4.3). The section proceeds with a detailed discussion of Hurtt’s
(2010) notion of professional skepticism as an innate personal disposition (Section 4.4). The
in-depth theoretical consideration of professional skepticism in auditing concludes with a syn-
thesis of the key insights concerning the influence of trait professional skepticism on auditors’
belief revisions in light of information order effects (Section 4.5).
Building on these broad interdisciplinary theoretical considerations, the empirical analysis is
presented in Section 5. After a general description of the research method (Section 5.1) and
the experimental design employed (Section 5.2), the research instrument utilized to capture
auditors’ belief revisions and their degrees of innate skepticism is outlined (Section 5.3).
These methodological considerations are followed by a description of the participants in the
experimental study (Section 5.4). At the end of this section, the results of a number of analyti-
cal procedures and statistical tests conducted to examine the research hypotheses and to pro-
vide additional explorative findings are presented (Section 5.5).
23
See Peecher/Solomon (2001): 194. Note that in contrast to the German literature on scientific approaches
and theories, in the relevant English literature, the prescriptive rather than the normative focus of research is
viewed as the last stage of scientific understanding and advance. See Peecher/Solomon (2001): 194.
4
The conclusion is contained in Section 6. The section provides a summary and discussion of
the basic insights and implications of the present work (Section 6.1) as well as an overview of
the limitations of the study and the potential directions for future research (Section 6.2).
5
2 Broad Theoretical Background
This section deals with some foundational psychological notions which provide valuable in-
sights into the cognitive foundations of human judgment as a prelude to the inquiry into the
central features and determinants of auditor judgment. It should be noted that the considera-
tions within this section are of general nature and are not intended to be all-encompassing.24
Rather, they aim to provide a broad theoretical basis for the study of information order effects
and professional skepticism in the context of auditors’ belief revisions.
Economics, the queen of social sciences, is built upon the fundamental notion of rationality,
which serves as the central behavioral assumption in explaining economic phenomena. 25 Ac-
cording to expected utility theory26, which is the core formal account of rationality, human
behavior can be viewed as a choice from a set of alternatives, each of which is characterized
by a designated value as well as a probability of occurrence. The economic agent effortlessly
and thoroughly processes the vast amounts of information27, assesses all available options in
terms of their likelihood and expected value, and eventually chooses the alternative that prom-
ises the maximal expected utility, i.e., goal achievement.28 As new information arrives, the
rational economic man is assumed to revise his previously held probabilistic positions in strict
accordance with logical rules. The core normative principle typically applied for updating
probabilities in light of new evidence is Bayes’ theorem, which represents a rigorous combi-
24
As claimed by Kaufman (1999): 363, the relevant literature in economics and psychology is immense.
Hence, it is barely possible to provide an overview that does more than merely scratch the surface. The same
applies to the voluminous auditing literature on judgment and decision making.
25
See, e.g., Hogarth/Reder (1986): 2. Overall, the rationality concept has been viewed as one of the most im-
portant achievements of the social sciences ever. See Simon (1980): 75. For an excellent and very insightful
review of the varieties of economic rationality from Adam Smith (1723-1790) to the present, refer to
Zouboulakis (2014): 5-139.
26
The origins of expected utility can be traced back to 1738 and the famous St. Petersburg essay of Daniel
Bernoulli (1700-1782). However, the real prominence of the notion of expected utility came only two centu-
ries later with the publication of the seminal work of von Neumann/Morgenstern (1944). See Abdellaoui
(2004): 15f; Kahneman (2003a): 164. For a discussion of the central aspects of expected utility theory, in-
cluding its violations, the most prominent of which are known as the Allais (1953) paradox and the Ellsberg
(1961) paradox, refer to Abdellaoui (2004): 15-30.
27
Information can be broadly defined as “zweckorientiertes Wissen” (Wittmann (1959): 14), i.e., a source of
knowledge for purposeful judgment and decision making. Information is an immaterial good with unique
characteristics, including relevance, validity, reliability, trustworthiness, availability, source, topicality, and
acquisition costs. See Ballwieser/Berger (1985): 1; Gemünden (1993): 847.
28
See Baron (2007): 24; Sontheimer (2006): 238. Essentially, goals can be viewed as criteria by which states
of affairs are evaluated. See Baron (2007): 23.
6
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2_2, © Springer Fachmedien Wiesbaden 2015
nation of contingent probabilities.29 Stated formally, the theorem runs along the following
lines: 30
ሺᑻுሻήሺுሻ
(1) ሺܪᑻܦሻ ൌ ሺᑻுሻήሺுሻାሺᑻுഥሻήሺுഥሻ
where
ሺܪሻ = a priori probability of event H;
ഥ ሻ = a priori probability of the complement of event H;
ሺܪ
ሺܦᑻܪሻ = probability of D given H;
ഥ ሻ = probability of D given the complement of H; and
ሺܦᑻܪ
ሺܪᑻܦሻ = probability of H given D (a posteriori probability of H).
Overall, the rationality concept and its major formal and logical pillars (e.g., the expected
utility paradigm and Bayes’ theorem) provide a reasonable and stringent account of human
behavior in that they reflect its purposefulness and goal-orientation.31 However, the unlimited
rationality assumed in economic theory does not reflect the manner in which individuals actu-
ally process information, update beliefs, form judgments, and arrive at decisions. As promi-
nently argued by Herbert Simon, as a consequence of the limited availability of information
as well as the restrictions of cognitive resources and computational abilities, human behavior
in the real world is “intendedly rational, but only boundedly so.”32
The concept of bounded rationality coined by Simon highlights the discrepancy between the
limitless human rationality that is at the heart of normative economic theory and the reality of
human behavior.33 The bounded rationality paradigm is built upon the notion that the capacity
of human cognition (involving both knowledge and computational ability) is not sufficient to
achieve the optimal outcomes posited in economic theory, especially under conditions of high
29
See Birnberg (1964): 108f.; Karni (2013): 5; Tisdell (1975): 266. As will be shown in Section 3.1, Bayes’
theorem represents the relevant normative benchmark for the process of belief revision in light of newly in-
coming evidence. Therefore, it is considered in more detail in the present section. Note in passing that the
theorem is named after Reverend Thomas Bayes (1702-1761) whose outrageous propositions and ground-
breaking ideas summarized in a work labeled “An Essay towards solving a Problem in the Doctrine of
Chances” were discovered post mortem by his friend Richard Price (1723-1791) and published in 1763.
30
See, e.g., Baron (2007): 32. For an extensive review of the Bayesian theory and a number of interdiscipli-
nary practical application of the Bayes’ rule, consult Damien et al. (2013).
31
See Simon (1980): 75.
32
Simon (2013): 88, italics in original. Similar arguments and quotes can also be found in his earlier publica-
tions (see, e.g., Simon (1955): 101; Simon (1957a): xxiv). It is instructive to note that the subsequent consid-
eration of the boundaries of human cognition does not aim to discard the fundamental normative models.
Rather, the main goal is to stress the importance of relaxing the utopic behavioral assumptions regarding
global rationality, thereby extending the descriptive and empirical validity of mainstream economics. See
Fromlet (2001): 64.
33
See Simon (1992): 3. For a concise consideration of the fundamental notion of bounded rationality, see
Simon (1997): 291-294. For critical remarks on this concept, see Marris (1992): 194-221.
7
complexity.34 Accordingly, in the real world, optimization is replaced by satisficing, i.e., a
search for a good and feasible solution rather than an optimal one.35
Importantly, the notion of bounded rationality does not imply that humans are consciously
and intentionally irrational, but merely that they possess limited cognitive capacities. 36 To
cope with cognitive resource constraints, people usually take mental shortcuts, i.e., they adopt
heuristics37 which simplify complex problems and provide rapid solutions at low cognitive
cost.38 The use of simplifying heuristics procedures typically results in intuitively reasonable
and acceptable solutions. However, in some cases, heuristics can also lead to systematic errors
in judgment and decision making, the so-called “biases”.39 In essence, biases represent devia-
tions from the normative standards of rationality.40
Overall, by assuming unlimited rationality, economic theory places severe demands on the
economic agent, while considering his/her internal structures and mental properties as a black
box.41 As aptly put by Simon, to explain real-world human behavior performed under condi-
tions of considerable complexity and dynamics of the environment, normative economic theo-
ry has to describe the economic agent “as something more than a featureless, adaptive organ-
ism; it must incorporate at least some description of the processes and mechanisms through
which the adaptation takes place.”42 In contrast to normative economic theory that prescribes
how individuals ought to behave under the orthodox assumption of perfect rationality, psy-
chology describes how people actually behave in the real world.43 The following section out-
lines the fundamental importance of the discipline of cognitive psychology for the under-
standing and exploration of human judgment and behavior.
34
See Simon (1957b): 198. In a similar vein, Gigerenzer (2007): 62f. argues that even for well-structured prob-
lems such as chess, the optimal solution and methods for achieving it are unknown. This uncertainty applies
all the more to less well-defined problems such as what stocks to invest in or whom to marry.
35
See Simon (1955): 108; Simon (1959): 262-264, 277; Simon (1997): 295-298.
36
See Simon (1992): 3.
37
For a broader discussion of the term “heuristic”, refer to Keren/Teigen (2007): 92f. For a general psycholog-
ical and accounting-related discussion of some central heuristics, including representativeness, availability
as well as anchoring and adjustment, see Belkaoui (1989): 203-231.
38
See Fiske/Taylor (2008): 12f., 197.
39
See Tversky/Kahneman (1974): 1124.
40
See Kahneman (2003b): 1449.
41
See Kaufman (1999): 363f.; Simon (1955): 103.
42
Simon (1959): 256. Henceforth, for the sake of readability, direct quotes are highlighted by italics. The in-
stances where italics were used in the original quotation are marked accordingly.
43
See March/Simon (1958): 138; McKenna (2000): 219; Over (2007): 3. It might be remarked that economists
have put considerable effort in defending the rationality paradigm and categorically distancing economic
theory from psychology. For a review of this issue, see Bruni/Sugden (2007): 146-171. Generally, econo-
mists criticize psychological research for its manner of generating lists of cognitive fallacies, while being
unable to provide a coherent alternative to normative models of rational human behavior and thought. In-
deed, cognitive psychological theories cannot achieve the coherence, elegance, and the formal precision of
normative models of rational judgment and choice. However, these qualities of the normative models are at
the expense of behavioral reality and descriptive validity. In contrast, psychology offers integrative models
with considerable descriptive power that can be applied to a wide range of behavioral phenomena in differ-
ent domains. See Kahneman (2003b): 1449.
8
2.1.2 Cognitive Psychology and its Contribution to Explaining Human Behavior
The term “psychology” stems from the Greed words “psyche” which means “mind”, “soul”,
“spirit”, and “logos” which means “knowledge”, “study”. Semantically, then, psychology is
the study of the human mind.44 Psychology can generally be divided into the subfields of ap-
plied and theoretical psychology. While the former focuses on the practical use of psychology
by individuals and society, and includes the subfields of clinical, counselling, environmental,
and educational psychology, the latter deals with the basic principles and mechanisms of how
the mind controls human behavior45. Theoretical psychology encompasses a wide array of
subfields, including perception, learning and memory, thinking, psycholinguistics as well as
cognitive, physiological, comparative, developmental, social, and personality psychology.46
The focus within this work is on cognitive psychology. The term “cognition” comes from the
Latin verb “cognoscere” which means “to become acquainted”, “to get to know”.47 Hence,
cognition refers to all the processes by which humans come to know the surrounding world.48
It encompasses virtually all human intellectual activities and mental processes, including sen-
sation, perception, attention, learning, memory, problem solving, thinking, imagining, and
language, among others.49 Consequently, as argued by Neisser (1967), “cognition is involved
in everything a human being might possibly do” and essentially, “every psychological phe-
nomena [sic!] is a cognitive phenomena [sic!].”50
Even though cognition is now widely recognized as a central and integral part of psychology,
the cognitive approach has not always been that prominent and non-contentious.51 Specifical-
ly, for nearly five decades (from ca. 1913 to ca. 1960) psychology was dominated by the be-
havioristic stream. The latter considered the behavior (i.e., the physical movements) of organ-
isms as the only real and scientifically justified subject of psychological study; Mental
processes, in contrast, were viewed as illusory, and the study of the human mind was consid-
ered unworthy.52 During the 1960s, however, cognition re-emerged as a scientifically legiti-
mate focus for psychological theory. This development is commonly referred to as the “cogni-
tive revolution” in psychology and was reinforced by important developments in the fields of
artificial intelligence, computer science, and human problem solving.53
44
See Gross (2009): 127.
45
Due to the inseparable relationship between mind and behavior, psychology has been typically referred to as
the science of human mind and behavior. See, e.g., the correspondingly titled works by Blair-Broeker/Ernst
(2007), Gross (2012) and Passer/Smith (2010).
46
See Wickelgren (1979): 3. For a concise historical overview of psychology as a science, see Wärneryd
(1999): 7-10.
47
See Anshakov/Gergely (2010): 1.
48
See Gross (2009): 128; Neisser (1967): 4; Neisser (1976): 1.
49
See Edelmann/Wittmann (2012): 109; Gross (2009): 128; Solso/MacLin/MacLin (2008): 10.
50
Neisser (1967): 4.
51
See Fiske/Taylor (2008): 21.
52
See Baars (1986): 4f., 9; Miller (2003): 141.
53
See Fiske/Taylor (2008): 9; Miller (2003): 142. For a thorough discussion of the cognitive revolution in
psychology, refer to Baars (1986).
9
In the last few decades, the cognitive approach has revolutionized and considerably advanced
the field of psychology and is nowadays an indispensable and essential part of theoretical
psychology. As will be demonstrated throughout the present work, the influence of the cogni-
tive approach reaches far beyond the field of psychology and has marked a number of other
scientific disciplines, among which is also the field of auditing.
In conclusion, it should be noted that both economics and cognitive psychology provide valu-
able insights into and explanations for auditing phenomena, so that the adoption of an inter-
disciplinary approach in auditing research appears inevitable for advancing the state of
knowledge and cognizance in this field. As argued by Libby (1991), it is important to bear in
mind that the majority of auditing phenomena explored by academic research are essentially
economic phenomena. Consequently, neglecting the economic foundation of individual,
group, and institutional behavior will yield a fairly fragmentary picture. Not less important is
to recognize, however, that auditing involves human decision-makers.54 Consequently, it is
crucial to consider individual factors in the study of judgment and decision making in auditing
parallel to the economic motives driving auditors’ behavior. With other words, through the
combination of the normative rigor of economic theory and the descriptive validity of psycho-
logical models and concepts, a more complete picture of auditing phenomena and auditors’
behavior can be achieved.
Building on the insight of the essential importance of cognitive aspects for the study of human
behavior, subsequently, the basic structure of the human cognition is outlined.55
The most essential feature of the human mental system is its modularity.56 There is a wide
spectrum of notions of modularity.57 In the present work, modularity is understood in the
sense that the human mind is constructed of distinct modules or subsystems, each of which
fulfils a specific function within the operation of the whole system.58 Overall, the mind has
been conceptualized as consisting of multiple parts, including cognition and emotion, reason
and intuition, consciousness and unconsciousness, automaticity and control, ego and id, just to
54
See Libby (1991): 19f.
55
Note that the following discussion involves more general features of the cognitive structure, particularly the
two prevalent types of cognitive processing. For an extensive neurological review of cognition, consult
Koziol/Budding (2009).
56
See Fodor (1983); Garfield (1991): 1; Gazzaniga (1985): 74, 77-80; Gilbert (1991): 109; Smith (2003): 108.
57
For a discussion of the range of notions on modularity, see Carruthers (2003): 67-71; Carruthers (2006): 1-
3.
58
See Carruthers (2006): 2. It might be remarked that this view is in line with the massive modularity hypoth-
esis advocated by Carruthers (2003): 68; Pinker (1994): 420; Pinker (1997): 27-31; Sperber (1996): 123-
129 and Tooby/Cosmides (1992): 94, 113, among others. According to this hypothesis, the entire cognitive
apparatus can be viewed as modular. That is, both lower-level and higher-level mental processes possess a
modular structure. See Visala (2011): 35. This notion contradicts the modularity view set forth by Fodor
(1983) according to which only the peripheral sensory and perceptual systems of the human mind are modu-
lar, while the higher-order cognitive systems are conceived as central and holistic. See Fodor (1983): 47-
119.
10
name a few. Consequently, a deeper understanding of mental phenomena is only possible by
considering the structure and interaction of the multiple parts of the mind rather than by an
isolated analysis of the individual components.59
The present work builds on recent developments60 in psychology which have proposed the so-
called “dual-system” or “dual-process” view on human cognition.61 This view highlights the
distinction between two general constituent blocs of the cognitive system (categories of cog-
nitive processing): intuition and reasoning. In general terms, intuition involves automatic,
effortless, quick, and reflexive processing. Intuitive impulses and associations normally come
to mind without conscious search or computation. In contrast, reasoning involves deliberate,
effortful, slow, and analytic processing.62
The characteristics of the two grand categories of cognitive operations are schematically
summarized in Figure 1. Following Stanovich (1999) and Kahneman and Frederick (2005),
the generic labels “System 1” and “System 2” are employed to reflect intuitive and deliberate
reasoning, respectively, and the term “system” is used to refer to a bunch of cognitive pro-
cesses (subsystems) that can be categorized by their speed, their deliberateness and the con-
tents on which they operate.63 As the classification criteria depicted in Figure 1 are all contin-
ua, the placement of strict border lines between the systems is not possible.64
59
See Gilbert (1999): 4.
60
Note that the idea of fragmenting the human mind and cognition into active (deliberate) and passive (auto-
matic) domains is not a new one. It can be traced back to the influential work of the famous French philoso-
pher and mathematician René Descartes (1596-1650). See Gilbert (1991): 108. In the last decade, this notion
was revived in psychological research and is now a fairly prominent and topical branch of the contemporary
psychological literature. Overall, a variety of differently nuanced dual process theories have been established
in psychology in the last years. Although different in focus and detail, these theories share the idea of sepa-
rating quick and associative cognitive operations from slow and controlled ones. See Evans (2008): 270;
Kahneman/Frederick (2005): 267.
61
For an excellent overview of dual-process theories of higher-order cognition, i.e., thinking, reasoning, judg-
ment and decision making, see Evans (2008): 256-271. The consideration of lower-order cognition, i.e., per-
ception, attention, development of motor skills, etc., is beyond the scope of Evan’s review as well as the pre-
sent work. For a critical consideration of dual-system theories, refer to Keren/Schul (2009).
62
See Kahneman (2003b): 1450; Kahneman/Frederick (2002): 51.
63
See Kahneman/Frederick (2005): 267; Stanovich (1999): 126. Note that the numeration of the two systems
follows their evolutionary development. See Weldon/Corbin/Reyna (2013): 52. Recently, dual-process theo-
rists increasingly highlight the need to replace the somewhat ambiguous “System 1”-“System 2”- terminolo-
gy with a new one which should be plural or more neutrally formulated to make clear that there is not a sin-
gle automatic and a single controlled system, but rather a set of subsystems and processes in the brain. For
such claims and a number of relevant references along these lines, see Stanovich (2011): 18f. For a list of al-
ternative labels used throughout the relevant dual-process theory literature to refer to System 1 and System 2
processing, including “on-line thinking” versus “off-line thinking”; “heuristic processing” versus “analytic
processing”; “associative system” versus “rule-based system”; “reflexive processing” versus “reflective pro-
cessing”; “stimulus bound processing” versus “higher order processing”, see Evans (2008): 257; Stanovich
(2004): 35.
64
See Kahneman/Frederick (2005): 288.
11
PERCEPTION INTUITION REASONING
System 1 System 2
Fast Slow
Parallel Serial
PROCESS
Automatic Controlled
Effortless Effortful
Associative Rule-governed
Slow-learning Flexible
Emotional Neutral
65
See Stanovich (2011): 19. For instance, individuals cannot help themselves but understand simple words and
phrases in their native language, nor can they refrain from knowing that 2+2=4. For such claims and a list of
further examples for System 1 processing, see Kahneman (2011): 21f.
66
See Kahneman (2003b): 1451f.; Stanovich (1999): 126; Stanovich/West (2000): 659.
67
As identified by Stanovich (1999): 126f., in contrast to System 2, which is considered to relate to analytic
intelligence, i.e., the kind of intelligence measured via psychometric test like IQ or SAT, System 1 rather re-
lates to interactional intelligence, i.e., intelligence in terms of utilizing a set of computational heuristic strat-
egies, tactics, and meta-rules. See also Prevignano/Thibault (2003): 158. For a discussion of the concept of
intelligence and a critical note on the use of IQ tests to measure intelligence, see Sternberg (1988). For a
brilliant discussion of the common misconception of intelligence as a proxy for rationality, see Stanovich
(2009). In this context, he uses the term “disrationalia” to refer to the phenomenon of irrationality despite
adequate intelligence. See Stanovich (2009): 7.
68
As indicated by Schneider/Shiffrin (1977): 2f., automatic processing generally utilizes the nearly limitless
long-term memory store, while controlled processing claims the limited short-term memory store. For a de-
tailed discussion of human memory, refer to Baddeley (1997) and Loftus/Loftus (1976). The distinction be-
tween short- and long-term memory is also touched upon in Section 2.4.2.3.1.
69
See Evans (2008): 257; Evans (2013a): 116f.; Stanovich (2009): 78. Importantly, this finding does not nec-
essarily imply that System 1 is less capable than (or inferior to) System 2. See Kahneman/Frederick (2002):
51.
70
See Kahneman/Frederick (2002): 51; Stanovich (1999): 126f.; Stanovich (2011): 20. For a list of instances
of typical System 2 operations, see Kahneman (2011): 22.
12
restricted by working memory capacity. Evolutionarily, System 2 is the more recent cognitive
system and is conceived to be either a uniquely human property or at least particularly devel-
oped in the humankind.71
As to the role of consciousness, it is generally assumed that System 2 processing is typically
conscious and requires a great deal of attention72, while System 1 processing can be either
conscious or unconscious.73 With regard to the error proneness of the two systems, it is im-
portant to note that even though System 1 typically makes use of heuristic strategies, cogni-
tive biases are equally attributed to the intuitive operations of System 1 as well as the analytic
processing of System 2. With other words, the greater deliberation and cognitive effort ger-
mane to System 2 processing does not necessarily guarantee bias-free responses.74 This is also
evident in light of the interplay between the two cognitive systems. Specifically, as a particu-
lar stimulus or problem is encountered, System 1 typically generates intuitive responses –
impressions, associations, stereotypes, and tendencies – which are transmitted and suggested
to System 2. The latter has a monitoring and corrective function, and it can endorse, amend,
or abolish the input from System 1.75 However, as empirical research has shown, this function
does not always detect and correct potential flaws in the input from System 1. In fact, System
2 processing is typically found to endorse the intuitive stimuli generated by System 1. 76 As
argued by Stanovich (2009), in order to override responses generated by System 1, System 2
must be capable of suppressing the System 1 stimulus and generating a superior (i.e., bias-
free) response for replacing the potentially flawed input.77 Consequently, the judgments that
people make and the pertinent mistakes that they commit largely depend on the input from
System 1 as well as on the functionality and effectiveness of the corrective mechanisms of
System 2.78 A second important insight emerging from the consideration of the interplay be-
tween the dual cognitive systems is that the majority of higher-order cognitive tasks (e.g.,
problem-solving, judgment, decision making) normally involve a combination of automatic
and controlled processing.79
71
See Evans (2008): 257; Evans (2013a): 116f.
72
For a general discussion of attention, including the origins of the concept, different forms of attention (selec-
tive vs. intensive), attention bottlenecks, and pertinent capacity models, refer to Kahneman (1973): 1-12.
73
See Chen/Chaiken (1999): 86. Moskowitz/Skurnik/Galinsky (1999): 33 argue, however, that even intent, will,
and control can operate beyond conscious awareness. Hence, System 2 processing may also be sometimes
unconscious.
74
See, e.g., Evans (2008): 267; Evans (2013a): 126f.; Evans (2013b): 113; Moskowitz/Skurnik/Galinsky
(1999): 13; Van Boven et al. (2013): 395f.
75
See Kahneman/Frederick (2002): 51. For a brief consideration of the interaction between System 1 and
System 2, see Kahneman (2011): 24f.
76
See Kahneman/Frederick (2002): 58f.
77
See Stanovich (2009): 23. Schneider/Shiffrin (1977): 2 argue that automatically generated responses are
relatively difficult to ignore, suppress, or modify.
78
See Kahneman (2003b): 1467. Impending factors for the quality of both automatic and controlled processing
include, among others, time pressure, stress, and distraction, i.e., the parallel computation of different cogni-
tive tasks. See Finucane et al. (2000): 5, 8; Gilbert (1991): 110; Kahneman/Frederick (2002): 57f.
79
See Bargh (1989): 4-7.
13
Finally, it is important to note that System 1 and System 2 have been theorized to operate
concurrently.80 As argued by Wilson (2002), the unique architecture of the human mind al-
lows for a plethora of cognitive tasks to be accomplished simultaneously by performing sub-
conscious processing of a substantial part of the informational stimuli, while consciously
working on other issues.81 Research on the two systems of human cognition indicates that
System 1 processing is prevalent in the real world. This observation can largely be attributed
to the cognitive economy and efficiency associated with System 1 processing.82
Overall, the system of human cognition can be viewed as a remarkable computational device
that is fairly efficient and adaptive to changes in its environment. However, this impressive
system differs significantly from the standard of rationality assumed in economic theory. 83
Drawing on the dual-system terminology, the rational economic agent can be described as
possessing a single cognitive system (rather than a modular one) that has the analytical and
computational ability of a flawless System 2 and the high speed and low effort of System 1.84
In reality, the human mental system has limited perceptual85, processing, and computational
capabilities and operates under conditions of uncertainty and sparse cognitive resources which
have to be distributed across a number of competing tasks and problems. 86 In consequence,
distortions in both intuitive and analytic cognitive processing are bound to arise. 87
The preceding considerations and arguments highlight the critical importance of taking into
account the structure and limitations of human cognition in order to attain a better understand-
ing of human behavior and thought. The general psychological foundations of the present
work are rounded off with a glance at the information processing approach. This approach has
proved particularly influential and prominent in cognitive psychology and has found a broad
application in auditing research in the last few decades.88
80
See Chen/Chaiken (1999): 76; Evans (2013a): 121; Kahneman/Frederick (2002): 51; Stanovich (2009): 22.
In this context, Moskowitz/Skurnik/Galinsky (1999): 33 emphasize that automatic and analytic processes are
conceptualized as “dual” rather than “dueling”.
81
See Wilson (2002): 5, 8-10.
82
See Chen/Chaiken (1999): 74; Moskowitz/Skurnik/Galinsky (1999): 28; Stanovich (2009): 22; Wilson
(2002): 4f. It has long been recognized that in order to cope with limited cognitive resources and the over-
whelming complexity of the decision environment, individuals typically follow “the principle of least ef-
fort”. That is, people process information, reason, and act in pursuit of their goals with the least amount of
cognitive effort and work possible. See Allport (1954): 173f.; Moskowitz/Skurnik/Galinsky (1999): 28. Indi-
vidual behavior is further guided by a desire to achieve a reasonable balance between cognitive effort exert-
ed and the satisfaction of the pertinent motivational constraints. See Chen/Chaiken (1999): 74.
83
See Kahneman (2003b): 1454.
84
See Kahneman (2003b): 1469.
85
The notion of perceptual limitations is vividly illustrated by Simon (1959): 273. He argues that every single
second, the surrounding environment produces millions of bits of new information, while the human percep-
tual system hardly manages to capture a thousand bits per second.
86
See Gilbert (1991): 109; Kahneman/Frederick (2005): 268.
87
See Simon (1959): 272.
88
See Galotti et al. (2009): 23; Morrow/Fiore (2013): 204.
14
2.3 Information Processing Approach
Broadly speaking, information processing relates to the cognitive operations through which
information is attained, attended to, interpreted, evaluated, aggregated, and used to draw in-
ferences, judge, and act. In this sense, information processing can be characterized as “not just
a passive response to stimuli but also an active process of constructing reality.”89
The psychological information processing approach utilizes a computer metaphor to the study
of human cognition. Specifically, it draws a parallel between the operations of human cogni-
tion and the serial manner in which computers process information. Essential to the infor-
mation processing approach is the notion that human cognition can be conceived as infor-
mation passing through a complex computational system, i.e., the human brain.90
In their simplest form, information processing models consist of an input (sensory91 infor-
mation), a process (perception, System 1 and System 2 processing), and an output (judgment
or decision).92 This basic structure is illustrated in Figure 2.
At each stage, there are several factors which may impact information processing.93 Specifi-
cally, input-related factors concern task and information set characteristics which influence
the way in which individuals absorb, weight, and integrate information. Instances for such
factors include, among others, task complexity, presentation sequence, presentation format,
aggregation level of data, type of task, and response mode.94 Some of these factors are dis-
cussed within Section 2.4.2.1. Of particular relevance to the present work are the task charac-
teristics of complexity, presentation sequence, and response mode which are discussed in de-
tail in Section 3.2.1.1 and Section 3.2.1.2.
Process-related factors, on the other hand, encompass both characteristics of the information
processing style (e.g., heuristic (System 1) versus analytic (System 2)) as well as features of
the information processor (e.g., ability, personality, cognitive structure, experience, involve-
ment, demographics).95 Generally, the use of a particular processing mode (analytic versus
heuristic) has been theorized to result from an analysis – sometimes explicit and sometime
89
Vertzberger (1990): 9.
90
See Fiske/Taylor (2008): 8; Galotti et al. (2009): 23; Sternberg (2009): 329. For a concise overview of the
brain processes and systems involved in information processing, refer to Eysenck/Keane (2010): 473-477.
91
Generally, there are five grand modalities of the sensory apparatus: vision (sight), audition (hearing), olfac-
tion (smell), somesthesis (touch) and gustation (taste). See Cardello (1996): 5; Noyes/Garland/Bruneau
(2004): 38. For a detailed discussion of these modalities, see May (2009): 11-77.
92
See Eysenck/Keane (2010): 2; Libby/Lewis (1977): 246; Noyes/Garland/Bruneau (2004): 36f.
93
See Libby/Lewis (1977): 247.
94
See Libby/Lewis (1977): 246f.
95
See Libby/Lewis (1977): 246f.
15
preconscious – of the costs associated with its use (e.g., cognitive effort 96) and the expected
benefits (e.g., promise of the strategy to yield a “correct” outcome, speed of arriving at an
outcome, and justifiability of the result).97 Consequently, all else being equal, with increasing
task complexity and ambiguity, individuals are hypothesized to revert to heuristic processing
as the costs of deliberative processing do not compensate for the expected benefits. The adap-
tive manner in which subjects adjust their beliefs in light of new information can be described
as a heuristic procedure employed to ease cognitive strain and promote efficiency. The cogni-
tive mechanisms underlying the process of belief updating are discussed in Section 3.2.1.3.
As indicated above, cognitive processing is also considerably determined by the characteris-
tics of the information processor. A general review of the most relevant personal factors is
provided in Section 2.4.2.3. Of main interest in the present study is the personality character-
istic of skepticism which is discussed in detail in Section 4.4.
Finally, output-related factors concern characteristics of the information processing outcome
(judgment or decision) which are typically employed to ex post evaluate the quality of the
underlying process. Instances of such factors include accuracy, speed, consistency, (lack of)
biases, and predictability.98 The central judgment quality criteria are discussed in Section
2.4.1. The present work focuses on auditors’ susceptibility to information order effects in the
context of belief updating and judgment formation. Accordingly, the quality of auditor judg-
ment (processes) is evaluated herein in terms of presence or lack of cognitive biases.
Following these basic psychological considerations, the broad theoretical fundament of the
present study is completed by a discussion of the central features of auditor judgment and a
concise review of the relevant literature on this topic.
2.4.1 Preliminaries
Judgment is a higher-order cognitive activity which is not unique to auditing settings, yet its
importance in auditing cannot be overemphasized. Indeed, judgment permeates the entire pro-
96
Following Libby/Lipe (1992): 252, cognitive effort is defined herein as an “input of energy to mental activi-
ties.”
97
See Beach/Mitchell (1978): 448; Payne (1982): 382f. It should be noted that in the computer science and
problem-solving domains, heuristics are viewed as explicit deliberate strategies, whereas in the classical
psychological heuristics and biases approach, they are rather considered to be automatic and unconsciously
operating mechanisms. See Keren/Teigen (2007): 93.
98
See Libby/Lewis (1977): 246f.
16
cess of auditing and plays a fundamental role every step of the way. 99 Notwithstanding its
crucial importance and the tremendous academic research on auditor judgment conducted in
the last decades,100 AICPA’s statement depicted above still appears valid today.101 With other
words, there is still some unease regarding the description, capture, and operationalization of
auditor judgment.102 The subsequent considerations lean upon the meta-analytical work of
Sarah Bonner to shed light on some of the central features of auditor judgment as a prelude to
the exploration of the interplay of information order effects and professional skepticism in
auditors’ belief revisions.
In consistency with Bonner (1999), herein judgment is broadly defined as the formation of
“an idea, opinion, or estimate about an object, an event, a state, or another type of phenome-
non.”103 Closely related to and typically discussed in conjunction with judgment is the cogni-
tive activity of decision making. The latter relates to “making up one’s mind about the issue at
hand and taking a course of action.”104 Bonner (1999) argues that “[d]ecisions typically fol-
low judgments and involve a choice among various alternatives based on judgments about
those alternatives and, possibly, preferences for factors such as risk and money.” Conse-
quently, an important distinction between judgment and decision consists in the fact that
“judgments reflect one’s belief, and decisions may reflect both belief and preferences.”105
Following this line of thought, there are two main alternative explanations for erroneous audit
outcomes: Flawed outcomes can result from intentional inconsistencies with one’s underlying
judgments. For instance, an unqualified opinion may be issued in spite of identified material
misstatements. Alternatively, audit deficiencies can result from reasonable and diligent, yet
99
See Bernstein (1967): 90; Solomon/Shields (1995): 139; Trotman (1998): 115; Trotman/Tan/Ang (2011):
279. For a fundamental contemplation on the nature of auditing, its normative frame, basic objectives, stag-
es, and techniques, refer to Marten/Quick/Ruhnke (2011) and Krommes (2011).
100
It is worth noting that Trotman/Tan/Ang (2011): 278 identified a total of 5,745 papers on judgment and deci-
sion making in accounting and auditing published in four major (high-impact) journals in the last five dec-
ades. Hence, it is far beyond the scope of the subsequent section to provide a comprehensive literature re-
view of the vast amount of judgment and decision making research in auditing. For such (monographic)
reviews, refer to Bonner (2008) and Schreiber (2000).
101
As will be demonstrated in Section 4, this statement likewise applies to professional skepticism.
102
See Anderson/Koonce/Marchant (1991): 44; Dennis (2013): 3; Libby/Lewis (1977): 245; Mason/Gibbins
(1991): 21. For a thorough review of the diverse meanings and dimensions attached to the term “judgment”
in auditing standards and professional guidance, academic research, and audit practice, see Dennis (2013): 3-
13.
103
Bonner (1999): 385. For a critical discussion of Bonner’s definition of the term “judgment”, see Dennis
(2013): 6.
104
Bonner (1999): 385. For a critical discussion of Bonner’s definition of the term “decision”, see Dennis
(2013): 7.
105
Bonner (1999): 385. This distinction between judgment and action is also essential to the Nelson model and
will be discussed in more detail in Section 4.3. Note, however, that a distinction between the terms “judg-
ment” and “decision” is not invariably evident in the relevant literature. Rather, these terms are frequently
used interchangeably. See Gibbins (1984): 106; Trotman/Tan/Ang (2011): 279. As a striking instance for
this contention, consider the definition of “judgment” provided by Wedemeyer (2010) 321, who employs
“the term ‘auditor judgment’ to describe any decision or evaluation made by anauditor, which influences or
governs the process and outcome of an audit of financialstatements” (italics added). Overall, judgment and
decision making have been acknowledged to go hand in hand. See, e.g., Dennis (2013): 10 as well as the
definition of professional judgment provided in ISA 200.13(k).
17
unconsciously biased judgments.106 Bazerman, Loewenstein and Moore (2002) argue that the
latter seems to be the ultimate problem with auditing. They further suggest that unlike inten-
tional misjudgments of the auditor, subconscious flaws cannot be mitigated by conditions of
high litigation and strong normative environment. Consequently, it is crucial to recognize the
possibility of bias, analyze its roots, and deliberately battle it.107 These considerations under-
line the need to raise awareness of cognitive traps and judgment biases in audit practice and to
further explore this issue in auditing research.
The branch of research concerned with the study of judgment and decision making in an au-
diting setting and the illumination of the cognitive processes, determinants, and constraints
underlying auditors’ judgments and choices is labeled “behavioral auditing” or “judgment and
decision making research in auditing”.108 In general terms, this stream of research aims at
providing a profound understanding, and based on that an improvement, of individual and
group judgments and decisions in an auditing setting.109 During the last few decades, substan-
tial behavioral research in auditing has been conducted. A brief overview over the central in-
sights and findings of this research is presented in the following section.
Before turning to the review of the relevant literature, some important issues need to be ad-
dressed, and several clarifications have to be made. To begin with, the audit process can gen-
erally be viewed from two different perspectives. At the macro level, auditing is considered as
an ongoing process encompassing the following sequential stages: audit engagement ac-
ceptance, planning, preliminary review, fieldwork (substantive audit procedures), and report-
ing.110 At the micro level, individual auditors’ ongoing judgment and decision making pro-
cesses are considered. Apparently, the micro- and the macro-dimension are closely
interrelated: Each separate stage of an audit builds on a number of individual auditors’ judg-
ments and decisions and makes a great deal of information processing necessary. This thesis
focuses on the psychologically-based micro-processual level of auditing. The present work
has evolved around the notion that cognitive processes and individual characteristics are not
to be treated as a black box but deserve an explicit and thorough consideration as significant
determinants of auditors’ behavior.111
The second general remark concerns the fact that auditors typically make judgments and deci-
sions under conditions of considerable complexity, ambiguity, uncertainty, and time pres-
sure.112 Given the vast amount of frequently ambiguous and usually not perfectly diagnostic
106
See Boylan (2008): 230.
107
See Bazerman/Loewenstein/Moore (2002): 97.
108
See Libby/Luft (1993): 425; Trotman/Tan/Ang (2011): 280. In consistency with prior research, these terms
are used interchangeably within this work.
109
See Ashton (1974): 143; Nelson/Tan (2005): 41; Trotman (1998): 116-118; Trotman/Tan/Ang (2011): 279.
110
For a more differentiated decomposition of the audit process, see Bonner/Pennington (1991): 2-12. More
particularly, the authors separate the overall audit process into 28 single steps and provide an excellent over-
view of the cognitive processes, problem representations, and knowledge requirements pertinent to each
stage.
111
See Schreiber (2000): 10f.; Schwind (2011): 6; Tsui/Gul (1996): 41.
112
See Anderson/Koonce/Marchant (1991): 44; Einhorn (1976): 200; Schreiber (2000): 3. For a discussion of
the influence of pressure in the context of auditing, see Section 2.4.2.2.5.
18
information underlying auditors’ judgments, the tight deadlines for performing an audit, and
auditors’ limited cognitive capacities, the sophisticated optimizing strategies prescribed by
normative economic theory appear utopic in audit reality. Instead of applying comprehensive
ideal-typical optimization procedures, auditors are expected to use satisficing strategies that
yield efficient and reasonable but not necessarily optimal outcomes due to the presence of a
number of cognitive biases and judgment traps pertinent to complex problem solving. 113 With
other words, when it comes to cognitive constrains, auditors make no exception of the general
human tendency to succumb to bias.114
A further point that needs to be addressed concerns the general focus of the present work.
More particularly, in consistency with the research emphasis in the relevant behavioral audit-
ing literature, the current study focuses on judgment rather than decision making in audit-
ing.115 There are several reasons for proceeding so: As touched upon above, judgments typi-
cally serve as the basis for decisions, i.e., judgments can be conceived as “a prelude to taking
action”.116 Hence, it is essential to explore and understand the processes and determinants
underlying judgment before focusing on the next (decisional) stage. In addition, although
judgment generally aids decision by diminishing the uncertainty via processes of deliberative
reasoning and information processing, as indicated above, it is not necessarily sufficient for
taking an action. Indeed, when it comes to choice, even strong beliefs and judgments may be
suppressed and disregarded, for instance because of the implications associated with a par-
ticular (critical) choice.117 That is, various factors can, and do, influence the connection be-
tween judgment and choice. The resulting interrelations are quite complex and hard to disen-
tangle and control for in an experimental analysis. Hence, auditor judgment constitutes the
research focus of the present study.
The final point that needs to be addressed herein concerns the issue of evaluating the quality
of auditor judgment. As the latter is neither a directly observable nor an easily definable con-
struct,118 it needs to be approximated in a reasonable manner. Judgment quality is typically
evaluated based on a comparison to an objective standard or reference point. In psychology,
judgment quality is frequently measured in terms of accuracy. However, in the context of
auditing, accuracy is not a widely applicable measure, as there rarely is an unambiguous cor-
113
For such claims in a more general context (that of the social perceiver), see Fiske/Taylor (2008): 164. The
bias-proneness due to time and other pressures as well as limited cognitive capacity and computational abil-
ity is also acknowledged by audit practitioners. See, e.g., KPMG (2011): 5.
114
For a review of the relevant auditing literature, see Koch/Wüstemann (2009): 9-21.
115
See Kotchetova/Salterio (2007): 551.
116
See Boylan (2008): 230; Solomon/Trotman (2003): 396.
117
See Ashton/Kennedy (2002): 226; Einhorn/Hogarth (1981): 20. Indeed, Ashton/Kennedy (2002): 228 find
that 29% of the participants in their study made choices which were inconsistent with the preceding judg-
ments.
118
For an excellent, thorough discussion of judgment quality in accounting and auditing, refer to Bonner
(2008): 23-51. Note that she dedicates a whole chapter of her book to the consideration and discussion of is-
sues related to the various perspectives, definitions, operationalizations, and dimensions of judgment quality,
which underscores the central importance of this topic and the ambiguity underlying the concept of judg-
ment quality in accounting and auditing.
19
rect answer for an audit task.119 As aptly put by Joyce (1976), “[o]ne of the difficulties in-
volved in studying the validity of auditors' judgments is the absence of a suitable criterion by
which to distinguish correct from incorrect judgments. Because strict guidelines for infor-
mation collection and evaluation do not exist, there are no clear-cut ‘right’ judgments availa-
ble with which to compare individual professional judgments in most audit tasks.”120 Given
the difficulties regarding the use of accuracy as a quality standard, alternative measurement
criteria (surrogates) have been employed in the relevant auditing literature. Specifically, the
quality of auditor judgment has been measured in terms of consensus, consistency, and stabil-
ity.121 Consensus generally refers to agreement among individuals.122 The rationale behind
this notion is that ideally, when confronted with the same set of circumstances, every audit
professional should choose the same audit tests and procedures and apply them equivalently
in arriving at (a consistent) overall judgment.123 Consistency, on the other hand, involves the
comparison with some criterion or norm. There are three types of consistency: consistency
with professional standards and regulations, logical consistency, and process consistency. The
former type of consistency relates to the degree to which auditor judgment corresponds to
auditing standards, rules of professional ethics, or other relevant regulations which can speci-
fy both correct answers and correct processes in a number of situations.124 Logical consisten-
cy, on the other hand, relates to the compliance with the standards of a normative system,
such as utility theory, probability theory, Bayes rule, logic, etc. Finally, process consistency
refers to the consistent application of judgment rules and professional standards by an indi-
vidual across time and/or across different problems and situations.125 Stability, the third major
judgment quality proxy, refers to the extent to which an individual makes analogical judg-
ments for analogical conditions at two distinct points in time.126 While the three precedingly
discussed central judgment quality proxies (consensus, consistency, and stability) address
119
See Solomon/Shields (1995): 151; Trotman/Tan/Ang (2011): 284.
120
Joyce (1976): 30.
121
See Johnson/Kaplan (1991): 99; Libby/Lewis (1982): 231; Trotman (1998): 117. The latter also explicitly
lists the lack of cognitive bias in individual judgment as a further judgment quality criterion. However, this
category appears redundant to the consistency and stability criteria, as a bias-free judgment is normally a one
which is consistent with normative theories and/or stable over time.
122
See Ashton (1985): 173. She points out that consensus is necessary but not sufficient for accuracy. While
lack of consensus might be indicative of inaccuracies in at least some of the responses, strong consensus
does not automatically imply correctness. Furthermore, recent auditing and neuroscientific studies embrace
an increasingly critical view on the use of consensus as a measure of judgment quality. For such claims and
further references on this issue, see Peecher/Solomon/Trotman (2013): 604f. Overall, lack of consensus can
be attributed to the fact that individuals think about a certain problem or task differently. That is, they use
different mental models and cognitive processing modes (intuitive versus analytical), organize, combine, and
weight information in distinct ways, and exhibit different variance in their judgments. See
Mumpower/Stewart (1996): 194f., 205. All these factors give raise to differences in individual judgment (a
dearth of interpersonal consensus) which are not necessarily to be equated with poor judgment quality.
123
See Hicks (1974): 40.
124
See Bonner (2008): 39-41. The role of standards and regulations in the context of auditor judgment is dis-
cussed in Section 2.4.2.2.6.
125
See Kleinmuntz (1990): 109.
126
See Johnson/Kaplan (1991): 99.
20
how well a particular judgment task is performed, i.e., relate to effectiveness127, judgment effi-
ciency128, i.e., the amount of resources exerted to make a judgment at a specific effectiveness
level, has also been suggested as a relevant quality dimension. Instances for efficiency-related
judgment quality criteria include the time and cost to conduct a particular task.129 However, as
will be demonstrated subsequently, the existing auditing literature has predominantly focused
on effectiveness-related judgment quality criteria (especially consistency) within the explora-
tion and discussion of auditor judgment.
Building on these general considerations, subsequently, an overview of the relevant academic
research on auditor judgment is provided.
2.4.2 Determinants
Following Bonner (2008), the subsequent literature review is organized in three major catego-
ries of factors determining auditor judgment: task characteristics, contextual factors, and indi-
vidual characteristics of the auditor.130 For the sake of clarity and systematization, the factors
reviewed below are analyzed individually. It is important to note, however, that these factors
are highly interdependent and essentially represent small mosaic pieces which only taken as a
whole reveal a more complete picture of the complexity and multifacetedness underlying au-
ditor judgment.
The notion that human information processing is highly contingent on general structural char-
acteristics of the task is well established in psychology.131 Overall, different tasks place dif-
ferent demands on an individual’s cognitive resources, abilities, knowledge, and effort exer-
tion, thereby impacting upon the information processing strategy utilized for performing the
task.132 In auditing research, task characteristics have likewise been recognized as important
determinants of auditors’ information processing and judgment performance. The task factors
that have experienced the greatest research interest and validation in the auditing literature
127
In general terms, effectiveness refers to “[e]nsuring that the output from any given activity is achieving the
desired results”, Glynn (1985): 30. Thus, effectiveness concerns the relation of actual output to the desired
objective. See Higson (1997): 206; Higson (2003): 101f.
128
In general terms, efficiency can be defined as “[s]eeking to ensure that the maximum output is obtained from
the resources devoted (…), or alternatively, ensuing that only the minimum level resources are devoted to a
given level of output”, Glynn (1985): 29. With other words, efficiency concerns the proportion of input and
output. See Higson (1997): 206; Higson (2003): 101f.
129
See Bonner (2008): 41; Davis/Solomon (1989): 152.
130
See Bonner (2008): 23-255. For alternatively organized literature reviews of the topic of judgment and deci-
sion making in auditing, refer to Libby/Lewis (1977), Libby/Lewis (1982), Libby/Luft (1993), Nelson/Tan
(2005), Solomon/Shields (1995), Solomon/Trotman (2003), and Trotman/Tan/Ang (2011).
131
See, e.g., Einhorn/Hogarth (1981): 61; Payne (1982): 382; Payne/Bettman/Johnson (1993): 22.
132
See Libby/Luft (1993): 427.
21
include task complexity, information relevance, framing, and information order.133 These fac-
tors are briefly discussed and underpinned with relevant empirical findings and insights in the
following subsections.
There exist a variety of differently nuanced task complexity notions and conceptualizations in
the psychological literature.134 In auditing, task complexity has generally been conceptualized
in terms of task difficulty or task structure. Task difficulty relates to the cognitive load associ-
ated with a particular task. Basically, cognitive load is an increasing function of the number of
processing stages required by the task, the amount of information to be evaluated (i.e., infor-
mation load), and the number of alternative hypotheses to be considered.135 Task structure, on
the other hand, relates to the degree of specification of the problem at hand, the relevant vari-
ables, and the range of possible solutions.136 Overall, with growing task complexity, the level
of effortful and analytic (System 2) cognitive processing required to arrive at a judgment is
expected to increase. However, beyond some point, the substantial processing difficulties im-
posed by high task complexity are theorized to lead to the use of simplified, heuristic (System
1) processing.137 Hence, beyond a particular level, task complexity generally has a negative
impact on judgment quality. This negative effect can be attributed to the strains of short-term
memory and computational capacity, the mismatch with an individual’s knowledge structures
and/or the increasing cognitive costs (but not necessarily benefits) associated with performing
the task, which may induce individuals to deliberately alter their processing strategies to be
less costly, i.e., less effortful and extensive.138
Auditing research generally provides support for the adverse effect of task complexity on au-
ditors’ information processing and judgment quality (accuracy and consistency). Instances of
auditing studies documenting judgment quality detriments in light of high task complexity
include Chung and Monroe (2001), Simnett (1996), Tan, Ng and Mak (2002), and Wright
(1995).139 Relatedly, Tan and Kao (1999) find that the adverse effect of complexity on judg-
ment quality can only be mitigated under conditions of high accountability, knowledge, and
problem-solving ability on the part of the auditor.140
As task complexity plays a central role in Hogarth and Einhorn’s belief-adjustment model,
this task characteristic will be further elaborated in Section 3.2.1.1.
133
See Bonner (2008): 157.
134
See, e.g., Handzic (2004): 191-193.
135
See Bonner (1994): 214; Bonner (2008): 159f. For an extensive theoretical overview of cognitive load, see
Sweller/Ayres/Kalyuga (2011).
136
See Abdolmohammadi/Wright (1987): 4; Bonner (1994): 214; Bonner (2008): 159; Eysenck/Keane (2010):
460; Libby/Tan (1994): 704.
137
See Payne (1982): 386; Trotman/Wright (1996): 176.
138
See Bonner (2008): 161f.; Payne (1976): 384.
139
See Chung/Monroe (2001): 120f.; Simnett (1996): 712, 716; Tan/Ng/Mak (2002): 91; Wright (1995): 144.
140
See Tan/Kao (1999): 219-221. For a discussion of accountability effects, see Section 2.4.2.2.2.
22
2.4.2.1.2 Information Relevance
Information relevance, i.e., the degree to which the information available is pertinent to and
diagnostic for a particular judgment, is a further task-related factor that deserves consideration
within the discussion of auditor judgment. Two important questions emerge with regard to
information relevance in the context of auditing: (1) do auditors integrate all the information
and audit evidence141 relevant to the problem or task under consideration, and (2) do auditors
manage to disregard all irrelevant information when forming a judgment? Given human in-
formation processing limitations and cognitive fallacies as well as the imperfect information
auditors frequently encounter in reality, both questions are to negate.142 In consistency with
Bonner (2008) and prior auditing research, the considerations within this section focus on the
second issue, particularly illuminating the phenomena of anchoring on irrelevant information
and dilution by nondiagnostic information.143
Anchoring generally refers to the human tendency to rely too heavily (i.e., to anchor) on a
particular salient but potentially irrelevant item of information, thereby insufficiently taking
into account (i.e., adjusting for) subsequent evidence.144 In auditing, evidence of anchoring
has been reported in a variety of contexts. Specifically, instead of generating independent as-
sessments and judgments, auditors have been found to anchor on client-provided unverified
book values when conducting an analytical review or on management’s classifications of in-
ternal control problems (e.g., Biggs and Wild (1985), Earley, Hoffman and Joe (2008), Heintz
and White (1989), Kinney and Uecker (1982)).145 Furthermore, auditors have been found to
overrely on the recommendations of decision-supporting systems, thereby risking a too nar-
row view on a problem (e.g., Kowalczyk and Wolfe (1998)). Wright (1988a), on the other
hand, reports results suggesting that consultation of prior year working papers does not yield a
strong anchoring effect impairing audit effectiveness. However, he documents audit efficien-
cy detriments due to participants’ reliance on prior year working papers.146
Dilution refers to the phenomenon of irrelevant (nondiagnostic) information reducing the ef-
fect of relevant (diagnostic) information on human judgments.147 Empirical evidence on the
negative impact of the dilution effect on auditor judgment has been provided by Hackenbrack
141
According to ISA 200.13(b), the term “audit evidence” refers to the “[i]nformation used by the auditor in
arriving at the conclusions on which the auditor’s opinion is based.” There are two relevant dimensions of
audit evidence: a quality dimension (appropriateness of evidence, i.e., relevance and reliability) as well as a
quantity dimension (sufficiency of evidence). See ISA 200.13(b). Mautz/Sharaf (1961): 68 emphasize the
importance of evidence in human reasoning stating that “evidence gives us a rational basis for forming
judgments.” Throughout this work, when discussed in an auditing context, the terms “evidence” and “audit
evidence” are used interchangeably.
142
See Bonner (2008): 167; Joyce/Biddle (1981): 120.
143
See Bonner (2008): 167f.
144
See Chapman/Johnson (2002): 121; Payne/Bettman/Johnson (1993): 260; Soltani (2007): 297. For a thor-
ough discussion of anchoring, refer to Chapman/Johnson (2002).
145
See Biggs/Wild (1985): 629; Earley/Hoffman/Joe (2008): 1479; Heintz/White (1989): 35f.; Kinney/Uecker
(1982): 61f.
146
See Kowalczyk/Wolfe (1998): 165f.; Wright (1988a): 604.
147
See Nisbett/Zukier/Lemley (1981): 248.
23
(1992), Hoffman and Patton (1997), and Glover (1997), with the latter two studies finding
that accountability (i.e., an effort-enhancing and thus potentially debiasing mechanism) does
not diminish the dilution effect of nondiagnostic information.148 Shelton (1999), on the other
hand, demonstrates that experience mitigates the dilution effect.149
Both anchoring and dilution effects represent a threat to the quality of auditor judgment. Spe-
cifically, these cognitive phenomena can lead to a lack of interpersonal agreement because
auditors obtaining different normatively irrelevant anchors are likely to arrive at different
judgments distorted in the direction of the anchor, and auditors encountering both relevant
and irrelevant data are expected to make a different judgment than their counterparts facing
only relevant information. In addition, anchoring and dilution effects can induce responses
that do not correspond with normative and/or professional standards.150
2.4.2.1.3 Framing
A further task characteristic that can considerably affect (auditor) judgment is the framing of
the judgment task. Broadly stated, framing effects refer to the phenomenon that different se-
mantic descriptions (e.g., positive versus negative wording) of the very same problem induce
different perceptions and responses on the part of the problem solvers.151 Framing effects vio-
late the normative principle of description invariance.152 In addition, as subjects facing posi-
tively framed tasks tend to provide more positive judgments compared to subjects obtaining
negatively framed tasks, framing effects also typically lead to a lack of interpersonal consen-
sus in judgment.153
The relevance of framing effects to auditing is obvious: If differently framed but substantially
identical audit tasks and issues are perceived and acted upon differently by auditors due to
alternative assertion frames (e.g., “misstatement” versus “non-misstatement”; “viability” ver-
sus “failure”), this can put audit (judgment) quality at risk.154 The results of auditing studies
on framing effects are fairly mixed. Specifically, Asare (1992) and Shields, Solomon and
Waller (1987) do not find a significant influence of alternative frames on auditors’ judgments.
Ayers and Kaplan (1993) and Kida (1984), on the other hand, demonstrate that subjects oper-
ating under positive (negative) frames consider positive (negative) cues as more relevant to
the task under consideration. These framing-driven different cue weighting tendencies have
148
See Glover (1997): 222-224; Hackenbrack (1992): 135; Hoffman/Patton (1997): 235. Psychological re-
search also supports the notion that accountability is not effective in reducing the dilution effect. In fact, just
the opposite has been indicated, i.e., accountability has been found to act as a bias-amplifier. See, e.g.,
Tetlock/Boettger (1989): 396.
149
See Shelton (1999): 223.
150
See Bonner (2008): 169, 173.
151
See Keren (2011): 4; Tversky/Kahneman (1981): 453. For a meta-analysis of research on framing effects, see
Kühberger (1998): 23-48.
152
See Frisch (1993): 401. For a detailed discussion and instances of the violation of the invariance principle
by framing effects, see Tversky/Kahneman (1986): S253-S257.
153
See Bonner (2008): 179f.
154
See Ayers/Kaplan (1993): 113f. The concept of audit quality is touched upon in Section 4.2.2.
24
not been found to translate into auditors’ final judgments though. 155 A recent study by
Fukukawa and Mock (2011), however, provides evidence for strong framing effects in audi-
tors’ judgments. In addition, and more importantly, the authors also find that assertion fram-
ing is significantly related to the degree of professional skepticism exhibited by the partici-
pants, with auditors operating under a negative (positive) assertion frame displaying higher
(lower) risk assessments.156 Significant framing effects have also been reported by Emby
(1994), Mueller and Anderson (2002), and O’Clock and Devine (1995).157
The impact of information order on auditor judgment is among the most prominent and in-
sightful areas of behavioral research in auditing.158 Generally, information order effects refer
to the phenomenon that the very same set of information items presented and processed in a
different order yields different overall assessments and judgments of the issue under consider-
ation.159 Consequently, information order effects represent a threat to judgment quality
viewed both in terms of interpersonal consensus and correspondence with normative theories.
Particularly, as subjects encountering information in a certain order (A-B) are generally found
to provide substantially different responses than subjects facing the same information present-
ed in the opposite order (B-A), a lack of interpersonal agreement is bound to occur. Further-
more, order effects contradict the rationale of normative models such as Bayes’ rule that
judgments are updated solely on the basis of the diagnosticity and content of information, not
its normatively irrelevant presentation order.160
Overall, psychological and auditing research provides extensive evidence in support of the
notion that information order effects systematically, significantly, and adversely influence
individuals’ judgments and assessments. The comprehensive theoretical and empirical con-
sideration of information order effects is presented in Section 3.
In summary, task characteristics can clearly influence the manner in which auditors perceive
and process information and arrive at a judgment. In addition, given the growing complexity
germane to a variety of auditing tasks as well as the possibility that the client uses some fea-
tures of the task (e.g., the information presentation frame and/or order) strategically to steer
155
See Asare (1992): 388f.; Ayers/Kaplan (1993): 126-128; Kida (1984): 336f.; Shields/Solomon/Waller
(1987): 384.
156
See Fukukawa/Mock (2011): 16f. Note in passing that this understanding of professional skepticism is in
line with the presumptive doubt view on skepticism which is discussed in detail in Section 4.2.3.2.
157
See Emby (1994): 108-111; Mueller/Anderson (2002): 169f.; O’Clock/Devine (1995): 202-204. Interesting-
ly, O’Clock/Devine (1995) find that the framing effect is majorly related to the responses of auditors from
non-Big N audit firms.
158
See El-Masry/Hansen (2008): 35.
159
See Anderson (1981): 144; Hogarth/Einhorn (1992): 3.
160
See Bonner (2008): 185 as well as Section 3.1.
25
auditors’ judgment in the desired direction, task variables deserve profound consideration and
attention in both audit practice and academic research.161
As audit tasks are not performed in a vacuum, the consideration of task characteristics is fol-
lowed by a brief review of some of the central features of the environment which surrounds
auditors and shapes the context in which they conduct the audit. Following Bonner (2008) and
Libby and Luft (1993), the subsequent review entails the contextual factors of incentives, ac-
countability, feedback, team work, pressure, professional standards and regulations, and audit
technology.162 Basically, environmental factors affect auditor judgment by altering auditors’
motivation and cognitive effort163, which in turn is expected to reflect on judgment quality.164
2.4.2.2.1 Incentives
Incentives are a central and powerful factor inherent in the audit environment. Generally, it
can be distinguished between incentives which are directly tied to the current period perfor-
mance (monetary incentives) and incentives which are more mediate, long-range, and primari-
ly concern future periods (e.g., client retention, litigation, and reputation).165 Overall, incen-
tives influence auditors’ behavior by imposing pressure, which can either improve or
deteriorate judgment quality contingent on the extent of preexisting pressure as well as the
demands of the judgment task under consideration.166
With regard to direct monetary incentives, auditing research indicates that the provision of
financial rewards to participants motivates them to exert more cognitive effort (e.g., Awasthi
and Pratt (1990), Libby and Lipe (1992)) and greater judgment accuracy (e.g., Ashton
(1990)).167 However, the majority of auditing studies explore the influence of indirect incen-
tives on auditors’ behavior. In a nutshell, prior auditing research generally finds that auditors’
judgments and choices, particularly concerning controversial client reporting preferences, are
determined by the balance of two opposing categories of incentives: incentives related to cli-
ent retention (generally found to threat objectivity and audit quality) as well as incentives re-
161
See Bonner (2008): 197.
162
See Bonner (2008): 198; Libby/Luft (1993): 435-444. Trotman/Tan/Ang (2011): 304 list a number of further
(and more contemporary) environmental factors that potentially impact auditor judgment, including press
coverage of a client company, multiple client audit environments, complex accounting information systems
environment, external audit reliance on internal audit work, etc.
163
Following Libby/Lipe (1992): 252, cognitive effort is defined herein as an “input of energy to mental activi-
ties.”
164
See Bonner (2008): 200; Libby/Luft (1993): 435. For a discussion of the role of motivation in auditing, see
Bonner (2008): 84-89.
165
See Bonner (2008): 202f.
166
See Ashton (1990): 149. The effects of pressure on auditors’ behavior and performance are discussed in
Section 2.4.2.2.5.
167
See Ashton (1990) 161, 173; Awasthi/Pratt (1990): 806; Libby/Lipe (1992): 251.
26
lated to avoidance of the risk of costly litigation and reputational damages (generally found to
act as a disciplinary and quality-enhancing mechanism). Instances for studies along these
lines include Farmer, Rittenberg and Trompeter (1987), Hackenbrack and Nelson (1996), and
Nelson and Kinney (1997).168
The influence of (indirect) incentives on auditors’ behavior is discussed in more detail and in
a more differentiated manner in Section 4.3.2.3 within the consideration of the determinants
of professional skepticism in auditing.
2.4.2.2.2 Accountability
168
See Farmer/Rittenberg/Trompeter (1987): 8-10; Hackenbrack/Nelson (1996): 53f.; Nelson/Kinney (1997):
258. For a more detailed general discussion of the accounting and auditing judgment and decision making
literature on incentives, see Bonner (2008): 202-213.
169
See DeZoort/Lord (1997): 55; Peecher (1996): 126; Tetlock (1983): 74; Turner (2001): 684.
170
See Buchman/Tetlock/Reed (1996): 381; Lerner/Tetlock (1999): 256f.; Tetlock (1983): 75;
Tetlock/Skitka/Boettger (1989): 632f.
171
See Bonner (2008): 215.
172
See Ashton (1990): 167, 173; Gibbins/Newton (1994): 183; Johnson/Kaplan (1991): 105; Kennedy (1993):
243; Koonce/Anderson/Marchant (1995): 380f.; Messier/Quilliam (1992): 133-135; Tan/Kao (1999): 221.As
27
However, accountability does not appear to act as a universal remedy for all cognitive biases
as indicated by the findings of Glover (1997) and Hoffman and Patton (1997) concerning the
dilution effect as well as the results reported by Kennedy (1995) regarding the hindsight bi-
as.173 A further important implication of accountability in the context of auditing refers to the
tendency of subjects under scrutiny to provide more conservative judgments, as conservative
positions are generally viewed as more defensible when there are no clues regarding the supe-
rior’s opinion and preferences. Empirical support for this tendency has been provided by
Hoffman and Patton (1997), Lord (1992), and Turner (2001).174 In contrast, when supervi-
sors’ preferences are known, empirical findings indicate that subordinates strategically align
their responses to the supervisor-preferred conclusion regardless of its reasonableness and
direction (e.g., Buchman, Tetlock and Reed (1996), Peecher (1996), Turner (2001), Wilks
(2002)).175
2.4.2.2.3 Feedback
a matter of precision, Tan/Kao (1999) indicate that even though accountability enhances motivation and
cognitive effort, it does not improve performance per se. In fact, accountability exerts its positive influence
on judgment quality when the requisite individual knowledge and problem-solving abilities match the char-
acteristics of the task. See Tan/Kao (1999): 209f., 221.
173
See Glover (1997): 222-224; Hoffman/Patton (1997): 233-236; Kennedy (1995): 257, 265. Hindsight bias,
also known as “the curse of knowledge” phenomenon, refers to the general inability to ignore a posteriori
knowledge of the outcome of a particular (uncertain) event when evaluating the a priori judgments of others.
See Fischhoff (1975): 288; Hawkins/Hastie (1990). For an in-depth general review of the hindsight phenom-
enon, refer to Pohl (2004).
174
See Hoffman/Patton (1997): 229f., 233-236; Lord (1992): 97f., 103; Turner (2001): 702f. This aspect is also
addressed in Section 4.3.2.3.1.4.
175
See Buchman/Tetlock/Reed (1996): 394; Peecher (1996): 126; Turner (2001): 702; Wilks (2002): 66.
176
See Earley et al. (1990): 87.
177
See Earley et al. (1990): 89; Pritchard et al. (2013): 241. For an excellent in-depth review of the behavioral
determinants and implications of feedback in the context of accounting, refer to Luckett/Eggleton (1991).
28
Overall, psychological research has yielded a highly inconsistent picture regarding the influ-
ence of feedback on judgment quality. While the majority of studies indicate an overall im-
provement in performance following the provision of feedback due to increased attentiveness,
motivation, and effort, a considerable number of research works report the contrary finding,
i.e., a decline in judgment accuracy and overall performance after receiving feedback. 178 The
latter phenomenon can be interpreted in terms of feedback inducing pressure on individuals,
which is caused by the anticipation of the upcoming performance evaluation.179
In auditing, the empirical evidence consistently reveals a positive influence of feedback on
judgment accuracy and performance, especially when the feedback is not merely restricted to
the outcome, but is supplemented with information about the pertinent cognitive processes
and task features (e.g., Ashton (1990), Bonner and Pennington (1991), Bryant, Murthy and
Wheeler (2009), Earley (2001), Earley (2003), Leung and Trotman (2005)).180 In addition, as
demonstrated by Massey and Thorne (2006), the provision of feedback can promote higher
ethical reasoning and stronger focus on the public’s interest in audit conflict situations and
moral dilemmas.181
Finally, psychological theory posits that under certain circumstances, feedback can activate
learning processes and contribute to the acquisition of knowledge, as feedback can motivate
subjects to expand cognitive elaboration, processing, and retention in order to improve or
maintain their performance.182 Support for this notion in an auditing setting is provided by
Bonner and Walker (1994), Earley (2001), and Leung and Trotman (2005).183
A further relevant characteristic of the audit environment is its multi-person184 nature, or the
performance of audit work within interactive audit teams and groups. Generally, the term
“team” refers to the collective of auditors who work together to plan and conduct the audit.
Audit teams are usually composed of subjects at different rank levels in the firm hierarchy
178
See Kluger/DeNisi (1996): 254f., 278; Kluger/DeNisi (1998): 67.
179
See DeZoort/Lord (1997): 36. The (adverse) effects of pressure on auditors’ behavior and performance are
discussed in Section 2.4.2.2.5.
180
See Ashton (1990): 167, 173; Bonner/Pennington (1991): 32f.; Bryant/Murthy/Wheeler (2009): 52-54;
Earley (2001): 93f.; Earley (2003): 121f.; Leung/Trotman (2005): 550f.
181
See Massey/Thorne (2006): 112-114. For a general discussion of the concept of ethical reasoning and moral
development, see Section 2.4.2.3.7. For a consideration of the role of these factors in the context of profes-
sional skepticism, see Section 4.3.2.2.2.
182
See, e.g., Kluger/DeNisi (1996): 263. For a general and skepticism-related discussion of knowledge, see
Section 2.4.2.3.1 and Section 4.3.2.1, respectively.
183
See Bonner/Walker (1994): 173; Earley (2001): 90; Leung/Trotman (2005): 551.
184
Note that while the focus in the present work is clearly on individual judgment, the complementary consid-
eration of multi-auditor judgment appears important not only because of the substantial research interest in
this area, but also because of the insight that can be gained from contrasting individual to multi-person in-
formation processing and judgment. For a review of the psychological literature on individual versus group
performance, refer to Hill (1982).
29
(i.e., staff, seniors, managers, and partners)185 and are formed for longer periods of time. The
term “group”, on the other hand, is used to refer to a collective of auditors who collaborate to
solve a particular problem (e.g., a complex auditing issue). Audit groups are typically com-
posed of individuals at the same level in the firm hierarchy who work together for shorter
time periods and on more specific and delimitable projects and tasks.186 In consistency with
prior research and general usage, the terms “teams” and “groups” are used interchangeably
herein.
Overall, group interaction is conceived to facilitate the generation of new ideas, strategies,
and solutions, to enable learning and knowledge exchange among the group members, and to
contribute to the elimination of inconsistencies, random errors, and biases in individual judg-
ment. In this sense, multi-person judgment is likely to outperform individual judgment. How-
ever, there are also several aspects diminishing the process gains of team interaction. Specifi-
cally, team members may feel less responsible and motivated to exert cognitive effort in
problem solving (i.e., display social loafing187). In addition, subjects may anchor on the incor-
rect view of a team member or employ similar cognitive strategies resulting in biases that go
in the same direction, and so remain uncorrected in the group.188
Auditing research corroborates the heterogeneous effect of team work on judgment quality.
Specifically, a number of studies provide empirical support for the superiority of multi-person
judgment performance as compared to individual judgment performance in terms of increased
consensus, reduced judgment variability, and greater accuracy (e.g., Ahlawat (1999),
Carpenter (2007), Reckers and Schultz (1993), Schultz and Reckers (1981), Solomon (1982),
Trotman (1985), Trotman and Yetton (1985), Trotman, Yetton and Zimmer (1983)).189 How-
ever, group-assisted judgments appear ineffective in offsetting persistent cognitive biases
such as information order effects (e.g., Johnson (1995), Reckers and Schultz (1993)). Fur-
thermore, as demonstrated by Bedard, Biggs and Maroney (1998), group work can result in
inefficiencies and process losses leading to a failure to select an identified correct answer.
Finally, as shown by Johnson (1995), audit teams appear to exhibit attitude shifts toward
greater risk in judgment and choice.190
185
For a general description of the different positions at the audit firm, including a concise consideration of
both the required knowledge and experience as well as the scope of tasks performed and responsibilities per-
tinent to the particular team hierarchy level, see Cook/Winkle (1988): 25-27.
186
See Rich/Solomon/Trotman (1997a): 90; Solomon (1987): 2.
187
“Social loafing” is a term coined by Latané/Williams/Harkins (1979): 823 as a reference to the phenomenon
of declines in effort exhibited by individuals when working in a group as opposed to performing the task au-
tonomously.
188
See Bonner (2008): 235-239; Hill (1982): 531, 535.
189
See Ahlawat (1999): 84f.; Carpenter (2007): 1136f.; Reckers/Schultz (1993): 139; Schultz/Reckers (1981):
498-500; Solomon (1982): 706; Trotman (1985): 751; Trotman/Yetton (1985): 264f.;
Trotman/Yetton/Zimmer (1983): 289.
190
See Bedard/Biggs/Maroney (1998): 226; Johnson (1995): 152; Reckers/Schultz (1993): 140. The finding of
riskier choices made by groups in comparison to individuals is in line with psychological research on group
judgment and decision making. See, e.g., Hill (1982): 531; Stoner (1968): 442f.
30
2.4.2.2.5 Pressure
A further environmental factor of critical importance to auditing is the pressure faced by audi-
tors throughout the performance of the audit. In general terms, pressure can be described as a
stimulus which impacts on an individual’s perceptual and cognitive processes.191 There is a
great variety of pressures pertinent to auditing. DeZoort and Lord (1997) distinguish two ma-
jor categories of pressure: environmental pressure (e.g., client-related pressure and litigation)
as well as organizational pressure (e.g., workload, feedback, obedience to superiors, conform-
ity pressure). In addition, they highlight accountability, competition, and time pressure as
pressures that may come both from outside or within the audit firm.192
In a nutshell, pressure is theorized to influence auditors’ attitudes and judgment performance
in an “inverted-U” manner. Specifically, moderate levels of pressure are conceived to enhance
attentiveness, focus, and effort in information processing, thereby improving efficiency and
performance. However, beyond a certain point, pressure generally induces dysfunctional be-
havior and susceptibility to error, thus posing considerable threats to audit effectiveness and
judgment quality operationalized in terms of accuracy and correspondence to normative theo-
ries and/or professional standards.193 Overall, high pressure induces stress and increases the
cognitive demands of the task, thereby causing individuals to resign and revert to simplifying
heuristics in order to reduce complexity and cognitive load. 194 The deleterious influence of
high pressure on auditors’ cognitive processing has been highlighted by Gibbins (1984), who
argues that the considerable pressure inherent in the audit environment renders the opportuni-
ties for a profound, perfectly conscious and self-reflected information processing and judg-
ment relatively rare in the auditing domain.195
Overall, auditing research has provided substantial support for the theorized ambivalent im-
pact of pressure196 on the quality of auditor judgment. Specifically, while some studies indi-
cate a (partially) positive effect of pressure on judgment and information processing perfor-
mance under certain circumstances (e.g., Glover (1997), McDaniel (1990), Spilker and
Prawitt (1997)), other studies report a variety of pressure-induced dysfunctional responses,
191
See DeZoort/Lord (1997): 31.
192
See DeZoort/Lord (1997): 36. For an in-depth, systematical review of the individual pressure constructs
outlined above, see DeZoort/Lord (1997): 36-71.
193
See Bowrin/King (2010): 162; Choo (1986): 28; Choo (1995): 634f.; DeZoort/Lord (1997): 29, 35.
194
See DeZoort/Lord (1997): 31; Payne/Bettman/Johnson (1993): 38, 135. Empirical evidence for the stress-
inducing effect of increasing time pressure in auditing is provided by Margheim/Kelley/Pattison (2005): 30.
For an excellent inquiry into the antecedents, moderators, and psychological symptoms of stress in the con-
text of auditing, refer to Haskins/Baglioni/Cooper (1990). For a list of practical suggestions and strategies
for coping with stress in the auditing profession, see Gaertner/Ruhe (1981): 73f.
195
See Gibbins (1984): 120.
196
It might be remarked that the majority of the subsequently cited studies concern the influence of time pres-
sure on auditors’ behavior. DeZoort/Lord (1997): 39 distinguish between two general types of time pressure:
deadline pressure and time budget pressure. The former type of pressure is associated with deadlines, i.e., a
specified point in time by which a task must be completed. The second type of pressure relates to the speci-
fication of a particular tight time frame in which to accomplish a job. For an extensive overview of the broad
and very diverse pressure-related literature in auditing, refer to DeZoort/Lord (1997): 36-71. For a general
review of the influence of time pressure on human information processing, see Bronner (1973): 1-47.
31
such as premature sign-offs (e.g., Alderman and Deitrick (1982), Raghunathan (1991)), un-
derreporting of audit time (e.g., Akers and Eaton (2003), Akers, Horngren and Eaton (1998),
Kelley and Margheim (1990), Lightner, Adams and Lightner (1982), Lightner, Leisenring and
Winters (1983)), inappropriate obedience to superiors (e.g., DeZoort and Lord (1994), Lord
and DeZoort (2001)), and last but not least, less professionally skeptical behavior (e.g.,
Robinson (2011)).197 Collectively, these results indicate the pervasiveness and momentous-
ness of pressure effects in auditing. Pressure effects are further considered in Section 4.3.2.3
in the context of the discussion of the determinants of professional skepticism in auditing.
A further relevant aspect of the audit environment concerns the existence of a variety of pro-
fessional norms and regulations issued to standardize audit procedures and codify appropriate
behavior in a number of contexts.198 As touched upon in Section 2.4.1, professional standards
and regulations represent a benchmark for judgment quality and the adherence to them is crit-
ical to the auditing profession. In fact, violations of auditing standards and norms can have
severe consequences, including legal litigation, irreversible reputation losses, and even audit-
ing firm collapses, as evidenced by the spectacular Enron case, which has led to the demise of
Arthur Andersen, one of the largest then Big 5 auditing firms.199
The behavioral auditing literature on rules and regulation has largely focused on the influence
of the type of auditing standards (ambiguous versus well-specified) on the quality of auditor
judgment. Overall, analytical, archival, and experimental auditing research on this topic yields
a relative consistent and intuitively plausible picture: While well-specified auditing standards
limit discretion, increment judgment consensus among auditors, and reduce auditors’ suscep-
tibility to client-imposed pressure (e.g., Carcello, Vanstraelen and Willenborg (2009),
Goldman and Barlev (1974), Knapp (1985), Magee and Tseng (1990), Nichols and Price
(1976), Trompeter (1994)), vague auditing standards allow for latitude in judgment and deci-
sion making in consistency with the individual, not necessarily norm-conform, incentives
(e.g., Hackenbrack and Nelson (1996)).200 However, empirical findings also reveal that even
stringent professional standards cannot fully counteract the effect of strong incentives and
197
See Akers/Eaton (2003): 90-92; Akers/Horngren/Eaton (1998): 15f.; Alderman/Deitrick (1982): 60-63;
DeZoort/Lord (1994): 21; Glover (1997): 222-224; Kelley/Margheim (1990): 37f.; Lightner/Adams/Lightner
(1982): 8f.; Lightner/Leisenring/Winters (1983): 53-55; Lord/DeZoort (2001): 228-230; McDaniel (1990):
282f.; Raghunathan (1991): 73-75; Robinson (2011): 79-83; Spilker/Prawitt (1997): 172-174. Overall, only
very few studies (e.g., Bamber/Bylinski (1987): 136f.) fail to indicate a significant influence of time pressure
on judgment performance in auditing.
198
See Bonner (2008): 251.
199
For a discussion of the Enron-Andersen-gate, see Benston/Hartgraves (2002). For a theoretical and empiri-
cal analysis of its impact on the reputation of the auditing profession, see Möller/Sigillo (2010). The Enron
scandal is also touched upon in Section 4.2.2 in the context of professional skepticism.
200
See Carcello/Vanstraelen/Willenborg (2009): 1398, 1424f.; Goldman/Barlev (1974): 712, 714;
Hackenbrack/Nelson (1996): 54f.; Knapp (1985): 203f., 208; Magee/Tseng (1990): 317; Nichols/Price
(1976): 340f.; Trompeter (1994): 66.
32
self-interest which result in aggressive reporting and opportunistic behavior on the part of the
auditor (e.g., Cuccia, Hackenbrack and Nelson (1995), Libby and Kinney (2000)).201
As to the behavioral implications of regulatory changes, Hronsky and Houghton (2001)
demonstrate that even minor changes in the wording of auditing standards are recognized and
apprehended by auditors to imply distinct meanings. In addition, and more importantly, audi-
tors act upon these subtle differences in consistency with the direction intended by the regula-
tors. Similarly, Zimbelman (1997) demonstrates that novel auditing standards can induce
greater attention and effort in information processing, thereby potentially improving the quali-
ty of auditor judgment.202 Overall, auditing standards and regulations represent a crucial de-
terminant of auditors’ judgment and decision making.
A further noteworthy environmental factor in auditing concerns the implementation and use
of technical and other guidance and support to aid and supervise auditors’ information pro-
cessing and judgment. Audit technology encompasses not only information technology
(IT)203, but also a wide range of formal and non-formal decision aids, practical guidance, firm
policies, etc.204 Overall, audit technology plays an increasingly important role in contempo-
rary auditing.205
Broadly stated, audit technology imposes structure, automation, and standardization on the
audit process, thereby reducing complexity and the time necessary for conducting the audit. It
aims at mitigating individual deficiencies in cognitive capacity, computational and infor-
mation processing ability, and ultimately at increasing judgment and decision making perfor-
mance.206 The performance-facilitating effect of decision aids has been generally confirmed
by auditing research (e.g., Ashton (1990), Butler (1985), Eining, Jones and Loebbecke (1997),
Libby and Libby (1989)).207 Kachelmeier and Messier (1990), on the other hand, find that
201
See Cuccia/Hackenbrack/Nelson (1995): 228f.; Libby/Kinney (2000): 384f.
202
See Hronsky/Houghton (2001): 135; Zimbelman (1997): 93f.
203
A definition of the term “information technology” in an auditing context is provided in the footnote to SAS
109 (AU 314.15). According to this this definition, IT “encompasses automated means of originating, pro-
cessing, storing, and communicating information, and includes recording devices, communication systems,
computer systems (…), and other electronic devices.” Instances for particularly frequently used IT audit ap-
plications include electronic working papers, analytical plausibility check techniques, statistical sampling
procedures, audit report writing, and Internet search tools. See Janvrin/Bierstaker/Lowe (2008): 16.
204
See Libby/Luft (1993): 436.
205
For a review of the importance and implications of advancing information technologies for twenty-first
century assurance, see Elliott (2002): 139-146. It is important to note that even though audit technology pro-
vides substantial support to the auditor throughout the conduct of the auditing procedures and is indispensa-
ble in contemporary auditing, it cannot perform the critical (cognitive) tasks of reasoning, judgment, and de-
cision making, and thus cannot replace the auditor.
206
See Janvrin/Bierstaker/Lowe (2008): 3; Libby/Luft (1993): 435f. It is worth noting that prior research has
found that auditors’ intuitive (i.e., unaided) assessments of audit risk are inconsistent with the audit risk
model set forth in authoritative guidance. See, e.g., Daniel (1988): 179; Jiambalvo/Waller (1984): 86f.
207
See Ashton (1990): 161-164; Butler (1985): 521; Eining/Jones/Loebbecke (1997): 15-17; Libby/Libby
(1989): 741-743.
33
while reducing auditors’ tendency to utilize too small sample sizes in analytical testing proce-
dures, decision aids lead to greater variance and interpersonal inconsistency in judgment.208
Furthermore, several studies indicate that audit technology may have an adverse effect on the
quality of auditor judgment, as it may induce anchoring effects, limit effort, impede inde-
pendent, controlled and analytical (System 2) reasoning, and inhibit the consideration of pro-
cedures and arguments other than those suggested by the audit information technology. The
negative influence of decision-supporting systems and checklists on auditors’ judgments has
been both theoretically and empirically corroborated (e.g., Bryant, Murthy and Wheeler
(2009), Hogan et al. (2008), Kowalczyk and Wolfe (1998)).209
In summary, the preceding review reveals that environmental factors can, and do, significant-
ly affect auditor judgment both in a positive and an adverse way. Overall, some environmen-
tal variables (e.g., incentives to avoid litigation and reputation loss, accountability to superiors
with unknown preferences) generally enhance subjects’ motivation to pursue precise and jus-
tifiable conclusions, induce effort and deliberation in reasoning and information processing,
and facilitate judgment performance. However, other environmental factors (e.g., client reten-
tion incentives, conformity, obedience, and time pressure) have a detrimental effect on auditor
judgment. They motivate subjects to reason and act toward achieving a cognitively efficient,
desired (client-favorable, authority-preferred) outcome, thereby pursuing conformity, social
acceptance, and conflict avoidance rather than precision in judgment and choice.210
To recapitulate, the preceding sections show that it is critical to take into account that audit
judgments are not made in a vacuum, but are considerably influenced by the features of the
task and the characteristics of the audit environment. However, it is not less critical to bear in
mind that audit judgments are made by humans who possess a unique combination of individ-
ual characteristics. The subsequent section sheds light on some of the key personal character-
istics which significantly influence the manner in which auditors reason, process information,
and form judgments.
208
See Kachelmeier/Messier (1990): 223. Similarly mixed results regarding the influence of decision aids on
subjects’ judgments are reported by Seow (2009): 202f. in an accounting context.
209
See Bryant/Murthy/Wheeler (2009): 39f.; Hogan et al. (2008): 246; Kowalczyk/Wolfe (1998): 165f.
210
For similar claims, see Bonner (2008): 254f.
211
Guilford (1959): 13. Note that there are a variety of different views and conceptualizations of personality in
the relevant psychological literature. See, e.g., Engler (2014): 2f. Among the most prominent conceptual
views of personality is the so-called “Five-Factor Model”. This model posits that personality traits are struc-
tured in a hierarchical manner from broad and basic to more narrow and specific dispositions. The former,
also referred to as “higher-order traits”, include the five fundamental personality factors of neuroticism, ex-
traversion, openness, agreeableness, and conscientiousness. Under these five major categories, a number of
“lower-order traits” (e.g., anxiety, self-consciousness, assertiveness, trust, compliance, dutifulness, self-
34
during way in which one individual differs from others.”212 They represent “endogenous basic
tendencies”213 which encompass “cognitive, interpersonal, motivational, and stylistic aspects,
some adaptive, some pathological.”214 Personality traits have been generally found to influ-
ence an individual’s personal and professional behavior,215 and hence clearly deserve consid-
eration in the context of auditing. It is important to note, however, that traits are essentially
tendencies, i.e., they represent general dispositions rather than absolute determinants. With
other words, there are a variety of other (task-specific and environmental) factors beyond per-
sonality traits which influence human judgment and behavior.216
The review of personal factors (both traits and other individual characteristics) presented sub-
sequently encompasses the features that have been most frequently discussed in the relevant
auditing literature and that have been found to possess the greatest capacity to influence audi-
tors’ information processing and judgment. These factors are knowledge and expertise, abil-
ity, risk attitude, confidence, tolerance for ambiguity, cognitive style, moral development,
cultural background, and gender.217
discipline, etc.) are organized. See Costa/McCrae (1992): 653f.; Livesley/Jang/Vernon (2003): 67;
McCrae/Costa (1996): 72, 74. For a historical overview of the genesis of the Five-Factor Model, see
Digman (1996): 1-16. For a general theoretical and empirical discussion of the model, refer to McCrae/John
(1992): 175-207.
212
Guilford (1959): 6.
213
McCrae/Costa (1996): 72, italics in original.
214
McCrae/Costa (1996): 62. Note in passing that there is substantial evidence that personality traits are con-
siderably heritable, i.e., genetically determined. See Livesley/Jang/Vernon (2003): 78; McCrae/Costa
(1996): 72; Tellegen et al. (1988): 1035-1038. For a discussion of the biological determinants of personality
(including cortical arousal, monoamines, hormones, and genetics), see Zuckerman (2003): 85-110.
215
See Ajzen (2005): 19f., 24; Goffman (1959), quoted in Rose (2007): 217; Matthews/Deary/Whiteman (2003):
6; Wrightsman (1974): 40.
216
See McCrae/Costa (1990): 23.
217
See Bonner (2008): 54. Note that Bonner (2008) does not include moral development in her discussion of
determinants of judgment and decision making quality in the fields of accounting and auditing. However, as
ethics and morality play a central role in the context of professional skepticism in auditing, the personal fac-
tor of moral development is theoretically and empirically reviewed in this section.
218
See Bonner/Pennington (1991): 1; Libby/Luft (1993): 426.
219
See Bonner (2008): 58. Most generally, long-term memory refers to the relatively stable and permanent
memory system in which nearly unlimited amounts of data can be stored over very long periods of time. In
contrast, short-term memory concerns the more dynamic memory system in which (a limited amount of)
new data is contained for few (regularly between two and thirty) seconds and from which that data either is
transferred to long-term memory (via rehearsal) or is forgotten. In addition, short-term memory also serves
as a work space into which information is transferred back (i.e., retrieved) from long-term memory in the
cases where its use is required. See Kalat (2012): 394f.; Marschark (1997):150; Plotnik/Kouyoumdjian
(2013): 244f.
35
(i.e., obtained through education)).220 As an internal mental construct, knowledge is not di-
rectly observable and has therefore been typically approximated via experience.221 In auditing,
experience has been broadly described as “longevity in a particular position or in performing
a particular task.”222 Experience basically provides structure and thus a guide to an individu-
al’s cognitive processes.223 The comprehensiveness, organization, and content of the experi-
entially formed knowledge structures determine what hypotheses are generated, what infor-
mation is attended to, how this information is evaluated and integrated in an overall judgment,
and thus ultimately the quality of the judgment.224 Overall, knowledge is conceived to de-
crease the cognitive effort associated with performing the task and to enable subjects to more
efficiently and accurately process and analyze information.225
A further concept usually discussed along with knowledge and experience is expertise. In
general, expertise can be defined as “highly skilled, competent performance in one or more
task domains”226. Like knowledge, expertise is not directly observable, and it has also been
typically operationalized through experience.227 Generally, experts have been theorized to
exhibit more comprehensive knowledge, better memory organization, associativity, and prob-
lem-solving ability, and thus ultimately higher judgment quality in comparison with non-
experts.228
Auditing research has largely corroborated the positive effect of knowledge and expertise on
the quality of auditor judgment. Empirical evidence along these lines has been provided by
Bedard and Biggs (1991a,b), Libby and Frederick (1990), Low (2004), and Shelton (1999),
among many others.229
220
See Bédard (1989): 115; Bonner/Lewis (1990): 1; Christ (1993): 304. In addition to experience, Libby/Luft
(1993): 433 recognize ability as a key determinant of knowledge. Consequently, ability influences judgment
performance both directly (as will be discussed in Section 2.4.2.3.2) and indirectly (through knowledge ef-
fects).
221
See Libby/Luft (1993): 430. Note that different types of experience have been used as proxies for knowledge
in the relevant auditing literature: general audit experience, task-specific experience, and industry-specific
experience, with the latter two categories being generally found to possess a greater explanatory power re-
garding individual judgment performance compared to general experience. See, e.g., Bedard/Biggs (1991b):
87f.; Bonner/Lewis (1990): 16.
222
Davis/Solomon (1989): 151.
223
See Gibbins (1984): 105; Nisbett/Ross (1980): 7.
224
See Bonner (2008): 62; Christ (1993): 306; Libby (1985): 649. For a discussion of the dimensions of
knowledge content and empirical evidence on their association with judgment performance, see Choo
(1996): 341f., 351-353.
225
See Colbert (1989): 137. For a general discussion of a number of moderators of the positive effect of
knowledge on judgment quality, see Bonner (2008): 63-65, 71f.
226
Sternberg/Ben-Zeev (2001): 365. As argued by Bédard (1989): 116, expertise is an elusive concept which
lacks a universally accepted definition. For a number of alternative general-usage and technical definitions
of expertise, see Bédard (1989): 114 and Bédard/Chi (1993): 21f.
227
Note, however, that in some cases, experience might be an imprecise proxy for expertise. For instances of
the mismatch between experience and expertise, see Bédard (1989): 114.
228
See Bédard (1989): 115; Bédard/Chi (1993): 23; Frederick (1991): 244.
229
See Bedard/Biggs (1991a): 636, Bedard/Biggs (1991b): 87f. Libby/Frederick (1990): 350, 362f.; Low
(2004): 202, 214-216; Shelton (1999): 223. Note that the literature directly or indirectly exploring the influ-
ence of experience, expertise, and knowledge on auditors’ behavior is tremendous and way too diverse to be
captured even rudimentary within a single subsection. Detailed literature reviews on this topic are provided
36
Knowledge and expertise, however, might also involve some threats to the quality of auditor
judgment – they might induce automatic processing and a tendency to premature judgments
and decisions.230 Specifically, with growing experience and familiarity, auditors may (sub-
consciously) utilize mental shortcuts, strategies, and inferences that have proven useful and
adequate in past performance, but are not necessarily optimal to the present judgment task.231
With other words, experience may induce auditors to underprocess current information and
rush into conclusions retrieved from memory rather than embrace the present problem situa-
tion and search for novel solutions based on actual observations rather than solely on prior
experience and expectations.232 In addition, experience may lead to overconfidence in the
correctness and reasonableness of the own responses.233 This contention has been empirically
corroborated by Moeckel (1990).234
Overall, although knowledge, experience, and expertise are critical and usually beneficial to
the performance of high-quality audits, under a cognitive viewpoint, in some cases, they may
also have detrimental effects on the quality of auditor judgment.235 The influence of
knowledge effects on auditors’ behavior is further considered in Section 4.3.2.1 within the
discussion of the key determinants of professional skepticism in auditing.
2.4.2.3.2 Ability
A further personal factor which deserves consideration within the discussion of the determi-
nants of judgment (performance) in auditing concerns an individual’s ability. In the context of
human information processing, ability can be defined as “the capacity to complete infor-
mation encoding, retrieval and analysis”236. The psychological literature distinguishes be-
tween three major categories of cognitive abilities: analytical, creative, and practical abilities,
all of which are conceived as related to intelligence237. Analytical abilities generally relate to
an individual’s capacity to identify the existence of a particular problem, structure it, define
its essence, and develop a strategy for solving the problem. Hence, analytical abilities repre-
by Bédard (1989), Bédard/Chi (1993), Bonner (2008): 56-73, Bonner/Pennington (1991), Colbert (1989),
Schreiber (2000): 183-314, and Wright (1988b).
230
See Gibbins (1984): 108; Grenier (2013): 6.
231
See Fiske/Taylor (2008): 165. Note that the use of mental shortcuts (heuristics) may depend on the mood
and the current emotional state. The importance of mood, affect, and emotions in the context of auditors’
judgment and decision making as well as the need for future research on this topic have been recognized by
Birnberg (2011): 12 and Nelson/Tan (2005): 54. For an overview and discussion of psychological and audit-
ing research on the influence of emotion on human judgment, refer to Nolder (2012): 81-88.
232
See Moeckel (1990): 382.
233
See Einhorn (1976): 204.
234
See Moeckel (1990): 382. For a discussion of auditing studies indicating further negative effects of
knowledge, see Bonner (2008): 68f., 73.
235
It is worth noting that this research insight seems to have found its way into audit practice, as evidenced by
KPMG’s statement that when it comes to cognitive processing traps, experience might not always be “the
best teacher.” KPMG (2011): 2.
236
Libby/Luft (1993): 428.
237
For an excellent in-depth overview and discussion of the different definitions, approaches, and theories of
human intelligence, refer to Sternberg/Kaufman (1998): 480-498.
37
sent a crucial success factor in human information processing and problem solving. A further
important determinant of problem-solving performance concerns the capacity to generate di-
verse and potentially unpopular problem-solving ideas and paths. This is where creative abili-
ties are demanded. And finally, practical abilities are necessary to implement the generated
ideas effectively.238
In auditing, the main focus of research has been on auditors’ analytical ability which has been
argued to be to some extent innate and to some extent formed through experience. 239 Overall,
ability has been found to positively affect auditors’ judgment performance (e.g., Bierstaker
and Wright (2001), Bonner and Lewis (1990), Libby and Tan (1994)).240
The individual attitude toward risk has also been recognized as an important behavioral factor
in auditing. Generally, the term “risk attitude” refers to a “generic orientation (as a mind-set)
towards taking or avoiding a risk when deciding how to proceed in situations with uncertain
outcomes.”241 Basically, there are three categories of risk attitudes: risk aversion, i.e., an atti-
tude which involves risk avoidance; risk-seeking, i.e., an attitude which involves risk prefer-
ence; and risk neutrality, i.e., an attitude which involves risk indifference.242 The importance
of risk attitude in auditing arises from the fact that auditors typically make judgments and
decisions under conditions of uncertainty. Overall, individual differences in risk attitude may
lead to inconsistent judgments and actions among auditors.243
Empirical auditing research has not demonstrated much of interest in the study of the influ-
ence of risk attitude on auditors’ judgments and behaviors. Hence, virtually nothing is known
on whether and how different risk attitudes affect the quality of auditor judgment. Empirical
evidence on the role of risk attitude in auditing is provided by Chang and Hwang (2003) and
Koch and Schunk (2013). Chang and Hwang (2003) find that the auditors in their study are on
average risk neutral and that their decisions regarding aggressive reporting practices are not
significantly affected by individual risk attitudes. Koch and Schunk (2013), on the other hand,
indicate that risk attitudes depend on situational factors such as liability regimes. The authors
demonstrate that under conditions of unlimited liability, auditors are significantly more risk
averse than in the case of limited liability.244
238
See Sternberg/Kaufman (1998): 494.
239
See Bonner/Lewis (1990): 4.
240
See Bierstaker/Wright (2001): 58; Bonner/Lewis (1990): 1, 14-16; Libby/Tan (1994): 713.
241
Rohrmann (2005): 2.
242
See, e.g., Chavas (2004): 35; Hirschey (2009): 643.
243
See Bonner (2008): 98.
244
See Chang/Hwang (2003) 211, 214; Koch/Schunk (2013): 55.
38
2.4.2.3.4 Confidence
A further personal factor of central importance in the context of auditors’ judgment and deci-
sion making concerns the individual tolerance for ambiguity. Tolerance for ambiguity gener-
ally refers to an individual’s predisposition and attitude toward the ambiguity inherent in the
environment. It is theorized to be a continuous concept, with intolerance for ambiguity at the
245
See Bonner (2008): 93.
246
See Walters (2002): 181.
247
Bailey/Daily/Phillips (2011): 33.
248
For such claims and empirical evidence on the confidence-inflating role of industry-specific experience, see
Taylor (2000): 700, 707f. Gul (1983): 87, in contrast, finds an inverse relationship between general experi-
ence and accountants’ confidence.
249
See, e.g., Plous (1993): 217.
250
See Fischhoff/Slovic/Lichtenstein (1977): 552; Griffin/Tversky (1992): 411f.; Hoffrage (2004): 235.
251
See Ahlawat (1999): 76; Block/Harper (1991): 188; Einhorn/Hogarth (1978): 398-400; Mahajan (1992):
330; Pincus (1991): 39.
252
See Hoffrage (2004): 242.
253
See Boatsman/Moeckel/Pei (1997): 230; Kennedy/Peecher (1997): 286; Moeckel/Plumlee (1989): 665;
Owhoso/Weickgenannt (2009): 14f.; Pincus (1991): 52.
254
See Bonner (2008): 97.
39
one end of the continuum and tolerance for ambiguity at the other end.255 Overall, intolerance
for ambiguity refers to the predisposition to perceive psychological discomfort (or even
threat) in light of ill-defined, ambiguous problem situations characterized by uncertainty,
complexity, vagueness, incompleteness, inconsistency, and lack of structure of the pertinent
informational stimuli. This tendency is believed to affect the way in which information is
evaluated and processed. Specifically, ambiguity-intolerant subjects are theorized to be con-
cerned about the uncertainty and ambiguity underlying a particular judgment situation and to
attempt to resolve this tension by gathering and evaluating a greater amount of information on
which to found their judgments. In contrast, individuals exhibiting tolerance for ambiguity do
not perceive cognitive strain and discomfort in light of high uncertainty and are not expected
to feel the need to collect more evidence and to elaborate more on the information available in
order to reduce ambiguity.256
Despite the evident relevance of the feature of tolerance for ambiguity to judgment and deci-
sion making in auditing, which is a field that naturally involves high uncertainty and com-
plexity, studies exploring the influence of auditors’ ambiguity aversion or tolerance on the
their information processing attitudes have rarity value. Overall, the results of the relevant
auditing studies generally support the theoretical contentions outlined above. Specifically,
Pincus (1990) reports results showing that ambiguity-intolerant auditors are better able to
recognize material misstatements in client accounts in comparison to ambiguity-tolerant sub-
jects. Furthermore, Pincus (1991) finds that ambiguity-intolerant auditors exhibit less judg-
ment confidence in situations involving high uncertainty, which is expected to trigger more
informed and extensive inquiry and evidence collection. In addition, Makkawi and
Abdolmohammadi (2004) demonstrate that ambiguity-intolerant auditors plan significantly
more audit hours than their ambiguity-tolerant counterparts. Moreover, the individual degree
of tolerance for ambiguity has been found to be strongly related to auditors’ risk assessments.
Specifically, Makkawi and Rutledge (2000) document that ambiguity-intolerant auditors pro-
vide significantly higher audit risk assessments than ambiguity-tolerant subjects. Similarly,
the results obtained by Majid and Pragasam (1997) show that in the context of high contin-
gent liability, ambiguity-intolerant auditors are less likely to issue an unqualified audit opin-
ion in comparison with ambiguity-tolerant participants.257
A further personal factor recognized to influence human behavior in general and auditors’
judgment and decision making in particular relates to an individual’s cognitive style. There
255
See Budner (1962): 29; Dermer (1973): 513; Hartmann (2005): 245; Wee/Lim/Lee (1994): 37.
256
See Benett/Herold/Ashford (1990): 346f.; Budner (1962): 30; Dermer (1973): 512f.; Hartmann (2005): 246;
McGhee/Shields/Birnberg (1978): 683; Norton (1975): 608; Ylinen/Gullkvist (2012): 401, 408f. As will be
demonstrated in Section 4.4.2, the concept of tolerance for ambiguity is related to an individual’s profes-
sional skepticism.
257
See Majid/Pragasam (1997): 935, 940; Makkawi/Abdolmohammadi (2004): 147; Makkawi/Rutledge (2000):
69, 80; Pincus (1990): 158; Pincus (1991): 55. Similarly to Pincus (1991), Gul (1986): 104 also reports low-
er confidence levels for ambiguity-intolerant bankers compared to ambiguity-tolerant subjects.
40
exists a multitude of different definitions and notions of cognitive style in the psychological
literature.258 In the present work, cognitive style is viewed as encompassing an individual’s
enduring and consistent patterns of perception, gathering, storing, organizing, processing,
evaluating, and retrieving information, reasoning, problem-solving, and other related cogni-
tive operations.259 Generally, the construct of cognitive style is characterized by the following
essential features: First, it relates to the form and structure rather than the content and out-
come of thought and cognition. Second, cognitive style is a pervasive construct which relates
to human personality, not only cognition.260 Third, as a dimension of human personality, an
individual’s cognitive style is relatively stable over time. Forth, cognitive style is a bipolar
construct and each pole has a context-dependent adaptive value. This is an essential difference
to the concept of intelligence and other ability categories which are generally classified as
positive and desirable under all conditions.261 Subsequently, the two basic cognitive styles of
field dependence-independence and locus of control, which have elicited most attention with-
in the auditing literature, are briefly discussed.
258
For an overview of some of the foundational definitions and angles of consideration of the concept of cogni-
tive style, see Dani (1989): 17f.
259
See Hayes/Allinson (1998): 850; Messick (1976): 4f.; Pratt (1980): 502. For a discussion of the nature and
determinants of cognitive styles, see Witkin (1976): 38-47.
260
Note that personality and cognitive styles, although significantly related, are basically considered distinct
constructs. See Pratt (1980): 502.
261
See Witkin et al. (1977): 15f.
262
See Bennink/Spoelstra (1979): 480f.; Colman (2009): 282; Messick (1976): 5; Witkin et al. (1977): 7f., 10.
For an overview of psychological research on field dependence-independence exploring the influence of
gender, age, development, intelligence, creativity, learning capacity, etc. on an individual’s field dependence
or independence, see Dani (1989): 20-25.
263
See, e.g., Bennink/Spoelstra (1979): 488; Cochran/Davis (1987): 206f.; Robinson/Bennink (1978): 439. For
a list of further relevant psychological studies on individual performance differences on cognitive tasks be-
tween field-dependent and field-independent subjects, see Cochran/Davis (1987): 197f.
264
See Gul (1984): 273; Gul/Zaid (1981): 949f.
41
With regard to auditing, it can be argued that field independence, i.e., the tendency to impose
structure to complex and ambiguous problem settings as well as to abstract from the prevalent
context and focus on the relevant information cues throughout the conduct of the audit, repre-
sents an important skill of an auditor.265 The sparse auditing research on this cognitive charac-
teristic provides only partial support for the superior judgment quality of field-independent
auditors. Specifically, Pincus (1990) finds that field-independent auditors are better able to
identify material misstatements in client accounts than field-dependent auditors. In addition,
Chan (1995) reports that field-independent auditors are less sucpetible to information order
effects than their field-dependent counterparts.266 Bernardi (1994), on the other hand, does
not indicate differences in the individual fraud detection ability between field-dependent and
field-independent auditors.
Locus of control, the second basic cognitive style characteristic, relates to the individual pro-
pensity to perceive a causal relationships between the own behaviors and their outcomes, with
internal and external locus of control representing the two endpoints of the scale. Internal lo-
cus of control involves the propensity to believe that one has, at least to some extent, control
over life events and the own fate. Subjects exhibiting this characteristic are labeled “inter-
nals”. They generally see themselves in charge for the outcomes of their own behavior. Exter-
nal locus of control, in contrast, involves the tendency to believe that life events and the own
destiny are determined by external forces. Individuals of this cognitive type are called “exter-
nals”. They are reluctant to recognize the contingent relationship between their behaviors and
the consequences thereof. Instead, they readily attribute the outcomes of their own actions to
factors and circumstances extrinsic to themselves.267 These general predispositions also de-
termine the manner in which people reason, judge, and act. More particularly, by recognizing
the responsibility for the own judgments and actions and by being aware of the dependence of
the outcome on the resources exerted, internals are theorized to be more effortful, diligent,
thorough, and self-reflected in information processing than externals, and thus ultimately to
arrive at outcomes of higher quality.268 Following this line of thought and transferring it to the
field of auditing, it can be argued that externals, who rely on events and factors beyond the
self and their own capabilities, are potentially less autonomous, self-reflected, and critical in
their judgments and decision, identify less with the goals and responsibilities of the auditing
profession, and exhibit lower professional skepticism than do internals, who are capable of
rationalizing and controlling their own judgments and actions.269
265
See Bernardi (1994): 71.
266
See Bernardi (1994): 76f.; Chan (1995): 26; Pincus (1990): 157.
267
See Findley/Cooper (1983): 419; Jennings/Zeithaml (1983): 417; Julian/Katz (1968): 89; Rotter (1966): 1.
268
See Nasution/Östermark (2012): 167; Trevino (1986): 610.
269
Quadackers (2009): 28f. also identifies the importance of the cognitive style of locus of control in the con-
text of auditors’ professional skepticism. Moreover, as will be touched upon in Section 4.4.3, the Rotter Lo-
cus of Control Scale is among the psychometrical scales on which Hurtt (2010) has drawn her trait profes-
42
The theoretical contentions outlined above have been empirically supported by the results of
Tsui and Gul (1996) who document a lower likelihood for internal-locus-of-control auditors
to respond to management’s demands in an audit conflict situation as compared with external-
locus-of-control auditors. In addition, Donnelly, Quirin and O’Bryan (2003) find that inter-
nals exhibit lower acceptance of dysfunctional auditor behavior (e.g., premature sign-offs and
underreporting of time) and higher degrees of organizational commitment than externals. Cor-
roborating results regarding the relationship between locus of control and auditors’ attitude
toward dysfunctional behavior have been reported by O’Bryan, Quirin and Donnelly (2005).
Furthermore, Curtis and Taylor (2009) obtain a higher likelihood for internals to “blow the
whistle”270 on observed misbehavior in comparison with externals.271 In contrast, Bernardi
(1994) does not obtain significant differences in the fraud detection rate between internal-
locus-of-control and external-locus-of-control auditors. However, he reports that experienced
auditors exhibiting internal locus of control and high degrees of moral development provide
the highest and most accurate fraud detection rates. Similarly, even though Nasution and
Östermark (2012) do not obtain significant locus-of-control-related differences in auditors’
judgments made under conditions of high pressure, the authors indicate that internals are ten-
dentially less inclined to sign-off on a doubtful account and exhibit higher responsibility for
their judgments as compared with externals.272 In contrast, Quadackers (2009) finds that in-
ternals exhibit lower fraud likelihood assessments than externals. Finally, Hyatt and Prawitt
(2001) report that internals (externals) perform better at unstructured (structured) audit firms,
thereby highlighting the unique interaction between personality features and contextual fac-
tors as determinants of job performance.273
sional skepticism scale. The latter is the psychometrical instrument employed in the present study to meas-
ure auditors’ degree of professional skepticism.
270
“Whistleblowing” is a term which is figuratively used to refer to the act of alerting about malefactions,
fraud, or misstated information. See Brabeck (1984): 42. For a thorough discussion and analysis of whistle-
blowing in the context of auditing, see Alleyne/Hudaib/Pike (2013).
271
See Curtis/Taylor (2009): 206-210; Donnelly/Quirin/O’Bryan (2003): 102-104; O’Bryan/Quirin/Donnelly
(2005): 12; Tsui/Gul (1996): 46-48. Evidence that individuals exhibiting internal locus of control reason and
behave more ethically has also been provided by Trevino/Youngblood (1990): 384 in a non-auditing context.
272
See Bernardi (1994): 77; Nasution/Östermark (2012): 173, 175.
273
See Hyatt/Prawitt (2001): 268; Quadackers (2009): 48.
274
In consistency with the relevant psychological and auditing research (see, e.g., Bobek/Hageman/Radtke
(2013c): 5 and Colby/Kohlberg (1987): 23), the terms “ethical” and “moral” are used interchangeably
throughout the present work. Stated in very general terms, “[e]thics is the study of right and wrong con-
duct”, Bollom (1988): 55. The assessment of auditors’ adherence to the principles and standards of ethics is
based on laws, professional standards and codes of conduct which encompass professional ethics issues as
well as general (universal) ethical values and conventions. See Bollom (1988): 56.
43
justice and fairness. Overall, moral development determines how a person resolves conflicts,
ethical dilemmas, and problems in everyday life.275
The theoretical frame for the study of an individual’s morality-related attitudes and behaviors
is constituted by the multi-stage model of moral development initially developed by Kohlberg
(1958). The model distinguishes between three levels of moral development: pre-
conventional, conventional, and post-conventional.276 Basically, the three distinct levels can
be envisioned as involving different types of relationships between the self and societal norms
and expectations. At the pre-conventional level, rules and expectations are externally imposed
to the self. Moral reasoning and behavior at this level are guided by the consequent immediate
costs (punishment) or benefits (self-interest) at issue. At the conventional level, an individual
defines the self through relations to significant others and compliance with societal prescrip-
tions, conventions, and standards. The striving for approval, establishment of stable relation-
ships, and obeying of established societal norms are the driving forces behind moral reasoning
and action at this level. At the post-conventional level, an individual views established values
and norms as alterable if deviations rest on higher (more idealistic and universal) moral
grounds (e.g., common wealth and ultimate justice). At this level, ethical reasoning and be-
havior are guided by individual principles of conscience and idealism rather than de facto
norms.277 Over the life time, people proceed upwardly along these levels, but not every indi-
vidual reaches the highest dimensions of moral development. With progressive development,
the differentiation in moral reasoning and complexity of the concerns and arguments consid-
ered increases.278
Previous research in psychology has demonstrated that moral development significantly de-
termines an individual’s ethical reasoning and behavior. Overall, subjects at higher levels of
moral development have been found to exhibit more ethically-oriented judgments and integer
behavior (e.g., Brabeck (1984), Thoma, Rest and Barnett (1986), Trevino and Youngblood
(1990)).279
275
For similar claims as well as a discussion of the concept of morality and moral judgments, refer to Rest
(1979): 18-20. Note that the psychological treatment of morality and ethics differs from the philosophical
one, with the latter focusing on the development and postulation of moral and ethical standards, norms, and
ideals, see Ponemon (1990): 193. For a concise summary of the ten most influential and central ethical phi-
losophies, see Hosmer (1994): 21f. For an application and discussion of Hosmer’s framework in the context
of auditing, see Satava/Caldwell/Richards (2006): 275-279.
276
Note that each of these three levels of morality contains two stages, resulting in a total of six distinct stages
of moral development and ethical cognition. The six individual stages are as follows: (1) the morality of
obedience; (2) the morality of instrumental egoism (self-interest); (3) the morality of interpersonal concord-
ance (agreement); (4) the morality of law and duty to the social order; (5) the morality of societal consensus;
and (6) the morality of non-arbitrary social cooperation. For a detailed discussion of each individual stage of
the six-stage-model of moral development, refer to Rest (1979): 22-39.
277
See Kohlberg (1984): 44, 624-639; Ponemon (1990): 194; Rest et al. (1999): 41; Schatzberg et al. (2005):
243.
278
An important feature of the Kohlberg model is that it focuses on the process of moral reasoning rather than
the outcome of this act. See Ponemon (1990): 194f.
279
See Brabeck (1984): 52; Thoma/Rest/Barnett (1986): 133-148; Trevino/Youngblood (1990): 384. For an
excellent in-depth review of empirical research on moral development and ethical behavior, see Craft
(2013).
44
In auditing, the importance of morality and ethics has long been recognized.280 Given the mul-
ti-person and interactive nature of auditing and the variety of interest groups involved (the
client, the audit profession, the audit firm, and the public), colliding goals, conflicting inter-
ests, ethical issues, and moral dilemmas appear bound to arise.281 From a cognitive perspec-
tive, auditors’ ethical and moral dispositions determine the way in which they handle conflicts
(of interest) and dilemmas in professional practice.282
There is a multitude of auditing studies focusing on diverse behavioral manifestations of audi-
tors’ moral dispositions, including the identification of an ethical dilemma (e.g., Cohen, Pant
and Sharp (2001), Douglas, Davidson and Schwartz (2001), Patterson (2001), Ponemon
(1993)), the formation of ethical judgments (e.g., Ponemon (1990), Thorne (2000), Thorne,
Massey and Magnan (2003)), and the selection of ethical behaviors (e.g., Falk et al. (1999),
Ponemon (1992a)).283 Overall, the existing auditing research has predominantly documented a
positive association between auditors’ level of moral development and their propensity to
moral judgments and behaviors in situations of ethical conflict. Auditors’ ethical dispositions
and moral development are further discussed and underpinned with empirical results in Sec-
tion 4.3.2.2.2 within the consideration of the determinants of professional skepticism in audit-
ing.
2.4.2.3.8 Culture
A further person-related factor identified to play an important role in auditing concerns the
cultural background of the individual auditor. Overall, culture284 has been broadly recognized
to influence the personal and interpersonal values, ethical norms and attitudes in different
nations, and hence to impact on auditors’ mindsets, perceptions, judgments, and behaviors.285
Research dealing with the cultural aspects of human behavior builds on Hofstede’s (1980,
2001) theory, according to which variations in individuals’ attitudes across different cultures
and nations can be attributed to the following five dimensions: power distance, uncertainty
280
A clear indication for the profound importance of ethics in auditing is the fact that in their famous work on
the philosophy of auditing, Mautz and Sharaf (1961) dedicated an entire Chapter (Chapter 9) to this topic.
See Mautz/Sharaf (1961): 280-289.
281
Indeed, ethical dilemmas and moral conflicts are viewed as inherent to the nature of auditing, see, e.g.,
Shaub/Finn/Munter (1993): 146. As succinctly put by Westra (1986): 119, when thinking about ethics in au-
diting, one feels reminded of Goldini’s comedy “Servo di due padroni” – serving two demanding masters,
the client who hires and fires the auditor and the public to whom he/she is primarily responsible.
282
See Ponemon/Gabhart (1994): 101f.
283
For an overview of the main findings of these and a number of other relevant studies as well as an excellent
in-depth discussion of empirical research on auditors’ ethical reasoning, see Jones/Massey/Thorne (2003).
284
Culture is yet another central but ambiguous and elusive concept in the social sciences. See Patel/Psaros
(2000): 316. In this work, the definition by Hofstede (2001): 9f. is adopted according to which culture relates
to “the collective programming of the mind that distinguishes the members of one group or category of peo-
ple from another” and “is to human collectivity what personality is to an individual.” Accordingly, culture is
a multifaceted concept which includes “a whole range of components: knowledge, values, preferences, hab-
its and customs, traditional practices and behavior, implements and artefacts”, Child (1981): 323.
285
For a plethora of references in support of this notion, see Cohen/Pant/Sharp (1995): 38,
Jeffrey/Dilla/Weatherholt (2004): 554 and Patel/Harrison/McKinnon (2002): 3.
45
avoidance, individualism versus collectivism, masculinity versus femininity, and long- versus
short-term orientation. Power distance concerns dealing with issues of inequality in a society.
High power distance countries (e.g., the Asian countries) exhibit a high acceptance for ine-
quality, social stratification, and authoritative hierarchies. In contrast, low power distance
countries (e.g., the Western countries) emphasize egalitarianism. Uncertainty avoidance re-
lates to the extent of tolerance for ambiguity in a society. High uncertainty avoidance cultures
(e.g., Japan) strive for standards and structure to cope with uncertainty, while low uncertainty
avoidance countries (e.g., England) tolerate risk-taking to a greater extent. The individualism
versus collectivism dimension reflects whether the focus within a nation is fixed on the self,
individual development, autonomy, and internalization of ethical values (individualistic cul-
tures, e.g., the USA, North Europe), or on group affiliation and cohesion, fitting in with social
context, and values derived from the desire to “maintain face” (collectivistic cultures, e.g.,
Asia, Africa, Latin America). The masculinity versus femininity dimension concerns whether
societal values point toward success and performance (masculine cultures, e.g., West Europe,
Japan) or toward life quality, supportiveness, and relationships (feminine cultures, e.g., Portu-
gal, Spain, Scandinavia). Finally, long- versus short-term orientation relates to the horizon of
thinking and planning, with long-tern orientation cultures (e.g., Japan) focusing on the future,
sustainability, and foresight, while short-term orientation cultures (e.g., Russia) rather focus
on the present and immediateness of results.286
These cultural dimensions are also relevant to auditing.287 Above all, the dimension of indi-
vidualism versus collectivism is of particular importance, as it reflects culturally established
values and philosophies and is thus strongly morally nuanced.288 Individualism is fundamental
to auditing as it underlies the cornerstone concepts of independence and objectivity.289 Indi-
vidualistic cultural values (typical for the Western countries) positively relate to auditors’
autonomy, self-reliance in judgment formation, assertiveness, and propensity to scrutinize
management assertions.290 Related to this notion, and also involving ethical undertones, is the
issue of power distance. In particular, low power distance (again, characteristic for the West-
ern countries) is found to foster the application of high moral and ethical standards even in
light of considerable pressure from the client.291 Hence, individualism and low power distance
are cultural attributes that generally involve high autonomy, self-determination, and morality.
286
For a general discussion of these dimensions, see Hofstede (2001): 79-84, 145-148, 209-214, 279-284, 351-
355 and Perera (1989): 44-46. For a critical consideration of Hofstede’s theory of cultural value dimensions
and for a proposition of an alternative approach based on implicit theory of causality to explain cognitively
conditioned cultural differences in ethical reasoning and moral judgment, see Wong-On-Wing/Lui (2013):
15-20.
287
For an in-depth discussion of the role of culture in auditing with an emphasis of cross-cultural aspects likely
to lead to inconsistent application of ISA requirements, refer to Cowperthwaite (2010).
288
See Cohen/Pant/Sharp (1992): 691; Hofstede (1980): 213.
289
See Cohen/Pant/Sharp (1992): 692; Jeffrey/Dilla/Weatherholt (2004): 559; Patel/Psaros (2000): 319. The
concepts of independence and objectivity are touched upon in Section 4.
290
See Hughes et al. (2009): 32; Sim (2010): 56.
291
See Cohen/Pant/Sharp (1993): 8; Tsui (1996): 129.
46
In contrast, in collectivistic cultures where harmony, avoidance of confrontation, consensus,
group cohesion, and respect for authorities are highly treasured values (typically the Asian
countries), autonomy, self-determination, a questioning mindset, and willingness to challenge
client claims and integrity may be a culturally uneasy issue.292 Similarly, in high power dis-
tance cultures (again, typically the Asian countries), auditors may face difficulties to adopt
universal ethical standards, maintain an independent mindset, and withstand pressures from
authoritative and powerful clients or superiors.293
While the auditing literature has primarily focused on the cultural dimensions of individual-
ism versus collectivism and power distance, the dimensions of uncertainty avoidance and
long- versus short-term orientation appear not less relevant in the context of judgment and
decision making in auditing. Specifically, it can be argued that in high uncertainty avoidance
cultures, auditors will exhibit less tolerance for ambiguity and will generally require and
search for more information on which to found their judgments than will be the case in low
uncertainty avoidance cultures where subjects are expected to display high tolerance for am-
biguity and to make more premature and facile judgments.294 In addition, in long-term orien-
tation cultures, auditors are expected to consider the long-term consequences of their behav-
ior, thus making more conservative judgments and decisions than individuals from short-term
orientation cultures. Generally, under a short-term perspective, auditors are more likely to
concede to client preferences, thereby increasing the immediate benefits (e.g., client and fee
retention) but disregarding the long-term consequences (e.g., litigation and reputation loss) of
their behavior.
Finally, as to the last cultural dimension, masculine cultures have been argued to tolerate gen-
der-based inequality in the workplace as well as auditors’ aggressive, ethically questionable
behavior (e.g., low balling295), particularly if it promises success. Feminine cultures, in con-
trast, value equality between the genders and are less tolerant of aggressive competitive strat-
egies.296
The existing auditing research reports cross-cultural differences which generally corroborate
the theoretical contentions outlined above in a number of different contexts, including ethical
reasoning and perceptions (e.g., Arnold et al. (2007), Tsui (1996), Tsui and Windsor (2001)),
disclosure judgments (e.g., Tsakumis (2007)), risk assessments (e.g., O’Donnell and Prather-
Kinsey (2010)), whistle-blowing (e.g., Patel (2003)), and compliance with professional stand-
292
See Bik (2010): 111; Cohen/Pant/Sharp (1992): 696; Endrawes/Monroe (2012): 9-11; McKinnon (1984):
21-23; Yamamura et al. (1996): 349f.
293
See Cohen/Pant/Sharp (1993): 8; McKinnon (1984): 21-23; Patel/Harrison/McKinnon (2002): 11f.
294
See Cowperthwaite (2010): 182. For a general discussion of the tolerance for ambiguity construct and a
review of the relevant auditing research on this topic, see Section 2.4.2.3.5.
295
“Low balling” is a term coined by DeAngelo (1981a) to refer to the practice of offering audit services below
total costs in the initial engagement period with the prospect of earning fee premiums (the so-called quasi-
rents) in future periods. This practice can be viewed as resulting from the severe competition among audit
firms for obtaining new clients. See DeAngelo (1981a): 113-115. Low balling is also touched upon in Sec-
tion 4.3.2.3.2.1 within the consideration of factors potentially inhibiting auditors’ professional skepticism.
296
See Cohen/Pant/Sharp (1992): 697.
47
ards (e.g., Abdolmohammadi and Sarens (2011)).297 Only very few studies (e.g., Ho and
Chang (1994), Hughes et al. (2009)) fail to report culture-related differences in auditors’ re-
sponses.298 It is important to note, however, that culture-related research has a very serious
drawback which needs to be taken into account when interpreting the results of this research:
Essentially, culture is a compound of multiple different constructs, so that studies using cul-
ture as the (unidimensional) construct of interest potentially examine the influence of a num-
ber of completely confounded constructs and omitted variables.299 The role of culture will be
elaborated and underpinned with empirical results in Section 4.3.2.2.5 in the context of the
discussion of the key personal factors influencing auditors’ professional skepticism.
2.4.2.3.9 Gender
The final person-related behavioral determinant discussed herein is gender. In general terms,
gender relates to an individual’s femininity or masculinity.300 Overall, a number of cognitive,
attitudinal, and emotional differences have been recognized to exist between the genders.
Specifically, men have been found to possess greater problem-solving, reasoning, and spatial
cognitive abilities, whereas women possess greater verbal abilities and exhibit a more rapid
access and retrieval of information.301 In addition, women are tendentially more susceptible to
both positive and negative affect, less (over)confident, and more risk averse than men. These
gender-related differences have the potential to influence judgment quality.302
Prior research on gender effects in accounting and auditing generally indicates that female
subjects exhibit higher ethical standards and moral development303 (e.g., Ameen, Guffey and
McMillan (1996), Bernardi and Arnold (2004), Eynon et al. (1996), McManus and
Subramaniam (2009), Shaub (1994), Sweeney and Roberts (1997)), more comprehensive and
profound information processing in complex judgment domains (e.g., Chung and Monroe
(2001)), greater attendance to disconfirming evidence (e.g., Chung and Monroe (1998)), and
297
See Abdolmohammadi/Sarens (2011): 384; Arnold et al. (2007): 335; O’Donnell/Prather-Kinsey (2010):
560f.; Patel (2003): 89; Tsakumis (2007): 43f.; Tsui (1996): 129; Tsui/Windsor (2001): 148.
298
See Ho/Chang (1994): 196, 203; Hughes et al. (2009): 40.
299
See Bonner (2008): 104, 106.
300
Note that gender is a socially founded category, and as such it is alterable in nature. In comparison, sex is a
biologically founded (and generally definite) attribute. For such arguments as well as for a discussion of the
biologically and environmentally conditioned differences between the genders, see Breesch/Branson (2009):
79-82. In the relevant psychological and auditing literature, a subject’s sex is usually used as a measure of
his/her gender. The implicit assumption underlying this procedure is that all men (women) are necessarily of
male (female) gender. See Bonner (2008): 103. This assumption has also been adopted in the present work.
301
See Halpern (2013): 153f. For a broad discussion of gender-related differences in human cognition, refer to
Caplan et al. (1997).
302
For such claims and a number of relevant references along these lines, see Bonner (2008): 104. In addition,
men and women are typically found to differ in their social attitudes. For a discussion of gender differences
in social behavior, refer to Canary/Emmers-Sommer/Faulkner (1997) and Eagly (1987).
303
For contrary evidence suggesting that male auditors exhibit higher ethical attitudes than female auditors, see
Marques/Azevedo-Pereira (2009): 236. In addition, there are also numerous studies that do not identify a
significant gender-related difference in auditors’ ethical judgments and decisions. See, e.g., Akers/Eaton
(2003): 87; Ballantine/McCourt (2011): 196; Ponemon (1990): 205f.; Radtke (2000): 306f.; Stanga/Turpen
(1991): 746.
48
stronger risk aversion and conservatism in their judgments (e.g., Gold, Hunton and Gomaa
(2009), Hardies, Breesch and Branson (2014)) as compared with male subjects.304 Gender-
related differences are further discussed in the context of auditors’ professional skepticism
(Section 4.3.2.2.6).
Overall, the preceding review highlights the theoretical and practical importance of personal
characteristics as central determinants of auditors’ judgment and behavior. In summary, audi-
tor judgment is the results of a multitude of both automatic and elaborative cognitive process-
es which are influenced by the individual and interaction effects of task-specific, environmen-
tal, and personal factors.
Building on this broad theoretical foundation, the following section sheds light on a specific
type of cognitive activity essential to auditing – the process of belief revision – as well as on
the susceptibility of this cognitive activity to the pervasive influence of information order ef-
fects.
304
See Ameen/Guffey/McMillan (1996): 595f.; Bernardi/Arnold (2004): 362; Chung/Monroe (1998): 276;
Chung/Monroe (2001): 123; Eynon et al. (1996): 487; Gold/Hunton/Gomaa (2009): 3;
Hardies/Breesch/Branson (2014): 28; McManus/Subramaniam (2009): 638; Reheul et al. (2013): 1, 6f.;
Shaub (1994): 14; Sweeney/Roberts (1997): 345; Weeks et al. (1999): 307, 310.
49
3 Belief Revision and Information Order Effects
In a broad psychological sense, belief can be described as “the mental state or function of
cognizing reality.”305 Belief has been acknowledged as a crucial and omnipresent component
of human (intellectual) life. According to Russell (1921), believing is “the most ‘mental’
thing” people do and the exploration of the nature of belief is the focal point in the analysis of
the human mind.306 In a similar vein, Hume (1748) posits that a life without beliefs is impos-
sible and that holding and updating beliefs is an inevitable part of human nature.307
As with all central notions and terms in the social sciences, there exists no universally accept-
ed definition of the term “belief”. Indeed, the word “belief” has been used in a quite heteroge-
neous, differently nuanced manner in both ordinary usage and the existing literature.308 The
multifacetedness and variety of meanings and nuances attached to the term “belief” is depict-
ed in Figure 3 below.309
305
James (1890): 198.
306
See Russell (1921): 231.
307
See Hume (1748), quoted in Lammenranta (2008): 29. The fundamentality of belief is further reflected in
the fact that belief (or lack thereof) has been a central category and a subject of philosophical inquiry and re-
flection ever since the antiquity. See Carnota/Rodrigues (2010): 1.
308
See Parret (1983): 1. Note that there exist two major scientific views on belief: the epistemological (philo-
sophical) and the psychological view. The former uses the term “belief” in a more restrictive manner in the
context of the impossibility of knowing and being certain about the reality of facts or the external world. The
psychological view is wider in scope and does not consider belief and knowledge or certainty as mutually
exclusive categories. Rather, beliefs represent the personal degree of conviction regarding a particular issue,
with certainty representing the upper limit of an individual’s belief. See Ward (1933): 347. For a discussion
of the basic philosophical conceptions of belief, see Villoro (1998): 19-40.
309
Note that this figure is philosophically founded, and it is not meant to be exhaustive. Its inclusion merely
aims to demonstrate the variety and complexity of the concept of belief.
50
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2_3, © Springer Fachmedien Wiesbaden 2015
Belief is
nothing something
an explanatory
fiction
a matter of social
a way of conditioning and
speaking linguistic practice
content attitude
a strategy of systematiz-
ing and rationalizing the
data
in the mind
outside the mind
representational
format
ideal, de dicto
actual, abstract explicit mental rep-
physical referent resentation
referent de se implicit
natural
kinds de re conscious
experience
unconscious
biological state
needs and feeling,
functions sentiment
disposition
linguistics
social conventions occurrent state
practices
history
functional role
brain state
51
Beliefs can be generally conceived as notions that people rely on and which serve as the basis
for judgment and inference.310 In the auditing literature, beliefs have been defined as “states
of mind about facts that can be uncertain and vary in regard to justifiability.”311 According to
this definition, all audit expectations and assertions are genuinely beliefs.312
It has long been recognized that beliefs are cognitive constructs which are not fully enter-
tained in consciousness.313 Indeed, people often find themselves believing without even real-
izing how or why.314 That is, people do not deliberately and consciously set (and adapt) their
beliefs. Rather, one’s beliefs are determined by complex psychological processes operating
outside of full conscious awareness.315
Basically, the process of belief can be decomposed into the cognitive operations of mental
representation and assessment of information. These operations are processed in different
domains of the mind: while mental representation involves the automatic and effortless com-
prehension of ideas, objects, and the surrounding world (System 1 processing), the assessment
of information in order to accept or reject an idea is effortful, active, and controlled (System 2
processing).316 Generally, both belief sub-processes are susceptible to bias due to inherent
cognitive capacity boundaries and information processing limitations. In particular, at the
mental representation stage, bias can arise as a result of selective attention, disregarding of
fundamental environmental stimuli, neglecting interactions among stimuli and/or flawed pool-
ing of stimuli. At the assessment stage, bias can emerge as a result of anchoring on and rein-
forcing the potentially flawed mental representations. Even in the case of unbiased System 1
input, incremental biases can arise at this stage as a result of the utilization of mental shortcuts
in evidence processing as a means of relieving cognitive strain and coping with complexity.317
The interplay between System 1 and System 2 in the context of information order effects in
belief updating will be elaborated in Section 3.2.3.
Bayes’ theorem, the dominant normative model for evidential reasoning and belief updating
in light of new information, posits that a person’s final opinion is determined by the diagnos-
ticity of evidence, not the order in which information is presented and processed.318 In par-
310
See Villoro (1998): 34; Williams (2008): 91. It is instructive to note that the terms “belief” and “judgment”
have frequently been used interchangeably in the relevant philosophical, psychological, and auditing litera-
ture. The reasonableness of this approach is best reflected in the following statement by Stout (1896): 97:
“Judgment is the Yes-No consciousness ; under it I include every mode and degree of affirmation and denial
(…). I use the term Belief as a convenient variant for Judgment.”
311
Bell/Peecher/Solomon (2005): 21f., italics in original.
312
See Bell/Peecher/Solomon (2005): 22.
313
See Hume (1748), quoted in Lammenranta (2008): 29; Quine/Ullian (1978): 9f.; Zalabardo (2012): 1.
314
See James (1912): 21.
315
See Meidan (2004): 8.
316
See Gilbert (1991): 107f. Note that this conceptualization of the process of belief is attributed to the famous
French philosopher René Descartes (1596-1650), who partitioned the human mind into relatively passive
(automatic) and active (controlled) domains. See Gilbert (1991): 108.
317
For similar claims, see Lagnado (2011): 203. Note, however, that he discusses the cognitive operations of
mental representation and assessment in a more general context and does not employ a dual-process view.
318
See Asare/Messier (1991): 75; Hogarth/Einhorn (1992): 19; Pennington/Hastie (1986): 242. For a formal
consideration of Bayes’ theorem in the context of belief updating, see Melnick/Everitt (2008): 107-111. Note
52
ticular, in the case of a sequence of two pieces of evidence (X and Y), Bayes’ theorem yields
the following result regarding the probability of a hypothesis (H): P(H│X ∩ Y) = P(H│Y ∩
X). With other words, the Bayesian approach does not differentiate between variable infor-
mation orders.319 However, comprehensive empirical evidence has shown that human infor-
mation processing in real-world settings often diverges from the predictions of the Bayesian
model.320 The major reasons for this divergence are twofold: First, Bayesian inference re-
quires the utilization of conditional probabilities and the application of sophisticated mathe-
matical calculations. These criteria appear to considerably exceed human cognitive capacity
and computational capability. Second, Bayes’ theorem does not account for characteristics of
the task such as complexity, response mode, or the order of information presentation, all
which appear to be relevant to human belief revision.321 A more descriptive approach to belief
updating has been suggested by Hogarth and Einhorn (1992) in their belief-adjustment model
which is presented and discussed in the subsequent section.322
3.2.1 Foundations
Belief updating is an omnipresent and crucial aspect of human judgment and decision mak-
ing.323 The passage from one belief to another in light of new evidence reflects human ability
to adapt to changing informational conditions. Thus, belief updating can be characterized as a
constitutive activity of intelligence.324
A principle characteristic of the process of belief adjustment is its sequential nature. As aptly
put by Anderson (1981), “[i]n everyday life, information integration is a sequential process.
Information is received a piece at a time and integrated into a continuously evolving impres-
that the Bayesian approach has found a broad application in auditing research on evidence evaluation and in-
tegration. Bayesian statistics was used in a number of older studies (e.g., Abdolmohammadi (1985, 1986,
1987), Birnberg (1964), Corless (1972), Crosby (1981), Felix (1974), Kinney (1975), Knoblett (1970), Scott
(1973)) but also in some recent publications (e.g., Kochetova-Kozloski/Messier/Eilifsen (2011),
Laws/O’Hagan (2002), Meeden/Sargent (2007), Srivastava/Mock/Turner (2009), Stewart/Kinney (2013)).
319
See Trueblood/Busemeyer (2010): 1166. For a more detailed formal consideration of the inability of norma-
tive probability theory and Bayesian inference to consider information order effects, see
Trueblood/Busemeyer (2011): 1522.
320
See, e.g., Holt/Smith (2009): 125; Lagnado (2011): 219; Lipe (2008): 277f. For a comprehensive summary
of studies on information order effects which discredit the descriptive validity of the Bayesian model, con-
sult Hogarth/Einhorn (1992): 42-47.
321
See Kahle/Pinsker/Pennington (2005): 4; Lagnado (2011): 196.
322
For a more detailed discussion of the basic differences between the Bayesian approach and the Hogarth and
Einhorn (1992) model, see Krishnamoorthy/Mock/Washington (1999): 110. They indicate that in contrast to
normative (Bayesian) belief revision, the belief-adjustment model: (1) predicts that the order of information
presentation and processing affects belief updating; (2) accounts for the individual sensitivity to evidence;
(3) comprises an anchoring-and-adjustment mechanism which implies that the degree of belief adjustment is
conditional on the prior belief; and (4) allows non-additivity in belief updating (i.e., change in the degree of
support for a particular hypothesis without influencing support for its complement).
323
See Hogarth/Einhorn (1992): 2f.
324
See Payne/Bettman/Johnson (1993): 1; Rott/Williams (2001): 1.
53
sion. Each such impression (…) grows and changes over the course of time. At any point in
time, therefore, the current impression looks both forward and back.”325
Building on this notion as well as on the extensive psychological literature reviews on human
information processing and inference by Anderson (1981), Nisbett and Ross (1980), and
Slovic and Lichtenstein (1971), Hogarth and Einhorn (1992) synthesized and united the in-
sights of sixty relevant studies in their descriptive model of belief adjustment.326 The model
explicitly accounts for task characteristics, evidential features, and cognitive mechanisms
which remain neglected in normative theory of belief revision. Subsequently, the foundational
components of the belief-adjustment model are outlined.
Hogarth and Einhorn (1992) recognize that task characteristics affect the manner in which
people process and evaluate information in arriving at updated beliefs. The authors identify
the following three basic task features particularly relevant to the process of belief revision:
complexity, length of series, and response mode.
Hogarth and Einhorn (1992) posit that task complexity influences the cognitive strategies
adopted in information processing and belief adjustment. In particular, as complexity grows,
individuals are more likely to revert to heuristic processing in order to keep the task manage-
able and promote cognitive economy.327 The authors conceptualize complexity as an increas-
ing function of information load and lack of task familiarity. This view accounts for the nor-
matively neglected inherent limitations in human information processing capacities and for
the beneficial influence of expertise in the context of processing large amounts of infor-
mation. Basically, complexity is assumed if voluminous information sets consisting of several
hundreds of words or unfamiliar stimuli are involved.328
The second task variable discussed by Hogarth and Einhorn (1992), length of series, relates
to the number of information items to be processed in belief updating. With a growing num-
ber of cues and information accumulation, fatigue is expected to incur, and beliefs are likely
to become less influenced by new evidence (i.e., beliefs are expectedly primed toward the
initially processed items). As a benchmark, Hogarth and Einhorn (1992) consider a series
325
Anderson (1981): 144.
326
For a synopsis of the sixty psychological studies used as the basis for the derivation of the belief-adjustment
model, refer to Hogarth/Einhorn (1992): 42-47. Note that beyond psychology, systematic research in human
belief revision emerged in the late 1970s in several other disciplines, including philosophy, logic, computer
science, and artificial intelligence. Influential non-psychological research works on belief revision include
Alchourrón/Gärdenfors/Makinson (1985) (AGM Model), Harper (1976), Levi (1977), and Levi (1991). See
Rott/Williams (2001): 1. For a concise review of interdisciplinary belief-revision research, see Rott (2008):
514-530.
327
See Payne (1982): 386.
328
See Hogarth/Einhorn (1992): 4-6. Note that there is a multitude of alternative ways to conceptualize com-
plexity. For instance, Abdolmohammadi/Wright (1987): 3-5 view complexity as a function of task structure,
with ill-structured tasks being more complex and demanding than well-structured tasks. Payne (1982): 386,
on the other hand, views complexity as depending on the number of alternatives in the choice scenario, the
dimensions of information, and time pressure.
54
consisting of between two and twelve cues as short and a series of above 17 cues as long.329
As to the interrelation between complexity and length of series, it should be noted that a short
series of items does not necessarily imply simplicity, as the latter depends on information load
and/or familiarity with the task rather than solely on the number of cues. As a rule of thumb, a
long series of information cues automatically implies complexity. In the case of a short series
of cues, a supplemental consideration of information load and task familiarity is necessary in
order to infer complexity. If a short series of items is heavily loaded, i.e., full of detail, and/or
unfamiliar to the information processor, complexity is theorized to emerge. Otherwise, a short
series involves simplicity.330
Response mode, the third and last task variable addressed by Hogarth and Einhorn (1992),
relates to the manner in which beliefs and judgments are evoked. The researchers differentiate
between the following two types of response mode: Step-by-Step (SbS) and End-of-Sequence
(EoS). In the SbS mode, subjects disclose their beliefs after processing each evidence item in
a particular sequence. In comparison, in the EoS mode, participants only reveal their beliefs
after obtaining and processing all available information.331
The evidence characteristics relevant to belief updating involve direction, strength, type, and
presentation order. These features are briefly considered below.332
The characteristic of evidence direction relates to the sign of evidence encountered in the
course of belief revision. It can be distinguished between positive evidence, which typically
induces an increase in belief in a certain hypothesis, and negative evidence, which normally
leads to a decrease in belief in a hypothesis.333
A further evidential characteristic relevant to the process of belief revision concerns the
strength of evidence, i.e., the degree of impact of a piece of evidence on an individual’s be-
liefs. Hogarth and Einhorn (1992) distinguish between strong and weak evidence. Although
329
See Hogarth/Einhorn (1992): 4, 6.
330
See Kahle/Pinsker/Pennington (2005): 5.
331
See Hogarth/Einhorn (1992): 4f.
332
The brevity of consideration is due to the somewhat trivial nature of these characteristics. In addition, the
first three factors (direction, strength, and type) are also reflected in Bayesian inference, and thus do not rep-
resent an incremental contribution of the belief-adjustment model. See Ashton/Ashton (1988): 625.
333
See Hogarth/Einhorn (1992): 9. Note that this interpretation of evidence direction applies to evaluation type
tasks which, as will be demonstrated subsequently, are theorized to prevail in auditing. Furthermore, it is in-
structive to note that Hogarth/Einhorn (1992) use the pairs of terms “positive”-“confirming” and “negative”-
“disconfirming” interchangeably. In the context of their study, this intuitive procedure is appropriate because
the authors always use the aforementioned terms in relation to the hypothesis under consideration. In the au-
diting literature, however, such a consistency and precision in use are lacking. This makes a distinction be-
tween the terms “positive” and “confirming” as well as “negative” and “disconfirming” necessary. Specifi-
cally, in auditing, positive evidence relates to information hinting at advantageous characteristics of a client
or the client’s financial statements. In contrast, negative evidence suggests adverse features. Importantly, ei-
ther type of evidence can be confirming or disconfirming contingent on the direction of the hypothesis under
consideration. See Abou-Seada/Abdel-Kader (2003): 13.
55
the authors do not provide explicit classification guidelines, the illustrative examples that they
use suggest that strong evidence has an above-average influence on beliefs, while weak evi-
dence affects one’s position only modestly (below-average).334
By type of evidence, Hogarth and Einhorn (1992) refer to the composition of a set of evi-
dence items. There are two alternative types of evidence: consistent evidence, i.e., an entirely
positive or entirely negative set of evidence items, and mixed evidence, i.e., an evidence set
consisting of both positive and negative items.335 With regard to auditing, the obtainment of
purely consistent evidential cues is rather exceptional. With other words, auditors typically
encounter mixed and ambiguous information in the course of the audit. This notion is also
manifested in the predominant use of experimental settings involving inconsistent information
in auditing research.336
Finally, Hogarth and Einhorn (1992) take into account the evidential characteristic of presen-
tation order. They distinguish between the following theoretical orders of evidence presenta-
tion: weak-strong versus strong-weak (used especially in the case of consistent evidence) and
positive-negative versus negative-positive (used in the case of mixed evidence).337 The latter
sort of varying information order has been predominantly used in auditing research.
In addition to the task and evidence characteristics outlined above, the process of belief updat-
ing is influenced by three central cognitive mechanisms: encoding, processing, and adjust-
ment. A consideration of the major aspects relating to these mechanisms is provided subse-
quently.
By encoding, Hogarth and Einhorn (1992) refer to the notion that information is evaluated
against a reference point before beliefs are adjusted. Building on this insight, the researchers
differentiate between evaluation and estimation modes. In the evaluation mode, evidence is
viewed as bipolar (positive or negative) vis-à-vis a particular hypothesis regardless of the de-
gree of current belief in the hypothesis. Thus, evidence encoded as confirming the hypothesis
naturally leads to an increase in beliefs in the hypothesis, whereas information encoded as
disconfirming results in a decrease in beliefs. In comparison, in the estimation mode, evidence
is viewed as unipolar and is evaluated relative to the current position (degree of belief). Thus,
the direction of belief revision depends on the relative position of the new information com-
pared to the present belief. Overall, the evaluation mode involves information processing in
an additive manner, whereas the estimation mode involves information processing in an aver-
aging manner.338
334
See Hogarth/Einhorn (1992): 10.
335
See Hogarth/Einhorn (1992): 16.
336
For a review of the relevant empirical literature on auditors’ belief updating, see Section 3.4.
337
See Hogarth/Einhorn (1992): 21f.
338
See Hogarth/Einhorn (1992): 9f. These contentions imply that the evaluation mode involves a constant ref-
erence point, i.e., the hypothesis under consideration, while the estimation mode entails a “moving” refer-
56
To illustrate the different implications of both modes in the context of auditing, consider the
going concern task where auditors evaluate the appropriateness of the going concern assess-
ment made by the entity’s management. Auditors’ current belief in the adequacy of the going
concern assumption can be conceptualized as a value on the continuum between “false” (= 0)
and “true” (= 1). In an evaluation mode scenario, evidence indicating the viability of the firm
(positive information) will increase auditors’ belief in the appropriateness of the going con-
cern assumption, whereas evidence indicating entity’s financial and liquidity problems (nega-
tive information) will decrease this belief irrespective of auditors’ current opinion or evidence
strength. In contrast, if an estimation mode is assumed, the direction of belief updating will
depend on the relative position of new evidence to the current belief of the auditor. Hence,
evidence indicating the viability of the firm only leads to an upward belief revision if it is to
be located beyond the present state of belief. Otherwise, if the evidence is not strong enough
to “outrage” current beliefs, it will lead to a downward belief adjustment regarding the appro-
priateness of the going concern assumption. To make this difference clearer, assume that the
auditor begins with a neutral position (0.5) and obtains two pieces of evidence in support of
the going concern assumption. These cues differ in strength and have values of, say, +0.7 and
+0.3 if measured on the bipolar evaluation scale, and 0.9 and 0.6 if assessed on the unipolar
estimation scale. Now consider the following alternative information processing orders:
weak-strong and strong-weak. In the evaluation mode, both pieces of evidence will induce
incremental upward belief adjustments irrespective of information order and evidence
strength because both items are encoded as positive in relation to the going concern assump-
tion. In the estimation mode, in contrast, auditors’ final position (degree of belief in the rea-
sonableness of the going concern assumption) depends on the order of information pro-
cessing. In particular, if weak evidence is processed prior to strong evidence both information
items will trigger upward belief updates because 0.6 > 0.5 and 0.9 > 0.55 (the average of 0.6
and 0.5). With the opposite information order, however, different results emerge. Specifically,
an upward belief revision will follow the processing of the strong evidence cue because 0.9 >
0.5. The weak evidence supporting the going concern assumption, however, will lead to a
decrease in belief in the firm’s viability because 0.6 < 0.7 (the average of 0.9 and 0.5). Ac-
cordingly, even though both modes involve an anchoring-and-adjustment mechanism, they
yield different adaptation patterns and results.339
ence point, i.e., the changing prior opinion. Hogarth/Einhorn (1992) conceptualize the reference point (R) in
evaluation scenarios as equal to zero and in estimation scenarios as equal to the prior belief or anchor (Sk-1).
This distinction is important for the model specification presented in the next section. Furthermore, it is in-
structive to note that Hogarth/Einhorn’s (1992) proposition that people integrate evidence cues in either an
additive or an averaging manner into a judgment is consistent with the general tenor of empirical research on
cue aggregation. The overall finding of this research is that professionals and non-professionals likewise
tend to use simpler linear organizing principles (addition or averaging) rather than complex non-linear con-
figural or patterning strategies. See Arkes/Hammond (1986): 10.
339
This comparison follows the generic example provided by Hogarth/Einhorn (1992): 9f. With regard to the
conceptual measurement scales used in both modes to encode evidence, the authors state that in evaluation
tasks, evidence assessment is captured on a bipolar scale (range: from -1 to +1), but the influence of this
(subjective) evidence evaluation on the current degree of belief is measured on a unipolar scale (range: from
0 to 1). In contrast, estimation involves two distinct unipolar scales (range: from 0 to 1 each): one for (sub-
jective) evidence evaluation and one for the impact of this evaluation on subjects’ degree of belief. See
57
As the foregoing example illustrates, the evaluation mode appears to better reflect the evi-
dence evaluation process in auditing where a new information item is typically assessed as
corroborating or refuting an audit assertion prior to belief adjustment. With other words, au-
diting generally involves directional interpretation of evidence. This contention is further sup-
ported by the fact that belief revision research in auditing has (almost) exclusively focused on
the evaluation mode.340
Processing is a further cognitive mechanism relevant to belief updating. Hogarth and Einhorn
(1992) distinguish between sequential (SbS) and simultaneous (EoS) processing. A SbS pro-
cessing involves incremental belief adjustments following each new piece of information. In
comparison, an EoS processing involves a single belief adjustment based on the cumulative
impact of the entire information set. Importantly, both processing modes involve different
cognitive demands. Specifically, the simultaneous processing and aggregation of multiple
information cues before integrating them with the anchor in the EoS process is a mentally
demanding issue. It involves the storage of large amounts of successive information in
memory as well as the simultaneous evaluation and integration of multiple stimuli within a
single overall assessment. In contrast, the gradual integration of each information item in the
SbS process involves far less cognitive strain and information processing load.341
As the analogue terminology reveals, processing modes are closely related to the response
mode categories outlined in Section 3.2.1.1. However, the manner in which evidence is pro-
cessed does not solely depend on the response mode required. Similarly to task complexity,
inferences on the processing mode can only be made after considering the cognitive load of
the task. Specifically, Hogarth and Einhorn (1992) posit that while a SbS response mode nat-
urally invokes SbS processing, an EoS response mode can induce either EoS or SbS pro-
cessing, depending on task complexity and the mental resource demands involved. Complet-
ing a task which requires an EoS response in a SbS processing manner involves making
piecemeal belief adjustments but explicitly disclosing only the final position. This constella-
tion is theorized to occur when the memory and information processing load associated with
evidence aggregation exceeds the cognitive capacities of the decision maker. Hence, with
Hogarth/Einhorn (1992): 11. However, the authors do not elaborate on the transformation of an evaluation
scale onto an estimation scale. Hence, the values used in their example and also in the going concern analo-
gy outlined above appear to have been arbitrarily chosen rather than derived based on some formula or spe-
cific calculations and transformations.
340
See Asare/Messier (1991): 78; Bamber/Ramsay/Tubbs (1997): 252; Kahle/Pinsker/Pennington (2005): 6;
Messier/Tubbs (1994): 60. As a matter of fact, auditing has even been explicitly defined as an evaluation
process. See, e.g., Schandl (1978): 4. The only study that has employed an estimation (audit planning) task
in addition to an evaluation (internal control assessment) task is the one conducted by Kerr/Ward (1994).
They provide strong support for Hogarth/Einhorn’s (1992) supposition that in evaluation tasks, the current
position does not influence the encoding of a new evidence item, whereas in estimation tasks, encoding is
derived on the basis of the difference between the current position and the new piece of information. The
corresponding evidence integration rules (additive for evaluation-type tasks and averaging for estimation-
type tasks) have also been confirmed. See Kerr/Ward (1994): 38-40.
341
See Hogarth/Einhorn (1992): 12f.
58
increasing complexity and amount of information, a sequential (SbS) processing is more like-
ly to be used regardless of the response mode required in order to ease cognitive strain.342
The last essential cognitive mechanism underlying belief updating is adjustment, i.e., the pro-
cess by which adaptation to new information is carried out. Two aspects are particularly rele-
vant in the context of adjustment: The first one concerns the notion that the magnitude of ad-
justment depends on the degree of prior belief (anchor) in a manner consistent with a contrast
effect343. The latter implies that strongly held positions (i.e., large anchors) are affected
(“hurt”) to a greater extent than weaker ones by the same piece of negative evidence. Analo-
gously, weak initial beliefs (i.e., small anchors) are affected (“revalued”) to a greater extent
than are strong ones by the same piece of positive evidence. With other words, the contrast
assumption implies that for negative evidence, belief revisions are proportional and for posi-
tive evidence inversely proportional to the present position (anchor). With regard to mixed
information stimuli (positive and negative cues), the contrast assumption implicates that evi-
dence will trigger a greater belief revision when it contrasts, i.e., is opposite in sign to, prior
evidence. The contrast assumption is fundamental to the belief-adjustment model. It is the
primary source for the model’s predictions.344 Importantly, the contrast effect essentially op-
erates beyond conscious awareness.345
The second important aspect regarding the cognitive mechanism of adjustment concerns its
dependence on the individual sensitivity to both positive (β) and negative (α) evidence. Ho-
garth and Einhorn (1992) acknowledge that the individual attitude toward evidence may be
determined by both individual variables (e.g., personal dispositions) and contextual factors
(e.g., level of commitment due to investments in a particular belief). Importantly, the authors
do not conceptualize the individual sensitivities toward negative and positive information as
complementary categories, i.e., an individual can be sensitive to both types of information
(positive and negative), and a high sensitivity to negative evidence does not imply an insensi-
342
See Hogarth/Einhorn (1992): 12f.
343
In the social sciences, contrast effects in individual judgment have been long recognized and are a well-
documented phenomenon (e.g., Fernberger (1920): 145f., 149f.; Manis (1967): 326, 330, 332; Preston
(1936b): 401; Simpson/Ostrom (1976): 625, 628; White (1964): 231). Their emergence and impact on judg-
ment can be generally explained as follows: Contrasting stimuli are more salient, attract more attention, re-
quire more effortful information processing, and are more slowly and deeply encoded, which leads to im-
pressions biased in the direction of the discrepant stimuli. See Belmore (1987): 481; Maurer/Alexander
(1991): 3f., 7; Messier/Tubbs (1994): 59; Preston (1936a): 630.
344
See Hogarth/Einhorn (1992): 26, 38; Messier/Tubbs (1994): 59; Pei/Reckers/Wyndelts (1990): 123. In this
context, it is worth noting that the pre-publication versions of the belief-adjustment model were labeled
“contrast/surprise model” (see Einhorn/Hogarth (1985)) and “contrast-inertia model” (see Einhorn/Hogarth
(1987)). In both cases, the contrast effect underlying the process of belief adjustment stands in the fore-
ground. In addition, it should be noted that research on the belief-adjustment model which has been pub-
lished prior to 1992 has actually employed the older versions of Hogarth/Einhorn’s (1992) model. Despite
some differences in terminology and variable labeling, the final model and the two preliminary versions are
qualitatively similar and yield analogue order-effect predictions. For such claims, see also
Pei/Reckers/Wyndelts (1992): 177.
345
See Fernberger (1920): 145f.
59
tivity to positive information and vice versa.346 According to this conceptualization, the fol-
lowing categories of individual attitudes toward evidence can be outlined (Figure 4):347
1 Advocate Highly
Sensitive
ß = Sensitivity to
positive evidence
Semi-Sensitive
0 Insensitive Skeptic
0 α = Sensitivity to negative evidence 1
346
See Hogarth/Einhorn (1992): 14f., 20.
347
See Hogarth/Einhorn (1992): 40f. Note that in a recent study, Bobek/Hageman/Radtke (2013a) test whether
advocacy as measured by the Modified Client Advocacy Scale by Pinsker/Pennington/Schafer (2009) and
skepticism as measured by the Hurtt scale (Section 4.4) are indeed opposing constructs. They authors find
that the two constructs are distinct from each other but cannot be viewed as the opposite ends of a continu-
um. See Bobek/Hageman/Radtke (2013a): 10-12. Note, however, that the operationalization of advocacy and
skepticism chosen by the authors differs from Hogarth/Einhorn’s (1992) notion of advocacy and skepticism.
Hence, the findings of Bobek/Hageman/Radtke (2013a) do not call into question the validity of the advoca-
cy-skepticism-continuum suggested by Hogarth/Einhorn (1992).
348
The notion of skepticism set forth by Hogarth and Einhorn (1992) will also be touched upon in Section
4.2.3.2.
349
See Hogarth/Einhorn (1992): 40.
350
To the best of the author’s knowledge, there exist only two (auditing) studies which provide direct evidence
on participants’ sensitivities to evidence: the study by Bamber/Ramsay/Tubbs (1997) as well as the study by
Guiral/Esteo (2006). The results of these studies are discussed in Section 3.4.
60
The array of task, process, and evidence features relevant to the belief-adjustment model is
summarized in Table 1.
Types Characteristics
Task
Evidence
Process
Based on these preliminaries, in the subsequent section, the belief-adjustment model is for-
mally specified.
The belief-adjustment model is based on the notion that people process information in a man-
ner of sequential anchoring and adjustment351, in that a person’s beliefs start from an initial
351
As to the relation between the belief-adjustment model and the anchoring and adjustment heuristic set forth
by Tversky/Kahneman (1974), there is some controversy in the literature. While some authors consider the
Hogarth and Einhorn model as an application of this heuristic (see, e.g., Pinsker (2011): 163;
61
anchor and are adjusted sequentially upwards or downwards in light of new information.352
Importantly, the cognitive process of anchoring and adjustment operates essentially subcon-
sciously.353
Formally stated, the belief-adjustment model runs along the following lines:354
where
ܵ = degree of belief in some hypothesis after evaluating k pieces of evidence (0 ≤ ܵ ≤ 1);
ܵିଵ = the anchor or prior opinion;
ݏሺݔ ሻ = subjective evaluation of the kth piece of evidence;
ܴ = the reference point against which the impact of the kth piece of evidence is evaluated; and
ݓ = the adjustment weight for the kth piece of evidence (0 ≤ ݓ ≤ 1).
As indicated in Section 3.2.1.3, the cognitive mechanism of evaluation which involves direc-
tional interpretation of audit evidence and additive evidence processing appears to be more
realistic in auditing than estimation. Consequently, and in consistency with the prior auditing
literature, the focus of consideration within this section is on the evaluation form of the belief-
adjustment model.355 This form implies a bipolar (positive versus negative) evaluation of evi-
dence relative to the hypothesis under consideration such that - 1 ≤ ݏሺݔ ሻ ≤ + 1 and ܴ = 0.
Thus, equation (2) takes the following form:356
The adjustment weight for the kth piece of evidence (ݓ ) has been further specified as:357
Pinsker/Pennington/Schafer (2009): 93; Trotman/Tan/Ang (2011): 296), other researchers view the anchor-
ing-and-adjustment strategy underlying the model and the heuristic as distinct and independent categories
(see, e.g., Ashton/Ashton (1988): 625; Pei/Reckers/Wyndelts (1992): 190).
352
Note that this insight is hardly new. Specifically, Weld/Roff (1938) find that judicial judgments are not only
influenced by the evidence obtained but are also strictly conditional on the preceding judgment. With other
words, judgments are not formed in an absolute manner. Rather, they are built upon previous beliefs and
judgments. See Weld/Roff (1938): 625f.
353
See Abou-Seada/Abdel-Kader (2003): 6.
354
See Hogarth/Einhorn (1992): 8.
355
For a formal consideration of the estimation form of the model as well as the corresponding order effect
predictions, refer to Hogarth/Einhorn (1992): 9f., 15f., 48f.
356
See Hogarth/Einhorn (1992): 9.
357
See Hogarth/Einhorn (1992): 14.
62
where
ߙ = sensitivity to negative evidence (0 ≤ ߙ ≤1);
ߚ = sensitivity to positive evidence (0 ≤ ߚ ≤1); and
ߙ, ߚ = constants.
Hence, by substitution of equations (4a) and (4b) into equation (3), the SbS process model can
be mathematically represented as:358
Verbally stated, the SbS-evaluation form of the belief-adjustment model implies that when
encountering a piece of new information, individuals mentally encode it as positive or nega-
tive vis-à-vis the hypothesis under consideration, subjectively evaluate the item according to
its perceived strength, and subsequently integrate it in an additive manner with their current
beliefs. The so updated belief then acts as the new anchor when a further information cue is
obtained. This is an iterative process which is repeated until all pieces of information are pro-
cessed and a final belief (position) is achieved.359
With regard to the EoS processing, the belief-adjustment model takes the following general
form:360
where
ܵ = initial strength of belief; and
ݏሺݔଵ ǡ Ǥ Ǥ Ǥ ǡ ݔ ሻ= function, potentially weighted average, of the individual subjective evaluations
of information cues following the initial belief.
Subsequently, the implications and predictions of the belief-adjustment model are presented
and discussed.
358
See Hogarth/Einhorn (1992): 14.
359
See Lagnado (2011): 205.
360
See Hogarth/Einhorn (1992): 12. Note that the authors do not elaborate on the EoS processing – neither
theoretically nor formally. It is conceivable that the substitution of equations (4a) and (4b) into equation (6)
has not been conducted (in analogy to the SbS processing) because with EoS processing, only the average
effect of all evidence items is considered, so that the individual evaluation of the single cues and subjects’
sensitivities become irrelevant. Overall, Hogarth/Einhorn (1992) focus on SbS rather than EoS processing.
This is also evident in the Appendix to their seminal publication where analytical and mathematical proofs
for the order effect predictions are only provided for the SbS process.
63
3.2.3 Model Implications and Predictions
A major implication of the belief-adjustment model is that in a variety of situations and judg-
ment domains, people are susceptible to information order effects which, as might be recalled
from Section 2.4.2.1.4, relate to the dependence of an individual’s assessments and judgments
on the order in which evidence is encountered and processed.361 In essence, order effects can
be classified as weighting biases, i.e., biases which arise from the over- or underweighting of
information items contingent on the presentation order of a series of evidence. 362 The specific
order-effect predictions of the belief-adjustment model emerge from the interaction of the task
characteristics, evidence features, and cognitive subprocesses outlined in the previous sec-
tion.363
Overall, there are three possible outcomes: no information order effects, primacy effects, and
recency effects. Within the first category, information order does not affect subjects’ evidence
evaluation, belief formation, and judgment responses. Primacy effects, on the other hand, im-
ply an overweighting of information processed early in the sequence, while recency effects
imply an overweighting of information processed later in the series, so that this information
disproportionately affects a person’s beliefs.364 Prior research in psychology generally attrib-
utes primacy effects in long information series to a decrease in attention for late information
due to fatigue and increasing cognitive load. In addition, initially presented information has
been theorized to receive the greatest amount of attention and cognitive processing because it
is utilized to form preliminary impressions.365 Recency effects, on the other hand, have been
attributed to an increase in attention for late information when it contrasts evidence faced ear-
lier in the sequence. Hence, recency essentially results from the salience and deeper encoding
of contrasting stimuli.366 Viewed under a memory lens, primacy effects can be attributed to
the fact that items presented early in the information set can be rehearsed to a larger extent
and thus have a greater chance to be transferred into some more permanent memory storage.
Recency effects, on the other hand, can be attributed to the fact that items processed later in
the information series, particularly when they are contrasting, unexpected, and conspicuous,
are perceived as more informative and are thus highly elaborated in working memory, which
makes them more available to a search of long-term memory.367
361
See Anderson (1981): 144; Hogarth/Einhorn (1992): 3. Note that the influence of information order effects
on human assessments, impressions, and belief formations has long been recognized in psychology. See,
e.g., Asch (1946): 270-272; Fernberger (1920): 132f.
362
See Kahneman/Frederick (2005): 270.
363
See Hogarth/Einhorn (1992): 1.
364
See Hogarth/Einhorn (1992): 1f.
365
See Belmore (1987): 480, 486.
366
See Anderson (1981): 183, 273; Fiske (1980): 903f.
367
See Hastie (1980); Plotnik/Kouyoumdjian (2013): 245. For a list of further explanations for primacy and
recency effects and relevant psychological evidence, see Eagly/Chaiken (1993): 265.
64
Under a dual-process perspective, information order effects can be conceived as the result of
the joint failure of System 1 and System 2.368 Specifically, due to the complexity inherent in
the environment and the impossibility of human perception and cognition to fully and system-
atically apprehend and process all relevant stimuli and comprehend the complex interrelations
among them, simplifying cognitive processing strategies are likely to be utilized. In particular,
the generation of mental representations (System 1 processing) of the evidential stimuli en-
countered in a particular task is likely to involve selective attention to and focus on some sali-
ent but potentially normatively irrelevant aspects of the information set.369 In the context of
belief updating in light of newly incoming evidence, preconscious System 1 processing can
yield mental representations of the strength and importance of evidence which are determined
by the salience and serial position of information rather than by its substance only.370 These
intuitively disproportionately cued impressions are likely to be translated into subsequent as-
sessments and judgments at System 2 level. This contention is in line with the prevailing
opinion in the relevant dual-process literature which posits that human judgments and infer-
ences are largely determined by System 1 automatically generated stimuli, with System 2 ana-
lytical processing doing nothing more than merely endorsing and translating the input from
System 1 into explicit judgments.371 Even though the disproportional mental representations
of evidence weight generated by System 1 can theoretically be corrected at System 2 level,
this scenario appears rather unlikely because in order to block impulses from System 1, Sys-
tem 2 has to actively engage in suppressing the input from System 1.372 This requires a con-
scious awareness of the biasing influence of information order effects. However, as cognitive
biases naturally operate subconsciously, the flawed impulses from System 1 are expected to
be endorsed within the integration of the inconsistent pieces of evidence into a final judgment
at System 2 level.373
Following this general consideration of the nature and cognitive roots of information order
effects, the implications of the belief-adjustment model are considered in more detail subse-
quently. Before focusing on the particular order-effect predictions of the Hogarth and Einhorn
model, some general clarifications regarding the implications of the different cognitive pro-
cessing modes – sequential (SbS) and simultaneous (EoS) – are needed. Specifically, sequen-
tial processing is predicted to induce no order effects for consistent evidence. In the case of
mixed evidence, however, the SbS processing mode is always theorized to lead to recency
effects, unless an individual is completely insensitive to positive or negative evidence (α = 0
or β = 0). In addition, the size of recency effects is predicted to be an increasing function of
368
This notion follows the argument of Kahneman/Frederick (2002): 52 that cognitive errors and biases typi-
cally arise when both cognitive systems fail.
369
For such contentions in a general context, see Evans (2013b): 138.
370
Evidence for the fact that the contrast effect, which – as might be recalled from Section 3.2.1.3 – is the pri-
mary source of the order-effect predictions of the belief-adjustment model, operates at the stage of mental
representation has been provided by Simpson/Ostrom (1976): 628.
371
See, e.g., Evans (2013b): 138; Kahneman/Frederick (2002): 58f.
372
For such contentions in a general context, see Evans (2013b): 138.
373
For such contentions in a general context, see Kahneman (2011): 28.
65
both evidence strength and high general sensitivity to evidence.374 In contrast, simultaneous
processing is theorized to induce primacy effects in individuals’ belief revisions irrespective
of the nature of evidence (consistent or mixed).375
The order-effect predictions of the belief-adjustment model are summarized in Table 2376.
As shown in Table 2, the order-effect predictions of the belief-adjustment model result from
the interaction of type of evidence (consistent versus mixed), complexity (simple versus com-
plex), length of series (short versus long), and response mode (EoS versus SbS). With regard
to the latter, is instructive to recall Hogarth and Einhorn’s (1992) contention discussed in
Section 3.2.1.3 that the SbS response mode exclusively induces SbS processing. In contrast,
the EoS response mode will only invoke EoS processing when the task under consideration
comprises short series of simple (i.e., cognitively less demanding) evidence. This scenario is
depicted in the top row of Table 2. In contrast, in the case of short series of complex (cogni-
tively demanding) evidence or long, naturally heavy loaded series of information, SbS pro-
cessing is used in spite of EoS response mode task requirements. These conditions are reflect-
ed in the last two rows of Table 2.377
In brief, in the case of short series of simple evidence evaluated and processed in an EoS
manner, the belief-adjustment mode exclusively predicts primacy effects. In contrast, recency
effects are predicted for short series of simple, mixed evidence evaluated and processed in a
SbS fashion. On the other hand, if the evidence encountered is consistently positive or con-
sistently negative, no order effects are assumed to occur with SbS response (and processing)
mode. These conditions are presented in the first row of Table 2.378
374
See Hogarth/Einhorn (1992): 16. For a mathematical derivation of these predictions, refer to
Hogarth/Einhorn (1992): 49f.
375
See Hogarth/Einhorn (1992): 16.
376
For clarity of exposition, and in consistency with the state of the art in auditing research on belief updating
(see, e.g., Kahle/Pinsker/Pennington (2005): 8), merely the order-effect predictions for the evaluation-type
model (R = 0) are presented in Table 2. Order-effect predictions for all kinds of evidence (mixed and con-
sistent) under the estimation mode (R = Sk-1) are identical to the predictions for mixed evidence processed in
evaluation manner (left-hand branch of predictions in Table 2).
377
See Hogarth/Einhorn (1992): 18.
378
See Hogarth/Einhorn (1992): 18.
66
In the constellations involving short series of complex, mixed evidence, the belief-adjustment
model predicts recency effects regardless of the response mode. The rationale for this predic-
tion is that under conditions of complexity and high cognitive load, SbS processing will be
utilized even when the task explicitly requires an EoS response. For a short series of complex,
consistent evidence, no order effects are expected. These conditions are reflected in the sec-
ond row of Table 2.379
Finally, in the case of long series of evidence, the belief-adjustment model predicts (a tenden-
cy toward) primacy effects. That is, even though the SbS procedure used in processing large
amounts of information normally invokes recency effects, with growing information accumu-
lation, the individual sensitivity to new evidence (α, β) is assumed to decrease, thereby giving
rise to primacy. Hence, Hogarth and Einhorn (1992) argue that while recency effects may
theoretically occur in long information sets, this chance rapidly decreases with growing length
of the series. These conditions are depicted in the third row of Table 2.380
The present study is situated within the context of a short series of complex, mixed evidence
and sequential (SbS) responses to (and processing of) information. This approach is chosen
because it appears to be most representative of the real-world audit process in general and the
evaluation of the going concern assumption in particular. Specifically, in evaluating a firm’s
going concern ability, particularly in the context of economically turbulent conditions, an au-
ditor rarely obtains exclusively consistent information regarding the entity’s viability. Fur-
thermore, auditing standards explicitly require that both conditions and events that cast poten-
tial doubt on the going concern assumption as well as mitigating factors are considered (ISA
570.A2; IDW PS 270.11; SAS 126, AU-C 570.9, A2-A3). In addition, human judgment in
general and auditor judgment in particular naturally proceed incrementally rather than in one
go (simultaneously).381 Accordingly, a sequential obtaining and integration of evidence into
an evolving impression of the firm’s viability throughout the audit is more likely than the re-
ceipt of all the required information in a bundle or the postponement of preliminary assess-
ment of the going concern assumption until all relevant data is available and can be evaluated
in an aggregated manner.382 With regard to the length of the series, it can be argued that audi-
tors operate under conditions of immense time and deadline pressure, which renders the ob-
taining and processing of particularly lengthy sets of evidence (involving more than 17 pieces
of information) rather unlikely.383 In the context of going concern judgments, it has been rec-
ognized that even though the search for and evaluation of information relevant to the going
concern assessment continues during the entire audit process, auditors generally obtain only a
limited number of evidence items indicating firm’s viability or the lack thereof.384 Finally, the
379
See Hogarth/Einhorn (1992): 18.
380
See Hogarth/Einhorn (1992): 18.
381
See Ashton/Ashton (1988): 624 (with a number of relevant psychological references along these lines);
Gibbins (1984): 110.
382
See Asare (1992): 381.
383
See Kahle/Pinsker/Pennington (2005): 9f., Trotman/Wright (2000): 170.
384
See Asare (1992): 381.
67
use of a short series of mixed evidence presented and processed in a sequential manner is in
line with the great majority of the auditing studies using the belief-adjustment model.385
With regard to task complexity, the going concern case used in the present experimental in-
vestigation is classified as a complex task according to Hogarth and Einhorn’s (1992) concep-
tualization of complexity because it involves heavy informational and cognitive load. Specifi-
cally, the case involves extensive background information (620 words general information
plus balance sheet, income statement and cash flow data) and four additional pieces of infor-
mation (ranging from 27 to 75 words with a mean of 47.5 words) which had to be processed
within short time.386 In addition, the going concern judgment has been broadly recognized as
a highly complex and critical audit task.387
Collectively, the conditions of a short series of complex, inconsistent, and sequentially pre-
sented and processed information establish the occurrence of recency effects in auditors’ be-
lief revisions (Table 2, red-circled cell). With other words, in arriving at an assessment of the
going concern ability of the firm under consideration, auditors are expected to overweight the
most recently received information in the sequence. Overall, based on the predictions of the
belief-adjustment model, the following hypothesis is stated:388
H1 : Auditors who evaluate mitigating factors followed by contrary information will exhibit
greater downward belief revision (or less upward belief revision) than those who
evaluate the same information presented in the reverse order c.p.
Before discussing the belief-adjustment model and its implications in the context of auditing,
the subsequent section provides a review of the relevant general (non-auditing) research on
Hogarth and Einhorn’s belief-adjustment model. This review aims at demonstrating the va-
lidity of the model and the pervasiveness of recency effects in a number of judgment domains.
The belief-adjustment model has sparked considerable research interest in a number of differ-
ent research fields. Interestingly, auditing research was the first field of research to use the
model and to provide interdisciplinary corroboration for its validity. In particular, Ashton and
Ashton (1988) was the very first study to independently389 investigate and corroborate the
385
See Asare/Messier (1991): 79; Kahle/Pinsker/Pennington (2005): 9.
386
In comparison, in their validating experiments, Hogarth/Einhorn (1992): 21 classified background infor-
mation stems of averagely 88 words and additional pieces of evidence of averagely 52 words each as com-
plex. Hence, as the information amount and load in the present study is way beyond the scenarios employed
by Hogarth/Einhorn (1992), the classification of the going concern assessment as a complex task is also in
line with the reference study.
387
See, e.g., Ashton/Kennedy (2002): 221f.; Chow/McNamee/Plumlee (1987): 128f.
388
Note that the focus on auditors’ belief revisions rather than final assessments is methodologically driven.
This issue is addressed in Section 5.2. This hypothesis formulation is based on Asare (1992): 383 and Cush-
ing/Ahlawat (1996): 114.
389
Note that in their seminal paper, Hogarth/Einhorn (1992) not only present the belief-adjustment model and
derive its predictions but also provide initial evidence in support of the model’s validity. The authors par-
68
predictions of the belief-adjustment model with experienced professional auditors. The exten-
sive research on auditors’ belief updating conducted ever since the pioneer study by Ashton
and Ashton (1988) is reviewed in Section 3.4. However, it can already be noted at this point
that the great academic interest in the study of auditors’ belief revision underpins the funda-
mentality of this cognitive process and the importance of the implications of the predicted
information order effects in auditing.
Subsequently, a selection390 of interdisciplinary belief-revision studies is presented, and the
major findings of this research are briefly discussed. The review proceeds from general to
accounting-specific research fields. The degree of detail of consideration is determined by the
relevance of the individual research branches to auditing.
Overall, the majority of belief-revision studies reviewed herein have focused on settings in-
volving short sequences of complex, mixed evidence. As discussed in the previous section,
according to the belief-adjustment model, recency effects are theorized to emerge in this sce-
nario. In consistency with this prediction, recency effects have been found in a number of
judgment domains, including medical judgment (e.g., Chapman, Bergus and Elstein (1996)),
legal judgment (e.g., Costabile and Klein (2005), Kerstholt and Jackson (1998)), personnel
evaluations (e.g., Highhouse and Gallo (1997)), television picture quality evaluations (e.g.,
Aldridge et al. (1995), Hands and Avons (2001)), military judgments (e.g., Adelman, Tolcott
and Bresnick (1993)), and even in the context of blind date choice (e.g., Bruine de Bruin and
Keren (2003)).391
ticularly focus on the predictions of the model with regard to short series of complex evidence as this condi-
tion has not been extensively explored in the past, and it is also the area where the Hogarth and Einhorn
model makes novel predictions, i.e., the emergence of recency effects in the presence of mixed (positive and
negative) evidence and the absence of order effects in the case of consistent evidence. In a series of five ex-
periments using generic tasks which involved updating causal hypotheses by student participants, Ho-
garth/Einhorn (1992) obtain strong evidence corroborating the order effect predictions of the belief-
adjustment model. For a detailed description of the procedures and concrete results of these five experi-
ments, see Hogarth/Einhorn (1992): 20-33. Dillard/Kauffman/Spires (1991) replicated the four generic
evaluation-type task experiments by Hogarth/Einhorn (1992). Their results provide support for the model’s
predictions as well as for the validity of the fundamental contrast assumption. See Dillard/Kauffman/Spires
(1991): 623-630. A further replication study was conducted by Tubbs et al. (1993) who provide strong sup-
port for the occurrence of recency effects in the case of a short sequence of complex, mixed evidence. Inter-
estingly, and contrary to the predictions of the model, the authors also report patterns consistent with recen-
cy even in light of consistent evidence. In addition, Tubbs et al. (1993) find that even when participants are
given debiasing training, recency effects in their belief revisions still emerge. See Tubbs et al. (1993): 267f.
Thus, the authors provide strong support for the prevalence and robustness of recency effects in subjects’ be-
lief revisions.
390
As the relevant research on belief updating which employs the belief-adjustment model is immense, the
overview of research presented in this section does not make any claims to be comprehensive or complete.
Rather, it intends to provide some insights into the broad field of application and the substantial support for
the validity of the Hogarth and Einhorn model.
391
See Adelman/Tolcott/Bresnick (1993): 364; Aldridge et al. (1995): 154; Bruine de Bruin/Keren (2003): 100;
Chapman/Bergus/Elstein (1996): 209f.; Costabile/Klein (2005): 50; Hands/Avons (2001): 644, 646f.;
Highhouse/Gallo (1997): 42; Kerstholt/Jackson (1998): 452. Note that the studies by Aldridge et al. (1995)
and Costabile/Klein (2005) do not explicitly use the belief-adjustment model as their theoretical foundation;
however, as the results of these studies fit the model’s predictions, they are also viewed as providing support
for the validity of the model. In addition, it is noteworthy that while providing an overall support for the pre-
dictions of the belief-adjustment model, Adelman/Tolcott/Bresnick (1993) do not obtain results consistent
with the contrast assumption. In addition, they find that order effects explain only a small portion of the total
69
The belief-adjustment model has also experienced broad application and empirical validation
in the tax research domain. The pioneer studies in this area were conducted by Pei, Reckers
and Wyndelts (1990, 1992) who document significant recency effects in experienced tax pro-
fessionals’ belief revisions, both with respect to judgments of tax treatment adequacy and cli-
ent recommendations (i.e., decisions).392 Consistently, Gelardi (2010) reports strong overall
recency effects in tax students’ assessments. He further finds that recency is magnified with
increasing information load.393 The robustness and validity of the belief-adjustment model has
also been supported by Hite and Stock (1996). They demonstrate that the model’s predictions
hold even when subjects possess strong prior beliefs (anchors) regarding a particular tax posi-
tion. Such beliefs are recognized by Hogarth and Einhorn (1992) to induce ceiling and floor
effects, i.e., not to leave much room for upward (downward) belief adjustments in the case of
strong positive (negative) initial beliefs.394
In contrast to the aforementioned studies, Mayper, Anderson and Kilpatrick (1999) do not
find information order effects in participants’ responses in a tax working paper review scenar-
io. Rather, they obtain results suggesting that subjects anchor on the staff conclusion provided
in the working papers and tend to confirm their own initial position.395 While these results are
at odds with the predictions of the belief-adjustment model, it is conceivable that the findings
of the study are largely driven by the experimental design which provides participants with an
anchor (staff conclusion) in which the previously depicted information is already processed
(by another person) instead of allowing subjects to independently form beliefs. Hence, the
review task may not allow for order and anchoring effects to be disentangled.396
Cuccia and Mc Gill (2000), on the other hand, find that recency effects in tax judgments are
mitigated when tax professionals perform familiar tasks and have control over the order in
which inconsistent evidence is processed. In unfamiliar tasks, however, which are per defini-
tion complex in nature, subjects were found to exhibit recency effects in their judgments re-
gardless of the benefits of control over task structure and evidence evaluation sequence.397
variance in subjects’ responses (R2 < 0.10). In later studies, Adelman et al. (1996): 258 surprisingly and con-
trary to the predictions of the belief-adjustment model find primacy effects instead of recency for a scenario
involving the sequential presentation of simple, mixed data. In addition, in a study involving a long series of
information, Keltz/Adelman (2008): 27f. find that in contrast to the predictions of the Hogarth and Einhorn
model, recency rather than primacy effects arise. However, it should be noted that the model merely predicts
a “force toward primacy” which increases with growing number of items employed. As 21 very simple items
were employed in the study by Keltz/Adelman (2008), it can be expected that subjects’ tiredness and de-
creasing sensitivity to new evidence were not strong enough to produce primacy. In addition, Keltz/Adelman
(2008): 28 admit that due to the directional grouping of evidence items (seven positive, seven negative, and
seven neutral cues), a “chunking” effect might have occurred, so that subjects mentally processed three
complex evidence bundles instead of 21 simple single cues. Under this point of view, the emergence of re-
cency effect does not appear to be at odds with the rationale of the belief-adjustment model.
392
See Pei/Reckers/Wyndelts (1990): 136-139; Pei/Reckers/Wyndelts (1992): 188-190.
393
See Gelardi (2010): 42f.
394
See Hite/Stock (1996): 83f., 93f.
395
See Mayper/Anderson/Kilpatrick (1999): 158f.
396
This proposition is reinforced by the fact that there are no auditing studies which explore participants’ belief
revisions using a review task.
397
See Cuccia/Mc Gill (2000): 420, 426-429.
70
Dillard, Kauffman and Spires (1991) provide support for the intercontextual validity of the
belief-adjustment model by demonstrating that the predictions of the model generally hold
regardless of the judgment domain (generic vs. management accounting). In addition, they
corroborate the validity of the contrast assumption.398 Further evidence in support of the re-
cency-effect predictions of the Hogarth and Einhorn model in a management accounting con-
text has been provided by Rutledge (1995). He also empirically demonstrates that recency can
be alleviated by framing the information series in a manner which is inconsistent with the
direction of the expected recency effect.399 Further evidence substantiating the prevalence of
recency effects has been provided by LaSalle (1997) in the context of accounting students’
ethical judgments.400
Evidence in support of the validity of the belief-adjustment model has further been provided
in the context of financial disclosure. Specifically, Baird and Zelin (2000) and Theis, Yankova
and Eulerich (2012) demonstrate that the order-effect predictions of the model hold in scenar-
ios involving short information series, EoS responses, and both complex and simple evalua-
tion tasks.401 In addition, the Hogarth and Einhorn model has been found to provide a reason-
able account of the manner of information processing in disclosure scenarios involving long
information sets. In particular, Pinsker (2007) corroborates the validity of the processing and
adjustment mechanisms underlying the belief-adjustment model by demonstrating that se-
quential disclosure induces greater belief revisions compared to simultaneous disclosure be-
cause of the more salient contrast effect arising in the case of updating beliefs in a successive
manner. This result is reinforced in a follow-up study by Pinsker (2011). However, the latter
also finds that recency effects pervade subjects’ evaluations even in the case of a fairly long
series of information items (forty cues), whereas the belief-adjustment model predicts a force
toward primacy for such scenarios.402
The occurrence of recency effects in a series of contrasting, successively processed stimuli
has further been demonstrated by Guiral-Contreras, Gonzalo-Angulo and Rodgers (2007) in
the context of loan rating decisions.403 Finally, recency effects have been documented even in
experimental market settings. In particular, Tuttle, Coller and Burton (1997) report results
indicating that contrary to the efficient market hypothesis which posits that the market assimi-
lates new information immediately and immaculately, i.e., free of bias, systematic individual
information processing biases like recency can carry over to and “survive” at the market lev-
el.404 These findings suggest the descriptive superiority of the belief-adjustment model over
normative theory and highlight the robustness of recency phenomena.
398
See Dillard/Kauffman/Spires (1991): 623-630.
399
See Rutledge (1995): 34f.
400
See LaSalle (1997): 28-31. Remarkably, non-accounting students have not been found to exhibit recency
effects in their ethical judgments. LaSalle (1997) argues that this finding might be quite alarming if it mir-
rors accounting students’ inability to critically reflect their role in the profession.
401
See Baird/Zelin (2000): 76; Theis/Yankova/Eulerich (2012): 143-145.
402
See Pinsker (2007): 211; Pinsker (2011): 177, 179.
403
See Guiral-Contreras/Gonzalo-Angulo/Rodgers (2007): 301.
404
See Tuttle/Coller/Burton (1997): 101.
71
In summary, the results of the studies reviewed above provide substantial support for the de-
scriptive validity of the belief-adjustment model and the reasonableness of its predictions. A
striking common finding of a plethora of studies in diverse research areas is the salience and
robustness of recency effects in human information processing and inference. Collectively,
the empirical evidence reviewed in this section suggests that human information processing
and evidential reasoning indeed naturally follows an anchoring-and-adjustment (i.e., adaptive)
rather than a probabilistic (e.g., Bayesian) approach.405
In conclusion to the general consideration of the belief-adjustment model, the latter is subject-
ed to a critical evaluation. The main advantage of the model is its cognitive and computational
simplicity and economy.406 The underlying anchoring-and-adjustment strategy implies that at
any stage, merely the current position (anchor) and the effect of the new piece of information
have to be considered in order to arrive at an updated belief. With other words, the model “al-
lows one to keep a ‘running total’ of the effects of prior information while reducing memory
load.”407 The benefits arising from the use of the belief-revision approach in auditing have
been empirically demonstrated by Abou-Seada and Abdel-Kader (2003) who find that the
updating of beliefs in a manner of sequential anchoring and adjustment leads to audit efficien-
cy (i.e., completion of the audit in less time).408
However, like any other theoretical model, the belief-adjustment model also has limitations.
Its major drawback is the lacking consideration of interdependencies and correlations be-
tween different pieces of evidence. Hence, the fact that, for instance, in certain situations a
new piece of negative information can potentially influence a target hypothesis only by over-
riding a prior positive information item is not captured by the model. With other words, the
belief-adjustment model does not account for reinterpretations of prior information cues.409
The difficulties of the model to capture the effects of extreme new evidence which irrefutably
supersedes prior evidence have been empirically demonstrated by Trueblood and Busemeyer
(2010, 2011).410 Hogarth and Einhorn (1992) take this critical issue into account by stating
that the belief-adjustment model is based on the implicit assumption that “the outcomes of the
405
See Kerstholt/Jackson (1998): 452.
406
See Asare/Messier (1991): 76; Ashton/Ashton (1988): 625; Hogarth/Einhorn (1992): 36.
407
Einhorn/Hogarth (1985): 5.
408
See Abou-Seada/Abdel-Kader (2003): 109, 118.
409
See Adelman et al. (1996): 258; Adelman et al. (1997): 328; Lagnado (2011): 205f.
410
See Trueblood/Busemeyer (2010): 1169-1171; Trueblood/Busemeyer (2011): 1544-1549. In their studies, the
authors compare the predictive and explanatory power of the belief-adjustment model and the quantum in-
ference model (see Trueblood/Busemeyer (2011): 1522-1533 for a review) in the context of jury decision
making and conclude that while generally performing strongly, the belief-adjustment model is inferior in sit-
uations involving irrefutable evidence. Whether these findings are also generalizable to the field of auditing,
is an open empirical question.
72
coding process already include whatever conditioning the subject has done based on prior
evidence.”411
In addition to the issue of insufficient reflection of evidence reinterpretations, the qualitative
premises regarding the cognitive mechanisms underlying the process of belief adjustment,
which are essentially the basis for the order-effect predictions of the model, might be at issue.
To begin with encoding, although Hogarth and Einhorn (1992) implicitly assume that an in-
dividual will either use an evaluation or an estimation mode within the same task, this as-
sumption might be unwarranted, as the authors themselves admit. In particular, given that
minor contextual distinctions have been found to induce different processing modes, it is con-
ceivable that individuals change modes during information processing within the same task.412
With regard to the evaluation mode, which has been recognized as more relevant to auditing
than the estimation mode, the notion that encoding information as positive (negative) consist-
ently leads to upward (downward) belief revisions may not completely reflect real-world hu-
man belief updating. More specifically, it is conceivable that due to a “saturation” effect or
premature opinion formation and commitment to the individually reached position, subjects
do not change their beliefs although they perceive new information as positive or negative.
With regard to processing, there is no clear indication of when a transition from EoS to SbS
processing occurs in tasks involving an EoS response mode and substantial complexity and
cognitive demands. Hence, the derivation of order-effect predictions in the case of complex
tasks and EoS response mode might be somewhat arbitrary.413 Finally, it should be born in
mind that the contrast effect assumption relating to the cognitive subprocess of adjustment is
crucial for the order-effect predictions of the belief-adjustment model. Hence, evidence which
casts doubt on the validity of the contrast effect is also reducing the universality of the model.
These limitations and critical features notwithstanding, the belief-adjustment model possesses
considerable cognitive and descriptive appeal, and as indicated in the previous sections, it has
advanced the academic understanding of the processes and cognitive mechanisms underlying
belief revision and has inspired a wealth of multidisciplinary research. As mentioned at the
outset of Section 3.2.4, auditing research has demonstrated substantial interest in the explora-
tion of auditors’ belief revisions and has widely adopted the belief-adjustment model as the
theoretical foundation of the study of auditors’ judgments in a number of different contexts.
The following section sheds light on the rationale for this vast amount of research attention.
As indicated in Section 2.4, auditing is a highly complex multi-stage process conducted under
conditions of considerable uncertainty, ambiguity, pressure, and information load.414 As rec-
ognized by Mock et al. (1997), “[d]ealing with uncertainty involves implicitly or explicitly
411
Hogarth/Einhorn (1992): 39.
412
See Hogarth/Einhorn (1992): 37f.
413
See Hogarth/Einhorn (1992): 38.
414
See Rose/Rose (2003): 316.
73
expressing and updating beliefs about uncertain quantities, events or outcomes.”415 Thus,
belief adjustment plays a crucial role in the audit process.
The preceding sections indicate that people generally aggregate evidence and update beliefs in
an adaptive (sequential anchoring-and-adjustment) rather than normative (probabilistic) man-
ner. In auditing, this tendency might be even more salient because of the nature of auditor
judgment. Specifically, as aptly put by Gibbins (1984), naturally auditor “judgment proceeds
incrementally rather than by gathering full information and integrating it all before choosing
a response.”416 With other words, auditor judgment represents “an ongoing mental process
rather than a discrete act.”417 Moreover, auditing itself can be characterized as a sequential
process in which a number of individual judgments are successively integrated into a higher-
level composite judgment, i.e., the audit opinion.418 Overall, sequential information pro-
cessing plays a crucial role in auditing. In this context, the belief-adjustment model represents
a very relevant and reasonable descriptive account of auditors’ information processing and
belief updating.419 The model’s relevance and applicability to auditing has been empirically
demonstrated by Krishnamoorthy, Mock and Washington (1999) whose comparative exami-
nation of belief revision models in auditing reveals that the belief-adjustment model outper-
forms Bayesian inference, cascaded inference theory, and Dempster-Shafer theory of belief
functions as a descriptive model of auditors’ belief revisions. Specifically, the Hogarth and
Einhorn model is the only model found to capture both the direction and magnitude of audi-
tors’ belief adjustments.420
As implicated by the belief-adjustment model, the sequential processing of evidence and up-
dating of beliefs, although beneficial in terms of relieving cognitive load and demands on
memory and effort, can also lead to systematic information order effects in individual belief
revision.421 Ideally, auditors’ judgments are determined by the substance of evidence (i.e., its
diagnosticity, relevance, and reliability), not the order in which it happens to be received and
evaluated (i.e., its serial position).422
Generally, information order effects423 may have severe implications for both the efficiency
and the effectiveness of the audit, in that the order of presentation (1) may affect auditors’
search for additional evidence and the extent of audit testing, thus jeopardizing both the cost
and the timeliness of the audit, and (2) may also lead to identical information resulting in dif-
415
Mock et al. (1997): 123f.
416
Gibbins (1984): 110.
417
Gibbins (1984): 110.
418
See Felix/Kinney (1982): 246-248; Libby/Luft (1993): 435.
419
See Asare/Messier (1991): 76; Trotman (1999): 19.
420
See Krishnamoorthy/Mock/Washington (1999): 119f.
421
For such claims, see, among others, Asare (1992): 379 and Kahle/Pinsker/Pennington (2005): 2.
422
See Ashton/Kennedy (2002): 221; Kennedy (1993): 231.
423
It should be noted that order effects and recency bias concern the order of the presentation of information,
which may not necessarily coincide with the temporal (i.e., chronological) order in which the events that are
reported actually occurred. With regard to the latter, in many situations it reasonable to attach a greater
weight to temporally more recent (after year-end) evidence. See Favere-Marchesi (2006): 71.
74
ferent judgments and different actions by virtue of it being processed in a different order.
With other words, order effects can cause over- or underauditing and reduce the accuracy and
quality of auditor judgment.424
Another issue relevant to the discussion of information order effects in the context of auditing
is the fact that auditors frequently obtain management-provided information and evidence
which may be intentionally or unintentionally biased.425 Overall, the client may structure and
present evidence in such a manner and sequence that it favors auditors’ unduly optimistic
view on the issue under consideration. Management’s control over evidence and the fact that
people are generally inclined to intentionally influence the environment in pursuit of their
own interest and purposes426 highlight the necessity for auditors to be aware of the deleterious
influence of order effects and the ability of the client to induce them.
After indicating the fundamentality of the process of belief updating as well as the theoretical
importance of the belief-adjustment model in auditing, an overview of the relevant auditing
literature is provided.
The belief-adjustment model has inspired profound experimental research in auditing. For
clarification purposes, some basic points relating to the relevant auditing literature on belief
revision are addressed prior to the discussion of the central order-effect studies. First, as indi-
cated in Section 3.2.3, the auditing literature on belief revision has majorly focused on scenar-
ios involving a short series of mixed evidence presented and evaluated sequentially (SbS) as
these constellations are consistent with the incremental nature of auditor judgment as well as
the ambiguity, time and capacity constraints underlying the audit process. Second, auditors’
belief revisions have been explored in the context of various audit tasks, including going con-
cern evaluations, internal control assessments, accounts receivable estimations, fraud assess-
ments, material misstatement judgments, and inventory write-down evaluations.427 Third,
while early research using the belief-adjustment model was primarily concerned with testing
the model’s validity and order-effect predictions, later studies have focused on exploring the
interrelations between various task and context variables relevant to the audit environment as
well as on identifying and exploring potential debiasing mechanisms.428 Forth, the auditing
literature reviewed below has been somewhat inconsistent in the use of the relevant terminol-
ogy. Specifically, the terms “response mode”, “presentation mode”, and “processing mode”
424
See Ashton/Ashton (1988): 638; Krull/Reckers/Wong-On-Wing (1993): 144; Messier/Tubbs (1994): 58;
Tubbs/Messier/Knechel (1990): 459. Note that judgment quality in the context of cognitive biases is typical-
ly evaluated in terms of logical consistency, i.e., correspondence to normative theories. However, other
judgment quality proxies may also be appropriate depending on the nature of the particular bias. For a dis-
cussion of judgment quality criteria, see Section 2.4.1.
425
See Bell/Peecher/Solomon (2005): 19; Silvoso et al. (1973): 31.
426
See Fiske/Taylor (2008): 16.
427
See Kahle/Pinsker/Pennington (2005): 9f.; Trotman/Wright (2000): 170.
428
See Asare/Messier (1991): 82f.
75
have generally been used synonymously in the existing auditing literature. 429 This also applies
to the terms “attitudes toward evidence” and “sensitivity to evidence” as well as “positive
(negative) evidence” and “confirming (disconfirming) evidence”.430 The present work at-
tempts to use these terms as unambiguously and uniformly as possible. Nevertheless, the pos-
sibility of remaining ambiguities should be noted. Subsequently, a brief 431 review of the rele-
vant experimental studies dealing with auditors’ belief adjustments and susceptibility to
information order effects is presented.
As previously indicated, initial support for the predictions of the belief-adjustment model in
the context of auditing was provided by Ashton and Ashton (1988). The researchers demon-
strate that in the presence of inconsistent evidence, auditors’ belief revisions are systematical-
ly and significantly influenced by the order of information presentation, resulting in judg-
ments reflecting a recency bias432. For consistent evidence, the authors do not indicate
information order effects. In addition, Ashton and Ashton (1988) find that sequential pro-
cessing leads to more extreme belief revisions than simultaneous processing. 433 Overall, these
findings provide strong support for the validity of the Hogarth and Einhorn model.
In conformity with Ashton and Ashton (1988), Tubbs, Messier and Knechel (1990) also obtain
no order effects for consistent evidence and recency effects for mixed evidence. 434 Further
empirical evidence for the emergence of recency in auditors’ belief revisions in light of a
short series of complex, inconsistent, and sequentially presented and processed information is
provided by Asare (1992) and Pei, Reed and Koch (1992).435 Remarkably, Asare (1992) indi-
429
See Kahle/Pinsker/Pennington (2005): 37. Although these terms are not strictly equivalent in meaning, with
regard to the SbS mode, they have identical implications. Hence, the interchangeable application appears
reasonable.
430
Note that whether a piece of evidence of which sign whatsoever is confirming or disconfirming, depends on
the framing of the hypothesis under consideration. Thus, an interchangeable use can be misleading.
431
Note that it is beyond the scope of the present work to provide a detailed consideration and discussion of
each study on auditors’ belief revision conducted in the last two decades. For an excellent and comprehen-
sive overview of belief-adjustment research in the field of accounting and auditing, refer to
Kahle/Pinsker/Pennington (2005). They organize their discussion around the following five categories of
factors studied in the context of auditors’ belief revisions: (1) individual factors (e.g., knowledge, cognitive
style, sensitivity to evidence); (2) environmental factors (e.g., incentives, pressures, inherent risk); (3) input
task factors (e.g., evidence strength and type, task complexity, response mode); (4) processing factors (e.g.,
processing mode, accountability requirements); and (5) output task factors (e.g., type and domain of judg-
ment (audit vs. tax)).
432
While considered under a normative viewpoint, recency represents a bias in information processing, some
researchers have pleaded for the reasonableness of recency phenomena as means of easing cognitive strain
and information overload. The proponents of the reasonableness of recency effects have argued that the in-
terpretation of recency as information processing bias is not justified. See, e.g., Pei/Reckers/Wyndelts
(1992): 190.
433
See Ashton/Ashton (1988): 638-640. The finding of more extreme belief revisions in the case of sequential
processing is corroborated by Ashton/Ashton (1990): 15f. In the context of mixed evidence, this result is at-
tributed to the fact that in the EoS mode, evidence is evaluated in a cumulative manner within a single be-
lief-adaptation process, so that the contrasting informational stimuli are averaged out prior to integration
with the current belief. In the SbS mode, on the other hand, each information cue is assessed individually
within an evolving process, which is a precondition for the arousal of greater contrast effects. See Kennedy
(1993): 235; Trotman/Wright (1996): 176.
434
See Tubbs/Messier/Knechel (1990): 459.
435
See Asare (1992): 388f.; Pei/Reed/Koch (1992): 181.
76
cates that the recency bias in auditors’ going concern judgments is carried over to the choice
of type of audit report (modified vs. unqualified) at issue. Specifically, auditors who obtain
confirmatory evidence at the end of a series exhibit stronger final beliefs and a greater confi-
dence in the client’s viability, and thus issue more unqualified opinions than those who evalu-
ate the same evidence in a reverse order with the unfavorable evidence processed last.436 As-
are’s (1992) findings are reinforced by Guiral and Esteo (2006).437 In contrast to the two
aforementioned studies, Messier (1992) reports results suggesting that recency effects in audi-
tors’ belief revisions and final judgments do not influence the extent of audit testing or the
audit report choice made.438
While the aforementioned studies uniformly find evidence of recency effects at least at the
judgment level, Anderson and Maletta (1999), Monroe and Ng (2000), and Reckers and
Schultz (1993) do not indicate order effects in auditors’ judgments made under conditions of
high risk.439 In a similar vein, Butt and Campbell (1989) do not document recency effects for
participants exhibiting high prior beliefs. Subjects with low priors, however, are found to suc-
cumb to recency. Based on these findings, the authors conclude that the extent of recency ef-
fects in subjects’ responses is inversely related to the strength of prior beliefs.440 In opposition
to the aforementioned studies, Rose and Rose (2003) find that high fraud risk exacerbates,
rather than mitigates, recency effects because of the increased attention and elaboration in
information processing and evidence evaluation under high-risk conditions. The increase in
attention and in the amount of mental processing potentially leads to greater complexity and
cognitive load, which is a preposition for the emergence of recency effects.441
A bunch of studies have focused on the exploration of the role of knowledge and expertise in
the context of auditors’ belief revisions. This research has addressed the question of whether
experienced auditors are less susceptible to information order effects than are novices.442 In
436
See Asare (1992): 390f. However, Adelman/Tolcott/Bresnick (1993): 366 reconstructed the R2-values for
Asare’s (1992) study and found that the belief revision experiment yielded an R 2 of 0.22, while the audit re-
port choice experiment only produced an R2 of 0.067.
437
See Guiral/Esteo (2006): 607-610.
438
See Messier (1992): 149.
439
See Anderson/Maletta (1999): 83; Monroe/Ng (2000): 161-164; Reckers/Schultz (1993): 140. For low-risk
conditions, both Anderson/Maletta (1999) and Reckers/Schultz (1993) report order effects (primacy and re-
cency, respectively) in subjects’ responses.
440
See Butt/Campbell (1989): 476-478. Unfortunately, the authors fail to provide an explanation for this pattern
of findings. As participants are not found to be confirmation-prone, an escalation of commitment to the prior
belief is apparently not the mechanism leading to lower recency effects in the case of high prior beliefs. It is
conceivable that subjects with a higher commitment to an initial position are generally more focused and ef-
fortful in information processing and belief formation, and thus are less susceptible to non-normative factors
like the serial position of information. However, it is also conceivable that the results observed are merely an
experimental artefact as Butt/Campbell (1989) neither employ a strict SbS nor a strict EoS response mode
but a mixture of both (i.e., subjects obtain ten items of additional evidence and are asked to make two belief
adjustments, each after facing a set of five positive and five negative cues). The authors also employ a rather
simplistic task involving only very general data. Under this condition, the belief-adjustment model predicts
primacy effects for the EoS response mode and recency effects for the SbS response mode. As the experi-
mental design involves a “hybrid” mode, no clear predictions can be derived and interpretations made.
441
See Rose/Rose (2003): 314, 319, 329-332.
442
As might be recalled from Section 2.4.2.3.1, auditors’ experience has been typically used as a surrogate for
knowledge and expertise.
77
general, experienced auditors are theorized to possess richer and better organized knowledge
structures than less experienced auditors which allow them to recognize patterns in the evalu-
ated evidence and identify (more) warning factors and circumstances in the course of the au-
dit. This is expected to lead to a more precise problem assessment (i.e., initial belief), less
biased weighting of new evidence, and ultimately to higher judgment quality.443 While it is
undisputed that knowledge and expertise are critical for the performance of high quality fi-
nancial statement audits, whether they can temper the pervasive, robust, and subconsciously
operating recency effect is an empirical question.
Overall, the auditing literature does not provide a uniform picture regarding the debiasing
influence of experience in the context of auditors’ belief revisions. In particular, while Kenne-
dy (1993), Messier and Tubbs (1994), and Trotman and Wright (1996) find recency effects
only in the (unassisted) judgments of inexperienced subjects (students or audit novices), but
not experienced auditors, the results obtained by Arnold et al. (2000), Guiral and Esteo
(2006), Krull, Reckers and Wong-On-Wing (1993), and Morton (2001) suggest that experi-
ence does not mitigate order effects. That is, the latter studies indicate that experts and non-
experts are likewise prone to recency.444 However, these conflicting results may be an exper-
imental artefact attributed to the wide range of different experience measures employed across
the different studies (e.g., general experience, task-specific experience, professional exams,
and professional role (auditor vs. student)).445
Beyond experience, auditing research has explored a number of other factors and mechanisms
theorized to diminish recency effects in belief updating. This branch of research has evolved
around the notion set forth by Kennedy (1993) that information order effects are effort-related
(rather than data-related) biases and thus can be diminished by inducing subjects’ cognitive
effort.446 In an empirical investigation of this proposition, Kennedy (1993) utilizes accounta-
bility (announcement of a review of the audit work performed) as an effort-inducing mecha-
nism and reports that it effectively rules out recency effects in the judgments of inexperienced
students.447 Further effort- and attention-enhancing mechanisms found to mitigate recency
include documentation requirements (e.g., Cushing and Ahlawat (1996)) and self-reviews
(e.g., Ashton and Kennedy (2002)).448
With regard to the influence of teamwork, the empirical evidence is inconsistent. While
Ahlawat (1999) reports results indicating the superiority of multi-auditor judgment quality
(lack of recency bias), cognitive capacity, memory, and confidence over individual perfor-
443
See Abou-Seada/Abdel-Kader (2003): 42; Kahle/Pinsker/Pennington (2005): 17.
444
See Arnold et al. (2000): 129; Guiral/Esteo (2006): 614; Kennedy (1993): 243; Messier/Tubbs (1994): 67;
Morton (2001): 115f.; Trotman/Wright (1996): 190f. Interestingly, Krull/Reckers/Wong-On-Wing (1993):
151 even find that experience amplifies recency.
445
See Kahle/Pinsker/Pennington (2005): 17f.
446
Kennedy (1993) conceptualizes effort-related biases as generally resulting from an individual’s natural reluc-
tance to expend additional cognitive effort and attention. Data-related biases, in contrast, have been viewed
as being caused by fallacies in an individual’s memory or knowledge structure and/or faulty informational
input. See Kennedy (1993): 233f.
447
See Kennedy (1993): 243.
448
See Ashton/Kennedy (2002): 226-229; Cushing/Ahlawat (1996): 117f.
78
mance, Johnson (1995) finds evidence of recency effects in auditors’ belief revisions at both
the individual and group level. In addition, he reports that as compared to individual belief
revision, group-generated responses are characterized through a greater riskiness in choice.
Finally, Reckers and Schultz (1993) demonstrate that despite the superiority of multi-auditor
judgment over individual judgment in terms of adherence to guidance presented in auditing
standards, audit groups and individuals are equally susceptible to recency.449
Further important insights concerning the alleviation of recency effects are presented by
Favere-Marchesi (2006). He demonstrates that awareness of the temporal (chronological)
order of evidence mitigates auditors’ susceptibility to effects generated by the presentation
order of evidence in that it allows auditors to recognize trends, i.e., patterns of change and
stability over time, which are particularly relevant to judgment and decision making in audit-
ing.450
Another central but quite controversial topic in the relevant auditing literature concerns audi-
tors’ sensitivity to evidence. A number of studies, including Ashton and Ashton (1988, 1990),
Butt and Campbell (1989), Cushing and Ahlawat (1996), Favere-Marchesi (2006), Monroe
and Ng (2000), and Reckers and Schultz (1993), find that auditors are more sensitive and re-
sponsive to negative evidence than positive data as indicated by the larger belief adjustment
following the processing of negative cues. The tenor of these studies is that the adoption of a
conservative evidence evaluation strategy corresponds to the professional environment which
exposes auditors to considerable litigation and reputation risks in the case of erroneous audit
judgments and opinions.451 Bamber, Ramsay and Tubbs (1997), on the other hand, find that
auditors exhibit confirmation proneness452 in evidence evaluation in that they respond more
sensitively to information that corroborates their initial hypothesis regardless of its directional
frame (positive (e.g., fair presentation) vs. negative (e.g., misstatement)). Importantly, this
confirmation tendency is found to emerge irrespective of experience level or the presence of
potential irregularities in the task scenario.453 Evidence on auditors’ confirmation proneness in
information processing and evidence evaluation has also been provided by Ayers and Kaplan
(1993), Church (1991), Morton (2001), Pei, Reed and Koch (1992), and Waller and Felix
449
See Ahlawat (1999): 84f.; Johnson (1995): 152; Reckers/Schultz (1993): 139f.
450
See Favere-Marchesi (2006): 70.
451
See, e.g., Ashton/Ashton (1988): 639f.; Ashton/Ashton (1990): 15f.; Butt/Campbell (1989): 478f.;
Cushing/Ahlawat (1996): 118; Favere-Marchesi (2006): 82; Monroe/Ng (2000): 164; Reckers/Schultz
(1993): 138. For such claims, see also Smith/Kida (1991): 484f.
452
In general terms, confirmation proneness (also referred to as “confirmation bias”) relates to the phenomenon
of information being searched for, evaluated, and memorized in a manner involving a systematic and uncon-
ditional corroboration of the own hypothesis or position. Confirmation proneness has been figuratively clas-
sified as an immunity-stimulating means against rejection of a hypothesis. See Oswald/Grosjean (2004): 79.
For a general (psychological) discussion of the confirmation phenomenon, refer to Oswald/Grosjean (2004):
79-94. For a comprehensive meta-analysis of confirmatory tendencies in the context of auditing, refer to
Church (1990): 81-104.
453
See Bamber/Ramsay/Tubbs (1997): 263. Overall, the authors provide substantial support for the descriptive
validity of the belief-adjustment model.
79
(1984).454 In addition, Krull, Reckers and Wong-On-Wing (1993) argue that in reality, materi-
al adjustments to audited accounts are relatively scarce, suggesting that auditors maintain a
positive rather than negative attitude toward the client’s financial statements and audit evi-
dence.455 In contrast to the aforementioned studies, Guiral and Esteo (2006) find that auditors
are moderately sensitive to both positive and negative evidence. 456 As argued by Krull, Reck-
ers and Wong-on-Wing (1993), a neutral (balanced) stance toward evidence might be indeed
the most adequate attitude for an auditor because an asymmetry toward negative evidence can
lead to overauditing, considerable inefficiencies, and potentially to client loss. An asymmetry
toward positive evidence, on the other hand, bears the risk of material misstatements and oth-
er problems going undetected which may lead to considerable reputational damages and legal
lawsuits.457
In summary, despite some inconsistency,458 the existing belief-revision literature in auditing
provides substantial support for the recency-effect predictions of the belief-adjustment model
for the sequential presentation and processing of a short series of complex, inconsistent evi-
dence.459 This overall result has crucial implications for audit practice, as it suggests that
when an audit task naturally involves a step-by-step information processing or considerable
complexity and cognitive load, the emergence of recency effects could be germane. 460 As in-
dicated previously, firm’s management normally controls and determines the order in which
information is presented to the auditor. This fact, along with the recommendation for man-
agement accountants provided in the academic literature to structure and present data so that
the best and strongest piece of information is provided last in order to achieve maximum im-
pact,461 highlights the necessity for auditors to bear in mind the potential effects of the order
of information presentation on their judgments and decisions.462
454
See Ayers/Kaplan (1993): 126; Church (1991): 531f.; Morton (2001): 115f.; Pei/Reed/Koch (1992): 180;
Waller/Felix (1984): 399. For a list of relevant psychological studies corroborating the general human ten-
dency of confirmation proneness, see Smith/Kida (1991): 483.
455
See Krull/Reckers/Wong-On-Wing (1993): 145.
456
See Guiral/Esteo (2006): 613.
457
See Krull/Reckers/Wong-On-Wing (1993): 147.
458
As recognized by Trotman/Wright (2000): 170f., due to the very diverse and incomparable design choices in
the auditing studies on information order effects, it is not possible to determine whether this inconsistency in
results is caused by genuine effects and moderating variables or merely represents an experimental artefact.
459
Indeed, as the analysis of 25 belief-revision studies conducted by Kahle/Pinsker/Pennington (2005) reveals,
in the absence of debiasing factors, recency has been found in 21 studies for at least some experimental con-
dition. See Kahle/Pinsker/Pennington (2005): 34. In this context, Ashton/Kennedy (2002): 221, 223 classify
recency as a “pervasive” and “robust” judgment bias.
460
See Kahle/Pinsker/Pennington (2005): 28.
461
See Dillard/Kauffman/Spires (1991): 632. It should be noted that this recommendation was made generally
and probably was not intended to encourage accounting managers to “manipulate” auditors’ judgments and
decisions. Nonetheless, managers may be well aware of the impact of information order effects and use them
in order to achieve a maximally favorable treatment.
462
Zhao/Harding (2013) argue and demonstrate that order effects can also be used in a positive way to enhance
the quality of auditors’ judgments. Specifically, they show that when evidence suggesting troublesome
changes in the client’s business environment is considered subsequent to (rather than preceding, as usual in
audit practice) positive evidence obtained from information system and substantive audit testing procedures,
auditors provide reasonably higher assessments of the risk of material misstatement. See Zhao/Harding
(2013): 503.
80
As a final note on this section, it should be noted that prior auditing research on belief updat-
ing has largely focused on task variables, contextual factors, and the role of knowledge,
whereas personal dispositions463 have remained relatively unexplored.464 The only study ex-
ploring the influence of a personality trait (cognitive style) on auditors’ belief revisions is
conducted by Chan (1995). He finds that information order interacts with the individual cog-
nitive style to influence auditors’ belief adjustments. In particular, the results reported by
Chan (1995) indicate that auditors classified as field dependent exhibit greater recency bias in
their responses compared to their field-independent counterparts.465 The preliminary findings
presented by Chan (1995) hint at the relevance of traits in the context of auditors’ belief revi-
sions. In this context, Kahle, Pinsker and Pennington (2005) emphasize the need for future
research on the influence of personal factors on auditors’ information processing and belief
adjustment.466 The present study answers this call for research by exploring the effect of trait
professional skepticism on auditors’ belief revisions.
The following section focuses on the fundamental concept of professional skepticism in audit-
ing. It provides a broad theoretical and empirical consideration of the notion of skepticism as
a prelude to the discussion of the role of trait professional skepticism in context of auditors’
belief revisions.
463
As indicated in Section 2.4.2.3, personal dispositions or traits generally relate to relatively stable and endur-
ing aspects of an individual. Note that the terms “(personality) traits”, “personal dispositions”, “dispositional
characteristics”, and “innate characteristics” are used interchangeably in the present work.
464
See Kahle/Pinsker/Pennington (2005): 12.
465
See Chan (1995): 26f. However, it should be noted that the sample size used by Chan (1995) consists of
twenty auditors only, which is way too small to obtain statistical significance and reliable inferences. In ad-
dition, he does not refine upon the rationale for the hypothesis concerning the nature of the link between
field dependence-independence and information order effects. As might be recalled from Section 2.4.2.3.6.1,
field-dependent individuals generally tend to process information in a global and unstructured manner, while
field-independent persons exhibit more analytical and structured information processing. By imposing struc-
ture to the task, complexity is expected to decrease and thus, biasing cognitive tendencies may decline as
well.
466
See Kahle/Pinsker/Pennington (2005): 12, 22.
81
4 The Concept of Professional Skepticism
467
Although genuinely a concept rooted in the field of (external) auditing, professional skepticism, or the abil-
ity to challenge claims, motives, arguments, and to critically evaluate evidence and information, is also a
crucial attribute to the other participants involved in the financial reporting process, including management,
oversight boards, audit committees, and internal auditors. See CAQ (2010): 19. In order to keep the scope of
the present section within reasonable bounds, the following considerations are limited to the field of external
auditing.
468
See Hurtt (2010): 149; Murray (2012): 36. Given the dynamics of recent developments in academic research
on professional skepticism initiated with the publication of the seminal papers by Nelson (2009) and Hurtt
(2010), it appears reasonable and appropriate to include unpublished research papers dealing with some rel-
evant aspects of skepticism in order to provide a more complete picture of the contemporary research on the
concept of professional skepticism.
469
See Doucet/Doucet (1996): 158.
470
See Nolder (2012): 3.
471
See, e.g., Rose (2007): 216. In philosophy, in contrast, “concept” refers to the meaning of words or expres-
sions. See Craig (2005): 135; Dennis (2013): 4.
472
The approach of decomposing the compound term “professional skepticism” into its components as well as
the lines of argumentation presented here follow the procedure applied and claims made by Dennis (2013):
27-30 in his conceptual inquiry into the nature of professional judgment in auditing.
82
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2_4, © Springer Fachmedien Wiesbaden 2015
among other things, acting with integrity473, objectivity474, competence, due care, confidenti-
ality, and in compliance with the relevant standards and regulations (IESBA 100.5). Further-
more, a broad array of intellectual, technical, personal, interpersonal, and organizational skills
is required for an individual to be qualified as a professional (IES 3.1-3). In addition, accord-
ing to the view of the Canadian Institute of Chartered Accountants (CICA), skepticism and
inquisitiveness are among the most important characteristics of a good professional auditor.475
In this sense, skepticism is an integral part of professionalism in auditing. A further important
aspect of professionalism in auditing concerns the commitment of the profession to serve the
public interest (IESBA 100.1). Indeed, placing public service ahead of self-interest and re-
ward is a distinguishing feature of professionalism.476 In a recent conceptual paper on the
fundamental issues of duty and ethics in auditing, Shaub and Braun (2014) synthesize the
following major attributes of professionalism: (1) possession of knowledge, intellectual skill,
and proficiency; (2) adherence to generally accepted values set forth in a code of conduct; and
(3) duty to act in the public interest.477
It should be noticed that applying professional skepticism is not expounded as the application
of skepticism by a professional. Consider in this context the very broad definition of a profes-
sional accountant provided by the IAASB (2013) as “[a]n individual who is a member of an
IFAC member body”478. This description is not very insightful in comprehending what it takes
for skepticism to be qualified as “professional”. Clearly, the compound “professional skepti-
cism” is not intended to mean skepticism exercised by a member of a professional accounting
body.479 Rather, being professional in exercising skepticism is a matter of complying with
high ethical standards, possessing a broad array of skills and qualities, and placing the public
interest above self-interest or the client’s interest.
Building on this terminological foundation, Section 4.2 provides an array of general and au-
dit-related definitions and considerations of (professional) skepticism.480 In Section 4.3, the
antecedents of professional skepticism as synthesized in the unifying model developed by
Nelson (2009) are discussed and underpinned with relevant research findings. Section 4.4
sheds light on the notion of skepticism as a trait and on the psychometric scale developed by
473
Cook/Winkle (1988): 77 define integrity as “moral soundness, especially when revealed in situations testing
one’s steadfastness to truth, purpose, responsibility, and trust.” Buckley (1978): 65 argues that integrity has
no value per se in the context of professionalism, as “honest idiots have no hope of becoming professionals”.
Rather, integrity fulfils the function to add value to the services provided by a competent and skilled profes-
sional auditor.
474
Cook/Winkle (1988): 77 define objectivity as “a state of mind enabling the professional to be impartial,
intellectually honest, and free from conflicts of interests.”
475
See CICA (1995): 35, quoted in Dennis (2013): 29.
476
See Carey (1956): 50; Pound (1953): 5.
477
See Shaub/Braun (2014): 4. For a differently nuanced discussion of the meaning and implications of profes-
sionalism in auditing, see Quadackers (2009): 14.
478
IAASB (2013): 29.
479
For such claims and a similar argumentation in the context of professional judgment, see Dennis (2013): 30.
480
Given the complexity and multidimensionality of the professional skepticism construct, the adoption of a
broad and discursive approach in studying skepticism is inevitable. See FRC (2012): 1.
83
Hurtt (2010) which are of central importance for the present study.481 In Section 4.5, the role
of trait professional skepticism in the context of auditors’ belief revisions is discussed.
This section deals with some definitional issues and the various perspectives of (professional)
skepticism. The section is organized around three basic lines of consideration: skepticism in a
general (humanitarian) sense, professional skepticism in a normative sense, and professional
skepticism as discussed in the academic auditing literature.
There is a great variety of skepticism notions, definitions, views, and arguments in the hu-
manities, especially in philosophy and consumer research. Subsequently, some central aspects
are highlighted. The following overview does not intend to provide an exhaustive survey of
all relevant varieties, aspects, and definitions of skepticism. Rather, it aims to embed skepti-
cism in a broader context, illuminate some important aspects regarding its nature, and high-
light the versatility of skepticism, thereby paving the way for the subsequent consideration of
professional skepticism in auditing.
The term “skepticism” stems from the Greek word “skeptikos” which means “to consider”,
“to examine”, “to observe carefully”. It is closely related to the term “skepsis” which means
“inquiry” and “doubt”.482 In ordinary usage, skepticism generally refers to an attitude of
doubt.483 It essentially involves questioning truth and knowledge claims, challenging asser-
tions and beliefs as well as striving for precision and clarity in definition, consistency in logic,
and sufficient reasonableness of arguments and evidence. Despite its controversial ac-
ceptance, skepticism plays a key role in philosophy, religion, politics, ethics, science, and
consumer research.484 Subsequently, some of the most important skepticism areas are briefly
outlined.
481
It should be noted that the mutual consideration and combination of different models and notions of profes-
sional skepticism is vital, because contemporary auditing research on skepticism, although progressively de-
veloping and quite insightful, is still at a stage that does not allow for a single model of professional skepti-
cism in auditing to be depicted. See Doucet/Doucet (1996): 159. In the present study, the models and notions
of Nelson (2009) and Hurtt (2010) are combined to provide the theoretical framework for the consideration
of the concept professional skepticism in auditing.
482
See Kurtz (1992): 21; Stough (1969): 3.
483
See http://www.merriam-webster.com/dictionary/skepticism. The term “skepticism” is generally used syn-
onymously with “distrust”, “dubiety”, “suspicion”, and “query”. Terms closely related to “skepticism” in-
clude “disbelief”, “incredulity”, and “concern”. The antonyms of “skepticism” encompass “credence”,
“faith”, “trust”, “belief”, “assurance”, “conviction”, and “certitude”. See http://www.merriam-webster.com/
thesaurus/skepticism.
484
See Kurtz (2001): 39.
84
Philosophical Skepticism485
The origins of (philosophical) skepticism can be traced back to ancient Greece.486 Ancient
skepticism evolved into two schools of thought: Pyrrhonism, a stream of thought inspired by
Pyrrho of Elis (360-270 BCE), and academic skepticism which emerged in Plato’s Academy,
with Carneades (213-128 BCE) as its most influential representative. Skeptical thought fur-
ther flourished in Rome, with Cicero (106-43 BCE) and Sextus Empiricus487 as its most nota-
ble adherents.488 Overall, ancient skepticism involves an existential critical489 questioning of
the possibility of knowledge490. Hence, skepticism represents the antithesis of dogmatism (the
belief that absolute truth exists and knowledge of the truth is possible).491
In the ancient civilization, skepticism was considered a valuable virtue 492 and recommended
as an attitude in life. It was argued that a life lived in skeptical doubt is superior to a life lived
in dogmatic certainty, as the former involves living with a calm, peaceful state of mind, free
from particular assumptions, beliefs, and judgments. The ancient Greeks referred to this state
of mental tranquility as “ataraxia”.493 However, several centuries later, skepticism became
subject of severe criticism, most notably by Saint Augustin (354-430), and during the Dark
Ages, an era dominated by unconditional faith, skepticism virtually eclipsed. During the Re-
naissance, it reemerged as an essential part of human thought and a vital ingredient of the sci-
entific approach. Skepticism played a central role in the writings of a number of famous
485
Broadly speaking, philosophy deals with “questions about the meaning, truth, and logical connections of
fundamental ideas that resist solution by the empirical sciences”, Woodhouse (2011): 3. Skepticism is an es-
sential part of philosophical thought as raising doubts and maintaining a critical position inevitably accom-
panies philosophical questioning. See Lom (2001): 31. The philosophical considerations provided here focus
on epistemological skepticism, i.e., skepticism concerned with central categories such as knowledge, belief,
and justification. The counterpart of epistemological skepticism is conceptual skepticism which principally
deals with issues of meaning and terminology. See Stough (1969): 1; Williams (2008): 82. The consideration
of the latter is beyond the scope of this section. In addition, it should be noted that the philosophical litera-
ture on skepticism is immense and equivocal with regard to the nature, scope, and different forms of philo-
sophical skepticism. As the provision of an in-depth, exhaustive philosophical analysis of skepticism is be-
yond the scope of this work, the following considerations are necessarily rudimentary and selective in
nature. They merely intend to introduce some basic philosophical accounts with regard to skepticism as a
prelude to the discussion of professional skepticism in auditing.
486
See Popkin (1979): xiii.
487
Sextus Empiricus was a prominent defender of Greek Skepticism who lived in the last half of the second and
the first quarter of the third century AD. Although there is no historical certainty about his origins and affili-
ation, it is believed that he was a Greek, a physician, and possibly a medical doctor. See Hallie (1985): 27.
Sextus understood skepticism as “an ability to place in antithesis, in any manner whatever, appearances and
judgments, and thus – because of the equality of force in the objects and arguments opposed – to come first
of all to a suspension of judgment and then to mental tranquillity”, quoted in Hallie (1985): 32f.
488
See Kurtz (1992): 31.
489
Note that in ordinary usage, the term “critical” is frequently negatively loaded. As argued by Woodhouse
(2011): 48, there is some negativity in the sense that being critical involves questioning, challenging claims
and assumptions, suspending beliefs, and exerting cognitive load. However, in essence, critical thought is
considered a positive and constructive enterprise because it is the key to problem solving, open-mindedness,
and progress. This contention can also be conveyed to skepticism.
490
In a philosophical sense, knowledge can generally be defined as “justified true belief”, Johnson (1978): 12.
491
See Popkin (1979): xxi
492
For a definition and discussion of the term “virtue”, see Morales-Sánchez/Cabello-Medina (2013): 720f.
493
See Burnyeat (1983): 121; Greco (2008): 4; Meidan (2004): 10.
85
thinkers, including Montaigne, Bayle, Descartes, Hume, Wittgenstein, Kant, Nietzsche, and
Freud.494
The brief outline of the central aspects of philosophical skepticism is rounded off with a clas-
sification of the main forms of philosophical skepticism. This classification follows Kurtz
(1992, 2001) and distinguishes between four basic categories of skepticism: total negative
skepticism, neutral skepticism, mitigated skepticism, and contextual skepticism.
Total negative skepticism is the most extreme form of skepticism. It involves a total denial of
all claims to knowledge, truth, and value.495 This kind of skepticism involves the notion that
there is no certainty and no reasonable and satisfactory basis for beliefs and assertions. Hence,
all that people encounter are subjective impressions and perceptions, and there is no reliable
evidence or proof that these sensations correspond to the external world496 or reality.497 Total
negative skepticism can be equated with nihilism or negative dogmatism.498
Neutral skepticism is a less dogmatic form of skepticism which involves a position in which
one neither accepts nor rejects anything. This state of mind results in an entire suspension of
judgment and belief (epoché).499
Mitigated skepticism builds on the insight that even though knowledge is ephemeral and ulti-
mate truths about reality and the external world are utopic, the development of beliefs is a
central demand of life. Hence, even if beliefs merely rest upon probabilities, experience, or
other cognitive constructs rather than certainties and ultimate truths, one needs to update be-
liefs and develop knowledge as a pragmatic requirement of life.500
Contextual skepticism, the last and most contemporary major form of skepticism according to
Kurtz, can be distinguished from the other forms outlined above through its positive and con-
structive nature. Contextual skepticism does not involve a universal critical position to all
494
See Kurtz (1992): 31f. For a detailed overview of the historical origins and development of skepticism dur-
ing the centuries, refer to Hallie (1985): 9-28, Kurtz (1992): 32-74, and Lagerlund (2010). For a discussion
of modern (as opposed to ancient) skepticism and the influential work of Descartes (1596-1650), Locke
(1632-1704), Berkeley (1685-1753), Hume (1711-1776), and Kant (1724-1804), consult Popkin/Stroll
(2002): 59-101. For a discussion of contemporary (twentieth century) skepticism and the seminal work of
Russel (1872-1970), Moore (1873-1958), Wittgenstein (1889-1951), and Derrida (1930-2004), refer to
Popkin/Stroll (2002): 103-143.
495
This dogmatic form of skepticism is based on the famous Socrates’ phrase: “I know that I know nothing.”
See Popkin/Stroll (2002): 54.
496
This term refers to the world that exists outside of an individual’s own mind. See Huemer (2001): 7.
497
See Kurtz (1992): 23.
498
See Kurtz (2001): 40f. The unconditional rejection of the possibility of knowledge is the credo of academic
skepticism. Notice that this form of skepticism is claimed to be self-contradictory, as arguing that no
knowledge, certainty, and truth are possible, academic skeptics have already made a claim. See Johnson
(1978): 22f. However, there are also many good reasons to embrace and study the notion of total skepticism.
For an overview, see Johnson (1978): 22f.
499
See Kurtz (2001): 41. This neutral state of suspension of judgment and belief reflects the Pyrrhonian notion
of skepticism.
500
See Kurtz (2001): 42. The most prominent proponent of mitigated skepticism is the famous Scottish philos-
opher David Hume (1711-1776). Note, however, that the notion of mitigated skepticism arose twenty centu-
ries earlier, in the second century BCE, and was defended by the Greek Philosopher Carneades. See Kurtz
(2001): 42.
86
possible issues. Rather, it is restricted to the context under consideration.501 It takes into ac-
count that some assertions are unjustified, improbable, and incorrect. Consequently, contextu-
al skepticism involves reflective doubt which results in an increased demand of evidence and
reasonable argumentation for the evaluation of hypotheses and claims. This form of skepti-
cism advances the quest of knowledge and constructive informed inquiry. In contrast to neu-
tral skepticism, contextual skepticism builds upon the notion that human knowledge is possi-
ble, and that it can even be reliable. Furthermore, it differs from mitigated skepticism in that it
does not highlight total uncertainty. Rather, it focuses on the impressive ability of the human
mind and cognition to think, infer, and comprehend.502
With regard to auditing, the last skepticism form appears most relevant and applicable as it
involves reflective situational doubt and critical informed inquiry. In contrast, total negative
skepticism, which involves universal denial, disbelief, indecision, and diffidence,503 seems
incompatible with the objective of auditing to provide an overall opinion (i.e., judgment) re-
garding the compliance of the client’s financial statements with the applicable norms and
standards.504
Religious Skepticism
As indicated above, there is a variety of skepticism prospects. Over the ages, skeptics have
questioned the possibility of knowledge, the justification of beliefs, the reality of the external
world, human mind, the past, the future, induction, causation, etc.505 Considering this wide
array of abstract and material objects of doubt, it is not surprising that the existence of God
and religious506 claims have also been disavowed. This issue is covered by the genre of reli-
gious skepticism, which has a long history in the skeptical tradition.507
Religious skepticism is concerned with doubts about theistic religious commitment, resulting
in skepticism about God’s reality.508 Religious skeptics typically deny that any religious claim
is true and argue that there is no sufficient and adequate evidence supporting the belief in
501
See Kurtz (2001): 44.
502
See Kurtz (2001): 44f.
503
See Naess (1968): 53. He lists many other negatively-loaded characteristics attached to skeptics beyond the
ones mentioned above. However, he claims that this criticism on classical skepticism is groundless and mis-
concepted, at least with regard to the writings of Sextus Empiricus.
504
The fundamental overall objectives of the auditor, i.e., obtaining reasonable assurance about the accuracy
(freedom from material misstatement) of the financial reporting statements and their conformity to the appli-
cable financial reporting rules as well as the reporting on the financial statements in accordance with the re-
sults of the conducted audit and auditor’s findings, are stated in ISA 200.11.
505
See Moser (2008): 200.
506
According to Kurtz (1992): 193, “religion” generally refers to “those systems of faith that venerate an occult,
supernatural, or divine being”. However, there is a substantial disagreement regarding the meaning of the
term “religion” and a number of alternative (broader) definitions.
507
For instance, skeptical thoughts on God and his existence can be found in the ancient writings of Sextus
Empiricus. See Hallie (1985): 175-215.
508
See Moser (2008): 202. He acknowledges that each person attaches a different understanding to the term
“God”, which makes a generally accepted definition, or a common notion, of God utopic. Nonetheless, the
term “God” is usually used to refer to an “authoritatively and morally perfect creator who is worthy of wor-
ship, including full adoration, love, and trust.” Moser (2008): 202.
87
God’s reality. Under these circumstances, the cognitively appropriate approach is to raise
doubts, i.e., suspend judgment that God or religious reality actually exist.509
Political Skepticism
Political skepticism includes raising doubts about the virtue and moral principles of politi-
cians and the government. It mirrors a general distrust of those who rule and exercise pow-
er.510
Moral Skepticism
Moral skepticism questions the meaning and justifiability of major ethical categories, such as
“right” and “wrong”, “good” and “bad”, “justice” and “unjustice”. It posits that no ethical
standards, values, and claims are known to be true and correct.511
Scientific Skepticism
Skepticism is vital in science,512 as it implies that theories, postulates, arguments, interpreta-
tions, empirical evidence, etc. are not merely taken for granted but are subjected to critical
thought and investigation. The tension between skepticism and belief is an essential ingredi-
ent of scientific advance, as researchers regularly have to justify their findings, positions, and
beliefs against the doubts and concerns of the peers.513 Scientific skepticism can broadly be
described as an ongoing, systematic, and critical informed questioning. In this sense, it exhib-
its many parallels to auditing.514
Advertising Skepticism
Skepticism is also a central topic in the field of consumer research where the attitude with
which a person faces marketers’ motives and advertising claims and promises is a key deter-
minant of consumer choice and behavior. Similarly to the fields discussed above, advertising
skepticism has been defined in terms of an individual’s tendency to question and disbelief
509
See Moser (2008): 203; Schellenberg (2007): 3. For a more comprehensive philosophical and theistically-
oriented overview of religious skepticism, refer to Moser (2008), Popkin/Stroll (2002): 145-187 and
Schellenberg (2007).
510
See Lom (2001): 2. The most prominent and influential adherent of political skepticism is Nietzsche. For a
detailed overview of his skeptical view on politics and the state, refer to Shaw (2010). For a general philo-
sophical discussion of political skepticism, refer to Popkin/Stroll (2002): 245-289.
511
See Cornman/Lehrer/Pappas (1992): 299; Peterson (2008): 216. For a detailed account and references on
the issue of moral skepticism, refer to Popkin/Stroll (2002): 189-244, Sinnott-Armstrong (2007) and
Superson (2009).
512
See FRC (2012): 6f. Essentially, science is a cognitive achievement of the highest order. It involves a bunch
of intellectual qualities and highly complex cognitive processes such as reasoning, logic, analytical and ab-
stract thinking, problem recognition and solving, hypothesis formulation and testing, among others. See
Feist (2006): ix.
513
See Feist (2006): 225.
514
For such claims and a discussion of the analogies as well as the discrepancies between the scientific method
and the audit, refer to FRC (2012): 6-8.
88
claims.515 In addition, and more importantly, in consumer research, skepticism has been con-
ceptualized as involving both a situational and a dispositional component. Situational skepti-
cism represents a momentary state of disbelief aroused by contextual stimuli, whereas dispo-
sitional skepticism reflects an individual’s enduring and cross-situationally consistent
tendency to doubt and suspicion.516 As will be shown in Section 4.4, this view of skepticism
is also shared by Hurtt (2010) and is central to her research work.
Overall, skepticism is a central notion in the humanities. As will be shown subsequently, its
role in auditing is not less foundational than in the fields considered above. In a nutshell,
skepticism is essentially a state of mind which involves critical questioning and inquiry and
often induces deferral of belief and judgment.517
515
See Forehand/Grier (2003): 350; Koslow (2000): 248; Obermiller/Spangenberg (2000): 312.
516
See Forehand/Grier (2003): 349.
517
See FRC (2012): 3.
518
See IAASB (2012): 1; Peecher/Solomon/Trotman (2013): 597f.
519
See, e.g., Adams (1978): 39; Rezaee/Riley (2010): 130; Soltani (2007): 520; Wahlen/Baginski/Bradshaw
(2010): 41.
520
See King (1991): 108. For an in-depth discussion of the importance of auditing for the capital market, see
Qandil (2014): 50-72.
521
See ASIC (2012): 5; CAQ (2011): 2; CPAB (2012): 3. Note that audit quality is an omnipresent term in the
auditing literature. However, like professional skepticism, audit quality is a very complex and multifaceted
concept which cannot possibly be captured in a single definition, view, or measure. For such claims with re-
spect to audit quality, see Francis (2011a): 127 and Knechel et al. (2013): 385, and with regard to profes-
sional skepticism, see Hurtt et al. (2013): 45 and Toba (2011): 109. For an in-depth terminological, concep-
tual, and empirical review of audit quality and the substantial body of research on this topic, consult Francis
(2011a), Knechel et al. (2013) and Qandil (2014).
522
See FRC (2010): 10; Hurtt et al. (2013): 71; IAASB (2012): 1; Murray (2012): 36; PCAOB (2012): 1;
Shaub/Lawrence (2002): 168. As indicated by Carcello/Hermanson/McGrath (1992): 8, the notion of pro-
fessional skepticism as an important attribute of audit quality is not a new appearance that merely results
from the accounting debacles in the last decades.
89
skeptical, inquisitive, and rigorous, can we have confidence that financial statements speak
the truth.”523
Professional skepticism plays a fundamental role in auditing. This is particularly evident in
light of the fact that auditing is the only profession in which skepticism is explicitly legally
codified.524 Professional skepticism is considered a crucial, not to say the key, audit skill;
“[w]ith it, an auditor is a powerful independent safeguard. Without it, the auditor risks being
reduced to a mere compliance ‘rubber stamper’.”525 Overall, the importance of professional
skepticism in planning and performing an audit cannot possibly be overemphasized.526
Subsequently, the central definitions and regulations concerning professional skepticism are
presented and discussed.527 In consistency with the extant literature on professional skepti-
cism in auditing, the focus of the following considerations is on the two world-wide most
prevalent and influential sets of auditing standards – the International Standards on Auditing
(ISA) and the US auditing standards (SAS).
ISA 200.13(l) defines professional skepticism as “[a]n attitude528 that includes a questioning
mind, being alert to conditions which may indicate possible misstatement due to error or
fraud, and a critical assessment of audit evidence.”529 According to this definition, profes-
sional skepticism is essentially „a behavioral issue“530. It is a state of mind that implies a
questioning stance in evidence evaluation, information processing, judgment and decision
making.531 Professional skepticism involves challenging information and audit evidence ob-
523
Bogle (2000), emphasis added.
524
See Quadackers (2009): 14.
525
Fischer, quoted in Murray (2012): 37.
526
See AUASB (2012): 1; IAASB (2012): 1.
527
It should be noted that in order to keep the discussion of professional skepticism within reasonable bounds,
the subsequent normative consideration is restricted to the currently valid auditing standards and profession-
al guidance as well as recent developments and professional bodies’ releases on that topic. For a historical
overview of the development and transformation of professional skepticism in auditing with an emphasis on
the American Auditing Standards and US professional bodies’ releases and comments, see Toba (2011): 86-
92.
528
The view of skepticism as an attitude is also common in ordinary usage as indicated by the following dic-
tionary entry: http://oxforddictionaries.com/definition/english/scepticism?q=skepticism. In a psychological
sense, an attitude can be defined as “a mental and neural state of readiness, organized through experience,
exerting a directing and dynamic influence upon the individual’s response to all objects and situation with
which it is related”, Allport (1935): 810. It represents “an enduring organization of motivational, emotional,
perceptual, and cognitive processes with respect to some aspect of the individual's world”,
Krech/Crutchfield (1948): 152. Consequently, an attitude is a latent construct that determines an individual’s
beliefs, judgments, and behavior. Following this line of thought, attitudes and traits are very similar con-
structs. They are both viewed as “relatively stable, enduring dispositions”, with attitudes assumed to be
more malleable and versatile than traits. See Ajzen (2005): 6.
529
Note that the German auditing standards (IDW PS) do not provide an explicit definition of professional
skepticism. In the IDW PS, professional skepticism is referred to as “kritische Grundhaltung” (literally:
“critical tenor” or “critical attitude”). Whether this translation involves some subtle differences in meaning
or nuance, is an open question which has not been addressed to date in academic research in Germany.
Deckers/Hermann (2013): 2317 indicate that in contrast to the USA, in Germany and Europe there is still a
lack of sufficiently strong awareness of the fundamental importance and need for further emphasis of the
concept of professional skepticism in auditing.
530
NZAuASB (2013): 4.
531
See IAASB (2012): 3.
90
tained throughout the audit as opposed to performing the audit “in an unthinking box-ticking
way.”532 This general definition of skepticism has been criticized for being rather vague and
unspecific as to what constitutes reasonable degrees of questioning and critical inquiry, and
how exactly a professionally skeptical attitude is demonstrated and documented in different
audit settings.533 The IAASB acknowledges that it is virtually impossible for a construct of
such fundamentality, multidimensionality, and richness to be fully captured by a single defini-
tion. In addition, the board recognizes that professionally skeptical behavior can be demon-
strated in a number of different, not narrowly specified, ways. These caveats notwithstanding,
the definition contained in ISA 200.13(l) constitutes and highlights some central elements
which facilitate the understanding of the meaning of professional skepticism in auditing.534
Professional skepticism is – along with ethical requirements, professional judgment, sufficient
appropriate audit evidence, and performance of the audit in compliance with the relevant au-
diting standards – one of the essential requirements for the conduct of an audit according to
the ISAs.535 ISA 200.15 regulates that “[t]he auditor shall plan and perform an audit with
professional skepticism recognizing that circumstances may exist that cause the financial
statements to be materially misstated.”536 Importantly, despite the lack of explicit reference to
professional skepticism within each single ISA, skepticism is relevant and crucial throughout
the entire audit process.537
With regard to auditors’ positive past experience with a client, ISA 200.A22 sets straight that
a complete disregard of the client’s credibility and integrity experienced in the past cannot be
expected. Nonetheless, a belief in the client’s fidelity and trustworthiness “does not relieve
the auditor of the need to maintain professional skepticism or allow the auditor to be satisfied
with less than persuasive audit evidence when obtaining reasonable assurance.”538 In a simi-
lar vein, the Center for Audit Quality (CAQ) argues that in fulfilling their responsibilities to
serve the public interest, “auditors perform their engagements with a skeptical mindset, and
they cannot hesitate to challenge management’s assertions whenever those assertions run
counter to the audit evidence and the auditor’s own judgment.”539
532
NZAuASB (2013): 4. This view is also shared by Trompeter et al. (2013): 304. The potentially dysfunctional
effects arising from use of checklists (limited thinking, effort, and attention) are emphasized by
Bryant/Murthy/Wheeler (2009): 39f. and Hogan et al. (2008): 246.
533
See Glover/Prawitt (2013): 2.
534
See IAASB (2012): 3.
535
See ISA 200.14-24.
536
The corresponding German regulation (IDW PS 200.17) reads as follows: “Die Abschlussprüfung ist mit
einer kritischen Grundhaltung zu planen und durchzuführen; die erlangten Prüfungsnachweise sind kritisch
zu würdigen. Der Abschlussprüfer muss sich stets darüber im Klaren sein, dass Umstände (Fehler, Täu-
schungen, Vermögensschädigungen oder sonstige Gesetzesverstöße) existieren können, aufgrund derer der
Jahresabschluss und der Lagebericht wesentliche falsche Aussagen enthalten.”
537
See IAASB (2012): 7.
538
The corresponding German regulation (IDW PS 200.17) reads as follows: the auditor “kann (..) nicht ohne
weiteres im Vertrauen auf die Glaubwürdigkeit der gesetzlichen Vertreter bspw. von der Richtigkeit ihrer
Auskünfte ausgehen, sondern muss sich diese belegen lassen und die Überzeugungskraft dieser Nachweise
würdigen.”
539
CAQ (2011): 4.
91
Similarly to ISA 200.13(l), SAS 1 (AU 230.7) defines professional skepticism as “an attitude
that includes a questioning mind and a critical assessment of audit evidence” and highlights
its essentiality for exercising due professional care. In consistency with the ISAs, the US au-
diting standards also require that the audit is planned and conducted with an attitude of pro-
fessional skepticism (SAS 109, AU 314.19) and that the auditor refrains from assuming cli-
ent’s (dis)honesty and strives for nothing less than persuasive evidence (SAS 1, AU 230.9).
Professional skepticism is closely related to the cornerstone concept of independence. 540 The
IESBA Code differentiates two dimensions of independence – independence of mind and in-
dependence in appearance. The former is defined as “[t]he state of mind that permits the ex-
pression of a conclusion without being affected by influences that compromise professional
judgment, thereby allowing an individual to act with integrity and exercise objectivity and
professional skepticism.”541 The second dimension, independence in appearance, involves
“[t]he avoidance of facts and circumstances that are so significant that a reasonable and in-
formed third party would be likely to conclude (…) that a firm’s, or a member of the audit
team’s, integrity, objectivity or professional skepticism has been compromised.”542 In gen-
eral, independence in appearance can be thought of as the behavioral manifestation of the la-
tent independence-of-mind construct.543 Overall, independence is a necessary condition for
professional skepticism as only an impartial, autonomous, and objective auditor can critically
challenge management’s assertions and positions and place the public interest above self-
interest or the client’s interest.
According to Nelson (2009), the view of professional skepticism reflected in the aforemen-
tioned auditing standards can be classified as neutral. That is, the auditor does not take infor-
mation, evidence, management explanations and claims at their face value but also does not
assume biases and misstatements ex ante.544
A slightly different view of skepticism is found in the standards that deal with the issue of
fraud545. As argued by Nelson (2009), while these standards still refer to the neutral view of
professional skepticism codified in ISA 200.15 and SAS 1 (AU 230.9), they put a somewhat
540
See IAASB (2012): 1. According to DeAngelo (1981b): 186, independence is an integral component of audit
quality. Moreover, it has even been labeled “the most fundamental and vital asset possessed by the auditing
profession”, Johnstone/Sutton/Warfield (2001): 2.
541
IESBA Code 290.6(a), emphasis added.
542
IESBA Code 290.6(b), emphasis added.
543
As will be shown in Section 4.4, there is a strong parallel between the notion of independence and the notion
of skepticism not only regarding their fundamental importance in auditing but also with regard to their simi-
larly compound nature comprising both a mindset dimension (“skepticism in mind” or trait professional
skepticism) and a behavioral dimension (“skepticism in appearance” or state professional skepticism mani-
fested in skeptical behavior). In addition, both constructs are multifaceted, with independence comprising
numerous dimensions and attributes such as autonomous state of mind, idealism, freedom, conscientious-
ness, etc. See Farmer/Rittenberg/Trompeter (1987): 5.
544
See Nelson (2009): 2f. The notion of neutrality underlying the auditing standards has also been recognized
by O’Malley (2000): 76.
545
According to ISA 240.11(a), fraud can be defined as “[a]n intentional act by one or more individuals among
management, those charged with governance, employees, or third parties, involving the use of deception to
obtain an unjust or illegal advantage.” For an excellent review and synthesis of fraud-related auditing re-
search, refer to Hogan et al. (2008) and Trompeter et al. (2013).
92
differently nuanced emphasis on it – an emphasis which tendentially points toward a pre-
sumptive doubt perspective of professional skepticism. This view is similar to the one adopted
in forensic auditing where assumptions, statements, and data are generally considered doubt-
ful, unless there is sufficient persuasive evidence to the contrary. 546 In particular, both ISA
240 (A15-16, Appendix 2) and SAS 99 (AU 316.53-54) suggest directing inquiries about the
(potential) existence of fraud to other persons than those charged with governance (e.g., oper-
ating personnel not directly participating in the accounting and financial reporting process, in-
house legal counsels, etc.) in order to obtain valuable information which may not otherwise be
communicated, and to perform some audit procedures and tests on a surprise or unannounced
basis.547
Beyond fraud, professional skepticism is especially important in complex, significant and/or
highly judgmental areas of the audit, such as accounting estimates, going concern assess-
ments, impairment testing, and related party transactions.548 The present study is settled in the
context of a going concern assessment where an attitude of professional skepticism is essen-
tial.549
Bell, Peecher and Solomon (2005) indicate that the notion of professional skepticism in pro-
fessional guidance and societal perception appears to progressively shift from neutrality to-
ward a presumptive doubt perspective.550 This tendency is evident in the increasingly vocal
demands for auditors to raise the bar with regard to the application and maintenance of pro-
fessional skepticism.551
As a final note on this section, it should be noted that the results of audit quality reviews
around the world uniformly report a lack of sufficient professional skepticism in a number of
judgment domains in auditing. Insufficient or lacking skepticism has been identified as one of
the main reasons for audit failures and deficiencies.552 For instance, the SEC enforcement
546
See Bell/Peecher/Solomon (2005): 22; Bunge (1991): 76; Nelson (2009): 3. A more detailed consideration of
the neutrality and presumptive doubt views of professional skepticism is provided in the following section
within the discussion of the different skepticism perspectives found in the relevant auditing literature.
547
This line of argumentation follows Bell/Peecher/Solomon (2005): 22. The crucial importance of professional
skepticism for deterring and detecting fraud in financial statements has also been highlighted by the Center
for Audit Quality which has termed skepticism the inherent “enemy of fraud”, CAQ (2010): 19. A further apt
qualification of professional skepticism in the context of fraud examination is the one expressed by Wells
(2003): 80 who views skepticism as an amelioration of the auditor’s “sixth sense”.
548
See AUASB (2012): 3f.; IAASB (2012): 10f.; PCAOB (2012): 1. Due to the significance of these key audit
issues, they regularly represent an area of particular focus within the audit quality inspections of oversight
bodies. In addition, emphasis is put on complex judgments and audit tasks like revenue recognition, recover-
ability of deferred tax assets, and group audit considerations. See FRC (2013): 17.
549
The going concern assessment has been classified as one of the most critical and important audit judgments
because the assumption of viability and continued existence is crucial to the interpretation of financial
statement information. See Dilla et al. (1991): 114.
550
See Bell/Peecher/Solomon (2005): 66.
551
Consider in this context, for instance, the discussion paper issued by the UK’s Financial Reporting Council
(FRC (2010)) labeled “Professional Skepticism: Raising the Bar”. Note, however, that the requirement to
“raise the bar” and take on a presumptive doubt mindset bears the risk of audit inefficiencies and deteriorat-
ed working relationships with the client. This point is addressed in Section 4.2.3.2.
552
See, e.g., ACRA (2012): 4f., 32; APAK (2011): 3, 13; ASIC (2012): 4, 13; CPAB (2012): 4, 9, 21; FRC
(2013): 5. A prominent instance for the consequences of a lack of professional skepticism is the Enron deba-
93
actions against auditors in the period 1987-1997 show that the lack of an appropriate level of
professional skepticism is among the top three audit deficiencies (present in 61% of the cases
alleged). Poor skepticism was only preceded by deficient audit evidence (80% of the cases)
and lacking due professional care (71% of the cases).553 Similar results are reported in a re-
cent study by Messier, Kozloski and Kochetova-Kozloski (2010) who identified 28 instances
of enforcement actions against engagement quality reviewers conducted by the SEC and
PCAOB since 1993. In 22 out of the 28 identified cases (78.6%), a lack of sufficient profes-
sional skepticism was alleged.554 Overall, the lack of professional skepticism is clearly a sig-
nificant issue in auditing. It represents a common finding and source of considerable concern
in the context of audit quality inspections worldwide.555 These findings suggest that despite its
prominence throughout the auditing standards and the strong emphasis by auditing standard
setting bodies,556 professional skepticism remains a concept that is difficult to fully grasp,
comprehend, and implement.557 In this context, in its Green Paper on audit policy, the Euro-
pean Commission raises the question of whether and how professional skepticism should be
reinforced.558
In light of the foregoing, the conduct of a systematic theoretical and empirical inquiry into the
nature, determinants, and constraints of professional skepticism in auditing is becoming more
critical than ever. The following sections provide an overview of the state of the art of aca-
demic research on professional skepticism in auditing. The overview begins with a discussion
of the central definitions and views of professional skepticism provided in the relevant audit-
ing literature (Section 4.2.3). Building on that, the conceptual framework of determinants and
constraints of professional skepticism in auditing developed by Nelson (2009) (Section 4.3) as
well as the insightful notion of the nature of professional skepticism offered by Hurtt (2010)
(Section 4.4) are presented and discussed.
cle. In this case, strong indications were evident suggesting that the auditors in charge from Arthur Andersen
failed to exercise the necessary professional skepticism and to adopt a critical attitude to management’s as-
sertions and presentations. Instead of acting skeptically, they appear to have taken at face value manage-
ment’s statements regarding the related-party transactions and complex valuations. In the Enron case, famil-
iarity and fee dependence seem to have played a catalyzing role. See Benston/Hartgraves (2002): 126f. For a
discussion of the adverse effects of familiarity and fee dependence on auditors’ professional skepticism, see
Section 4.3.2.3.2.
553
See Beasley/Carcello/Hermanson (2001): 65. As demonstrated by Toba (2011): 86-92, the identification of
insufficient application of professional skepticism throughout the audit is not a recent phenomenon. In fact,
professional skepticism deficiencies consistently surround auditing from the emergence of the concept till
today.
554
See Messier/Kozloski/Kochetova-Kozloski (2010): 234, 243-245. Interestingly, only eight out of the 28 iden-
tified cases (28.6%) involved Big 5 audit firms.
555
See IFIAR (2013): 9.
556
See, e.g., IAASB (2012): 1; PCAOB (2012): 7.
557
Toba (2011) attributes the difficulties in the implementation and proper application of professional skepti-
cism in auditing to (1) a reluctance to increase professional responsibility; (2) a growing complexity in the
audit environment; and (3) a positivistic evidential approach inherent to auditing. For a discussion of the in-
dividual categories, see Toba (2011): 92-95.
558
See European Commission (2010a): 9.
94
4.2.3 Professional Skepticism as in the Auditing Literature
Although the academic auditing literature has dedicated substantial effort and attention to
exploring and specifying professional skepticism, a succinct and generally accepted definition
and view of skepticism are still lacking.559 Indeed, different nuances and emphases have been
attached to the term “professional skepticism” in the auditing literature, resulting in a hetero-
geneous use and understanding of what professional skepticism essentially means. At the one
end of the spectrum are definitions and descriptions that view professional skepticism as a
neutral construct; at the other end of the spectrum, skepticism is conceptualized in terms of
presumptive doubt.560 Subsequently, the two predominant views of professional skepticism in
the auditing literature are discussed. At the end of this section, other noteworthy definitions
and views of professional skepticism found in the relevant auditing literature are briefly out-
lined.
As touched upon in the preceding section, the neutrality perspective of skepticism is argued to
underlie the majority of current auditing standards.561 In essence, this perspective implies that
the auditor neither blindly trusts nor distrusts the entity’s management. Under this notion,
audit evidence needs to be obtained by the auditor in order to verify and corroborate manage-
ment’s representations and assertions but also to ensure that opposing explanations can be
eliminated. The neutrality perspective can be generalized as “trust but verify”.562 Subsequent-
ly, the main literature sources employing a neutrality approach to professional skepticism are
presented and discussed.
Above all, Hurtt (2010) embraces the neutrality notion of professional skepticism. She defines
skepticism as “a multi-dimensional construct that characterizes the propensity of an individu-
al to defer concluding until the evidence provides sufficient support for one alterna-
tive/explanation over others.”563 Her definition is neutral in terms of balancing out confirma-
tion and distrust proneness with regard to management’s assertions and claims. Instead of a
priori assuming bias or unbiasedness, this position involves concentrating on the sufficiency
and persuasiveness of audit evidence.564 Hurtt’s (2010) neutrality definition and view of pro-
fessional skepticism has been adopted by Hurtt, Eining and Plumlee (2010).
559
See Doucet/Doucet (1996): 159; Nelson (2009): 4; Toba (2011): 86
560
See Bell/Peecher/Solomon (2005): 21f.; Nelson (2009): 2-4; Quadackers/Groot/Wright (2012): 3.
561
See Nelson (2009): 2f.
562
See Grumet (2003): 9; Quadackers/Groot/Wright (2012): 6; Reidy/Theobald (2011): 5; Toba (2011): 84.
Note that “trust but verify” is a phrase originally stemming from a Russian national saying which the US
President Ronald Reagen adopted and frequently used in the context of the nuclear arms negotiations be-
tween the USA and the Soviet Union during the Cold War period. See Grumet (2003): 9.
563
Hurtt (2010): 151.
564
See Hurtt/Eining/Plumlee (2010): 5f.
95
A review of older auditing literature sources reveals that despite its recently gained promi-
nence, the idea of professional skepticism as a neutral concept is not really new. Specifically,
Chazen and Solomon (1975) argue that the auditor is neither an advocate nor an antagonist
but rather a fair-minded umpire. They elaborate on this stating that auditors must internalize
“that they are not mechanical voucherers or tick-markers (…). They must keep their eyes and
ears open with a bright, shiny sense of imaginative inquisitiveness and healthy skepticism.”565
Similarly, Thomas, Ward and Henke (1991) describe professional skepticism as “an attitude
of an impartial referee”566. In a similar vein, Boynton and Kell (1996) posit that “auditors
should neither disbelief management’s assertions nor glibly accept them without concern for
their truthfulness.”567 Shaub and Lawrence (2002) argue along similar lines, vividly stating
that “the auditor is required to find the balance between paranoia and naïveté, a difficult area
of professional judgment.”568
The neutral perspective has also been adopted by Cushing (2000) who equates skepticism
with a lack of bias – either in a positive or negative direction – in auditors’ beliefs. Following
this line of thought, being skeptical involves the avoidance of trusting or suspicious attitudes
and beliefs.569
In addition, defining professional skepticism in terms of informed objectivity is also con-
sistent with the neutrality perspective. This view has been embraced by Copeland (1996),
Fornelli and Desmond (2011), and Heninger (2001).570
Empirical support for the neutrality view is provided by Rasso (2013) who finds that auditors
adopting a deliberative mindset exhibit more skeptical judgments and actions than subjects
operating under an implemental mindset. A deliberative mindset involves impartiality, objec-
tivity, open-mindedness (i.e., lack of commitment to an alternative), and depth in evidence
processing and evaluation. Under this mindset, individuals consider broad, multifaceted inter-
pretations of evidence and diverse alternatives in order to understand the “why” of a problem.
An implemental mindset, in contrast, involves a settlement to a preferred alternative and a
more focused, specific consideration of evidence, particularly evidence favoring the individu-
al position.571
565
Chazen/Solomon (1975): 69. The use of the adjective “healthy” in connection with skepticism can be inter-
preted in terms of the potential negative influence of an overly skeptical attitude. An alternative (and more
polarized) explanation is provided by Toba (2011): 104 who views herein an indication of the “sickness” of
the skepticism concept in audit theory and practice.
566
Thomas/Ward/Henke (1991): 113.
567
Boynton/Kell (1996): 38.
568
Shaub/Lawrence (2002): 169.
569
See Cushing (2000): 1. This notion is seized in Section 4.5.
570
See Copeland (1996): 37; Fornelli/Desmond (2011); Heninger (2001): 115.
571
See Rasso (2013): 8f., 29. He also finds that in comparison to an implemental mindset, a deliberative mind-
set involves a greater perceived difficulty of a problem/task. This might force deliberative individuals to ex-
ert more cognitive effort to cope with the complexity of the task, which is expected to boost judgment quali-
ty. Alternatively, it can lead to a premature resignation and ending of evidence search and processing, thus
adversely affecting judgment quality. See Rasso (2013): 5.
96
4.2.3.2 Presumptive Doubt Perspective
As indicated in Section 4.2.2, the positions of some audit boards and regulators as well as
societal expectations increasingly reflect a tendency to view professional skepticism under a
presumptive doubt lens.572 This perspective refers to an attitude of disbelief573 regarding man-
agement’s honesty and the unbiasedness of client’s assertions and claims unless there is suffi-
cient reasonable evidence to resolve these doubts. This notion can be generalized as “think the
worst, and prove the best” 574. The presumptive doubt notion basically corresponds to a foren-
sic approach to auditing. It implies a focus on evidence indicative of material misstatement
and fraud rather than non-error-related evidence.575 Overall, the presumptive doubt perspec-
tive has found a very broad support and application in the relevant auditing literature. Subse-
quently, the major studies embracing this perspective are reviewed.
The presumptive doubt view is most prominently manifested in the professional skepticism
definition provided by Nelson (2009) as “a heightened assessment of the risk that an assertion
is incorrect, conditional on the information available to the auditor.”576 He elaborates on this
definition inferring that it implicates that auditors who possess a high level of professional
skepticism tend to require more sound and convincing evidence – both in terms of quantity
and quality – in order to view an assertion as reasonable, correct, and justified.577 However,
Nelson (2009) also acknowledges that the maintenance of an attitude of presumptive doubt
comes at a cost – it may induce considerable audit inefficiencies (overauditing, deadline and
budget issues) and client’s bad blood.578 These caveats notwithstanding, Nelson’s (2009) def-
inition of professional skepticism has been adopted in a number of recent studies, including
Bowlin, Hobson and Piercey (2014), Brown-Liburd, Cohen and Trompeter (2013), Carpenter,
Durtschi and Gaynor (2011), Carpenter and Reimers (2013), Fukukawa and Mock (2011),
Harding and Trotman (2011), Peytcheva (2014)579, Robinson (2011), and Robinson, Curtis
and Robertson (2013).
572
See Bell/Peecher/Solomon (2005): 66.
573
Note in passing that in there is some controversy in the philosophical literature as to whether “belief” and
“disbelief” are opposing categories. While some authors contend that “doubt”, rather than “disbelief”, is the
opposite of “belief” (see, e.g., Lundholm (1936): 5), others view “disbelief” and “belief” as the end points of
a continuum (see, e.g., Quine/Ullian (1978): 12).
574
Fischer, quoted in Murray (2012): 37.
575
See Bunge (1991): 76; Nelson (2009): 3.
576
Nelson (2009): 1.
577
See Nelson (2009): 4. Quadackers (2009): 16 argues that, strictly speaking, Nelson’s definition cited above
does not literally and essentially represent a definition/description of the concept of skepticism but rather re-
flects its related appearances (i.e., (skeptical) judgments and decisions).
578
See Nelson (2009): 4. In this context, Murray (2012): 36 emphasizes the importance of achieving the right
balance between doubt and practicality. This view is also shared by the FRC (2010): 5. In a recent study,
Carpenter/Reimers (2013): 58f. do not identify efficiency losses due to an emphasis on professional skepti-
cism.
579
Note that in addition to presumptive doubt, Peytcheva (2014) considers professional skepticism under a
social contract theory lens. Drawing on this theory, she argues that when individuals (auditors) face situa-
tions in which others (client’s management or personnel) receive benefits in exchange for obligations, a
cheater-detection state of mind is automatically activated which involves the explicit consideration of the
motives and incentives of the others to violate a social contract as well as an active search for evidence of
97
Like neutrality, the presumptive doubt perspective is also present in the older literature. In
particular, Hogarth and Einhorn (1992) define a skeptic as an individual who possesses a
high sensitivity to negative evidence but neglects positive evidence. Similarly, McMillan and
White (1993) define professional skepticism as an attitude involving a focus on evidence hint-
ing at error and misstatement.580
Also consistent with the presumptive doubt perspective is the branch of research which de-
fines professional skepticism in terms of distrust. Before focusing on distrust, it is important
to shed light on the concept from which it originates, i.e., the concept of trust.581 As defined
by Rotter (1967), trust involves “an expectancy (…) that the word, promise, verbal or written
statement of another individual or group can be relied upon.”582 Trust plays an essential role
in human life, both personal and professional.583 It is also vital in the context of auditing.584
As argued by Shaub (1996), trust is an integral and indispensable part of the audit process,
given that a full audit of all transactions is usually not possible and practicable, and hence the
auditor has to rely on the truthfulness of the unaudited part of the client’s financial files and
records.585 As the results of a recent survey show, auditors indeed believe it is essential to
trust the auditee.586 However, an overly trust587 in the client may have an impeding influence
deception. Peytcheva argues that the cheater-detection mindset essentially represents a state of enhanced
skepticism. For a detailed discussion of the rationale of social contract theory and its applicability to audit-
ing, see Peytcheva (2014): 32-35.
580
See Hogarth/Einhorn (1992): 40; McMillan/White (1993): 443f. As might be recalled from Section 3.4, a
battery of auditing studies provide empirical support for auditors’ greater sensitivity to negative evidence in
information integration and belief revision (e.g., Ashton/Ashton (1988), Ashton/Ashton (1990),
Butt/Campbell (1989), Cushing/Ahlawat (1996), Favere-Marchesi (2006), McMillan/White (1993),
Monroe/Ng (2000), Reckers/Schultz (1993)). Studies reporting contrary findings, i.e., providing evidence of
higher sensitivity to positive information and confirmation proneness in auditors’ judgments and actions, in-
clude, e.g., Bamber/Ramsay/Tubbs (1997), Ayers/Kaplan (1993), Church (1991), Morton (2001),
Pei/Reed/Koch (1992), and Waller/Felix (1984).
581
For an excellent general discussion of the concepts of trust and suspicion as well as the nature of their inter-
relation, refer to Deutsch (1958): 265-278 and Kee/Knox (1970): 357-365. Kee/Knox (1970) raise some fun-
damental (yet open) questions, e.g., are trust and suspicion the extreme end points of a continuum, is distrust
actually the opposite of trust, or are distrust and suspicion essentially one and the same thing? See Kee/Knox
(1970): 357f. Webb/Worchel (1986): 215 view trust and distrust as the two extremes of a single dimension.
582
Rotter (1967): 651.
583
See Webb/Worchel (1986): 213. They argue that without trust, life would be uneasy and unthinkable, and
people would turn into paranoids, i.e., individuals who are suspicious of anything and anyone.
584
Following the framework presented by Lewicki/Bunker (1996), the most adequate form of trust in the con-
text of auditing is calculus-based trust, that is, trust that is founded on a rational account of the relative costs
and benefits arising from trusting or distrusting behaviors. This kind of trust is fragile and adaptive. It fol-
lows the principle “trust until otherwise appears justified”. In contrast, knowledge-based trust which rests on
prior experience and familiarity, and identification-based trust which involves anticipating and sensitizing
the desired outcomes of the client both counteract professional skepticism requirements, as they involve a
sense of confirmation proneness and lack of critical reflection. See FRC (2010): 23.
585
See Shaub (1996): 154. Shaub (2004): 170 distinguishes between two major types of trust with different
implications: rational trust and emotional trust. For a discussion of these trust categories, see Shaub (2004):
173-179.
586
See Rennie/Kopp/Lemon (2010): 288f.
587
Extreme trust is referred to as “pathological trust”, “gullibility”, or “credulity”. See Deutsch (1958): 278.
Note that, on the whole, trust is a fragile psychological state which can be much easily and quickly de-
stroyed than built. For a discussion of this issue and further references, refer to Kramer (1999): 593f.
98
on professional skepticism and jeopardize audit quality. Accordingly, distrust is viewed as a
more fundamental and critical input to audit quality than trust.588
As argued by Shaub (1996), distrust and professional skepticism are closely related catego-
ries.589 This notion is also reflected in the definition of professional skepticism provided by
Shaub and Lawrence (1996) as the “willing[ness] to doubt, question or disagree with client
assertions or generally accepted conclusions”590. The authors emphasize the crucial role of
professional skepticism in auditing which consists in “tempering auditor-client trust with an
appropriate measure of suspicion.”591 In the same vein, Shaub and Lawrence (1999) argue
that professional skepticism – understood as the interaction between trust and its opposite –
influences auditing at the most fundamental level.592 The distrust view is also embraced by
Choo and Tan (2000), Rose (2007) as well as Rose and Rose (2003).593
It is worth noting that the focus of the distrust perspective is on the auditor-client relation-
ship, with auditors’ attitudes toward evidence being of secondary importance. This perspec-
tive builds on the notion that the exercise of professional skepticism and suspicion can pre-
vent, or at least reduce, the deleterious consequences of deceptive, hazardous, and self-
interested behavior of client’s personnel and management. In comparison, the focus of the
presumptive doubt perspective is on the risks auditors are confronted with in the conduct of an
audit, such as intentional misstatements, biased information, and “half-truths” obtained from
client’s management. This view involves an auditor’s conservative attitude toward evi-
dence.594
In summary, neutrality and presumptive doubt are the two most well-established perspectives
of professional skepticism. However, as argued by Glover and Prawitt (2013), none of them
is necessarily optimal in all audit situations,595 and thus none can be viewed as superior to the
other. Empirical research directly comparing the two notions suggests that they indeed in-
volve different implications for audit practice. Specifically, while Quadackers, Groot and
Wright (2012) find that the presumptive doubt perspective is more reflective of auditors’
skeptical behavior than the neutrality view (especially in high audit risk scenarios), Cohen,
588
See Shaub (1996): 156.
589
See Shaub (1996): 155f. For a contrary opinion, see Burton (1980): 52 who argues that professional skepti-
cism does not involve a posture of distrust. He further states that the audit should not be based on an as-
sumption that client’s management is concealing facts and throwing dust in the auditor’s eyes, and thus
should not be conducted in an adversary manner. Interestingly, a similar view could not be identified in the
contemporary auditing literature, which suggests that a change of focus from a positive or neutral view of
skepticism toward presumptive doubt and distrust has indeed taken place lately.
590
Shaub/Lawrence (1996): 126. They view professional skepticism as being determined by an individual’s
ethical dispositions (personality traits), experience, and situational variables. As will be shown in the next
section, this view is in line with the Nelson model of determinants and antecedents of professional skepti-
cism in auditing.
591
Shaub/Lawrence (1996): 125.
592
See Shaub/Lawrence (1999): 62.
593
Note in passing that Rose/Rose (2003): 318f. find that more suspicious auditors exert more elaborative evi-
dence evaluation and information processing in comparison to less suspicious subjects.
594
See Hurtt/Eining/Plumlee (2010): 5; Shaub (1996): 156; Shaub/Lawrence (1996): 125.
595
See Glover/Prawitt (2013): 3.
99
Dalton and Harp (2014) report that presumptive doubt is negatively associated with job atti-
tudes (person-job fit, professional identification, and organizational trust) and involves higher
turnover intentions on the part of the auditors. In contrast, neutrality is found to positively
affect job attitudes and involve more intentions to continue working within the auditing pro-
fession.596
As argued previously, professional skepticism is a very diverse and complex construct that is
quite difficult to define and capture. This notion is also reflected by the fact that in addition to
the definitions and views of professional skepticism grouped around neutrality and presump-
tive doubt, a bunch of other perspectives of skepticism can be identified in the relevant audit-
ing literature. Following, the most notable of these different views of professional skepticism
are briefly outlined.
Drawing on attitude theory, Nolder (2012) views professional skepticism as a construct that is
externally (socially) imposed to the auditor.597 In this context, she suggests the adoption of a
more holistic approach in the study of professional skepticism which goes beyond individual
aspects and recognizes the role of audit teams, firms, the auditing profession, and society (col-
lective attitudes).598
On a more abstract level, Toba (2011) defines professional skepticism as a “hybrid concept
constituting the epistemic599 and psychological aspects of cognition.”600 He elaborates on this
definition stating that skepticism encompasses both a knowledge aspect (relating to an indi-
vidual’s attitude toward evidence) and a behavioral aspect (questioning mind, i.e., a personal
predisposition).601 However, this definition appears somewhat tautological because an indi-
vidual’s approach toward evidence evaluation can also be considered as a behavioral aspect
that arises from an individual’s predisposition to raise questions and doubts vis-à-vis some
assertions and evidential matters. Notwithstanding these caveats, the definition provided by
Toba (2011) highlights the multifaceted cognitive nature of professional skepticism.
The importance of cognitive aspects for the understanding and proper application of profes-
sional skepticism has also been recognized by Plumlee, Rixom and Rosman (2012) who adopt
596
See Cohen/Dalton/Harp (2014): 24; Quadackers/Groot/Wright (2012): 29f. The latter study is described in
more detail in Section 4.3.2.2.4 and in Section 4.4.4. For a discussion of the concepts of person-job fit, pro-
fessional identification, organizational trust, and turnover intentions in connection with the two skepticism
notions, see Cohen/Dalton/Harp (2014): 8-15. Note in passing that while an attitude of presumptive doubt is
increasingly expected from auditors, Lee/Welker/Wang (2013): 213 find that auditors’ responses in a decep-
tion detection scenario fit an attitude of presumptive trust.
597
See Nolder (2012): 4f., 34f. Note that this view is in contrast to Hurtt’s (2010) understanding of professional
skepticism as an innate personal trait. Hurtt’s notion of skepticism will be discussed in great detail in Sec-
tion 4.4.
598
See Nolder (2012): 34f.
599
Note in passing that epistemology is the theory of knowledge. See Butcharov (1998): 65.
600
Toba (2011): 83, italics in original.
601
See Toba (2011): 83f.
100
a cognitive process view of skepticism. In particular, the authors describe professional skepti-
cism as “a coherent, diagnostic-reasoning process”602, the completion of which requires a
wide array of cognitive skills. The authors emphasize two specific skills believed to signifi-
cantly influence an auditor’s ability to judge, reason, and act in a skeptical manner: divergent
and convergent thinking603. Divergent thinking involves the generation of alternative explana-
tions for an unusual evidence or audit findings, whereas convergent thinking refers to the crit-
ical evaluation of the reasonableness of the generated hypotheses and explanations. These
problem-solving skills are essential for coping with complexity, ambiguity, and lack of struc-
ture in the audit domain. In this context, Plumlee, Rixom and Rosman (2012) argue that the
extant evidence of lacking professional skepticism identified in audit practice can be attribut-
ed to the failure of auditing standards and authoritative guidance to provide adequate instruc-
tions on the cognitive aspects of information processing and evidence evaluation which un-
derlie professional skepticism. The researchers demonstrate empirically that providing
auditors with specific guidance on the proper application and sequencing of thinking modes
(divergent thinking followed by convergent thinking) fosters the generation of a more rich set
of alternative explanations as well as a more thorough evaluation of the derived alternatives.
This finding is in line with the normative requirement of a questioning mind and a critical
audit approach.604
A different aspect of professional skepticism is addressed by Bell, Peecher and Solomon
(2005). They emphasize the need for far-reaching skepticism as a prerequisite for high audit
quality – skepticism with regard to management’s assumptions, evidence, but also regarding
the appropriateness of an auditor’s own (sometimes fallible) beliefs and judgment processes.
While the focus of regulatory, practical, and academic attention has clearly been placed on the
former kind of skepticism, i.e., outward skepticism, the latter type of skepticism, i.e., inward
skepticism, although not less fundamental, has largely been neglected. Inward skepticism re-
fers to auditors’ preventive self-criticism in anticipation of the claims and objections that oth-
ers might express against one’s beliefs, judgments, and actions.605 In general, inward (or self-
602
Plumlee/Rixom/Rosman (2012): 3.
603
The notion of divergent and convergent thinking was developed by J.P. Guilford (Guilford (1950, 1967))
within his conceptualization of creativity and human intelligence. Divergent thinking is closely related to the
concept of creativity. It represents a mental process involving the generation of novel ideas on a certain task,
with a focus on diversity and quantity of the outcome rather than correctness of the ideas and solutions gen-
erated. Divergent thinking encompasses the following aspects of creativity: fluency, flexibility, originality,
and elaboration. It involves a number of complex cognitive processes, including deduction, induction, syn-
thesis, differentiation, critical analysis, and problem solving. See Cohen/Ambrose (1999): 11; Razumnikova
(2012): 1028. While divergent thinking naturally involves reasoning in different directions and the genera-
tion of a multitude of ideas and potential solutions to a problem, convergent thinking targets at identifying a
single (or very few) correct, best possible, or conventional solution(s) from a pool of possible alternatives.
See Plotnik/Kouyoumdjian (2010): 310; Runco (2009): 252. Convergent thinking can be viewed as a funda-
mental precondition of logical inference. The quality of the convergent thought processes depends on the
availability of reasonable declarative knowledge. See Hammar (2012): 811. For a concise comparison of di-
vergent and convergent thinking, see Cash (2011): 93 and Cropley (2006): 391f.
604
See Plumlee/Rixom/Rosman (2012): 1-4. For a detailed description of the training modules and experimental
procedures, see Plumlee/Rixom/Rosman (2012): 17-19, 37.
605
See Bell/Peecher/Solomon (2005): 33f. Note that the notion of preemptive self-criticism originates from the
psychological literature on accountability effects. See Tetlock/Skitka/Boettger (1989): 632f. Empirical sup-
port for the benefits of an attitude of self-criticism has been provided by Grenier (2013): 4. The requirement
101
directed) skepticism involves critical, thorough, versatile, counterfactual, and flexible think-
ing.606 In the context of growing complexity, uncertainty, and risk of exposure peculiar to the
audit environment, inward skepticism appears crucial for avoiding judgment traps, mitigating
judgment flaws, and strengthening the justifiability of auditors’ beliefs and positions.
Finally, in a recent publication, Glover and Prawitt (2013) suggest that viewing professional
skepticism as a continuum which ranges from neutrality over presumptive doubt to complete
doubt in dependence of the particular audit situation and risk environment might be more
beneficial than strictly focusing on any one of the individual skepticism perspectives outlined
above.607 The adoption of a situationally dependent, rather than generalized or aggregated,
view of professional skepticism is further advocated by Robinson (2011).608
In summary, the preceding considerations suggest that a considerable definitional and concep-
tual unease and heterogeneity exist with regard to professional skepticism in auditing. Over-
all, professional skepticism can be characterized as a complex, broad, multifaceted, vague,
and elusive construct which still involves more questions than answers.609 In this line of
thought, Doucet and Doucet (1996) qualify attempts to conceptualize professional skepticism
as “a daunting task, to say the least”610. Nevertheless, endeavors toward reaching a more co-
herent conceptualization of professional skepticism are critical for gaining a deeper under-
standing of the nature of this fundamental construct. In the following section, Nelson’s (2009)
model which seeks to provide some remedy for the existing conceptual ambiguity by synthe-
sizing the current theoretical and empirical knowledge in a unifying framework of the deter-
minants of professional skepticism in auditing is presented and discussed.
Recognizing the need for a more systematic and profound understanding of the very complex
and entangled construct of professional skepticism in auditing, Nelson (2009) synthesized the
insights from previous interdisciplinary research into a model of determinants of professional
skepticism in auditing (hereafter: the Nelson model). The model is depicted in Figure 5. In a
nutshell, the Nelson model indicates that auditors’ knowledge, traits, and incentives dynami-
for auditors to be aware of and alert to the ample risks they are exposed to in their profession as well as the
possible legal or third party scrutiny of their beliefs, judgments, audit procedures, and working papers has
long been recognized in the auditing literature. Given that auditors’ a priori formed judgments and conduct-
ed work are regularly critically questioned in the context of hindsight, i.e., with the distinct benefit of rele-
vant a posteriori knowledge, auditors have to be (self-)critical in mind, in style, and in approach. See
Chazen/Solomon (1975): 67f.
606
See Grenier (2013): 7, 9.
607
See Glover/Prawitt (2013): 3.
608
See Robinson (2011): 18.
609
See Doucet/Doucet (1996): 158; Hurtt et al. (2013): 45, 72; Toba (2011): 109.
610
Doucet/Doucet (1996): 158.
102
cally interact to influence the degree of professional skepticism reflected in auditors’ judg-
ments and actions.611
3 4 5 8 9 10
Incentives
Traits
Knowledge
Audit Expe-
13 rience and 12
Training
Figure 5: The Nelson Model (Source: Nelson (2009): 5)
At the heart of the Nelson model is the notion that skeptical judgment and skeptical action are
two distinct components that need to be considered separately (link 1). In general, skeptical
behavior results from skeptical judgment. However, as will be discussed subsequently, not
every skeptical judgment necessarily translates into a skeptical action.612
Audit evidence613, i.e., all information obtained and considered throughout the audit, repre-
sents a central input to skeptical judgment (link 2). A further important input is knowledge
(link 3), which on its part is determined by traits (link 6) and experience (including training)
(link 7). Traits generally refer to stable, enduring individual attributes of an auditor, such as
intelligence, problem-solving ability, ethical predispositions, skeptical distributions, and the
inclination to doubt. Beyond their indirect influence on skeptical judgment (via knowledge),
traits also affect the judgment process directly (link 4). In addition, auditors’ incentives also
represent an important determinant of skeptical judgment (link 5).
611
See Nelson (2009): 1.
612
See Nelson (2009): 5. The distinction between skeptical judgment and skeptical action is further addressed
by Shaub/Lawrence (2002): 171f.
613
For a detailed discussion of audit evidence, including a consideration of the characteristics of evidence (rel-
evance, authenticity, verifiability), its nature (books of account, mathematical computations, physical exist-
ence, documents, letters of representation, oral statements, etc.), and methods of gathering (examination, in-
spection, confirmation, recomputation, observation, inquiry, and comparison), consult Cook/Winkle (1988):
311-333.
103
Whether a skeptical judgment leads to a skeptical action depends not only on the underlying
skeptical judgment (link 1) but also on the presence and constellation of certain incentives
(e.g., deadline pressures) (link 8), traits (e.g., autonomy) (link 9), and knowledge (link 10).
Finally, skeptical action can influence the extent or nature of evidence cognitively available to
the auditor (link 11), which then flows into the auditor’s experiential basis (link 12) and trans-
forms into evidential input (link 13) to future judgments and decisions. Accordingly, the Nel-
son model is circular in nature.614
Subsequently, the fundamental determinants of professional skepticism as indicated by Nel-
son (2009) are discussed and underpinned with relevant theoretical and empirical insights.615
The subsequent review is not intended to be exhaustive or to replicate the contents discussed
by Nelson (2009). Rather, it aims to provide a broader (meta-analytic) frame within which the
very complex and multifaceted construct of professional skepticism to be ranged in.
4.3.2 Determinants
4.3.2.1 Knowledge
614
The model description closely follows Nelson (2009): 5f.
615
Note that in a recent synthesis of academic research on professional skepticism, Hurtt et al. (2013) proposed
an alternative model for the inquiry into professional skepticism. Their model builds upon the model of
Nelson (2009) but employs a somewhat different approach. Specifically, they build two main categories:
skeptical judgment and skeptical action, each of which is analyzed on the basis of the following skepticism
antecedents: auditor characteristics, evidential characteristics, client characteristics, and environmental influ-
ences. For a comparison of the components of the Nelson model and the alternative model developed by
Hurtt et al. (2013), see Hurtt et al. (2013): 49. Overall, both models are very thorough and insightful. As a
parallel in-depth consideration of both models is beyond the scope of the present work, the genuine and
more established model developed by Nelson (2009) builds the basis for the subsequent discussion of the de-
terminants of professional skepticism. Where necessary and appropriate, Nelson’s (2009) considerations are
extended by adding more dimensions, details and/or novel insights from recent studies recognized as rele-
vant. The most notable supplements are made transparent.
616
See AUASB (2012): 5; IAASB (2012): 1, 5; Nelson (2009): 2, 7f.
617
See Hurtt et al. (2013): 51.
618
See Bell et al. (1997): 69; Low (2004): 203; Nelson (2009): 7.
619
See Nelson (2009): 2, 7f.
104
There are numerous studies demonstrating the beneficial influence of auditors’ knowledge on
performance and judgment quality. For instance, research on this topic has indicated that ex-
perience (as a proxy for knowledge) increases the quality of generated hypotheses in identify-
ing material errors in the client’s accounts (e.g., Bedard and Biggs (1991a), Wright and
Wright (1997)), improves auditors’ risk assessments (e.g., Low (2004)), triggers more com-
plete and global problem representations620 (e.g., Christ (1993), Hammersley (2006)), and
enhances the ability to focus on relevant information (e.g., Shelton (1999)) and resist client
relations pressures (e.g., Asare, Cianci and Tsakumis (2009)).621 Overall, these experiential
aspects can serve to facilitate and increase auditors’ professional skepticism. Further empiri-
cal support for the benefits of experience for professional skepticism is provided by Bernardi
(1994), Carpenter (2007), Hammersley (2006), and Knapp and Knapp (2001) who find that
more experienced auditors exhibit greater effectiveness in the identification and proper as-
sessment of fraud and material misstatement risk in the client’s accounts than less experi-
enced auditors.622
However, there is also an array of studies which demonstrate an inverse relation between ex-
perience and professional skepticism. Instances for such studies include Shaub and Lawrence
(1996, 1999, 2002) and Payne and Ramsay (2005).623 A striking example is the study by
Shaub and Lawrence (2002) who find that the least experienced auditors in their study exhibit
the greatest tendencies to not only think but also act in a highly skeptical manner. In contrast,
the most experienced auditors are identified as low in both skeptical thought and action.624
Likewise, Taylor (2000) finds that auditors with no industry-specific experience are more
conservative (skeptical) in their inherent risk assessments and exhibit lower confidence in
their judgments in comparison with experienced auditors. Corroborating results are presented
by Wright and Wright (1997).625
620
Christ (1993): 306 defines problem representation as an individual’s comprehension and interpretation of a
specific problem situation which is evoked by the cognitive process of mapping the information available for
solving the problem into the existing knowledge structure associated with the particular type of problem.
621
See Asare/Cianci/Tsakumis (2009): 230-232; Bedard/Biggs (1991b): 87f.; Christ (1993): 318f.; Hammersley
(2006): 310f.; Low (2004): 214; Shelton (1999): 223; Wright/Wright (1997): 290.
622
See Bernardi (1994): 77; Carpenter (2007): 1132, 1134-1137; Hammersley (2006): 332; Knapp/Knapp
(2001): 25.
623
See Shaub/Lawrence (1996): 138-143; Shaub/Lawrence (1999): 62, 75f.; Shaub/Lawrence (2002): 183;
Payne/Ramsay (2005): 325f.
624
See Shaub/Lawrence (2002): 183. The authors refer to the tendency to exhibit high (low) skeptical thought
combined with high (low) skeptical action as “aggressive skepticism” (“reluctant skepticism”). They argue
that aggressive skepticism can produce inefficiencies, overauditing and strained relations with the client,
whereas reluctant skepticism bears the risk of reduced audit quality. Furthermore, the researchers also identi-
fy two additional categories of skepticism: “measured skepticism”, i.e., skepticism involving moderate skep-
tical thinking (which is activated in high-risk situations) underpinned with heightened tendency to act skep-
tically, and “conflicted skepticism”, i.e., skepticism involving pronounced skeptical thinking combined with
reluctance to act on it. The researchers consider measured skepticism superior to the other forms of skepti-
cism as it is involves the lowest cognitive and audit resources and the highest benefits. In contrast, conflicted
skepticism is viewed as the least desirable category because it involves recognition of a problem paired with
unwillingness to address the issue, which is most common for pragmatic individuals lacking in idealistic
values. See Shaub/Lawrence (2002): 172-175, 183f.
625
See Taylor (2000): 697; Wright/Wright (1997): 291f.
105
At first glance, the finding of more skeptical judgments and behaviors exhibited by inexperi-
enced and/or unspecialized auditors appears counterintuitive, as one may expect novices to
lack the knowledge required for a critical evidence evaluation and risk assessment.626 Howev-
er, under a dual-process theory perspective (Section 2.2), the documented inverse relationship
between experience and professional skepticism is not that surprising. In particular, due to
their profound experience and familiarity with specific tasks, industries, and issues, expert
and specialist auditors are expected to be confident in their own abilities and to process in-
formation in an automatic manner.627 Automatic thinking and processing (System 1), howev-
er, involves a narrow spectrum of alternatives considered, cursory reasoning, and rapid but
less thorough and reflected information processing. Consequently, experience and the related
automatism in judgment are expected to limit the depth and width of auditors’ information
processing and to inhibit inward skepticism. Importantly, this phenomenon is neurally condi-
tioned, persistent, and subconscious in nature. Novices, on the other hand, are theorized to
exhibit deliberation, effort, and thoroughness in information processing to compensate for the
lack of knowledge and to deal with the uncertainty underlying the judgment situation. This
condition is expected to foster inward skepticism. Generally, the latter can be viewed as a
cognitively demanding attitude which involves reflection, thoroughness, and self-insight, and
thus presupposes controlled thinking and processing (System 2). Overall, under a dual-
process theory perspective, experience might reduce information processing intensity and
inward criticism.628
In addition, it should be noted that most auditors lack direct experience with material mis-
statements and fraudulent reporting, which may lead to a decreasing perception of the possi-
bility of material misstatements and fraud over time.629 This fact, along with the automatic
information processing typical for experienced subjects, may lead to a “default” generation of
non-fraud and non-error explanations. Novices, in contrast, are expected to be less automatic
and confident in their judgment and decision making. Furthermore, they might be sensitized
to the deleterious consequences of fraud through high school and training courses and thus
potentially exhibit higher degrees of professional skepticism.630
Overall, research on the influence of experience on professional skepticism provides rather
inconclusive results. Consequently, future conceptual and empirical research on this topic is
clearly needed.631
In addition to experience and specialization, auditing standard setters, regulators, and academ-
ic research also highlight the importance of training as a knowledge- and skepticism-inducing
626
For such claims, see Shaub/Lawrence (1999): 67.
627
See Grenier (2013): 6f.
628
The dual-processing analysis of inward skepticism presented here follows Grenier (2013): 6-8.
629
See Montgomery et al. (2002): 64; Rose (2007): 215; Zimbelman (1997): 80.
630
See Fullerton/Durtschi (2004): 23; Grenier (2013): 7f.; Loebbecke/Eining/Willingham (1989): 25;
Payne/Ramsay (2005): 323f.; Shaub/Lawrence (1999): 67.
631
For such claims, see Hurtt et al. (2013): 52f.
106
mechanism.632 Auditing studies on this topic have focused on different aspects and kinds of
skepticism training. For instance, Fullerton and Durtschi (2004) demonstrate that training
internal auditors to be alert to the possibility and symptoms of fraud effectively compensates
for low innate (trait) skepticism, so that post training, even naturally less skeptical subjects
provide judgments reflecting a higher degree of (state) professional skepticism.633 Carpenter,
Durtschi and Gaynor (2011) examine the influence of a forensic accounting course on stu-
dents’ fraud-related judgments and document a significant improvement in subjects’ fraud
risk assessment abilities upon completion of the specialized forensic training. Importantly, the
forensic training appears to have a long-term positive influence on students’ mindset and
fraud assessment abilities as indicated by the follow-up tests conducted by the authors seven
months later. Drawing on Nelson’s (2009) notion of professional skepticism as involving pre-
sumptive doubt and higher risk assessments, the authors infer from the results outlined above
that a specialized forensic training can increase an individual’s level of professional skepti-
cism.634
Another branch of research highlights the importance of sensitizing and training auditors to be
skeptical of their own information processing, beliefs, and judgments, as these may be sub-
consciously biased. This relates to the notion of inward skepticism. As demonstrated by
Grenier (2013), instructing industry specialists to be self-critical and to bear in mind the
chance of bias in their own judgments results in more skeptical judgments than in the case of
instructing subjects to be critical of evidence and management’s representations. Overall,
Grenier’s results suggest that self-criticism (directed toward the own judgment processes)
outperforms outward skepticism (directed toward evidence and client’s assertions) in raising
experienced auditors’ levels of skeptical judgments.635
Harding and Trotman (2011), on the other hand, find that inducing outward skepticism is
more beneficial in fostering skeptical judgment than inducing inward skepticism. However, it
can be argued that these findings are an artefact of the specific research design of the study
where views on fraud of differing credibility (partner vs. client) are presented. This condition
is expected to interact more strongly with outward rather than inward skepticism. In addition,
the operationalization of inward skepticism chosen by Harding and Trotman (emphasis on
justification) differs from that in the study by Grenier (2013) (emphasis on self-insight).636
Finally, Plumlee, Rixom and Rosman (2012) find that training auditors in convergent and di-
vergent thinking results in more skeptical judgments. The benefits of training auditors to
structure their thought processes in particular ways have also been empirically demonstrated
632
For instances of auditing boards encouraging the use of training to foster professional skepticism, see ASIC
(2012): 22, IAASB (2012): 5f., and PCAOB (2012): 9. Examples of researchers’ pleadings for training as an
important skepticism-enhancing mechanism include Plumlee/Rixom/Rosman (2012): 3 and Shaub/Lawrence
(2002): 167.
633
See Fullerton/Durtschi (2004): 22f. The concept of skepticism as involving both a state and a trait dimen-
sion is discussed in Section 4.4.
634
See Carpenter/Durtschi/Gaynor (2011): 10, 19.
635
See Grenier (2013): 23.
636
See Harding/Trotman (2011): 28.
107
by Trotman, Simnett and Khalifa (2009). They show that instructing auditors to think back-
ward (i.e., beginning with an assumption of an undesirable outcome and searching for a prop-
er diagnose for the problem) leads to more reasonable fraud-related judgments.637
To sum up, the preceding overview clearly shows that training auditors to be inwardly and
outwardly skeptical and to alter their thinking and information processing toward multidimen-
sionality, creativity, and inquisitiveness appears to be a promising path for improving the ap-
plication and maintenance of professional skepticism in auditing.638
4.3.2.2 Traits
As demonstrated in Section 2.4.2.3, personal factors are central determinants of auditor judg-
ment (quality). In addition, their importance in the context of professional skepticism has
been widely recognized as well.639 Nelson (2009) particularly highlights the traits of problem-
solving ability, ethical reasoning, dispositional skepticism, and dispositional (dis)trust.640 A
review of the relevant literature suggests that the cultural background and gender of the indi-
vidual auditor can also affect his/her professional skepticism under certain conditions. Alt-
hough these attributes do not strictly represent personality traits, they clearly belong to the
factors that shape a person’s individuality and have an impact on his/her behavior. Hence,
culture and gender are also included in the discussion of person-related characteristics which
have the potential to influence auditors’ professional skepticism.
Nelson (2009) recognizes that the intelligence of an auditor can enhance his/her ability to
identify potential irregularities and misstatements in the client’s accounts and financial state-
ments. This intelligence-related individual characteristic is referred to as “problem-solving
ability”.641
637
See Plumlee/Rixom/Rosman (2012): 3f.; Trotman/Simnett/Khalifa (2009): 1135. By reference to the relevant
psychological research, Trotman/Simnett/Khalifa (2009) differentiate between backward and forward think-
ing. The former is retrospective in nature and involves searching for patterns, trying to connect seemingly
unrelated events into a sound story, and examining the causational chains behind the outcome. In contrast,
forward thinking is prospective and involves gathering and evaluating information in order to determine
(predict) an outcome. See Trotman/Simnett/Khalifa (2009): 1122.
638
See Hurtt et al. (2013): 54f.
639
See, e.g., AUASB (2012): 5; Glover/Prawitt (2013): 6; IAASB (2012): 1, 5; Nelson (2009): 2, 8-11.
640
See Nelson (2009): 8. As a matter of precision, Nelson (2009) uses the generic term “skepticism scales” to
include both the concept of trait skepticism (Hurtt (2010)) and the concept of (dis)trust (Shaub (1996)).
Within the present work, both concepts are discussed separately and under the labels “dispositional skepti-
cism” and “dispositional (dis)trust” which more accurately reflect the trait nature of these individual charac-
teristics.
641
See Nelson (2009): 8. Bonner/Lewis (1990): 6 define problem-solving ability as the skill “to recognize rela-
tionships, interpret data, and reason analytically”. They view this ability as a determinant of expertise. As a
matter of fact, the subsequently listed empirical studies which identify a positive influence of problem-
solving ability on performance also find a positive relation between experience/knowledge and performance.
108
Instances for studies indicating that auditors’ problem-solving ability contributes to improving
performance on complex and ill-structured skepticism-related tasks (e.g., identification of
errors causing ratio fluctuations, manipulation of earnings) include Bonner and Lewis (1990),
Bierstaker and Wright (2001), Libby and Tan (1994), and Tan and Kao (1999).642 Overall,
problem-solving ability has been found to be an important determinant of auditors’ judgment
performance and an essential attribute for progress and promotion within the auditing profes-
sion.643 To the extent to which problem-solving (analytical) skills increase auditors’ ability to
cognitively pervade complex and ambiguous audit issues and to recognize potential existing
problems and irregularities in audit evidence, it stimulates skeptical judgment. To date, how-
ever, empirical research specifically addressing this contention is still missing.644 Thus, the
comprehensive theoretical and empirical exploration of the link between problem-solving
ability and professional skepticism represents a promising avenue for future research.
The notion that auditors’ independence and professional skepticism are essentially ethical
issues is not a novel one.645 Ethics in the context of auditing involves bearing in mind that
auditors’ primary responsibility is to serve the public interest. It is crucial for auditors to align
their mindsets and behaviors with this ethical compass because a lack of sufficient integrity
and objectivity can jeopardize audit quality and lead to considerable reputational damages for
the whole auditing profession as in the case of the Arthur Andersen debacle.646 Hence, an au-
ditor’s morality and ethical orientation are critical quality-, independence-, and skepticism-
related attributes.647
642
See Bierstaker/Wright (2001): 58; Bonner/Lewis (1990): 13-16; Libby/Tan (1994): 713; Tan/Kao (1999):
220f. Note that the results of these studies may not be directly comparable, as different measures of prob-
lem-solving ability have been employed. For example, Bierstaker/Wright (2001): 50 use the Practical Prob-
lem Solving Ability Scale developed by Devolder (1993), whereas Bonner/Lewis (1990): 10 use excerpts
from the Graduate Record Examination (for a general description of the test, see Green/Wolf (2011): 3) to
measure problem-solving ability. Tan/Kao (1999): 216, on the other hand, utilize a set of questions from the
Primary Mental Abilities Test developed by Thurstone (1938) and the Triarchic Abilities Test designed by
Sternberg (1988), and Libby/Tan (1994) do not specify the measure of problem-solving ability they employ.
643
See Tan/Libby (1997): 102.
644
See Nelson (2009): 9.
645
See Falk et al. (1999): 399; Jones/Massey/Thorne (2003): 45.
646
See Duska (2005): 19, 28. In a similar vein, Kung/Huang (2013): 479f. argue that a stringent and powerful
regulation is a necessary condition for preventing corporate misdeeds and scandals, but it may not be suffi-
cient to solve the problem of corporate fraud and dysfunctional behavior. Hence, auditors’ high ethical
standards are essential for the provision of high audit quality and for strengthening (restoring) public trust in
the auditing profession. In this sense, strong ethical reasoning and an idealistic moral outlook can be viewed
as the “the last line of defense” for the auditing profession.
647
See, e.g., IAASB (2012): 5. As will be demonstrated in Section 4.4.2.5 and in Section 4.4.2.6, moral devel-
opment strongly relates to the autonomy and self-esteem dimensions of trait professional skepticism. Basi-
cally, high ethical standards and moral courage are essential for auditors to skeptically act on the evidence
obtained, thereby fulfilling their public interest responsibilities and counteracting pressure and persuasion at-
tempts coming from the client as well as objectivity-, independence, and skepticism-inhibiting incentives.
See Schatzberg et al. (2005): 233. For a general theoretical and empirical review of the concept of moral de-
velopment, see Section 2.4.2.3.7.
109
Prior auditing research indicates that higher levels of moral development are associated with
auditors’ greater sensitivity to management’s competence and integrity (e.g., Ponemon
(1993)) and generally yield more ethically-oriented judgments and behaviors. Such evidence
has been provided in the context of independence issues (e.g., Falk et al. (1999), Ponemon
and Gabhart (1990), Sweeney and Roberts (1997), Windsor and Ashkanasy (1995)), whistle-
blowing (e.g., Arnold and Ponemon (1991)), earnings management (e.g., Brandon et al.
(2007)), fraud detection (e.g., Bernardi (1994)), and work-related (time, peer, and client-
induced) pressure (e.g., Ponemon (1992b), Tsui (1996), Tsui and Gul (1996)).648 Overall,
auditors exhibiting higher degrees of moral development have been found to base their
judgments and behaviors on ethical standards and to place professional responsibility over
self-interest or client interest. Hence, auditors at high moral development levels are also
expected to judge and act in a professionally skeptical manner.
Similarly, auditors with an idealistic ethical philosophy649 have been found to exhibit greater
conservatism and intolerance for dubious client behavior (earnings management) (e.g., Elias
(2002), Kung and Huang (2013)) as well as a higher degree of commitment to the profession
(e.g., Shaub, Finn and Munter (1993)) than less idealistic auditors.650 Stronger professional
commitment, in turn, has been argued to prevent dysfunctional, non-skeptical behavior that
has the potential to harm the profession.651 Collectively, these findings suggest an association
between ethical orientation and auditors’ professional skepticism.
Note, however, that not all prior studies report a positive association between auditors’ moral
development and their ethical behavior. For instance, Lord and DeZoort (2001) find that audi-
tors exposed to obedience and conformity pressure from their colleagues display higher ac-
ceptance for material misstatements, and hence a lack of professional skepticism, regardless
of their level of moral development.652 Accordingly, strong pressures present in the audit en-
vironment may countervail an individual’s moral development. Schatzberg et al. (2005), on
the other hand, report a significant inverse relationship between moral development and sub-
648
See Arnold/Ponemon (1991): 12f.; Bernardi (1994): 78f.; Brandon et al. (2007): 59; Falk et al. (1999):
420f.; Ponemon (1992a): 185; Ponemon (1993): 3, 21; Ponemon/Gabhart (1990): 245-247; Sweeney/Roberts
(1997): 348; Tsui (1996): 128f.; Tsui/Gul (1996): 48; Windsor/Ashkanasy (1995): 715. Interestingly, an in-
dividual’s level of moral development has been found to decrease with increasing career level. See, e.g.,
Ponemon (1990): 191; Ponemon (1992b): 245-254; Ponemon/Gabhart (1990): 245-247. In a more recent
longitudinal study, however, Bernardi/Arnold (2004): 361-364 provide evidence for increasing levels of
moral development with promotion to high-rank positions at the audit firm. For an excellent systematic
overview of empirical research on auditors’ moral development and ethical judgments and behaviors, see
Jones/Massey/Thorne (2003).
649
In general, there are two higher-level ethical philosophies: idealism and relativism. The former relates to the
adoption of universal ethical principles and standards in social interaction as well as to a general concerned-
ness with the consequences of an action on the well-being of others. Relativism, in contrast, involves a more
pragmatic approach encompassing a rejection of absolute ethical principles as guidelines for moral reasoning
and action and a focus on the particular circumstances underlying the ethical issue. The different combina-
tions of high vs. low levels of these philosophies yield a 2 x 2 matrix with the following subcategories: situa-
tionism, absolutism, subjectivism, and exceptionism. For such claims as well as a description of the different
ethical ideologies, see Forsyth (1980): 175-177.
650
See Elias (2002): 40; Kung/Huang (2013): 496; Shaub/Finn/Munter (1993): 145, 164.
651
See Lord/DeZoort (2001): 220f.
652
See Lord/DeZoort (2001): 228-230.
110
jects’ tendency to truthfully report sensitive findings (net income overstatements). However,
the researchers utilize an abstract experimental audit market setting and student subjects,
which may limit the generalizability of their findings.653
Finally, recent auditing studies likewise provide mixed evidence regarding the relation be-
tween morality and ethics and auditors’ skeptical judgments. Specifically, Kerler and
Killough (2009) do not identify a connection between auditors’ moral reasoning and their
trust in the client.654 In addition, Bobek, Hageman and Radtke (2013b) do not obtain a direct
connection between auditors’ ethical behavior and their innate (trait) skepticism655. However,
the researchers find that when auditors perceive a situation as involving high ethical conflict,
they are less likely to tolerate the contentious accounting treatments of the client, i.e., they
display more ethical and critical behavior than in situations of perceived low moral intensity.
This can be viewed as indicative of increased (state) professional skepticism in situations of
intensive moral conflict.656 In contrast to the former study, Farag and Elias (2012) document
a significant positive relation between auditing students’ trait professional skepticism and
their ethical judgments.657
In summary, even though ethical orientation and moral development might not be sufficient to
induce ethical and skeptical behavior under all conditions, they appear to generally favor pro-
fessional skepticism.
The notion that skepticism is a personality trait is neither a novel idea658 nor a specific contri-
bution of auditing theory.659 However, in the last few years academic auditing research has
achieved considerable theoretical and empirical progress in this area, which is mainly attribut-
able to the influential work of Hurtt (2010). For the sake of completeness, the idea that pro-
fessional skepticism contains a dispositional dimension is touched upon here. The in-depth
consideration and discussion of this fundamental for the present work aspect is undertaken in
Section 4.4.
Overall, prior research has indicated that dispositional skepticism is positively related to audi-
tors’ skeptical judgments and behaviors. Specifically, auditors at higher trait skepticism levels
653
See Schatzberg et al. (2005): 258f.
654
See Kerler/Killough (2009): 109.
655
Note, however, that the authors use a shortened version of the Hurtt scale (i.e., only the 5-item questioning
mind subscale). See Bobek/Hageman/Radtke (2013b): 12. Hence, their measure does not actually capture the
entire trait skepticism construct as conceptualized by Hurtt (2010) but only reflects one dimension of this
construct.
656
See Bobek/Hageman/Radtke (2013b): 3, 18.
657
See Farag/Elias (2012): 194.
658
See, e.g., Chazen/Solomon (1975): 69; Choo/Tan (2000): 82.
659
In fact, the sense of an individual disposition is present in general dictionary entries to the word “skepti-
cism” (see, e.g., http://www.chambersharrap.co.uk/) as well as in the fields of philosophy, psychology, and
consumer behavior. For a detailed theoretical overview of the individual dispositions which constitute pro-
fessional skepticism, see Section 4.4.2.
111
have generally been found to need and search for more additional information on which to
found their judgments (e.g., Fullerton and Durtschi (2004), Hurtt, Eining and Plumlee
(2010), Robinson, Curtis and Robertson (2013)), to identify more inconsistencies in the in-
formation provided (e.g., Hurtt, Eining and Plumlee (2010), Robinson, Curtis and Robertson
(2013)), to develop more alternative hypotheses for a problem (e.g., Hurtt, Eining and
Plumlee (2010)), to anchor less on client explanations and assertions (e.g., Quadackers, Groot
and Wright (2012)), to be more sensitive to fraud red flags (e.g., Popova (2013)), and to apply
stringent ethical standards when evaluating morally critical issues (e.g., Farag and Elias
(2012)).660
As indicated in Section 4.2.3.2, professional skepticism can also be viewed under a distrust661
lens. Generally, individuals having a high propensity to distrust have been found to process
and evaluate information in a more thorough and critical manner than trusting subjects, a
stance which is considered favorable for exercising professional skepticism.662
Prior to the Hurtt scale, a psychometrically developed research instrument to capture profes-
sional skepticism (or some aspects of it) has been lacking. Hence, researchers have applied
existing psychological trust scales to derive inferences about professional skepticism.663 The
predictive ability of an individual’s level of dispositional trust (or lack thereof) for skeptical
attitudes, judgments, and behaviors has generally been corroborated in prior research. Specifi-
cally, utilizing a modified version of the Trust Scale by Rempel, Holmes and Zanna (1985),
Choo and Tan (2000) demonstrate that skepticism operationalized as low levels of trust (or
high levels of distrust) is a stable innate personality trait. In addition, they find that instruction
(training) interacts with skepticism to affect an individual’s ability to identify fraud.664 Simi-
larly, Rose (2007) find that auditors who exhibit low dispositional trust as measured by the
Philosophies of Human Nature Scale by Wrightsman (1974) are more sensitive to evidence of
aggressive reporting and more inclined to assume misstatement (fraud) than their high-trust
colleagues, which is interpreted as indicative of higher professional skepticism. The results
reported by Rose (2007) further suggest that dispositional trust has greater explanatory power
for the variance in auditors’ skeptical judgments than contextual risk factors designed to in-
duce skepticism (indication of abnormal sales).665
Valuable insights on the relationship between auditors’ skeptical dispositions and their skepti-
cal judgments are further provided by Quadackers, Groot and Wright (2012). In a direct com-
660
See Hurtt et al. (2013): 51 as well as the review of relevant research provided in Section 4.4.4.
661
In the relevant literature, the terms “distrust”, “doubt”, “suspicion”, and “disbelief” have been used inter-
changeably. This procedure is also adopted in the present study.
662
See Rose/Rose (2003): 313; Schul/Burnstein/Bardi (1996): 228.
663
See Hurtt (2010): 150.
664
See Choo/Tan (2000): 80.
665
See Rose (2007): 216.
112
parison between trait skepticism (measured by the Hurtt scale) and dispositional distrust
(measured by the inversed Rotter Interpersonal Trust Scale), the authors find that disposition-
al distrust better reflects auditors’ skeptical judgments than dispositional skepticism, particu-
larly in situations involving high control risk.666
Overall, although dispositional (dis)trust and trait skepticism are distinct personality features
(and concepts), they seem to be considerably related.667 Choo and Tan (2000) make this point
very clear by arguing that trust and skepticism are in essence “complementary personality
traits”.668
In summary, research conceptualizing and measuring professional skepticism as a personality
trait has generally indicated that higher levels of dispositional skepticism or lower levels of
dispositional trust are associated with more skeptical judgments and actions.669
4.3.2.2.5 Culture670
As indicated in Section 2.4.2.3.8, an auditor’s cultural background can affect his/her percep-
tions, attitudes, judgments, and behaviors in a number of different ways. As might be recalled
from this section, the cultural attributes of individualism and low power distance have been
contended to positively correlate with the autonomy, self-determination, and ethical orienta-
tion within a given country or cultural group. These attributes have the potential to reinforce
professional skepticism. In contrast, in collectivistic and high power distance cultures where
harmony, consensus, avoidance of confrontation, and respect for (obedience to) authorities are
typical, professional skepticism might be a culturally difficult attitude. Likewise relevant in
the context of professional skepticism are the cultural dimensions of uncertainty avoidance
and long- versus short-term orientation. Specifically, the cultural attributes of high uncertainty
avoidance and long-term orientation have been theorized to positively correlate with judg-
ment prudence and the amount of information and evidence gathered in forming a judgment.
Following these lines of thought, skepticism may be a more salient attribute in high uncertain-
ty avoidance and long-term orientation countries.
Empirical evidence on the relation between cultural attributes and professional skepticism is
rather sparse and ambiguous. Contrary to the expectation that US auditors (representing an
individualistic and low power distance culture) will raise more direct, critical, and challenging
questions and will test client’s assertions more extensively than Japanese auditors (represent-
666
See Quadackers/Groot/Wright (2012): 29.
667
See Aschauer et al. (2013): 12; Doucet/Doucet (1996): 162; Popova (2013): 146; Quadackers (2009): 16.
668
Choo/Tan (2000): 82. By reference to the relevant psychological literature, the authors attach a positive
general (i.e., not audit-related) meaning to trust by associating it with personality features like sincerity,
honesty, and duty, while viewing skepticism in a group of traits such as secrecy, cynicism, and even decep-
tion.
669
See Hurtt et al. (2013): 51.
670
Note that Nelson (2009) does not explicitly discuss cultural issues within his model of professional skepti-
cism. However, as indicated by Hurtt et al. (2013): 70f., culture is an important behavioral determinant
which has the potential to influence auditors’ professional skepticism.
113
ing a collectivistic and high power distance culture), Yamamura et al. (1996) do not obtain
significant culture-related differences in auditor behavior. The authors attribute this finding to
the high standardization among audit firms worldwide.671 Cohen, Pant and Sharp (1995), on
the other hand, find that Latin American auditors exhibit stronger ethical reasoning in evaluat-
ing hypothetical moral dilemmas than US auditors. The authors explain these (unexpected)
results by reference to the desire not to “lose face” by tolerating unethical behavior, which is
typical for the Latin American collectivistic and high power distance culture group.672 Patel,
Harrison and McKinnon (2002) also demonstrate that cultural differences related to individu-
alism-collectivism and power distance influence auditors’ skeptical judgments. They find that
Australian auditors are less willing to accede to client pressure in an audit conflict scenario
and are more critical of their colleagues’ decision to accede to the client (unethical behavior)
than are Indian and Chinese auditors.673 Similarly, Sim (2010) finds that in assessing the ef-
fectiveness of client’s internal control, Taiwanese auditors evaluate evidence in a manner al-
lowing them to “draw a good picture”, whereas Australian auditors do not exhibit a client-
favorable evidence interpretation.674 In addition, Endrawes and Monroe (2012) demonstrate
that in contrast to Australian auditors, Egyptian auditors avoid confrontation with the client.675
Further evidence on the relevance of culture for professional skepticism is provided by Bik
(2010) who reports some striking culture-related survey comments such as the following one:
“(…) the ability to challenge the client and professional skepticism is a huge question mark.
(…) We have questions whether you can ever do a truly effective audit in this region, because
of the lack of skepticism and the ‘bowing’ towards seniority and authority. (…) So, if you are
a junior auditor and you are talking to the general manager of your client, and the general
manager says something, you will always agree because he’s a more senior person. He just
won’t be counted wrong. (…) People tend to bow for pressure of their clients. I think it is
deeply rooted in the culture of the region.”676
In addition, even practitioners and regulatory board members appear to recognize the connec-
tion between cultural attributes and professional skepticism. Specifically, Craig Fisher, audit
director at Hayes Knight and member of the New Zeeland Auditing and Assurance Standards
Board, acknowledges that professional skepticism might be per se a hardly applicable attitude
in New Zeeland, as (in his view) the New Zealanders are a rather trusting nation.677
671
See Yamamura et al. (1996): 356f. Note that the authors use a somewhat different terminology than the one
outlined in Section 2.4.2.3.8 within the discussion of the cultural dimensions set forth by Hofstede. In partic-
ular, they use the terms “rank consciousness”, which essentially corresponds to power distance, and “group
orientation”, which essentially corresponds to individualism-collectivism.
672
See Cohen/Pant/Sharp (1995): 56f. Recent theoretical and empirical insights provided by Wong-On-
Wing/Lui (2013): 30f. suggest that these results may also be attributed to differences in the views of causali-
ty between the different nations (holistic in the Asian and South American countries versus analytic in the
US and the Western countries).
673
See Patel/Harrison/McKinnon (2002): 23.
674
See Sim (2010): 55.
675
See Endrawes/Monroe (2012): 19f.
676
Bik (2010): 112.
677
See Fisher, quoted in Murray (2012): 37.
114
Collectively, these theoretical and empirical insights reveal that cultural issues should be con-
sidered in the context of inquiries into the concept of professional skepticism. This field of
research is relatively unexplored and represents a fruitful avenue for future studies. In particu-
lar, in light of the growing complexity, ambiguity, and future orientation of accounting and
auditing, the dimensions of uncertainty avoidance and long- versus short-term orientation
appear central and thus clearly deserve to be addressed by future research. A further important
research area concerns the exploration of the nature and strength of the interrelations between
the individual cultural dimensions.
4.3.2.2.6 Gender678
As indicated in Section 2.4.2.3.9, prior research has documented the existence of gender-
related differences in auditors’ ethical dispositions, information processing style, sensitivity to
evidence, and risk attitudes. All of these behavioral aspects potentially relate to auditors’ pro-
fessional skepticism.
The existing auditing research on the relation between gender effects and professional skepti-
cism has yielded an inconsistent picture. In particular, Breesch and Branson (2009) find no
significant differences in auditors’ skeptical judgments and actions (detection of misstate-
ments and adequate reporting of the findings) between the genders.679 In contrast, Hardies,
Breesch and Branson (2014) demonstrate that female auditors are more likely to issue going
concern opinions to financially distressed firms than male auditors. This effect is amplified
with high client importance or risk. Hence, the authors conclude that female auditors are more
independent, risk averse, and concerned with serving the public interest than male auditors,
and thus provide higher audit quality.680 Consistently, Jonnergård, Stafsudd and Elg (2010)
indicate that male auditors focus on the commercial side of auditing (performance, profitabil-
ity, and revenue), whereas female auditors are more aligned with audit quality and the ethical
aspects of the profession.681 In addition, Fullerton and Durtschi (2004) document higher
scores on the Hurtt trait professional skepticism scale for female internal auditors than for
male participants.682 Bobek, Hageman and Radtke (2013c), on the other hand, find that male
auditors score higher on the questioning mind dimension of skepticism.683 However, the au-
thors also find that female participants are less likely to recommend conceding to a client’s
678
Note that Nelson (2009) does not explicitly discuss gender as an individual characteristic that may influence
auditors’ professional skepticism. However, a review of the relevant literature reveals a number of gender-
related behavioral differences which may also play an important role in the context of professional skepti-
cism in auditing.
679
See Breesch/Branson (2009): 96.
680
See Hardies/Breesch/Branson (2014): 7f., 28.
681
See Jonnergård/Stafsudd/Elg (2010): 737. This contention is echoed by Hardies/Breesch/Branson (2014): 7.
682
See Fullerton/Durtschi (2004): 23.
683
For a detailed discussion of the dimensions of trait professional skepticism, see Section 4.4.2.
115
questionable preferred position and employ a more reflexive approach to judgment and deci-
sion making.684
Overall, these findings suggest that female auditors appear to be more skeptically poled than
male auditors. To the extent to which these results are generalizable beyond the studies’ con-
text, they entail important implications for audit practice where the upper rank positions are
usually occupied by male auditors.685 Specifically, these results, if tenable, would generally
imply that it may be beneficial for audit firms to promote more females who can set a proper
skeptical tone at the top. Additional research on this issue is clearly needed to provide further
insights and a more conclusive picture of the links between gender and auditors’ skeptical
attitudes, judgments, and behaviors.
Finally, prior research also indicates that not only auditor gender but also client gender might
be of relevance for the inquiry into auditors’ professional skepticism. In particular, it has been
argued that client gender may affect the individual degree of trust (or lack thereof) attached to
the client’s explanations and claims. However, there is no clarity as to whether a male or fe-
male client representative is perceived as more trustworthy and persuasive. While Shaub
(1996) and Shaub and Lawrence (1996) argue that women generally enjoy greater trust than
man, and hence auditors may exhibit less skepticism and criticalness when obtaining infor-
mation and evidence from a female client, Gold, Hunton and Gomaa (2009) empirically
demonstrate the contrary effect. Specifically, their results indicate a clear male favorability
tendency for both genders, i.e., both male and female auditors were found to be more inclined
to accede to the client’s advocated position when the client representative was male. This
finding can be attributed to gender stereotyping according to which men are generally viewed
as more capable, successful, self-confident, and assertive than women.686
Collectively, the empirical research outlined above, although not particularly conclusive, gen-
erally suggests that gender effects involving both inherent individual auditor differences and
client-related stereotyping issues might affect auditors’ professional skepticism. Further re-
search along these lines is clearly needed. The present study provides additional empirical
evidence on the relationship between gender and auditors’ (trait) professional skepticism. The
results of this inquiry are presented in Section 5.5.2.
4.3.2.3 Incentives
Incentives are a further key determinant of professional skepticism in the Nelson model. They
can influence skeptical judgment and behavior through affecting the cost-benefit ratio of
(un)skeptical attitudes. Overall, there is a variety of incentives relevant to auditing in general
684
See Bobek/Hageman/Radtke (2013c): 4, 19, 21f., 30.
685
Indeed, as pictured by Doucet/Hooks (1999): 72f., while the distribution between male and female entrants
to the auditing profession (staff auditors) is virtually balanced out, the senior manager and partner positions
at the audit firms are dominated by male auditors (68% male senior managers and 84% male partners).
686
See Gold/Hunton/Gomaa (2009): 4f., 11; Shaub (1996): 160f., 168; Shaub/Lawrence (1996): 131.
116
and professional skepticism in particular. Subsequently, some of the most significant incen-
tives fostering or compromising auditors’ professional skepticism are discussed.
Oversight is a meaningful institutional mechanism which aims at controlling audit quality and
increasing public trust in the value and reliability of auditing services.687 In general terms,
regulatory scrutiny is hypothesized to enhance professional skepticism by raising the cost
(both financial and reputational) of deficient audit (judgment) quality. However, empirical
research directly examining the links between oversight intensity and auditors’ skeptical atti-
tude is still missing. While it is conceivable that enforcement actions have a deterrent effect
on the auditing profession and thus act as skepticism stimulators, it is also possible that the
influence of oversight on auditors’ skeptical behavior is limited by factors like social dis-
tance.688 Future research on this issue is clearly needed to provide insights into the nature and
strength of the relation between oversight and auditors’ professional skepticism.
Reputation is a key success factor in auditing. In fact, it is what adds value to the credibility-
enhancing (attestation) function of the audit.689 Hence, reputational damages which typically
result from accounting and/or fraud scandals and litigation procedures represent a powerful
incentive that can influence auditors’ skeptical judgments and behaviors.690 Similarly to regu-
latory oversight, litigation and reputation loss have the potential to reinforce auditors’ inde-
pendence, objectivity, professional skepticism, and in the end to improve audit quality by in-
creasing the cost of dysfunctional (unskeptical and/or unobjective) behavior and audit failure
(e.g., impending substantial audit fee reductions and client losses).691
687
See Johnstone/Sutton/Warfield (2001): 7. For a normative and institutional discussion of oversight and en-
forcement in Germany, see Köhler et al. (2008): 129-133. For a concise review of audit regulation in the
USA, see Moore et al. (2006): 13-15.
688
See Hurtt et al. (2013): 68; Nelson (2009): 11.
689
See Kinney (1994): 81; McCracken (2003): 165f.
690
See McCracken (2003): 177. For a review of the auditing literature on litigation, see Latham/Linville (1998):
175-207. For a formal-analytic consideration of reputation losses, see Bigus (2011): 297-304. He demon-
strates that the prospect of substantial reputation damages renders litigation concerns irrelevant. In addition,
while litigation is a factor of great relevance and importance in the USA, it is much less meaningful in con-
tinental Europe where the present study is settled. Hence, reputational concerns may be a more powerful
skepticism-inducing incentive for German auditors than litigation. Note, however, that under conditions of
uncertainty and ambiguity, reputational concerns might not be sufficient to prevent deviations from objectiv-
ity and independence, as indicated by the findings of an experimental market study by
Mayhew/Schatzberg/Sevcik (2001): 66.
691
See Bigus (2011): 288; Davis/Simon (1992): 66f.; Firth (1990): 386; Johnstone/Sutton/Warfield (2001): 3;
Palmrose (1988): 57; Skinner/Srinivasan (2012): 1759f.; Venkataraman/Weber/Willenborg (2008): 1317.
117
The scant empirical research on the link between litigation and reputation loss concerns on the
one side and professional skepticism on the other side generally corroborates the positive in-
fluence of these incentives on auditors’ skeptical judgments and choices. In particular, Asare,
Cianci and Tsakumis (2009) demonstrate that reputation loss consciousness improves audi-
tors’ ability to resist client retention pressures and to provide impartial and conservative
judgments. Similarly, Blay (2005) finds that auditors facing high litigation risk provide more
conservative (less client-favorable) going concern judgments and audit opinions than do audi-
tors facing high client loss risk. In addition, he indicates that participants in the high litigation
risk condition focus on negative information, while subjects in the high client retention pres-
sure condition strongly attend to positive information. Further support for the conservatism-
enhancing influence of litigation risk on auditors’ judgments and choices is provided by
Braun (2001), Farmer, Rittenberg and Trompeter (1987) as well as Hackenbrack and Nelson
(1996).692 Collectively, these results support the notion that reputation and litigation threats
are powerful incentives that advance professional skepticism.
Litigation and reputational concerns are theorized to increase as an auditor advances in his/her
career because of the accountability to a larger audience (including the public) and the greater
personal and economic consequences (e.g., tarnished reputation, client and capital loss) being
at stake.693 However, similarly to the argumentation provided for oversight, it is also conceiv-
able that due to overconfidence and social distance, individuals consider it less likely for
themselves to fall prey to litigation and/or reputational scandals.
Audit firm culture can be described as the tone at the top, that is, the message that an audit
company’s management forwards to the firm’s personnel concerning the primary addressee of
auditors’ professional duty. Audit firm culture can be broadly divided into three categories:
public duty culture (auditors serving the public interest), risk management culture (auditors
occupying a neutral position between the public and client interest), and client advocacy cul-
ture (auditors serving the client interest).694 Clearly, the latter category is at odds with the re-
quirement for auditors to exercise sufficient professional skepticism.
Indeed, as demonstrated by Peecher (1996) and Turner (2001), a tone at the top which favors
client interest and desired audit outcomes can result in biased judgments and insufficient pro-
fessional skepticism at staff level.695 In contrast, supervisors’ position which straightforwardly
emphasizes the importance of serving the public interest is most likely to encourage the appli-
cation of professional skepticism throughout the audit. As the results reported by Carpenter
and Reimers (2013) show, the proper tone at the top that emphasizes audit effectiveness rather
692
See Asare/Cianci/Tsakumis (2009): 226; Blay (2005): 777-780, 785; Braun (2001): 93f.;
Farmer/Rittenberg/Trompeter (1987): 8, 10; Hackenbrack/Nelson (1996): 54f.
693
See Asare/Cianci/Tsakumis (2009): 225; Hurtt et al. (2013): 68; McCracken (2003): 177.
694
See Johnstone/Sutton/Warfield (2001): 9.
695
See Peecher (1996): 138f.; Turner (2001): 702.
118
than efficiency, thus taking into account the responsibility of the auditing profession to the
public, indeed induces skeptical judgments and decisions in a fraud assessment context. Con-
sistent results are reported by Dennis and Johnstone (2014).696
4.3.2.3.1.4 Accountability697
Related to the incentive of skepticism-favoring audit firm culture is the issue of accountability
to superiors with unknown or skeptical preferences. As discussed in Section 2.4.2.2.2, ac-
countability to a supervisor with unknown preferences is expected to induce conservatism,
thoroughness, and a skeptical attitude in reasoning and information processing. This conten-
tion has been empirically supported by Hoffman and Patton (1997) who find that when audi-
tors are unaware of the preferences of the supervisor to whom they are accountable, they pro-
vide more conservative (i.e., higher) fraud risk assessments because skeptical judgments are
perceived as more defensible than unskeptical positions. Similar results are reported by Lord
(1992) and Turner (2001).698
In addition to the case of unknown reviewer’s 699 preferences, accountability can also benefit
professional skepticism when the views of the supervisor are well known and clearly pointing
toward a critical, inquisitive attitude in evidence gathering and evaluation. As demonstrated
by Turner (2001), when auditors are accountable to reviewers with preferences for audit effi-
ciency and client credence, they perform a narrow, client-prompted evidence search and ex-
amination. In contrast, when reviewers have a preference for audit effectiveness and profes-
sional skepticism, reviewees’ responses involve a much more critical, extensive, and
independent search for and evaluation of audit evidence. The importance of a superior’s
(partner’s) emphasis on professional skepticism as a skepticism-inducing mechanism has also
been empirically demonstrated by Carpenter and Reimers (2013). Similarly, Buchman,
Tetlock and Reed (1996) demonstrate that auditors accountable to a conservative audit partner
process more information items in arriving at a judgment and choose more frequently a quali-
fied opinion than auditors who have to justify their position to the client or are not held ac-
countable.700
In addition, Peecher et al. (2010) find that supervisors’ attitude (tendency) toward a client-
preferred conclusion affects subordinates’ judgments by evoking motivated reasoning701.
696
See Carpenter/Reimers (2013): 65f.; Dennis/Johnstone (2014): 6, 33.
697
Note that Nelson (2009) does not explicitly discuss accountability as an incentive that stimulates profession-
al skepticism, probably due to the overlap of accountability and audit firm culture aspects. In consistency
with Hurtt et al. (2013): 61f., and because of the extant literature highlighting the importance of accountabil-
ity as an effort- and judgment-quality-enhancing mechanism, this aspect is discussed separately from audit
firm culture.
698
See Hoffman/Patton (1997): 229f., 233-236; Lord (1992): 97f., 103; Turner (2001): 685.
699
For a general discussion of the review process in auditing, see Rich/Solomon/Trotman (1997b): 483-485. For
an overview of the relevant research on the audit review process, see Bonner (2008): 243-246.
700
See Buchman/Tetlock/Reed (1996): 394; Carpenter/Reimers (2013): 65f.; Turner (2001): 683.
701
Motivated reasoning refers to the psychological tendency to search for and overestimate information and
evidence that corroborate favored assertions and conclusions (in the auditing context typically client-
119
Moreover, subordinates’ distorted judgments are eventually incorporated into the final audit
team judgment. With other words, supervisors can first trigger and then anchor on subordi-
nates’ unskeptical judgments.702
Collectively, these studies highlight the importance of accountability to supervisors with un-
known or clearly skeptical preferences for inducing professional skepticism on the part of the
subordinates. These findings have important implications for audit practice. They indicate that
audit firms may be well advised to encourage and promote auditors exhibiting skeptical mind-
sets who can convey their skeptical attitude to subordinates and the audit team. Otherwise,
when supervisors favor client credence and advocacy, their lack of skepticism is likely to
translate into the judgments and actions of their subordinates. In this case, accountability may
act as a skepticism-inhibitor rather than skepticism-instigator.703
In summary, oversight, litigation, reputation loss concerns, and accountability represent in-
centives that can stimulate professional skepticism by increasing the cost of audit failure. It
should be noted, however, that these incentives target on intentional misconduct, dysfunction-
al behavior, and corruption, but they do not necessarily counteract subconscious cognitive
bias which can, under certain conditions, have a very strong deleterious impact on auditors’
behavior.704
A potential inherent barrier to professional skepticism lies in the economics of auditing and
the incentives arising from the nature of the auditor-client relationship. Specifically, while
required to serve the informational needs of shareholders and the broad public, auditors are
appointed and paid by the entity, have interest to maintain the engagement, and over time de-
velop an increasing familiarity with and commitment705 to the client. Consequently, the rela-
tionship between the auditor and the client is much more seizable, personal, and direct as
compared with the rather distant and impersonal connection to shareholders and the public.
The inherent economic and personal bond between the auditor and the client induces consid-
preferred treatments). In its extreme form, motivated reasoning may turn into “motivated blindness” – the
tendency to only see what is convenient to see and overlook the counter-facts. This phenomenon particularly
applies to situations of high ethical conflict (e.g., the Enron case where the economic dependency of Arthur
Anderson boosted motivational blindness of striking fraud red flags). See Bazerman/Tenbrunsel (2011a):
81f.; Bazerman/Tenbrunsel (2011b): 61.
702
See Peecher et al. (2010): 1782f.
703
See Hurtt et al. (2013): 62; Turner (2001): 702.
704
See Bazerman et al. (2006): 45f. As argued by Moore et al. (2006): 16-18, unconscious cognitive biases can
affect the quality of auditors’ judgments and behaviors to a much greater extent than intentional corruption
and misconduct. Biases particularly relevant in the context of auditor-client conflicts of interest include,
among others, motivated reasoning, selective perception, and escalation of commitment.
705
For an excellent review of the antecedents and consequences of commitment in the auditor-client relation-
ship, see de Ruyter/Wetzels (1999).
120
erable conflicts of interest and gives rise to a number of incentives which have the potential to
compromise auditors’ objectivity, independence, and professional skepticism.706
Audit firms operate in a highly competitive environment, which imposes considerable pres-
sure on audit companies to retain their clients, to respond to clients’ wishes and preferences in
order to keep the clients pleased and secure long-term future audit and non-audit fee streams,
and to increase audit efficiency. However, efficiency enhancements and close economic and
personal bonds with the client might be achieved at the cost of audit effort, effectiveness, and
professional skepticism.707
Prior research generally shows that client pressure708 affects auditors’ judgment and decision
making in a manner consistent with impaired professional skepticism. Specifically, several
studies indicate that when client loss risk is emphasized or an important client is concerned,
auditors are more willing to acquiesce to the frequently controversial client-preferred treat-
ments and reporting choices rather than maintain a critical attitude (e.g., Blay (2005), Farmer,
Rittenberg and Trompeter (1987), Nelson, Elliott and TarpIey (2002), Roberts (2010)).709
Similarly, prior research has found that fee pressure imposed by the client as well as the cli-
ent’s explicitly stated preferences also adversely affect auditors’ judgments and decisions
(e.g., Gramling (1999), Haynes, Jenkins and Nutt (1998), Houston (1999)). In addition, client
importance is found to increase auditors’ identification with the client, with this cognitive
attachment having adverse influence on auditors’ objectivity (e.g., Bamber and Iyer
(2007)).710
706
See Bazerman et al. (2006): 45; Moore et al. (2006): 15f.; Murray (2012): 36.
707
See Gavious (2007): 461; Johnstone/Sutton/Warfield (2001): 3; Nelson (2009): 11. As an answer to the
growing fee pressure, the auditing profession increasingly utilizes various checklists and templates which
provide a structured and time-saving opportunity to get the audit work efficiently conducted. However, this
“tick the box, get it done” mentality poses a threat to the thorough and consistent application of professional
skepticism in auditing. See Fischer, quoted in Murray (2012): 37. In addition, time pressure and extensive
work load can also lead to breakdowns in skepticism applied in the course of the audit because under these
conditions, auditors are more likely to willingly accept information and evidence at their face value, without
subjecting them to critical inquiry and testing, simply in order to get the audit work done upon deadline. See
Hurtt et al. (2013): 63.
708
In the auditing literature, several related terms have been used interchangeably to refer to the economic
bonding between the auditor and the client and the resultant conflicts of interests, e.g., “client pressure”,
“client importance”, “client influence”, and “fee dependence”. See Kao/Li/Zhang (2013): 142. Overall, it has
been long recognized that there is a power asymmetry in the auditor-client relationship in favor of the client
who also has an incentive to employ its power to induce the auditor to accede to the client-preferred out-
come. While the client’s power and influence is viewed as high, auditors’ ability to withstand client pressure
has been classified as rather limited. For such contentions and a framework for the analysis of the auditor-
client power distribution, see Goldman/Barlev (1974): 712.
709
See Blay (2005): 777-780; Farmer/Rittenberg/Trompeter (1987): 8; Nelson/Elliott/TarpIey (2002): 182,
196f.; Roberts (2010): 30.
710
See Bamber/Iyer (2007): 15-19; Gramling (1999): 126; Haynes/Jenkins/Nutt (1998): 88, 100f.; Houston
(1999): 81f.
121
However, recent archival research (e.g., Chen, Sun and Wu (2010), Kao, Li and Zhang (2013),
Li (2009)) suggests that regulatory acts like the Sarbanes-Oxley Act of 2002 (SOX) in the
USA or major institutional improvements in China have enhanced auditors’ ability to with-
stand client pressure.711 These results suggest that tough audit regulation creates incentives
inducing auditors to place the cost of audit failure and shredded reputation above the econom-
ic benefits associated with a continuous audit engagement and good relationship with the cli-
ent.712 Hence, regulation can, at least to some extent, counteract the adverse effects of client
importance and retention pressure on professional skepticism.
A further relevant branch of research concerns the influence of fees for non-audit services
(i.e., an indicator of the strength of the economic bond between the auditor and the client hint-
ing at potential independence and skepticism threats) on audit quality713 and professional
skepticism.714 The results of the extant archival research on this topic are quite inconsistent.
While some studies indicate that non-audit fees jeopardize audit quality and professional
skepticism (e.g., Frankel, Johnson and Nelson (2002), Habib (2012), Kinney, Palmrose and
Scholz (2004)), numerous other studies do not indicate an impairment of audit quality associ-
ated with the provision of non-audit services (e.g., Antle et al. (2006), Ashbaugh, LaFond and
Mayhew (2003), Ratzinger-Sakel (2013), Willoughby, Carmona and Momparler (2012)).715
The experimental evidence on the association between non-audit fees and auditors’ profes-
sional skepticism likewise provides an inconclusive picture. Specifically, Beeler and Hunton
(2002) find that non-audit fee prospects and low balling716 adversely influence auditors’ evi-
dence assessment and going concern evaluations, thereby representing a threat to auditors’
independence and professional skepticism. In contrast, Asare, Cohen and Trompeter (2005)
711
See Chen/Sun/Wu (2010): 128f.; Kao/Li/Zhang (2013): 158-160; Li (2009): 225. Note, however, that the
results of these archival studies may have been triggered by the severe media and public scrutiny in the af-
termath of the epic accounting debacles at the beginning of the twenty-first century and the issuance of
SOX. See Kao/Li/Zhang (2013): 145.
712
See Chen/Sun/Wu (2010): 127.
713
Note in passing that due to the fact that audit quality is an unobservable construct, proxies have been em-
ployed in auditing research to measure audit quality. In the studies listed subsequently, the following surro-
gates of audit quality have been used: auditors’ propensity to issue going concern opinions for financially
distressed companies, the amount and sign of abnormal accruals, the extent of just beating or missing im-
portant earnings targets, and the extent of fraudulent financial reporting. See Carcello/Nagy (2004): 55;
Carey/Simnett (2006): 673.
714
The critical role of non-audit fees in the context of professional skepticism in auditing has been underscored
by the results of a recent survey conducted by Aschauer et al. (2013): 13f. who find that in the presence of
high amounts of non-audit fees, even the client perceives auditors’ professional skepticism as compromised.
715
See Antle et al. (2006): 238; Ashbaugh/LaFond/Mayhew (2003): 613; Frankel/Johnson/Nelson (2002): 98-
100; Habib (2012): 242-244; Kinney/Palmrose/Scholz (2004): 584f.; Ratzinger-Sakel (2013): 130;
Willoughby/Carmona/Momparler (2012): 425. Of these studies, the one conducted by Habib (2012) de-
serves particular attention, as it represents a meta-analysis of 45 existing studies on the association between
non-audit fees and audit quality. Although his results should be considered in light of the caveat of great di-
versity and limited comparability of the reviewed prior research, his analysis shows an overall detrimental
effect of non-audit fees on audit quality and independence (professional skepticism).
716
Note that while regulating authorities and a substantial part of the academic literature view low balling as a
potential threat to auditors’ independence and professional skepticism, Lee/Gu (1998): 536, 550f., using a
formal analysis, argue and demonstrate that when then owner rather than the management of the firm has the
right to hire and fire the auditor, low balling can actually improve independence and even represent a less
costly substitute for litigation.
122
report results suggesting that auditors’ judgments and client acceptance decisions are based on
objective client risk criteria rather than on the potential to obtain non-audit fees from prospec-
tive clients.717
Overall, conflicting incentives arising from the nature of the auditor-client relationship repre-
sent a key factor in the analysis of audit quality. Whether and to what extent client importance
and fee dependence actually impair auditors’ professional skepticism, independence, and ob-
jectivity is a fairly controversial question which cannot be straightforwardly answered on the
basis of the existing empirical results. Hence, it is important that more systematic research
(both theoretical and empirical) is conducted on the nature and strength of the interrelation
between client-related pressures and professional skepticism issues.
4.3.2.3.2.2 Tenure
717
See Asare/Cohen/Trompeter (2005): 491-493; Beeler/Hunton (2002): 44.
718
See Bedard/Johnstone (2010): 49; Gavious (2007): 461; Lim/Tan (2010): 926; PCAOB (2012): 2.
719
See Deis/Giroux (1992): 465; Mautz/Sharaf (1961): 208; Shaub/Lawrence (2002): 169.
720
See Bedard/Johnstone (2010): 47; Carcello/Nagy (2004): 57; Chih-Ying/Chan-Jane/Yu-Chen (2008): 439f.;
Geiger/Raghunandan (2002): 73-75; Johnson/Khurana/Reynolds (2002): 654f.; Knechel/Vanstraelen
(2007): 114; Manry/Mock/Turner (2008): 570; Myers/Myers/Omer (2003): 796. Note that these studies have
been conducted using archival data and focus on the macro (audit firm) level as opposed to experimental re-
search focusing on the micro (individual auditor) level.
721
See Lim/Tan (2010): 923, 926. In a recent study, Aschauer et al. (2013): 3, 13 likewise do not indicate a
negative influence of tenure on auditors’ professional skepticism as perceived by the client. However, it is
questionable whether the client’s view represents the most adequate and objective measure of auditors’ pro-
fessional skepticism.
123
ed by Carey and Simnett (2006), Junaidi, Miharjo and Hartadi (2012) as well as Ye, Carson
and Simnett (2011).722 In addition, Bamber and Iyer (2007) find that long audit tenure, along
with client importance and positive image, strengthens auditors’ identification with the client
and increases auditors’ tendency to acquiesce to the client’s preferences and desires, thereby
impairing auditors’ objectivity and professional skepticism.723
With regard to the notion that audit firm rotation can potentially counteract the skepticism-
and independence-inhibiting influence of tenure, empirical results have likewise been incon-
sistent. While some experimental studies (e.g., Dopuch, King and Schwartz (2001), Wang and
Tuttle (2009)) find that under imposed audit firm rotation regimes, auditors are less likely to
concede to the preferences of the client, other studies (e.g., Bowlin, Hobson and Piercey
(2014)) indicate impairments to professional skepticism and audit effort as a result of manda-
tory rotation requirements.724
With regard to the influence of individual auditor rotation, Tan (1995) reports results suggest-
ing that staff rotation triggers a novel, fresh viewpoint as well as greater inquisitiveness, vigi-
lance and less dependence on prior year audit procedures and findings. Consistently, Favere-
Marchesi and Emby (2005) find that partner rotation results in more conservative and vigilant
goodwill assessments.725 Collectively, the two experimental studies indicate that personnel
rotation can strengthen professional skepticism both at staff and partner level.
In summary, auditing research on the nature of the relationship between tenure and audit qual-
ity, independence, and professional skepticism has produced an inconsistent picture. Addi-
tional research along these lines appears crucial, particularly in light of the current regulatory
debate concerning mandatory rotation of audit firms as a means of strengthening auditor in-
dependence.726
4.3.2.3.2.3 Familiarity
Like long audit tenure, familiarity with the client appears inevitable for auditors’ deep under-
standing of the client’s business and the planning and performance of audits of high efficiency
and effectiveness. Yet this very same aspect of the auditor-client relationship is also viewed
as a considerable threat to professional skepticism, as the maintenance of a close, familiar
relationship with the client might give raise to personal and economic bonds and undue trust
722
See Carey/Simnett (2006): 673f.; Junaidi/Miharjo/Hartadi (2012): 311-313; Ye/Carson/Simnett (2011):
144f. Again, the cited research is archival in nature.
723
See Bamber/Iyer (2007): 15-19. For such claims, see also Nelson (2009): 12.
724
See Bowlin/Hobson/Piercey (2014): 28f.; Dopuch/King/Schwartz (2001): 115f.; Hurtt et al. (2013): 67;
Wang/Tuttle (2009): 239.
725
See Favere-Marchesi/Emby (2005): 6; Tan (1995): 114f.
726
See European Commission (2010a): 12.
124
in the client, thereby inhibiting auditors’ ability to judge critically and act with objectivity and
integrity.727
Empirical research on the influence of familiarity issues on auditors’ ability to maintain pro-
fessional skepticism is very scant and less than consistent. Specifically, Bamber and Iyer
(2007) find that auditors’ familiarity with the client leads to greater cognitive identification
with the client and a stronger tendency to acquiesce to the client-preferred treatments, which
can be viewed as indicative of impaired skepticism. Kerler and Killough (2009), on the other
hand, find that auditors are capable of maintaining an appropriate level of professional skepti-
cism in a fraud assessment context regardless of their positive past experience with the client
and beliefs in the client’s integrity and trustworthiness.728 Additional research on the influ-
ence of familiarity and commitment on auditors’ ability and willingness to exercise sufficient
professional skepticism is clearly need in order to gain insights into this highly relevant but
very weekly explored topic.
In summary, client retention pressure and fee dependence, long audit tenure, and familiarity
with the client constitute incentives inherent to auditing which have the potential to compro-
mise auditors’ professional skepticism. Overall, professional skepticism can be viewed as a
costly attitude in terms of cognitive effort, amount of audit testing, audit budgets, adherence
to deadlines as well as client relations. These aspects are typically much more perceptible and
obvious to auditors and thus are likely to influence their judgments and decisions in a greater
extent compared to the rather abstract benefits of behaving in a professionally skeptical man-
ner.729 One possible approach to alleviate this problem would be to raise the cost associated
with dysfunctional unskeptical behavior. An alternative approach, and – as increasingly ar-
gued in the current auditing literature – a potentially more promising one, would be to design
and implement incentive frameworks which emphasize rewards over penalties and the quality
of judgment processes over results. Such frameworks are believed to enhance auditors’ moti-
vation to strive for high judgment quality and consistently apply professional skepticism.730
4.3.2.4 Judgment
The Nelson model posits that auditors’ skeptical judgments result from a complex and dy-
namic interplay of auditors’ knowledge, personality traits, and various incentives.731 In addi-
tion, evidential characteristics (e.g., nature, sequence, and source of evidence) and client-
specific factors (e.g., integrity, riskiness, preferences, and persuasion tactics of the client) also
727
See Bamber/Iyer (2007): 1f.; Doucet/Doucet (1996): 165; Johnstone/Sutton/Warfield (2001): 5;
Kerler/Killough (2009): 109f.; Murray (2012): 36.
728
See Bamber/Iyer (2007): 15-19; Kerler/Killough (2009): 109. However, it can be argued that past experience
with the client does not necessarily equate with long tenure. Hence, both studies potentially measure differ-
ent constructs.
729
See Brazel et al. (2013): 3f.
730
See Bazerman/Tenbrunsel (2011b): 64; Hurtt et al. (2013): 55; Peecher/Solomon/Trotman (2013).
731
See Nelson (2009): 5, 13, 17.
125
influence the degree of professional skepticism reflected in auditors’ judgments.732 Overall,
the exploration of the determinants and behavioral manifestations733 of auditors’ skeptical
judgments has attracted considerable research interest in auditing and has resulted in a wealth
of empirical studies.734
In his discussion of the role of judgment in the context of professional skepticism, Nelson
(2009) adopts a process view and particularly emphasizes a further less salient but not less
fundamental factor theorized to affect auditors’ skeptical judgment. He draws attention on the
various cognitive fallacies and traps found to significantly influence human information pro-
cessing in general and auditors’ judgment and decision making in particular. Drawing on psy-
chological and auditing research, Nelson (2009) indicates that cognitive fallacies, such as
availability effects, anchoring tendencies, motivated reasoning, confirmation proneness, and
recency effects, can systematically and adversely influence the manner in which auditors
gather, retrieve, process, evaluate, and integrate information and evidence, generate hypothe-
ses, and form beliefs. These cognitive traps can lead to judgments of impaired quality and
professional skepticism.735 Importantly, these cognitive biases operate at a subconscious level
and are thus potentially unaffected by skepticism-enhancing incentives.736 Recently, the
PCAOB has also drawn attention on the fact that unconscious cognitive biases can induce
auditors to search for, process, weight, and recall evidence and information in a manner corre-
sponding to the preferences and desired outcomes of the client rather than requirements of due
professional skepticism.737
Of the aforementioned cognitive biases, Nelson (2009) particularly emphasizes information
order (recency) effects as highly relevant to professional skepticism because of their pervasive
influence on auditors’ weighting and integration of evidence.738 The relevance of information
order effects in the context of professional skepticism is reinforced by the fact that the client
typically determines the order in which evidence is provided to the auditor and hence can ex-
ercise an undue influence on auditors’ views and judgments. In particular, the client may pre-
732
Note in passing that although the Nelson model entails evidential input as an immediate determinant of
skeptical judgment, Nelson (2009) does not explicitly discuss this component within his seminal paper. A
theoretical and empirical consideration of the influence of evidential characteristics and client-specific fea-
tures on auditors’ skeptical judgments is provided by Hurtt et al. (2013): 57-61.
733
Skeptical judgments have been operationalized in different ways in the existing auditing literature, e.g., as a
measurement of auditors’ ability to recognize irregularities and contradictions in the evidence presented, to
generate reasonable alternative explanations for a potential issue, to concentrate on fraud evidence, to ques-
tion clients’ explanations regarding unusual or unexpected evidence or circumstances, and to consider the
questionable actions of client’s management unethical. See Hurtt et al. (2013): 47, 51.
734
Indeed, Hurtt et al. (2013) identify a total of 42 studies on auditors’ skeptical judgment conducted within the
last decade. As a great deal of these studies have already been discussed in the context of auditors’
knowledge, traits, and incentives as determinants of professional skepticism, a repeated consideration within
this subsection is considered dispensable. For a detailed summary of auditing studies on skeptical judgment,
see Hurtt et al. (2013): 80-91.
735
See Nelson (2009): 13f. for a concise summary of selected cognitive biases and relevant studies on their
influence on auditors’ skeptical judgments.
736
See Bazerman et al. (2006): 45f.
737
See PCAOB (2012): 7.
738
See Nelson (2009): 14.
126
sent information in a manner inducing a strong contrast effect and a disproportionately strong
weighting of positive (client-favorable) information on the part of the auditor, so that the lat-
ter may form an unduly positive and optimistic beliefs and judgments regarding the entity’s
position, whereas under a normative viewpoint, a much more critical position might be ap-
propriate.
The present study explores the question whether the hypothesized influence of information
order effects on auditors’ judgments can be countervailed by their skeptical dispositions.
Hence, building on the Nelson model, the current study examines the influence of information
presentation order (evidential input739) and auditors’ dispositional skepticism (trait) on audi-
tors’ belief revisions (judgment) (link 2 and link 4, respectively). This aspect is elaborated in
Section 4.5.
4.3.2.5 Action
Following the Nelson model, skeptical judgment is a primary determinant of skeptical ac-
tion.740 Indeed, skeptical judgment is a necessary condition for skeptical behavior, as auditors
can only act with skepticism when they have recognized an existing issue, contradiction, or
risk factor in the first place.741 Skeptical actions arise when an auditor behaves in accordance
with the underlying skeptical judgment. Skeptical behavior involves, for instance, the conduct
of additional audit testing procedures or the confrontation of the client with critical questions
and findings.742 Importantly, while skeptical decisions naturally arise from skeptical judg-
ments, not every skeptical belief and thought initiates skeptical action in the real world. Ap-
parently, an individual threshold level must be reached first in order for a skeptical judgment
to trigger a skeptical action.743 This individual threshold depends on the incentives, traits,
knowledge aspects, and cognitive influences discussed above. In addition, client characteris-
tics (e.g., corporate governance quality, management integrity, risk characteristics) may also
influence auditors’ skeptical behavior.744
Overall, research on skeptical action has not been as voluminous and diverse as that on skep-
tical judgment, probably because as a major input to skeptical action, skeptical judgment
needs to be profoundly analyzed and well understood before valid inferences about skeptical
739
Recall that Nelson (2009) does not elaborate on the issue of evidential input as a determinant of auditors’
professional skepticism. Rather, he focuses on knowledge, traits, incentives, judgment, and action. Insights
into the importance of evidential characteristics in the context of auditor judgment were provided in Section
2 and Section 3. A consideration of evidential characteristics in the context of auditors’ professional skepti-
cism is provided by Hurtt et al. (2013): 57f.
740
See Nelson (2009): 5, 15.
741
See Hurtt et al. (2013): 63.
742
Hurtt (2010): 165 identifies the following major aspects of skeptical behavior: intensive information search,
high detection of contradictions and inconsistencies in arguments and evidence, ability to generate alterna-
tive explanations, and pronounced scrutiny of source credibility when facing interpersonal information.
743
See Shaub/Lawrence (1999): 63; Shaub/Lawrence (2002): 170.
744
See Hurtt et al. (2013): 65-67; Nelson (2009): 5, 15f.
127
action can be made. In light of the fact that audit regulators have largely focused on skeptical
action in their inspections, future research on auditors’ skeptical behavior is clearly needed.745
In summary, Nelson’s (2009) conceptual analysis of the key determinants of professional
skepticism in auditing represents an important step toward the achievement of a more precise,
rich, and structured view of this fundamental concept. A common finding with regard to vir-
tually all individual determinants is that the results of the existing empirical studies are rather
inconsistent and do not allow for the derivation of generalizable and straightforward infer-
ences regarding the influence of the different determinants on auditors’ professional skepti-
cism. This may be attributed to the fact that – as shown in Section 4.2.3 – the same label
(“professional skepticism”) has been attached to essentially different constructs in prior audit-
ing research.746 Future research would greatly benefit from the development of a consistent
definition of and measurement instrument for professional skepticism.
Subsequently, Hurtt’s (2010) conceptualization of professional skepticism as comprising a
trait dimension and the measurement scale she developed to capture trait skepticism are pre-
sented and discussed. Hurtt’s (2010) notion of trait professional skepticism and the Hurtt
scale are foundational for the present study.
4.4 Hurtt’s Notion of Trait Professional Skepticism and the Hurtt Scale
The considerations provided in the preceding sections reveal a remarkable insight: Although
broadly acknowledged as a cornerstone of auditing, the concept of professional skepticism is
still “a topic where there are more questions than answers.”747 This contention is less than
surprising in light of the fact that until very recently, there has been virtually no systematic
and profound research on the nature, underpinnings, and dimensions of professional skepti-
cism. Consequently, precise measurement approaches to professional skepticism have also
been lacking.748 Instead of aiming to better understand and precisely operationalize profes-
sional skepticism, prior auditing research has readily employed existing one-dimensional
scales designed to measure some related constructs such as (dis)trust, suspicion, or independ-
ence.749
In 2010, R. Kathy Hurtt published a seminal paper on professional skepticism in which she
not only provides some very essential and valuable insights into the nature of skepticism but
also presents a scale specifically developed to ex ante measure an individual’s level of innate
745
See Hurtt et al. (2013): 71. They identify a total of 11 studies exploring auditors’ skeptical action conducted
within the last decade. Again, some of these studies have already been discussed in the context of auditors’
knowledge, traits, and incentives as determinants of professional skepticism. For a detailed summary of the
relevant studies on skeptical action, see Hurtt et al. (2013): 91-94.
746
See Hurtt (2010): 150.
747
Hurtt et al. (2013): 72.
748
See Doucet/Doucet (1996): 158.
749
See Hurtt (2010): 149f.
128
professional skepticism. This research instrument is designed on the basis of individual char-
acteristics indicated as relevant to skepticism in auditing standards, psychology, philosophy,
and consumer behavior research.750 The consideration of interdisciplinary views of skepticism
aims to enrich the mainstream auditing understanding of professional skepticism. With the
publication of her influential research work, Hurtt (2010) has significantly advanced the aca-
demic state of knowledge and insight into the fundamental concept of professional skepti-
cism.
In essence, Hurtt (2010) views professional skepticism as a multifaceted individual character-
istic. According to this notion, professional skepticism involves both a trait dimension as well
as a state dimension.751 As discussed in Section 2.4.2.3, traits are facets defining human per-
sonality. They represent enduring and temporally stable internal attributes which characterize
an individual across various situations. In contrast, states are alterable, temporary, and occa-
sion-specific conditions induced by contextual factors. They characterize a person in a certain
situation. Overall, traits and states are two interrelated behavioral dispositions which combine
to influence human cognition, emotion, thoughts, beliefs, attitudes, judgments, and actions.752
Consequently, trait professional skepticism represents a temporally stable and cross-
situationally consistent personality attribute of an individual auditor, whereas state profes-
sional skepticism reflects a momentary condition caused by situational factors. Hurtt (2010)
posits that both trait and state skepticism combine to influence an auditor’s skeptical mindset
which then affects his/her behavior in the manner depicted in Figure 6. According to this
view, it is the complex and dynamic interplay between the person and the situation, which
determines auditors’ skeptical behavior.
State Skepticism
Skepticism Scale
Moderating Variables
(e.g., engagement cir-
cumstances)
Figure 6: Hurtt’s Notion of Professional Skepticism (Source: Hurtt (2010): 150)
This model of interaction between trait and state professional skepticism as determinants of
auditors’ skeptical behavior, although different in focus and emphasis, is compatible with the
750
See Hurtt (2010): 150.
751
See Hurtt (2010): 150. Note that the notion of skepticism as a multidimensional construct is well-established
in consumer behavior research. See Boush/Friestad/Rose (1994): 170; Tan (2002): 60f.
752
See Buss (1989): 1378; Church (2000): 651f.; Eid/Luhmann (2012): 2323; Robinson/Curtis/Robertson
(2013): 6f.; Schmitt/Steyer (1993): 520, 522; Steyer/Schmitt/Eid (1999): 390f.
129
Nelson model which posits that traits such as dispositional skepticism and the balance of in-
centives and factors of the audit environment determine the extent to which auditors’ judg-
ments and actions reflect professional skepticism.
The main focus of Hurtt’s (2010) research is on trait professional skepticism. She identifies
the following six characteristics that constitute trait professional skepticism: questioning
mind, suspension of judgment, search for knowledge, interpersonal understanding, self-
esteem, and autonomy.753 The first three of the professional skepticism characteristics listed
above involve a clear focus on audit evidence. They implicate an individual auditor’s affinity
and willingness to search for and thoroughly examine sufficient amounts of persuasive audit
evidence before arriving at a judgment and/or a decision. The last three characteristics address
some other aspects that are relevant in the context of judgment and decision making in audit-
ing: an auditor’s ability to critically consider the competence, motivation, and incentives of
client’s management and personnel (interpersonal understanding) as well as auditor’s capabil-
ity and courage to act on the evidence obtained, even if this would displease the client (self-
esteem and autonomy).754 Subsequently, the individual constituent elements of trait profes-
sional skepticism are presented and discussed.
753
See Hurtt (2010): 151.
754
See Hurtt (2010): 152. Note that the classification of the skepticism components as evidence-related and
non-evidence-related (other) characteristics is somewhat arbitrary, as will be shown in Section 4.4.2.4. Basi-
cally, all components are closely interrelated and only their joint consideration provides a full picture of trait
professional skepticism.
755
The considerations within this section follow and extend the arguments provided by Hurtt (2010): 152f.
756
See AUASB (2012): 3.
757
Chambers Universal Learners’ Dictionary (1994): 647.
130
stance, Shaub (1996) defines a skeptic as an individual “who instinctively or habitually
doubts, questions, or disagrees with assertions or generally accepted conclusions.”758 Simi-
larly, Murray (2012) argues that doubt is the very essence of skepticism and contends that
“doubt stimulates informed challenge and inquiry” and “professional scepticism encourages a
structured consideration of an alternative point of view.”759 The CAQ (2010), on the other
hand, argues that people generally possess a natural predisposition to trust and are inclined to
believe in the integrity and soundness of the institutions and organizations with which they
are affiliated. This trusting predisposition, however, collides with requirements of vigilance
and professional skepticism. Consequently, given the trusting human nature, it is even more
crucial to consciously maintain a critical and questioning mindset and adopt an attitude of
skepticism.760
The questioning mind dimension of professional skepticism relates to the psychological con-
cept of tolerance for ambiguity discussed in Section 2.4.2.3.5. Specifically, individuals pos-
sessing intolerance for ambiguity are expected to cope with uncertainty in complex problem
situations by adopting a questioning stance and seeking for a more solid and extensive eviden-
tial basis on which to build their judgments and decisions.761 In social psychology, the con-
cept of suspicion corroborates the questioning mind element of skepticism. Specifically, sus-
picion implies refraining from taking claims or behaviors without further questioning and
entertaining manifold, reasonably rival arguments and explanations.762
As discussed in Section 4.2.1, the notion of raising questions and doubts as an essential con-
stituent part of skepticism is well-established and broadly accepted in philosophy. This notion
is reflected in the etymology of the term “skepticism” which, as might be recalled, means “to
consider”, “to examine”, “to inquire”.763 Skepticism involves questioning the appropriateness
of one’s grounds for holding certain beliefs.764 Importantly, this questioning stance does not
merely concern beliefs and positions held by others; it also relates to the skeptic’s own asser-
tions and judgments which are unexceptionally subjected to doubt.765 Overall, to call things
into question and to raise doubts is considered the quintessence of skepticism.766
Consistently, the view that skepticism involves questioning and doubting is also present in the
field of media and consumer behavior research. Instances for studies dealing with this aspect
758
Shaub (1996): 155. For a more detailed discussion of the distrust view of professional skepticism in the
auditing literature, refer to Section 4.2.3.2.
759
Murray (2012): 36.
760
See CAQ (2010): 19.
761
See McGhee/Shields/Birnberg (1978): 683.
762
See Hilton/Fein/Miller (1993): 503. Note that the authors provide a very multifaceted description of the
concept of suspicion which also involves suspension of judgment and interpersonal understanding aspects.
763
See Kurtz (1992): 21; Stough (1969): 3.
764
See Strawson (1985): 2.
765
See McGinn (1989): 6; Strawson (1985): 2.
766
See Bunge (1991): 131; Fogelin (1994): 3.
131
include Cozzens and Contractor (1987), Ford, Smith and Swasy (1990), Koslow (2000),
Mangleburg and Bristol (1998) as well as Obermiller and Spangenberg (1998).767
In summary, the notion that a questioning mindset is a fundamental facet of skepticism is
widely and interdisciplinarily acknowledged. The critical importance of a questioning mindset
is succinctly illustrated by Edward Hodnett’s argument that only by asking the right ques-
tions, the right answers can be obtained.768
The next integral part of professional skepticism is the feature of suspension of judgment. In
general terms, professional skepticism involves withholding judgment and conclusion until
there is an appropriate amount of audit evidence to corroborate or refute the assertions under
consideration. Similarly to questioning mind, the suspension of judgment component is also
reflected in the auditing standards. Although the latter do not explicitly use the term “suspen-
sion of judgment”, the strong requirements regarding the nature and extent of audit evidence
needed for well-founded judgments and for obtaining reasonable assurance hint at the im-
portance of withholding judgment until extensive persuasive evidence is obtained. Specifical-
ly, ISA 200.13(l) and SAS 1 (AU 230.7) both describe professional skepticism as involving a
“critical assessment of audit evidence”. In addition, ISA 200.A22 as well as SAS 1 (AU
230.9) entail a requirement for auditors not to be “satisfied with less than persuasive evi-
dence”. Hence, auditing standards make clear statements regarding the essentiality of obtain-
ing reasonable, sufficient, and credible evidence as a necessary condition for auditors’ judg-
ment and decision making. The necessity of suspending judgment until reaching a reasonable
level of persuasive audit evidence as an integral part of effective auditing has also been rec-
ognized in the relevant auditing literature (e.g., Chazen and Solomon (1975), Mautz and
Sharaf (1961)).770
In psychology, the definition of suspicion as “a state of suspended judgment”771 provides sup-
port for the notion of withholding judgment as a constituent part of skepticism. The suspen-
sion of judgment feature is further closely associated with the psychological concept of toler-
ance for ambiguity. In particular, ambiguity intolerant individuals are theorized to strive for
avoiding or reducing uncertainty in complex problem domains by suspending judgment until
the informational basis allows for a well-founded judgment to be made.772 A further relevant
psychological concept in this context is need for cognitive closure. The latter reflects an indi-
vidual’s desire to quickly come to a definite judgment or conclusion on a certain issue in or-
767
For a more detailed consideration of this stream of research, see Hurtt (2010): 152.
768
This famous contention of the British poet was found in Fornelli/Desmond (2011).
769
The considerations within this section follow and extend the contentions provided by Hurtt (2010): 153.
770
See Chazen/Solomon (1975): 68; Mautz/Sharaf (1961): 34.
771
Hilton/Fein/Miller (1993): 502.
772
See McGhee/Shields/Birnberg (1978): 683. For a more detailed consideration of the tolerance for ambiguity
construct, refer to Section 2.4.2.3.5.
132
der to dispose of ambiguity and confusion. The need for cognitive closure involves an indi-
vidual’s predisposition to lack interest in new (additional) information and to be satisfied with
any evidence in arriving at a judgment, not necessarily the most adequate, relevant, and per-
suasive one. Consequently, the need for cognitive closure is generally associated with limited
information search as well as an increased tendency to apply simplified heuristic procedures
in information processing, problem solving, judgment, and decision making.773 With regard to
auditing, high levels of need for closure may induce auditors to prematurely discontinue the
search for alternative explanations and additional audit evidence. In contrast, low disposition-
al need for closure may stimulate the consideration of alternative hypotheses and explana-
tions, critical inquiry, extended information search, and withholding belief until the informa-
tional basis warrants a well-founded judgment.774
Further support for the suspension of judgment characteristic as a central component of skep-
ticism is provided in the field of philosophy. As mentioned in Section 4.2.1, suspension of
judgment and belief (epoché) is the quintessence of Pyrrhonian skepticism.775 More contem-
porary philosophical views likewise corroborate the fundamental interrelation between sus-
pension of judgment and skepticism as indicated by the insight of Kurtz (1992) that “suspen-
sion of judgments (...) is a necessary ingredient of skeptical inquiry”776. Similar contentions
are provided by Bunge (1991) and Naess (1968).777
Overall, the multidisciplinary overview provided above clearly indicates a strong relationship
between suspension of judgment and (professional) skepticism. Importantly, the suspension of
judgment aspect does not imply a skeptic’s general aversion to form beliefs and judgments.
Rather, it implies that a skeptic takes his/her time and exerts effort in judgment and decision
making to ensure that beliefs and conclusions are well-founded, i.e., based on sufficient per-
suasive evidence.
773
See Bailey/Daily/Phillips (2011): 27f.; Kruglanski (1989): 14; Suedfeld/Tetlock (2001): 296f.; Van
Hiel/Mervielde (2003): 560.
774
See Bailey/Daily/Phillips (2011): 28f. The authors find that auditors have significantly lower levels of distri-
butional need for cognitive closure as compared with student subjects used in psychological research, and
that low need for closure leads to a greater amount and higher quality of hypotheses generated as well as
more deliberate information processing.
775
See Lammenranta (2008): 13.
776
Kurtz (1992): 41.
777
See Bunge (1991): 131; Naess (1968): 28.
778
The considerations within this section follow and extend the arguments provided by Hurtt (2010): 153f.
133
acteristic of search for knowledge is not explicitly discussed in auditing standards, the rele-
vant literature clearly recognizes it as an important facet of skepticism.
In contrast to questioning mind which rather involves doubt or disbelief nuances, the charac-
teristic of search for knowledge has a sense of general inquisitiveness, interest, and curiosity.
Overall, an individual’s skepticism is not merely restricted to raising doubts, making critical
inquiries, and postponing judgment until sufficient evidence is obtained. Beyond these fea-
tures, the quest for knowledge is an important aspect of skepticism.
In the auditing literature, support for this view has been provided by Chazen and Solomon
(1975) who discuss skepticism along with curiosity, inquisitiveness, and imagination as es-
sential skills in the auditing profession.779 Similarly, the Canadian Institute of Chartered Ac-
countants (CICA) considers innate skepticism and inquisitiveness as critical factors for being
a good professional auditor.780 The search for knowledge facet of skepticism has also been
corroborated by Love and Lawson (2009) who describe professional skepticism as an “in-
quisitive attitude”781.
The search for knowledge dimension of skepticism is also well recognized within the field of
philosophy. In particular, Naess (1968) characterizes a skeptic as an open-minded inquirer
who is always willing to explore new perspectives and notions.782 Similarly, Popkin and
Stroll (2002) argue that skepticism involves a deep look beyond the obvious.783 Johnson
(1978) contends that skepticism comprises the quest for knowledge for knowledge’s sake.784
Kurtz (2001) argues that skeptical inquiry is foundational for any judgment and quest for in-
sight.785 Bunge (1991) argues that skepticism and neophobia (the fear of the new, including
new experiences, arguments, and knowledge) are mutually exclusive concepts.786
In psychology, the search for knowledge characteristic has been explored within the concept
of curiosity. Curiosity reflects an individual’s eagerness for knowledge. It promotes explora-
tion, insight, intellectual growth, and the arousal of individual interests.787
Overall, questioning mind, suspension of judgment, and search for knowledge are important
facets of professional skepticism as indicated in the auditing standards and the relevant inter-
disciplinary literature. Collectively, these features reflect auditors’ predisposition to maintain
a critical attitude and to gather, process, and evaluate sufficient amounts of persuasive audit
evidence before they believe, judge, or decide. Subsequently, the non-evidence related aspects
of skepticism – interpersonal understanding, self-esteem, and autonomy – are discussed.
779
See Chazen/Solomon (1975): 68f. The importance of being curious in auditing has also been recognized by
Mautz/Sharaf (1961): 19.
780
See CICA (1995): 35, quoted in Dennis (2013): 29.
781
Love/Lawson (2009): 32.
782
See Naess (1968): 5, 26.
783
See Popkin/Stroll (2002): 36.
784
See Johnson (1978): 14.
785
See Kurtz (2001): 45.
786
See Bunge (1991): 131.
787
For a number of relevant references along these lines, see Litman (2005): 793 and Litman/Silvia (2006): 318.
134
4.4.2.4 Interpersonal Understanding788
Although Hurtt (2010) does not classify interpersonal understanding among the evidence-
related professional skepticism characteristics, she admits that it is also a considerable aspect
of vigilantly processing and evaluating evidence, as scrutinizing and comprehending the mo-
tivation, objectivity, honesty, and integrity of the persons who provide information is essential
for maintaining a skeptical attitude. ISA 200.A22 and SAS 122 (AU-C 200.A26) both require
that “a belief that management and those charged with governance are honest and have in-
tegrity does not relieve the auditor of the need to maintain professional skepticism”.789 Fur-
thermore, ISA 240.A1 and SAS 99 (AU 316.85.A2) recognize that a number of incentives,
pressures, and opportunities (reinforced by certain ethical values and mindsets) exist that can
induce client’s management and personnel to provide biased evidence and/or to involve in
fraudulent activities. It is important to be aware of these threats and mind the personal motiva-
tion of those charged with governance.
The interpersonal understanding aspect of skepticism is also reflected in the relevant auditing
literature. For instance, in their discussion of professional skepticism in auditing, Shaub and
Lawrence (1996) acknowledge the importance for an auditor to mind the potential incentives
and motives of the client.790 Similarly, the Center for Audit Quality (CAQ) highlights the ne-
cessity to consider the personal dimension while communicating with the client. In particular,
the CAQ suggests that scrutinizing the body language of the interlocutor may provide valua-
ble insights beyond the verbal content of information.791 Accordingly, a mismatch between
the verbal and the non-verbal content of the information provided by the client may be an
alarming issue which possibly hints at bias or irregularity and thus necessitates a pronounced-
ly critical consideration on the part of the auditor.
As with the other skepticism characteristics, there is multidisciplinary support for the inter-
personal understanding dimension of skepticism. Particularly, consumer behavior research
identifies that advertising skepticism does not only comprise challenging advertising promises
and claims but also involves a scrutiny of the incentives and targets of the advertisers. 792 In
psychology, social competence (encompassing social skills and social interaction ability) has
been acknowledged as an important ingredient of skilful problem solving, goal achievement,
and interpersonal relationships.793 Following this line of thought, social competence can also
be expected to reinforce the individual ability to identify errors and bias in the arguments of
the others. In addition, the concept of suspicion defined as questioning “the motives that un-
788
The considerations within this section follow and extend the contentions provided by Hurtt (2010): 154.
789
Despite these normative requirements, some empirical studies demonstrate that auditors are not particularly
sensitive to client integrity. See, e.g., Bernardi (1994): 75; Kaplan/Reckers (1984): 13. For contrary find-
ings, see Goodwin (1999): 13.
790
See Shaub/Lawrence (1996): 125.
791
See CAQ (2010): 24f.
792
See Obermiller/Spangenberg (1998): 160.
793
See Wentzel (2005): 280. An individual’s social competence is typically measured via the Texas Social Be-
havior Inventory Scale developed by Helmreich/Stapp (1974).
135
derlie a person’s behavior or (..) the genuineness of that behavior”794 also relates to the inter-
personal understanding dimension of skepticism. Research findings indicate that individuals
with disposition toward suspicion and distrust generally tend to more elaborative and critical
information processing.795 Finally, the philosophical understanding of skepticism as an in-
depth examination and inquiry also (indirectly) involves the notion of scrutinizing the person-
al incentives, motives, and objectives underlying a certain claim or assertion.796
Overall, the tendency to consider information in light of the personal motivation of the indi-
vidual from whom it is obtained, along with an attitude of ongoing questioning, withholding
judgment, and questing for knowledge, equips the auditor with the flexibility and criticalness
needed to effectively process information and assess evidence provided by the client. The last
two components of professional skepticism, autonomy and self-esteem, are more general in
nature and capture basic personal characteristics of an auditor which are foundational for a
skeptical mindset. Subsequently, these two aspects of skepticism are discussed.
4.4.2.5 Autonomy797
The term “autonomous” stems from the Greek words “auto” (i.e., self) and “nomus” (i.e., law)
and literally means “self-governed”, “not controlled by outside forces”.798 Hence, the attribute
of autonomy reflects an individual’s independence, self-reliance, and ability to act self-
containedly instead of allowing external influences, opinions, and claims to unduly affect
his/her beliefs and position. This aspect is critical to skepticism.
The autonomy aspect of professional skepticism is normatively anchored in the independence
requirements of ISA 200.14 and AICPA ET 101.1 as well as in the definition of independence
of mind provided in the IESBA Code 290.6(a) as a “state of mind that permits the expression
of a conclusion without being affected by influences that compromise professional judgment,
thereby allowing an individual to act with integrity and exercise objectivity and professional
skepticism”.
In the auditing literature, corroboration for the importance of auditors’ autonomy is provided
by Mautz and Sharaf (1961) who emphasize the essentiality of professional courage, pru-
dence, and impartiality throughout the audit.799 In addition, Bobek, Hageman and Radtke
(2013a) demonstrate that autonomy and client advocacy (i.e., the opposite of skepticism) are
significantly inversely related.800
794
Hilton/Fein/Miller (1993): 502.
795
See Rose/Rose (2003): 313; Schul/Burnstein/Bardi (1996): 228.
796
In this context, Hurtt (2010): 154 refers to the writings of Burnyeat (1983), Hallie (1985), Hookway (1990),
Johnson (1978), Kurtz (1992), McGinn (1989), and Popkin (1979).
797
Note that the considerations within this section are differently nuanced as compared with the arguments
provided by Hurtt (2010): 154f.
798
See Moors/De Houwer (2007): 29.
799
See Mautz/Sharaf (1961): 35, 136. This aspect is also acknowledged by Hurtt (2010): 154f.
800
See Bobek/Hageman/Radtke (2013a): 12.
136
The notion of autonomy as a constituent part of skepticism is also recognized in the field of
philosophy where autonomy has been characterized through freedom of the will and sover-
eignty (absence of external influence) in belief, thought, and action. These attributes are also
critical to skepticism.801 In addition, consumer behavior research suggests that susceptibility
to interpersonal influence, i.e., the inverse of autonomy, is negatively associated with skepti-
cism.802
4.4.2.6 Self-Esteem803
Closely related to autonomy is the attribute of self-esteem which is the sixth and last constitu-
ent part of skepticism. Broadly stated, for maintaining objectivity, independence of mind, and
fortitude, a certain degree of self-esteem is necessary in order to resist persuasion attempts
and pressures.
Auditing standards do not explicitly address the self-esteem dimension of skepticism. Howev-
er, it can be argued that assertiveness and self-esteem are among the manifold central charac-
teristics required of auditors in order to fulfil their professional duties. This notion has been
corroborated by Chazen and Solomon (1975) who consider skepticism and the ability to resist
pressures from client’s management as critical ingredients of audit quality.804 This view is
also shared by Craig Fischer, Member of the New Zeeland Auditing and Assurance Standards
Board, who aptly argues that professional skepticism “is about having the self-confidence and
strength of character to retain an enquiring mind and to resist the temptation to just accept
the easy answer – especially when management presents it to you on a plate.”805
Support for the self-esteem aspect of skepticism can also be found in the field of philosophy
where confidence has been acknowledged as essential for skeptical inquiry and self-esteem
has been described as a state of inner peace and tranquility. This state of mind reflects the
ideal of philosophical skepticism – ataraxia (calmness of the mind).806
In psychology and consumer behavior research, self-esteem has been described as involving
the attributes of self-confidence, self-worth, and faith in the own skills, abilities, and judg-
ments. These characteristics are inversely associated with conformity and convincibility.
Hence, skepticism encompasses the confidence and courage to challenge others’ claims and
conclusions and to withstand persuasive attempts rather than promptly believe whatever is
stated.807
Consequently, self-esteem can be viewed as a key characteristic of a skeptical auditor, as it
enables him/her to maintain a critical mindset and challenge the client’s arguments and asser-
801
See Dworkin (1988): 6; Hookway (1990): 146f.
802
See Boush/Friestad/Rose (1994): 167.
803
The considerations within this section follow and extend the contentions provided by Hurtt (2010): 155.
804
See Chazen/Solomon (1975): 68.
805
Fischer, quoted in Murray (2012): 37.
806
See Burnyeat (1983): 121; Hookway (1990): 234; Lom (2001): 33.
807
See Boush/Friestad/Rose (1994): 167.
137
tions. Without a sufficient level of self-esteem, an auditor is likely to readily accept the posi-
tion of the client and ignore remaining doubts or unanswered questions, which may put at risk
the effectiveness of the audit. Hence, although not explicitly mentioned in the auditing stand-
ards, self-esteem is a crucial feature in auditing in general and in the context of professional
skepticism in particular.
Collectively, these six attributes constitute trait professional skepticism and build the theoreti-
cal foundation of the Hurtt scale. The scale is of fundamental importance to the present study
because it is the instrument employed to capture and measure auditors’ dispositional skepti-
cism. Following, the genesis of the Hurtt scale is outlined.
Based on the broad interdisciplinary theoretical foundation described above and following the
procedure suggested by Churchill (1979), Hurtt (2010) employed a multi-stage iterative ap-
proach in order to develop a scale of high reliability, validity, and quality. Subsequently, the
scale development and validation procedure employed by Hurtt (2010) is described.
808
Specifically, the following psychological scales and constructs served as the base frame for the development
of the initial item pool: To capture the characteristic of a questioning mind, Hurtt employed the State-Trait
Personality Inventory (Spielberger (1995)) as well as the Interpersonal Trust, Trustworthiness and Gullibil-
ity Scale (Rotter (1980)). The suspension of judgment component of professional skepticism was reflected
by the Ambiguity Tolerance Scale (MacDonald (1970)) and the Need for Cognitive Closure Scale
(Kruglanski (1990)). The search for knowledge dimension of skepticism was taken into account by using the
State-Trait Curiosity Inventory (Naylor (1981)) as well as the State-Trait Personality Inventory (Spielberger
(1995)). The interpersonal understanding attribute was also captured through the State-Trait Curiosity Inven-
tory (Naylor (1981)). The items related to self-esteem were generated on the basis of the relevant literature
(Forsman/Johnson (1996); Fleming/Courtney (1984)) as well as via the Texas Social Behavior Inventory
(Helmreich/Stapp (1974)). The last attribute of professional skepticism, autonomy, was reflected using the
Locus of Control Scale (Rotter (1966)). See Hurtt (2010): 158f. Note, however, that Hurtt (2010) does not
give reasons for the particular scale choices she made. This gives rise to the question by which criteria she
decided to employ some scales while leaving other potentially appropriate measures like the Interpersonal
Trust Scale (Rotter (1967)) or the Self-Esteem Scale (Rosenberg (1965)) disregarded.
809
See Hurtt (2010): 151.
810
Broadly stated, Likert scales, also labeled “summated-rating scales”, measure the degree of agreement with
an issue or item stem. See Litwin (1995): 49; Wrightsman (1974): 49.
138
number of response points was chosen in order to gain valuable insights regarding the direc-
tion of participants’ responses.811
In order to avoid construct underrepresentation, i.e., a too narrow specification of a construct,
which may ultimately put at risk construct validity, Hurtt (2010) requested 25 experienced
auditors to provide their personal views and beliefs about what constitutes professional skep-
ticism and what is its counterpart. She then compared the responses obtained from the audit
professionals with the theoretically derived constructs and scale items to determine whether
some important aspects and dimensions of skepticism had been missing in her construct and
to assure that her theoretical vision of professional skepticism corresponded to the view of
professional auditors. The comparison did not reveal any new aspects or underpinnings of
professional skepticism beyond the ones already considered by Hurtt (2010).812
Pretest
Subsequent to the construct specification and the generation of the initial item pool, Hurtt
(2010) asked three academic experts with rich professional background to review the scale
items for consistency, plausibility, relevance, and reasonableness of classification of cues to
skepticism subconstructs. This procedure ultimately aimed at enhancing the content validity
of the scale. Based on the experts’ responses, Hurtt (2010) excluded items classified as irrele-
vant, redundant or inconsistent, resulting in a 49-item scale813 which was subjected to a pre-
test with 89 business students. The administration of the scale to student subjects is justified
by the fact that the scale is designed to measure an individual disposition and hence is like-
wise applicable to professional and non-professionals.814
Hurtt (2010) analyzed the pretest results employing factor analysis in order to detect and
eliminate items that did not correlate with (load on) the theorized constructs. Twenty-six such
items were identified, resulting in a reduced scale containing 23 cues. The remaining items
exhibited good factor loadings of 0.70 or higher on the six identified skepticism subcon-
structs, where values exceeding 0.65 are considered substantial.815 The 23-item scale yielded
a Cronbach’s alpha value of 0.82, where values above 0.80 are considered very good. 816 In
essence, Cronbach’s alpha represents a measure of the internal consistency and homogeneity
811
See Hurtt (2010): 156. For a critical note on the use of response scales involving an even number of scale
points and thus forcing participants to pick a side, see Peytcheva (2014): 38f.
812
See Hurtt (2010): 156f.
813
Obviously, the review process conducted by the tree academic members has had a great influence on the
final item set, as only 49 of the initially generated 220 cues remained part of the scale after the academic ex-
perts’ feedback. That is, more than three-fourth of the originally identified cues have been eliminated in the
course of the experts’ content review. While it is plausible that a scale containing 220 items is way too ex-
tensive and thus hardly feasible and practicable, it appears that some additional validating procedures and
care could have been undertaken in this review process in order to avoid subjectivity and arbitrariness
claims.
814
See Hurtt (2010): 157.
815
See DeVellis (2003): 126.
816
See DeVellis (2003): 95f.; Hurtt (2010): 157. DeVellis (2003): 95f. indicates the following ranges for alpha
values: between 0.60 and 0.65, undesirable; between 0.65 and 0.70, minimally acceptable; between 0.70 and
0.80, respectable; and between 0.80 and 0.90, very good.
139
of a scale and is thus a reliability (and quality) indicator. It provides clues about the extent to
which the items of a psychometric test complement each other in reflecting and measuring the
same underlying (latent) construct.817
As indicated by DeVellis (2003), a replication of the procedure described above on an inde-
pendent sample is essential for ensuring that “the results obtained were not one-time chance
occurrence.”818 Before proceeding to the replication stage, Hurtt (2010) developed a number
of additional items for the six theoretical subconstructs of professional skepticism, as some of
the items had apparently been confusing to the participants in the pretest. This updating pro-
cedure resulted in a 40-item refined measure of trait professional skepticism. Note, however,
that the ultimate aim of Hurtt (2010) was to arrive at a 30-item scale in which each skepticism
factor is reflected by an equal number of cues (i.e., five cues per scepticism subconstruct). She
considers the equal item weighting reasonable and feasible, since no particular theoretical or
empirical evidence exists in the relevant literature suggesting that any single dimension of
skepticism is more substantial that the other.819 Hurtt (2010) intentionally subjected more
items to pilot testing (40) than the amount desired for the final scale (30) in order to account
for the possibility of inconsistent, irrelevant, or invalid items.820
817
See DeVellis (2003): 28; Litwin (1995): 21, 24. For a detailed (and more formal) consideration of this coeffi-
cient, refer to Cronbach (1951).
818
DeVellis (2003): 136.
819
Note, however, that Hurtt (2010) does not provide any particular arguments for the desired scale length of
30 items.
820
See Hurtt (2010): 157.
821
See Hurtt (2010): 157, 159.
822
See Hurtt (2010): 161.
823
See DeVellis (2003): 43f.; Litwin (1995): 8.
140
who participated in the pilot study to complete the scale824 a second time. She obtained 147
responses, of which only 92 were complete and usable for the subsequent analysis. Overall,
the results indicated a significant and very strong correlation between subjects’ total skepti-
cism scores on the first and second test (r = 0.89, p < 0.001).825
In the course of the re-test, the 15 additional cues were scrutinized in terms of the appropri-
ateness of their factor loadings on the six skepticism components. Based on the results of this
factor analysis, Hurtt (2010) picked the five “missing” items with the best factor loadings to
complement the established core item pool of 25 items and to finally arrive at the desired 30-
item scale with equally represented subconstructs. An individual’s score on this (final) scale
can range from 30 to 180 points. The students participating in the re-test exhibited a mean
score of 132.7 with a standard deviation of 15.9. Their scores fell within the range from 77 to
175.826
Validity Test
At the end of the instrument development process, the final scale was administrated to prac-
ticing auditors from a Big 4 auditing company in order to test the validity of the scale in a
professional setting. On average, the auditors participating in the survey reached a score of
138.6 points on the skepticism scale, with results ranging from 111 to 173 and a standard de-
viation of 12.6. Compared to the student sample, professional auditors exhibited a slightly
higher (5.9 points) mean score on the scale indicative of higher levels of trait professional
skepticism. In addition, the responses within the auditor sample did not vary that strongly as
compared with the student group. In consistency with the results obtained with the student
subjects, an individual’s score on the scale was not found to be driven by any of his/her de-
mographical characteristics (gender, age, experience, or qualification). The mean age of the
auditors who participated in the study was 28.2 years; the average audit experience amounted
to 58 months. The majority of the subjects were male (n = 112) and possessed a CPA title (n =
179).827
With regard to the internal consistency and validity of the scale, this testing loop also yielded
good results (Cronbach’s alpha of 0.86). The results of the factor analysis conducted at this
administration level revealed that the final 30 items loaded on the six identified skepticism
subconstructs, but two cues did not correspond to the factors they were hypothesized to meas-
ure. In addition, one of the items in question yielded a factor loading of 0.49, which is mini-
mally below the critical benchmark of 0.50 suggested in the relevant literature.828 Hurtt
824
Note that Hurtt (2010): 161 does not clarify which version of the scale the participants at the re-test stage
actually completed. Table 2 on page 160 in Hurtt’s (2010) paper suggests that the students completed the re-
updated 40-item scale which contains 15 new items as compared to the pilot test scale. Hence, the compari-
son of the inter-temporal stability of subjects’ scores can only concern the 25 core items which proved ap-
propriately loading on the six skepticism dimensions at the pilot test stage. For the additional cues, no basis
for comparison appears to exist.
825
See Fink (1995): 36; Hurtt (2010): 161; Litwin (1995): 8.
826
See Hurtt (2010): 161.
827
See Hurtt (2010): 162.
828
See DeVellis (2003): 95f.
141
(2010) does not interpret this matter as alarming with regard to the validity of the scale as a
whole. She argues that the internal consistency of the items is largely reasonable, so that the
improper factor loadings of the two cues might be attributed to the particular sample popula-
tion rather than to a general deficiency of the scale.829
As a final validation procedure, Hurtt (2010) re-tested the scale’s temporal stability applying
the test-retest technique to the auditor sample. On average, the participants completed the re-
peated study 22 days after the first run, with a range from 2 to 56 days and a modal value of
15 days. Responses to the repeated skepticism survey were obtained from 88 auditors who
achieved a mean skepticism score of 135.4 points, with a range from 105 to 177 and a stand-
ard deviation of 14.7. These results are very similar to subjects’ responses on the previous
administration of the instrument. The correlation coefficient between the test and re-test skep-
ticism scores on the scale amounted to 0.85 (p < 0.01). In addition, the internal consistency
between the test and re-test scores was also very good, as indicated by the obtained
Cronbach’s alpha value of 0.91. Overall, these findings provide a substantial support for the
validity of the Hurtt scale as a theoretically well-founded psychometric instrument with a rea-
sonable inter-item and temporal stability.830
Nevertheless, some caveats should be considered with regard to the Hurtt measurement in-
strument. First, the scale rests on two major assumptions: that all six skepticism subconstructs
are equally important and that high levels of a certain characteristic can compensate for low
levels of another. However, these assumptions may be untenable. Specifically, it is possible
that some trait skepticism subconstructs possess greater predictive and explanatory power
regarding auditors’ skeptical behavior than others. In addition, it is conceivable that the con-
stituent characteristics combine in a non-compensatory manner, so that deficits in some foun-
dational aspects of skepticism cannot be compensated by another attributes.831 Second, the
Hurtt scale may be inadvertently capturing individual differences in tacit (i.e., non-
verbalizable and non-transferrable) knowledge or some other potentially relevant but not an-
ticipated psychological constructs.832 Third, as a self-report measure of personality aspects,
the Hurtt scale bears the inherent threat that subjects’ responses are potentially biased into a
socially desirable direction.833
These caveats notwithstanding, the Hurtt scale is the first thoroughly developed and validated
psychometric instrument which enables auditing researchers to ex ante measure (and explore)
auditors’ trait professional skepticism. While prior research has mainly focused on investigat-
ing state skepticism induced by situational factors such as fraud red flags or indications of
829
See Hurtt (2010): 162. It should be noted, however, that attributing some potential problems and misconcep-
tions to the characteristics of a particular sample while claiming generalizability, validity, and replicability
of the scale might be a slippery slope.
830
See Hurtt (2010): 162, 164. The final Hurtt scale is presented in Appendix 1.
831
See Hurtt (2010): 166.
832
See Hurtt/Eining/Plumlee (2010): 26.
833
See Schmitt/Steyer (1993): 519. Indeed, in their study on linking trait skepticism to skeptical behaviors,
Hurtt/Eining/Plumlee (2010): 18, 21 obtain a significant correlation between participants’ scores on the
Hurtt scale and their scores on the Social Desirability Scale developed by Crowne/Marlowe (1960).
142
aggressive reporting,834 still very little is known about auditors’ innate skepticism and how it
interacts with contextual and other factors to influence auditors’ judgment and behavior. In
this context, the Hurtt scale offers a unique opportunity to gain further insights and to develop
a better understanding of the nature of professional skepticism in auditing. Consequently, the
scale represents a considerable advance in the current state of academic theory and research
on auditors’ professional skepticism. Subsequently, the findings of the contemporary studies
on professional skepticism using the Hurtt scale (or its predecessors) are presented and dis-
cussed.
The first study which utilized the Hurtt scale835 and provided empirical support for its validity
was conducted by Fullerton and Durtschi (2004). They report a Cronbach’s alpha coefficient
for the whole scale of 0.96, which indicates a very good inter-item validity. In addition, they
find that the trait professional skepticism construct as captured by the Hurtt scale has a rea-
sonable predictive validity for skeptical judgments. In particular, their results demonstrate that
when encountering fraud red flags, internal auditors with high trait skepticism exhibit a signif-
icantly greater need for additional information than subjects with low trait skepticism.836 Sim-
ilarly, using a sample of student participants, Farag and Elias (2012) and Popova (2013) also
confirm both the theoretical and behavioral predictive validity of the Hurtt instrument in the
context of ethical reasoning and fraud, respectively.837 Peytcheva (2014) likewise corrobo-
rates the internal consistency and reliability of the Hurtt scale. Furthermore, she finds that trait
skepticism significantly predicts students’ cognitive performance in a hypothesis testing task.
However, no significant relationship between dispositional skepticism and cognitive perfor-
mance is obtained for auditors.838
Further support for the reliability and consistency of the Hurtt scale is provided by Bobek,
Hageman and Radtke (2013a) who report mean trait skepticism scores of 132.5 (SD 10.6) and
Cronbach’s alpha for the entire scale of 0.838.839 These values are consistent with those ob-
tained by Hurtt (2010). Similarly, using a sample of Dutch auditors, Quadackers (2009) and
Quadackers, Groot and Wright (2012) obtain mean trait skepticism scores of 131.66 (SD:
10.71) and 133.09 (SD: 10.84) as well as Cronbach’s alpha values of 0.821 and 0.834, respec-
834
See Rose (2007): 217.
835
As a matter of fact, the authors used a pre-publication version of the Hurtt scale which, however, is identical
to the final (published) version of the scale.
836
See Fullerton/Durtschi (2004): 15, 19.
837
See Farag/Elias (2012): 192, 194; Popova (2013): 152f.
838
See Peytcheva (2014): 41, 45. It is noteworthy that in deviation from the Hurtt scale which, as might be
recalled from the previous section, includes 6-point response scales, Peytcheva (2014) utilizes 7-point re-
sponse scales. She justifies her scale design choice by the desire to provide subjects with a neutral midpoint
instead of forcing potentially neutral participants to choose a side. By transforming the scores to a 100-point
scale as suggest by Hurtt (2010): 161, the comparability of results (scores obtained) is ensured. See
Peytcheva (2014): 38f., 45.
839
See Bobek/Hageman/Radtke (2013a): 11, 22.
143
tively. However, the authors only partially corroborate the notion that trait skepticism as re-
flected by the Hurtt scale significantly determines auditors’ skeptical judgments and behav-
iors. Specifically, the researchers report results suggesting that the Hurtt scores are weaker
predictors of auditors’ skeptical responses compared to the scores on the inverse Interpersonal
Trust Scale by Rotter (1967). This pattern of results particularly relates to high audit risk sce-
narios.840 In contrast, Hurtt, Eining and Plumlee (2010) find that under conditions of in-
creased risk, subjects’ skeptical behaviors are systematically influenced by the individual lev-
el of trait skepticism as measured by the Hurtt scale.841
Carpenter and Reimers (2013), on the other hand, report results showing no significant rela-
tionship between trait professional skepticism as measured via the Hurtt scale and auditors’
fraud-related skeptical judgments and decisions. These findings cast doubt on the predictive
and explanatory power of the trait skepticism construct for the variance in auditors’ behavior.
However, Carpenter and Reimers’ results corroborate the dispositional nature of trait skepti-
cism suggesting that it is relatively stable over time and is unaffected by situational factors
and incentives.842 Finally, Robinson, Curtis and Robertson (2013) obtain results suggesting
that while dispositional (trait) and situational (state) skepticism individually and mutually
influence auditors’ skeptical behavior, the effect of state skepticism is stronger.843
Overall, the substantial contemporary research on auditors’ professional skepticism corrobo-
rates the internal reliability and validity of the Hurtt scale and provides support for the notion
that trait skepticism is a stable, distinct individual disposition. With regard to the ability of the
trait skepticism construct to explain and predict auditors’ (skeptical) behavior, the empirical
results are fairly mixed. The present study provides additional empirical evidence along these
lines. It extends prior research by examining the effect of trait professional skepticism on au-
ditors’ judgments in light of a pervasive cognitive fallacy. The key conceptual aspects of this
investigation are discussed in the subsequent section.
840
See Quadackers (2009): 42, 44, 50-52, 92; Quadackers/Groot/Wright (2012): 29f., 45, 47. Skeptical judg-
ments and behaviors are operationalized in these studies via measures relating to auditors’ trust in the cli-
ent’s explanations, fraud assessments, generation and nature (error vs. non-error) of alternative explanations,
and budgeted audit hours. In addition to interpersonal trust and trait skepticism, Quadackers (2009) also ex-
plores the characteristics of need for closure and locus of control. However, he does not find a strong associ-
ation between these features and auditors’ skeptical judgments, behaviors, or degree of dispositional skepti-
cism.
841
See Hurtt/Eining/Plumlee (2010): 22-27. The researchers utilize the extent of evidence search, detection of
contradictions and errors as well as the generation of alternative explanations as proxies for skeptical behav-
ior. Note, however, that an extensive evidence search, if related to the examination of information that is ir-
relevant to the particular issue under scrutiny, is likely to impair audit quality both in terms of efficiency and
effectiveness. See Peytcheva (2014): 42.
842
See Carpenter/Reimers (2013): 64f.
843
See Robinson/Curtis/Robertson (2013): 19, 23f., 27. The authors operationalize skeptical behavior via
measures concerning the extent of audit testing, search for additional evidence as well as the identification of
inconsistencies in evidence. With regard to the observed stronger influence of state skepticism on auditors’
behavior in comparison to trait skepticism, the authors conclude that this is an encouraging finding since
traits are stable and inflexible components of one’s personality and thus are not easily changeable. In con-
trast, states are context-dependent and much more adjustable and malleable. Hence, it might be more pro-
ductive and beneficial for audit firms and regulators to focus on identifying and implementing effective state
skepticism prompts (incentives) rather than personality traits in order to successfully promote professional
skepticism in audit practice.
144
4.5 The Influence of Trait Skepticism on Auditors’ Belief Revisions
As indicated in Section 4.3.2.4, Nelson (2009) particularly highlights the importance and rele-
vance to professional skepticism of research focusing on auditors’ belief updating. 844 Prior
studies on this topic have adopted the skepticism notion set forth by Hogarth and Einhorn
(1992), i.e., skepticism as an extreme sensitivity to negative evidence combined with insensi-
tivity to positive evidence, and have reported fairly inconsistent results.845 With the Hurtt no-
tion and scale, a more universal, multifaceted, and psychometrically refined concept and
measure of auditors’ trait professional skepticism are employed in the present study. This
allows for shedding light on the role of personality traits in the context of auditors’ belief ad-
justments. As indicated in Section 3.4, the relationship between individual dispositions and
auditors’ belief revisions has remained virtually unexplored in the auditing literature. Beyond
the contribution to the belief revision literature in auditing, the present study makes an im-
portant contribution to the professional skepticism literature. Specifically, it provides valuable
insights into the behavioral manifestations of trait professional skepticism in an important yet
still unexplored setting which involves (1) a critical forward-looking audit judgment (going
concern) which, as argued in Section 4.2.2, calls for substantial professional skepticism, (2)
limitedly available information846, and (3) no particular skepticism-inducing contextual fac-
tors (e.g., fraud signs, seeded errors and/or inconsistencies in the data set).847
The study of the effect of dispositional skepticism on auditors’ belief revisions encompasses
two dimensions. First, the individual effect of trait skepticism on auditors’ belief adjustments
is explored. This line of investigation follows Cushing (2000) who argues that “skepticism
affects auditor behavior primarily by influencing the formation and revision of auditor be-
liefs” and that “auditors who are skeptical will tend to form different beliefs (…), and there-
fore will tend to process information (e.g., audit evidence) differently in updating their beliefs,
than will auditors who are not”848. Similarly, Abou-Seada and Abdel-Kader (2003) recognize
844
See Nelson (2009): 14.
845
Specifically, Bamber/Ramsay/Tubbs (1997): 262 find that auditors are confirmation prone, i.e., they are
highly sensitive to evidence which corroborates the hypothesis under which they operate. Contrary results
are obtained by McMillan/White (1993): 444 in light of an absolute measure of subjects’ belief revision (S 4–
S0). However, when a relative measure which accounts for scale effects potentially resulting from the
strength of subjects’ initial belief is employed, virtually no difference in auditors’ sensitivity to confirming
and disconfirming evidence is found. Guiral/Esteo (2006): 612, on the other hand, find that auditors are
moderately more sensitive to positive evidence than negative evidence, which they interpret as a lack of pro-
fessional skepticism.
846
Note that in order to confidently derive valid causal inferences on the influence of information order effects
on auditors’ belief updates, and in consistency with prior belief revision research, the present study uses a
fixed setting in which subjects are presented with predetermined sequences of information cues and cannot
obtain additional audit evidence but have to found their going concern assessments merely on the infor-
mation available.
847
The critical importance of the exploration of the strength of the link between skepticism and its behavioral
manifestations has been acknowledged by Doucet/Doucet (1996): 162.
848
Cushing (2000): 1.
145
skepticism as one of the key factors affecting auditors’ belief revisions.849 Based on these
contentions, the following hypothesis is stated:
H2 : Auditors with high levels of trait professional skepticism will exhibit different belief
revisions than auditors with low levels of trait professional skepticism c.p.
Second, beyond the main effect of skepticism, the interaction effect between trait skepticism
and information order is explored. With other words, the present study explores whether the
influence of information order effects on auditors’ belief revisions varies across different lev-
els of trait professional skepticism. Basically, two alternative scenarios are conceivable. These
scenarios are discussed subsequently.
849
See Abou-Seada/Abdel-Kader (2003): 33. Other factors identified by the authors to affect the process of
belief revision (or the formation of initial beliefs and the attitude toward positive and negative evidence) in
auditing include confirmation proneness, the diagnostic content of evidence, expertise, motivational and
cognitive factors. See Abou-Seada/Abdel-Kader (2003): 33.
850
See Fahrer (2012): 31f.; Hurtt et al. (2013): 72.
851
See Schreiber (2000): 173.
852
See Majors/Shefchik/Vitalis (2014): 2. However, as will be argued below, the exertion of higher cognitive
effort does not necessarily involve higher judgment quality.
146
tive and profound information processing and ultimately results in less extreme and biased
beliefs and judgments.853
The second mechanism via which skepticism can potentially mitigate recency in auditors’
belief revisions relates to the suspension of judgment facet of trait professional skepticism.
Specifically, as argued in Section 4.4.2.2, skeptical auditors are expected to withhold beliefs
when facing insufficient information rather than to prematurely “jump” into conclusions. By
so doing, auditors can eventually evaluate all available information in a simultaneous manner
(rather than sequentially), so that the effect of contrasting information stimuli, and thus the
extent of recency effects, may be reduced. To date, however, empirical support for the sus-
pension of information processing and belief updating until judgments can be made with
greater certainty is still lacking.854 This point is also addressed within the discussion of Sce-
nario 2.
The third mechanism by which skepticism can potentially reduce recency relates to auditors’
attitude toward evidence. Specifically, in the case of highly asymmetrical individual sensitivi-
ty to evidence (i.e., skepticism or advocacy in the sense used by Hogarth and Einhorn
(1992)), the belief-adjustment model predicts that order effects will become negligible.855
Under this perspective, in its extreme form, skepticism may rule out recency.856
Overall, there are several good reasons to expect an alleviating effect of skepticism on the
emergence of information order effects in auditors’ belief revisions. However, as will be
demonstrated subsequently, there are some not less meaningful clues for assuming the oppo-
site.
853
See Fein/Hilton (1994): 196; Hilton/Fein/Miller (1993): 501; Kramer (1999): 587; Schul/Burnstein/Bardi
(1996): 228, 246f.
854
See Hilton/Fein/Miller (1993): 503; Schul/Burnstein/Bardi (1996): 248.
855
See Pei/Reckers/Wyndelts (1990): 128.
856
However, in this case, negative effects like audit inefficiency and a hostile relationship with the client may
arise. This view corresponds to the presumptive doubt perspective of professional skepticism.
857
For such arguments and relevant psychological references along these lines, see Robinson/Curtis/Robertson
(2013): 7. The results of their study also suggest that the state component of professional skepticism (i.e.,
situationally induced skepticism) more strongly influences auditors’ behavior than the trait (i.e., disposition-
al) component. See Robinson/Curtis/Robertson (2013): 27.
858
The robustness and pervasiveness of recency effects in belief updating, especially in the absence of debi-
asing contextual factors, has been long and broadly recognized in psychology (e.g., Slovic/Lichtenstein
(1971): 693) and auditing (e.g., Kahle/Pinsker/Pennington (2005): 10). Additional support for the notion that
cognitive biases may counteract and dominate professional skepticism in auditors’ judgments is provided by
Payne/Ramsay (2005): 326 for anchoring effects.
147
Nelson (2009) provides theoretical support for this idea by arguing that cognitive biases like
recency can systematically and adversely affect auditors’ professional skepticism.859 In addi-
tion, Rose and Rose (2003) demonstrate that higher degrees of suspicion (i.e., a facet of skep-
ticism) are associated with a more profound and effortful evidence evaluation, which, howev-
er, instead of reducing recency, as would be expected according to Kennedy (1993), has been
found to reinforce the bias in auditors’ judgments. This phenomenon can be attributed to the
fact that elaboration involves drawing more inferences which are stored in long-term memory,
which increases the complexity and cognitive load during information processing. Following
the predictions of the belief adjustment model, high complexity is expected to induce recency
effects.860 Overall, the results reported by Rose and Rose (2003) suggest that higher cognitive
effort is not necessarily beneficial in terms of judgment quality under all conditions. In addi-
tion, these results suggest that the classification of recency as an effort-related bias in the
sense used by Kennedy (1993) might not be tenable, as apparently the bias operates beyond
conscious awareness and does not relate, as assumed in prior auditing research, to cognitive
“laziness”.
Further noteworthy findings along these lines have been reported by Majors, Shefchik and
Vitalis (2014) who find that higher degrees of trait skepticism trigger performance detriments
and compromise audit effectiveness due to skeptics’ ongoing questioning and critical stance
which deplete subjects’ cognitive resources available for information processing. 861 These
results suggest that high professional skepticism comes at cost and may, under certain condi-
tions, reinforce rather than mitigate cognitive bias.
In addition, Ashton and Kennedy (2002) recognize that it is virtually impossible to suspend
belief and judgment until there is sufficient and persuasive evidence on which to found one’s
position because “forming a judgment is a natural process when evidence is accumulated
over time.”862 They further argue that by withholding judgment, auditors would collect evi-
dence without processing it, which might jeopardize the possibility for timely follow up,
elaboration on and resolving of certain audit issues.863 With regard to recency, it is conceiva-
ble that the postponement of judgment could induce auditors to attach greater weight and im-
portance to evidence presented at the end of the sequence while largely ignoring evidence
faced at the outset. Following this line of thought, high degrees of trait professional skepti-
cism are likely to amplify (rather than mitigate) recency effects in auditors’ belief revisions.
In summary, there is a lack of theoretical and empirical clarity with regard to the strength and
direction of the interrelation between recency effects and dispositional skepticism in auditors’
859
See Nelson (2009): 2, 21. Although not particularly specified by Nelson (2009), cognitive biases are likely to
affect auditors’ state professional skepticism (i.e., a temporary condition evoked by environmental circum-
stances) and its behavioral manifestations (outward signs) rather than individuals’ trait professional skepti-
cism which is a stable and enduring characteristic and thus is not easily alterable by situational or other fac-
tors. The notion of cognitive biases and traps representing a threat to professional skepticism is echoed by
Glover/Prawitt (2013): 8-11.
860
See Rose/Rose (2003): 313f., 319.
861
See Majors/Shefchik/Vitalis (2014): 22f.
862
Ashton/Kennedy (2002): 224.
863
See Ashton/Kennedy (2002): 224.
148
belief revisions. As the formulation of an unambiguous and theoretically, empirically and/or
logically well-founded hypothesis is not possible, the interaction between information order
effects and trait professional skepticism is explored via the following research question:
Will information order affect subjects with high trait skepticism differently than subjects with
low trait skepticism?
These theoretical considerations build the frame for the empirical inquiry into the influence of
information order effects and trait professional skepticism on auditors’ belief revisions. In the
subsequent section, the procedures and results of the empirical analysis are presented and dis-
cussed.
149
5 Empirical Analysis
The research method employed within this study is that of a controlled experiment. Following
Peecher and Solomon (2001), the term “experiment” is used to refer to “a method of inquiry
in which the researcher randomly assigns participants within a controlled setting that allows
reproduction of some phenomenon (e.g., an individual’s decision process), actively manipu-
lates antecedents hypothesized to affect the nature of the phenomenon and then makes various
measurements of the phenomenon, often measurements that could not be made in a natural
setting”.864 The experiment has been viewed as a strong and influential research method that
enables the researcher to explain, rather than merely explore and describe, behavioral phe-
nomena, thereby achieving the highest degree of scientific understanding.865
There are numerous benefits arising from the application of experimentation. The main ad-
vantage of this research method is that it allows strong causal (as opposed to correlational)
inferences about the relationships underlying observed phenomena to be drawn.866 This is
achieved through deliberate manipulation of the independent variable(s), control of other po-
tentially influential factors, and random assignment of participants to treatment conditions.
This approach enables the experimenter to isolate effects, i.e., disentangle interrelated factors
that interact in natural settings to influence behavior, and explore under which conditions and
via which processes particular behavioral phenomena occur. With other words, experimenta-
tion allows for exploring why and how, rather than whether or not, specific phenomena arise.
In addition, experiments allow for testing conditions that do not yet exist or would be impos-
sible to directly observe and study in real-world settings. Hence, the experimental method is
considered as one of the most powerful research techniques.867
In general, there are two major types of experiments: laboratory and field experiments. The
former type is characterized through tight experimental control (rather than the research loca-
tion), which enables the researcher to reduce the noise caused by extraneous variables, where-
as the latter involves a more naturalistic (i.e., less controlled) setting in which people fre-
quently do not even know that they are subject to scientific investigation.868 The present study
employs a laboratory experiment in which participants are randomly assigned to treatment
conditions and a number of factors potentially influencing their belief revisions are controlled
for. While laboratory experiments have frequently been criticized for their artificiality and
864
Peecher/Solomon (2001): 195, bolded terms have been emphasized (italicized) in original.
865
See Peecher/Solomon (2001): 195.
866
See Aronson/Wilson/Brewer (1998): 102; Loewenstein (1999): F26; Solso/MacLin/MacLin (2008): 23;
Trotman/Tan/Ang (2011): 280.
867
See Libby/Bloomfield/Nelson (2002): 778; Libby/Luft (1993): 428; Peecher/Solomon (2001): 195; Reynolds
(1971): 156f.; Solomon/Trotman (2003): 395f.; Swieringa/Weick (1982): 63; Trotman (2011): 205;
Trotman/Tan/Ang (2011): 280. Furthermore, experiments are generally conducted under controlled condi-
tions, i.e., under the supervision of the researcher, so that they typically yield good response rates. See
Northrop/Arsneault (2007): 231.
868
See Aronson/Wilson/Brewer (1998): 105f.; Roth (1988): 974. For a general discussion of the advantages and
disadvantages of these two types of experiment, see Coolican (2013): 115-121.
150
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2_5, © Springer Fachmedien Wiesbaden 2015
lack of generalizability of findings beyond the context of the particular study, their contribu-
tion to increasing the understanding of the processes underlying human behavior achieved via
experimental control and specification of cause-effect-relationships cannot be overempha-
sized.869
The trade-off between experimental control and its inevitable by-product simplification on the
one hand and realism and extrapolability of the experimental results beyond the particularities
of a study on the other hand concerns two crucial aspects of scientific experiments – their
internal and external validity, respectively. Generally, internal validity has been acknowl-
edged as “the sine qua non of good experimental research.”870 It concerns the confidence with
which causal inferences can be drawn based on the research results. That is, internal validity
relates to the degree to which observed variation in the dependent measure can be straight-
forwardly attributed to the variations in the manipulated factor(s). Internal validity is typically
enhanced by random assignment of participants to experimental conditions. In essence, ran-
domization is an equalizing procedure which ensures that participants are distributed reasona-
bly evenly across conditions in terms of their personal characteristics, past experiences, etc.871
External validity, on the other hand, concerns the robustness of an observed phenomenon, i.e.,
the degree to which an inferred causal relationship can be generalized872 beyond the specific
experimental context and extrapolated to other persons, settings, and tasks. External validity
is enhanced through realism of the experimental setting and the use of participants who are
reasonably representative of the population in question.873 It should be remarked, however,
that in contrast to internal validity, external validity is not always an ultimate goal in experi-
mental research.874
A further relevant point that needs to be addressed within the general methodological consid-
eration of experimentation concerns the existence of two major and distinct types of realism
which involve different implications: experimental realism and mundane realism.875 Experi-
mental realism relates to whether participants attend to, take seriously, and are impacted by
869
See Dobbins/Lane/Steiner (1988): 282.
870
Aronson/Wilson/Brewer (1998): 129.
871
See Aronson/Wilson/Brewer (1998): 104, 129; Keppel (1973): 25. Note that perfect equivalence of subjects
among treatment groups is an ideal condition which is only possible in the “long run”, i.e., with the use of a
sufficiently large sample size. Real-world experiments, however, are conducted in the “short run”, which
means that there is no guarantee of (perfect) equivalence of subjects’ characteristics among groups. See
Keppel (1973): 25f.
872
It is instructive to note that there are two types of generalization: statistical generalization which is only
justified in light of random sampling and non-statistical generalization which is justified when the extension
of results beyond the bounds of the experiment appears reasonable in light of the knowledge and insights
gained in prior research in the particular area. See Keppel (1973): 28f. This argument is picked up on in Sec-
tion 6.2.
873
See Aronson/Wilson/Brewer (1998): 130f.; Peecher/Solomon (2001): 195. For a brilliant in-depth discussion
of validity issues in the context of experimentation, including internal and external validity but also addi-
tional types of validity such as statistical conclusion validity and construct validity, consult
Shadish/Cook/Campbell (2002): 34-102.
874
See, e.g., Aronson/Wilson/Brewer (1998): 132f.; Mook (1983): 379-387.
875
Aronson/Wilson/Akert (1994): 58 suggest a third type of realism which they label “psychological realism”.
This kind of realism relates to the degree to which the psychological processes elicited in an experiment re-
flect the relevant psychological processes naturally occurring in the real-world.
151
experimental stimuli, whereas mundane realism concerns the extent to which experiments
mimic real-world situations in their full contextual and informational richness. In general,
experimental realism has been acknowledged as highly desirable. Mundane realism, in com-
parison, has been viewed as neither necessary nor sufficient for internal and external validity.
In fact, in some cases, mundane realism has been argued to compromise an experiment by
distracting subjects’ attention away from the primary variables of interest, thereby diminish-
ing experimental control and inducing noise in the experimental setting.876 Hence, the present
study was designed to be reasonably realistic in order to elicit subjects’ motivation and in-
volvement with the study (experimental realism), while involving the necessary simplification
and abstraction (i.e., lack of mundane realism) in order to ensure tight experimental control,
temporal feasibility, and experimental efficiency.
Building on these general methodological considerations, in the subsequent section the exper-
imental design of the current study is discussed.
The experimental design of the present study contains one manipulated and one measured877
independent variable, both applied in a between-subject878 manner. Particularly, in consisten-
cy with prior research, information presentation order was manipulated at two levels: In the
first treatment condition, two pieces of positive information were followed by two pieces of
negative information ( െ െ). In the second treatment condition, the information cues were
presented in the exact opposite order (െ െ ). Within the positive and the negative cue
clusters, presentation order was held constant among the treatment groups.879
In contrast to the order of information presentation which was determined (i.e., controlled) by
the researcher, professional skepticism is a stable personality trait that is not manipulable.880
Hence, in the current study participants’ trait professional skepticism was measured (rather
than manipulated) using the Hurtt scale and all subjects were classified into two categories of
skepticism, high versus low, based on a median split. The resulting experimental design can
be classified as a 2 x 2 randomized post-hoc block between-subjects design.881 This procedure
876
See Peecher/Solomon (2001): 197; Swieringa/Weick (1982): 80f.
877
In the relevant research design literature, measured variables have also been referred to as “quasi-
independent” because in contrast to typical independent variables, they are not manipulated. However, like
their manipulated counterparts, measured variables also act as group-building factors in ANOVA proce-
dures. See Gravetter/Wallnau (2012): 347, 394. Alternative labels for measured variables include “non-
experimental (attribute) variables” (Kerlinger/Lee (2000): 367) and “blocking factors” (Hair/Black/Babin
(2010): 440).
878
A between-subjects design involves multiple conditions and different participants in each condition. See
Shadish/Cook/Campbell (2002): 109.
879
This procedure follows the design of Experiment 4 by Hogarth/Einhorn (1992): 28 and has found a broad
application in auditing research on belief adjustment. See Kahle/Pinsker/Pennington (2005): 12.
880
See Peecher/Solomon (2001): 195.
881
See Keppel (1973): 502, 511.
152
yielded the following four treatment-block combinations (henceforth referred to as experi-
mental groups or cells):
In consistency with prior auditing research,883 the dependent variable used in this study was
an individual’s belief revision measured as the difference between the last (S4) and the initial
(S0) belief (going concern likelihood assessment) of each participant. As posited by Hogarth
and Einhorn (1992), the consideration of the difference between the final judgment and the
stem (rather than the final judgment alone) is crucial for avoiding the threat of initial differ-
ences in beliefs masking order effects.884 In this context, a recency effect emerges if subjects’
responses reveal significantly greater downward (or smaller upward) belief revisions in the
െ െ treatment condition than in the െ െ condition. With other words, evidence of
recency would be obtained if S4–S0 ( െ െ) < S4–S0 (െ െ ).885 In strict normative
terms, in the case of an equal number of approximately equally strong positive and negative
information items, the final belief (S4) should approach the starting position (S0),886 i.e., S4–S0
( െ െ) = S4–S0 (െ െ ) = 0.
Beyond the variables of primary interest (i.e., the independent variables of information order
and trait professional skepticism), a number of demographic, evidence-, and effort-related
variables were collected in order to control for potential confounding effects. These variables
are listed in Section 5.3.3.
The subsequent section provides a detailed description of how the experimental design out-
lined above was implemented in the research instrument and which procedures were under-
taken to validate the experimental instrument. In addition, the administration of the study is
outlined.
As indicated in Section 3.2.3, the audit task utilized in the present study was that of a going
concern assessment. It may be recalled that the task choice was based on the insight that the
going concern evaluation essentially represents an ongoing sequential belief-adjustment pro-
882
The general distribution of participants among experimental groups is presented in Section 5.4.
883
See Kahle/Pinsker/Pennington (2005): 13-16 for an overview.
884
See Hogarth/Einhorn (1992): 23.
885
See Favere-Marchesi (2006): 74.
886
See Pinsker (2007): 199.
153
cess in which previously held beliefs are updated in light of new information which normally
arrives in a piecemeal fashion. In evaluating an entity’s going concern ability, auditors typi-
cally face mixed information comprising both evidence that refutes (negative evidence) and
evidence that corroborates (positive evidence) the going concern assumption. Overall, the
going concern evaluation has been acknowledged as one of the most complex, difficult, am-
biguous, information-intensive, and critical tasks in auditing.887 Hence, it provides an “ideal
context” for the investigation of recency effects in auditors’ belief revisions.888
Auditors’ going concern assessments were elicited using case materials based889 on a real au-
dit engagement of a Big 4 audit firm which were designed with the close cooperation of the
incumbent engagement partner. This approach was chosen in order to ensure the substantive
correctness, practical realism, adequacy, and consistency of the case materials.
Importantly, in contrast to the procedure typically employed in prior auditing research, herein
it was deliberately refrained from the adoption of an already existing case study. The rationale
for this was twofold: First, existing research materials had been designed to fit the specific
characteristics of the accounting and auditing regime in the particular country where the study
was conducted (typically the USA). Second, prior research instruments dealing with the issue
of belief updating in the context of auditors’ going concern judgments are not very recent and
topical, stemming without exception from the period before the most recent financial and
economic crisis in the years 2008 and 2009. This crisis has considerably hit the steel industry
where the present case study was located, and it has jeopardized the viability of a number of
companies, thereby providing a unique opportunity for the use of a topical, complex, and
highly uncertain setting for the study of auditors’ belief revisions.890 Overall, the development
of a novel, realistic, contextually well-suited contemporary case was deemed much more rea-
sonable and scientifically meaningful than the mere replication of potentially superseded prior
research instruments.
As the validity, consistency, and plausibility of the research instrument are of crucial im-
portance for obtaining meaningful and valid results, great care and deliberation were exerted
in developing the case study. Wherever possible and reasonable, measures and scales estab-
lished in prior research were employed to ensure highest possible validity and comparability.
For instance, as indicated in the previous section, in consistency with prior research, total be-
lief revision (S4–S0) was used as the dependent measure in the present study and subjects’
trait professional skepticism was measured using the well-established Hurtt scale. Further-
more, considerable value was placed on obtaining comprehensive feedback from representa-
887
See Carcello/Neal (2000): 455; Carmichael/Pany (1993): 49; Chow/McNamee/Plumlee (1987): 129;
Guiral/Esteo (2006): 603.
888
See Cushing/Ahlawat (1996): 115. This contention is underpinned by the fact that the going concern task has
been employed in a number of prior studies of recency effects in auditing, including Arnold et al. (2000),
Asare (1992), Ashton/Kennedy (2002), Cushing/Ahlawat (1996), Favere-Marchesi (2006), Guiral/Esteo
(2006), Kennedy (1993), Messier (1992), and Trotman/Wright (1996).
889
Note that to ensure the anonymity of the client firm, the original values were aggregated and modified but
the overall picture was substantially retained.
890
For relevant economic facts and statistics concerning the deleterious influence of the economic and financial
crisis on the steel industry, see European Commission (2009): 6 and European Commission (2010b): 8.
154
tives of audit practice and academic research. In particular, at the development stage of the
study, three university professors specialized in auditing research were asked to review the
case study and provide comments on the experimental instrument. Their remarks and sugges-
tions were implemented before the pilot test of the study was conducted. In addition, feedback
was also obtained at two PhD seminars conducted in Germany in 2011 where the project was
subjected to thorough critical discussion. Furthermore, three highly experienced audit practi-
tioners (two partners and one senior manager) reviewed the experimental materials to ensure
the plausibility, realism, and reasonableness of the case study. The comments obtained
through this review were integrated into the case materials. A detailed description of the final-
ized research instrument is provided in Section 5.3.3.
The experimental materials were written in German, as this was the official language used in
the training sessions within which the case study was administrated.891 Overall, it was ex-
pected that the participants in the experiment would be more comfortable, motivated, in-
volved, and quick in completing the case study in German rather than in English. In addition,
the particular language choice was expected to avoid the threat of terminological confusion
and confounding effects. To ensure linguistic uniformity of the research materials, the Hurtt
scale, which is originally available in English, was translated into German. For this purpose,
the parallel blind technique was employed.892 The discrepancies identified in the course of
this procedure were thoroughly discussed and resolved.
Before a pretest with audit professionals was conducted, the research instrument was subject-
ed to pilot testing with five academic staff members with practical experience in auditing.893
As a result of the pilot test, minor changes were made to the experimental instrument.
5.3.2 Pretest
The pretest was designed to support the selection of the four additional pieces of information
to be used in the final version of the case study, to pilot test the Hurtt scale with German audi-
tors, and last but not least to examine the understandability and plausibility of the case materi-
als. The pretest materials were structured as follows: At the outset, a general description of
the audit client and some aggregated financial statement data were provided. This baseline
information was followed by a set of ten pieces of additional information. For each of the ten
cues, subjects were asked to indicate the direction (positive, negative, or neutral894) and im-
891
For a more detailed description of the administration procedure, see Section 5.3.4.
892
For a description of this technique and a review of the strengths and weaknesses of alternative translation
methods, see Behling/Law (2000): 18-24. Quadackers/Groot/Wright (2012): 21 also applied the parallel
blind technique to translate the Hurtt scale into Dutch.
893
On the importance of pilot testing and pretesting, see Litwin (1995): 60f., 66f.
894
Note that while the ten cues were conceptualized to be strictly positive or negative, the neutrality of evi-
dence option was also included in order to avoid the issue of participants being forced to choose an option
that does not correspond to their perceptions. This approach is in line with prior research. See, e.g., Adam
(2007): 157f.; Cushing/Ahlawat (1996): 115f. This point is also addressed and underpinned with further rel-
evant references in Section 5.5.1.
155
portance895 of the individual items for the assessment the going concern ability of the firm.
The importance assessments were measured on a scale from 1 (“not at all important”) to 7
(“very important”). The initial pool of ten items consisted in part of real facts and in part of
fictional cues designed with the help of the engagement partner to fit into the basic scenario.
Following the cue assessment task, subjects were asked to complete the Hurtt scale without
being informed about the scale’s subject in order to prevent untruthful, socially-desirable re-
sponses. Finally, subjects answered a number of demographic and debriefing questions and
were requested to comment on the case study and to assess its understandability. The full pre-
test materials are available in Appendix 2.896
In total, 15 auditors at different levels within the audit firm (seniors897, managers, senior man-
agers, and partners) participated in the pretest.898 Of the 15 participants, two subjects missed
to provide an indication regarding the direction of the evidence for all ten items under review.
Hence, these subjects were excluded from the cue selection analysis. Table 3 provides an
overview of the key demographic characteristics of the 13 auditors who remained in the sam-
ple.
895
In consistency with Hite/Stock (1996): 88, 94, in the present work the terms “importance” and “strength” are
used interchangeably with regard to evidence or information.
896
Note that the pretest and the final version of the case materials are nearly identical, with the only difference
being that auditors in the pretest evaluated a set of ten additional cues without explicitly indicating their be-
liefs regarding the going concern ability of the firm, while auditors in the main test had to explicitly state
their initial beliefs and belief changes following the sequential presentation of four pieces of additional in-
formation. See also the description of the final case materials provided in the next section as well as the full
versions of the pretest and main test materials provided in Appendix 2 and 3, respectively. As will be shown
below, subjects in the pretest did not provide any critical remarks regarding the plausibility, understandabil-
ity, or extent of the case study, so that the retention of the baseline structure and content of the experimental
draft appeared reasonable and justified.
897
As will be argued in Section 5.4, although the final going concern assessment and the selection of the related
audit report type are typically tasks performed by more experienced, higher-ranked auditors, seniors are also
involved, at least at the preliminary stages, in the examination of the going concern assumption. Hence, sen-
ior subjects were retained in the sample. However, it was tested whether the overall findings changed if the
responses of the two seniors were excluded from the analysis. As this was not the case, the subsequently re-
ported results also include seniors’ cue direction and cue importance assessments.
898
The participants in the pretest were not included in the sample on which the final analyses reported in the
following sections were conducted.
156
Variable N Frequency Mean (SD) Range
Table 3 shows that the participants in the pretest had substantial general audit experience
(mean: 16.31 years), above-average experience with the going concern task (mean: 4.69 on a
scale ranging from 1 (“no experience at all”) to 7 (“very large experience”)), and little experi-
ence in auditing firms from the steel industry (mean: 2.00 on the same scale).
As indicated at the outset, the first major aim of the pretest was to support the cue selection
procedure by testing the directional plausibility and similarity in perceived importance of the
ten additional cues. An overview of the results of the cue assessment procedures is provided
in Table 4.
157
Item Brief Characterization Sign Misrated Neutral Mean SD
4 Major sales manager and the whole team leave the firm െ 1 0 6.08 0.86
An analysis of participants’ cue direction responses reveals that only 4 out of 130 observa-
tions (13 participants ൈ 10 observations per participant) (3.08%) were not in the intended di-
rection (i.e., a cue which was conceptualized to represent positive (negative) information was
rated as negative (positive)). This misclassification rate is comparable to the values reported
by prior auditing research.899 As the four misclassifications were conducted by four different
participants and concerned different cues, no systematical misconception regarding a single
cue could be identified. Remarkably, however, positive items were more frequently mistaken
for negative (three times in total) than conversely (only one time). In addition, participants
more frequently viewed positive cues as neutral (seven times in total). In comparison, nega-
tive cues were indicated only in two cases as neutral. Furthermore, subjects assigned greater
overall importance to negative information (the mean of all five negative cues amounts to
5.80) than to positive information (the mean of all five positive cues amounts to 4.72). How-
ever, it cannot be said whether these findings are attributable to auditors being more sensitive
to negative information than to positive information, to actual differences in the quality and
importance of the negative cues compared to the positive cues utilized in the pretest, or to
some other confounding effect.
With regard to participants’ cue importance assessments, an inspection of the results summa-
rized in Table 4 shows that compared to the other cues, item 8 and item 9 were viewed as
relatively weak by the participants in the pretest. Hence, these items were dropped from the
initial pool to ensure that the final cue selection was among items with above-average per-
ceived importance and thus sufficiently strong to induce belief revision. The remaining eight
items were then thoroughly discussed with the engagement partner of the audit on which the
899
See, e.g., Cushing/Ahlawat (1996): 117; Monroe/Ng (2000): 161. The misclassification rates obtained in
prior auditing studies are discussed in more detail in Section 5.5.1.
158
case materials were based, and the two positive and two negative cues which were most cru-
cial to the going concern assessment of the real-world prototype firm were selected.900 These
cues are listed below, with the item number used in Table 4 and the sign of the evidence in
parentheses:
1. Favorable external report on the economic outlook and the ability of the firm to con-
tinue as a going concern as well as banks’ willingness to negotiate over the renewal of
the firm’s credit lines901 (item 2) ();
2. Rejection of state funding (item 3) (െ);
3. Announcement of a cash capital increase of 200 MEUR until the end of the current
year (item 5) (); and
4. Bankruptcy of a major supplier and lack of alternatives902 (item 7) (െ).
As noted at the beginning of this section, the second major objective of the pretest was the
pilot test and validation of the Hurtt scale with professional auditors. To the author’s
knowledge, this was the first-time use of the scale with German auditors. All 15 auditors who
participated in the pretest filled in the scale in full and were thus included in this part of the
analysis. Subjects’ responses yielded a Cronbach’s alpha value of 0.922, thus hinting at a very
good internal consistency (reliability) of the scale. The mean Hurtt score amounted to 136.87,
which is significantly above the midpoint of the scale (it may be remembered that the scale
ranges from 30 to 180, which yields a theoretical midpoint of 105). The standard deviation of
the Hurtt score was 18.18 and subjects’ responses ranged from 105 to 163 and thus covered
approximately 38.7% of the full theoretical range. These results are in line with prior re-
search.903 In consistency with Quadackers, Groot and Wright (2012), one-sample Kolmogo-
rov-Smirnov test was performed on the Hurtt score and the histogram of the score was visual-
ly inspected to test for normal distribution.904 The approximate normal distribution of the
variable was confirmed both formally (asymptotic p-value (two-tailed) = 0.200) and visually
(the distribution of the scores was relatively symmetrical and bell-shaped, with the majority of
values centered around the mean and tapering out toward the tails).
The final major objective of the pretest was to check the understandability and reasonableness
of the case materials. Overall, subjects rated the understandability of the materials as above-
average (mean: 5.2 (SD: 1.08; range: 3-7) on a scale from 1 (“not at all understandable”) to 7
(“totally understandable”)). Although at the beginning of the pretest session subjects were
particularly encouraged to provide critical or other comments on the case study, no comments
900
This procedure follows Asare (1992): 386.
901
Note that Favere-Marchesi (2006): 73 used a similar cue, however, in the opposite direction (i.e., involving
a non-renewal of credit lines). This reinforces the critical importance of the issue of credit lines renewal in
the context of a firm’s going concern assessment and provides additional support for the cue choice made.
902
Note that this issue is explicitly stated in IDW PS 270.11 as an example of an operational circumstance that
potentially casts doubt on a firm’s ability to continue as a going concern. Hence, its direction and importance
is also normatively validated.
903
See, e.g., Hurtt (2010): 161f.; Popova (2013): 147; Quadackers/Groot/Wright (2012): 25, 45.
904
See Quadackers/Groot/Wright (2012): 25.
159
or remarks were obtained. Based on these results, the pretest was considered successful and
the case materials were viewed as understandable and reasonable.
Subsequently, the outcome of the development and validation procedures outlined above is
described in detail.
The case materials utilized for the inquiry into the influence of information order effects and
trait professional skepticism on auditors’ belief revisions were structured in two parts. In Part
1, a description of the client and the case scenario was provided and auditors’ repeated as-
sessments of the viability of the firm were elicited. The client, a company called “Premium
Steel AG”, was described as a leading manufacturer and distributor of quality steel products
for the automotive and plant construction industries, with around 10,500 employees and sev-
eral production sites in Germany. The company was depicted as currently experiencing a con-
siderable economic downturn905 due to the global financial and economic crisis and its severe
consequences for the automotive and plant construction industries as key consumers of steel
products. The portrayed substantial economic difficulties of the firm constitute a critical situa-
tion in which auditors are required to be particularly vigilant and skeptical.906
Following the general description of the client, participants were placed in the position of the
appointed auditor of the Premium Steel AG for the financial year ended 2009 and were told
that they were auditing the firm for a third year in succession. In addition, it was clarified that
the risk-based audit approach had been applied, the internal controls of the auditee were effec-
tive, and the firm had received an unqualified audit opinion in the last years. Participants were
then informed that the current-year audit was finished except for the final going concern eval-
uation. Thereafter, financial statement information was provided comprising an aggregated
balance sheet, income statement, and cash flow statement for the past two years (2007 and
2008) and the financial year under audit (2009).
Based on this information, subjects were asked to assess the likelihood that the Premium Steel
AG will continue in existence for the next twelve months. To reflect the nature of the likeli-
hood judgment required from the participants, a Likert-type response scale ranging from 0%
to 100% portioned into ten discrete intervals of ten percentage points each was employed for
eliciting participants’ initial beliefs (S0). At this point, participants were unaware that they
would receive further information on the case and the opportunity to update their previous
judgments.907 Importantly, in contrast to some older research works,908 in the present study
905
This result is reinforced by a review of the catalogue of 15 criteria indicating potential going concern prob-
lems identified by Mutchler (1986): 151, according to whom the fulfillment of even one single criterion can
cast significant doubt on a firm’s viability. Premium Steel AG meets 4 of the 15 negative criteria: reorgani-
zation, negative cash flow, current year loss, and negative income from operations.
906
See CPAB (2012): 3. In particular, in a situation of substantial economic downturn, client’s management is
potentially under increased pressure to represent the best possible picture of the company, i.e., to demon-
strate its going concern ability despite the considerable contrary economic conditions and factors.
907
See Ahlawat (1999): 78.
160
subjects were asked to provide self-generated initial beliefs instead of being given certain pri-
ors (initial likelihood values specified in the case materials). This procedure is central in terms
of capturing participants’ genuine initial beliefs. The assignment of priors, on the other hand,
can induce confounding effects (e.g., anchoring on the externally provided priors), thereby
compromising the results of the study.909
Next, the four additional information cues validated and selected in the pretest were presented
in a sequential manner, i.e., each new piece of evidence was presented on a separate page.910
After encountering each additional cue, subjects were requested to reindicate their assessment
of the likelihood that the Premium Steel AG will continue in existence for the next twelve
months. Importantly, subjects were instructed to complete each assessment in the predeter-
mined order and not to subsequently adjust already provided assessments. Again, participants’
responses were measured on an 11-point Likert-type response scale ranging from 0% to
100%. The resulting four sequential assessments are henceforth referred to as S1, S2, S3, and
S4 .
Part 2 of the experimental materials contained the Hurtt scale, manipulation checks as well as
demographic data and debriefing questions about the case. At the beginning of this part, sub-
jects were asked to complete the 30-item Hurtt scale. In consistency with prior auditing re-
search using psychometric scales, the Hurtt scale was presented subsequent to the case study
in order to avoid distortions of participants’ perceptions of the task, demand effects, and bi-
ased responses.911 To reduce the threat of potentially non-genuine responses, participants
were not informed about the subject of the scale. Following Hurtt’s (2010) instructions for
administration of the scale, the only explanation provided with regard to the scale was that it
contains a number of general statements which have to be assessed from the personal point of
view and that there are no right or wrong answers.912
908
Meant are here especially the following studies: Ashton/Ashton (1988, 1990), Bamber/Ramsey/Tubbs (1997),
Chan (1995), and Kennedy (1993).
909
See Asare (1992): 386; Butt/Campbell (1989): 474; Hite/Stock (1996): 86, 95; Pinsker/Pennington/Schafer
(2009): 98f. For a more detailed consideration of the anchoring phenomenon and the underlying psychologi-
cal processes and mechanisms, see Epley/Gilovich (2005, 2006).
910
Note that the presentation order of these four cues ( െ െ vs. െ െ ) represented the only difference in
the case materials provided to both treatment groups. As indicated in Section 5.2, and in consistency with
prior research (e.g., Ashton/Ashton (1988): 634; Ashton/Kennedy (2002): 225; Cushing/Ahlawat (1996): 115;
Monroe/Ng (2000): 157), the order of cues within the positive (mitigating) and negative (contrary) catego-
ries was held constant.
911
See, e.g., Bernardi (1994): 74f.; Peytcheva (2014): 38. Demand effects arise when subjects provide respons-
es in line with the anticipated expectations of the researcher and/or the experimental treatment. Closely re-
lated to this phenomenon is the so-called social desirability bias (also known as “halo effect”). This bias oc-
curs when participants respond unnaturally (and untruthfully) to sensitive questions in order to meet societal
norms and appear in a more desirable light. See Cohen/Pant/Sharp (1995): 41; Jones/Massey/Thorne (2003):
92; Randall/Fernandes (1991): 805f. Note in passing that Robinson/Curtis/Robertson (2013): 15 place the
Hurtt scale prior to the case study, which they explain by reference to the desire to ensure that trait skepti-
cism was not affected by the content of the case. However, as an individual disposition, trait skepticism is a
stable and enduring characteristic of an auditor’s personality, and thus it is not theorized to be easily mallea-
ble and context-dependent. Hence, the reasonableness of the above argumentation is at issue.
912
See Hurtt (2010): 168.
161
Subsequent to the Hurtt scale, the manipulation checks were presented. These procedures
aimed at verifying whether participants perceived the positive and negative pieces of addi-
tional information in consistency with the intended direction and whether the four cues were
viewed as similar in importance. The choice not to place the manipulation checks immediate-
ly after the manipulations was motivated by the desire to obtain a more objective and valid
assessment of the extent to which participants internalized the manipulations.913 In analogy to
the procedure adopted in the pretest, the manipulation checks consisted of two related ratings
for each piece of evidence: a classification of the direction of the cue as positive, negative, or
neutral914 information and an evaluation of its importance for the assessment of the going
concern ability of the Premium Steel AG on a scale from 1 (“no importance at all”) to 7
(“very large importance”).915
Upon completion of the manipulation checks, subjects were asked to provide demographic
information, including age, gender, general audit experience (in years), self-assessed916 expe-
rience with audits in the steel industry as well as with going concern evaluations, professional
certification, and position in the firm. The demographic questions were followed by a number
of evidence-, effort-, and skepticism-related questions. Specifically, participants were asked to
indicate to what extent they considered the information contained in the case study to be suf-
ficient for a conclusive evaluation of the going concern ability of the Premium Steel AG. Fol-
lowing this question, participants were asked to list the additional information that they would
need in order to conclusively evaluate the going concern ability of the Premium Steel AG.
913
This contention follows Grenier (2013): 15.
914
As indicated in the previous section, while the intended manipulation only involved positive and negative
information, the neutrality option was also included in order to provide participants with the full range of al-
ternatives and the possibility to provide genuine (rather than forced-choice) responses. This approach is in
line with Anderson/Maletta (1999): 82, Bamber/Ramsay/Tubbs (1997): 257f., and Blay (2005): 771 who all
used a bipolar manipulation check scale involving a neutral midpoint (0) to capture participants’ cue evalua-
tions. The procedure employed herein is also consistent with the manipulation check design utilized by
Cushing/Ahlawat (1996): 115f. who used the following cue classification options: positive evidence, nega-
tive evidence, or neither. In addition, Adam (2007): 157f. likewise recognizes that the direction (or tendency)
of information can be positive, negative, or neutral.
915
Note that with exception of the likelihood scale (0%-100%) on which subjects’ beliefs were measured and
the Hurtt scale which involves 6-point “forced-choice” response scales, all scales employed in the present
study ranged consistently from 1 to 7, with lower (higher) values indicating lower (higher) degrees of an at-
tribute or lower (higher) agreement with a statement. The use of a 7-point Likert-type scale follows the rec-
ommendation of Krosnick/Fabrigar (1997): 144f. Generally, a scale involving an odd number of response
alternatives offers participants the possibility to pick a neutral midpoint instead of forcing them to choose a
direction. However, it goes along with the risk of providing a convenient midpoint to which participants re-
luctant to exert cognitive effort could adhere. See, among others, Hurtt (2010): 156; Krosnick/Fabrigar
(1997): 147f.; Peytcheva (2014): 38f.
916
Self-ratings as measures of industry- and task-specific experience have been broadly applied in behavioral
auditing research. See, e.g., Cuccia/McGill (2000): 425, Green (2008): 247, Kayadelen (2007): 345, and
Zhao/Harding (2013): 492. Note that there also exist alternative measures of industry- and task-specific ex-
perience, e.g., the proportion of time spent in auditing clients within a certain industry over the past years
(e.g., Solomon/Shields/Whittington (1999): 197), the number of audit engagements within a certain industry
(e.g., Wright/Wright (1997): 278), or the percentage of clients facing going concern problems over the last
five years (e.g., Favere-Marchesi (2006): 72). However, these measures appeared more time-consuming to
elicit and were expected to produce non-uniform, difficultly comparable results, which would have compli-
cated the subsequent analyses. Hence, the use of self-assessments measured on uniform, predetermined
scales was preferred over alternative measures.
162
Next, subjects were asked to indicate the degree of certainty in their (own) assessment of the
going concern likelihood of the firm. Subsequently, participants were requested to provide
self-assessments of the attention, intensity of thought, and concentration exerted during the
going concern assessment task as well as the perceived difficulty of the evaluation. Subjects
were then asked to indicate whether they had participated in training courses or workshops on
professional skepticism in the past, and if this was the case, to indicate the number of these
events. In essence, the demographic and the other questions outlined above were conceptual-
ized to generate a bunch of control variables which were used in the supplemental and explor-
ative analyses reported in the next sections (particularly in Section 5.5.4.3 and in Section
5.5.5).
Finally, subjects were asked to rate the understandability of the case study, to provide com-
ments, and to indicate the time they needed to complete the study. The experimental materials
are available at full length in Appendix 3.
In conclusion, an important final note concerning the case study is in order. As is the case for
virtually all audit tasks and problems encountered in audit practice and/or studied in auditing
research, there is no “right” solution to the present case study.917 The only “correctness” crite-
rion that can be taken into account is normative theory. Hence, herein judgment quality is
evaluated under strictly normative-theoretical criteria.
5.3.4 Administration
Paper-and-pencil administration was selected for the conduct of the experimental study. This
procedure was preferred over online administration as it typically ensures higher participation
and stronger experimental control.
The experimental study was conducted in August and October 2012 during four annual train-
ing course sessions of a Big 4 audit firm. At the outset of each experimental session, support
for the project was expressed by an audit partner who encouraged active and diligent partici-
pation in the study. Overall, round 85% of all training course attenders participated in the ex-
perimental study.
In consistency with prior research, the instructions on the study and the case materials were
provided in the official language of the training sessions, i.e., German. At the beginning, a
brief verbal introduction was provided, including some very general remarks regarding the
study as well as few important instructions to be followed throughout the study. In particular,
subjects were asked to work independently, to complete the study in the predetermined order,
and not to revise their assessments in hindsight. The only comment regarding the subject of
the study was that it involved individual assessments of some economic issues. The verbally
provided instructions were summarized on the cover page of the experimental materials and
subjects were asked to read them through carefully before proceeding to the case study. In
917
See Ashton/Ashton (1990): 16; Trotman (2005): 78.
163
consistency with research convention, participants were assured anonymity and confidentiali-
ty.
To ensure the internal validity of the experimental study, subjects were randomly assigned to
treatment conditions (െ െ vs. െ െ). The randomization procedure was performed
using a computer-generated list which specified the sequence in which both versions of the
research instrument were administrated.
All experimental sessions were conducted under the vigilant supervision of the author. Over-
all, no notable questions were addressed during the experimental sessions. On average, it took
participants 21.54 minutes (SD: 5.92) to complete the study.
5.4 Participants
The study was completed by auditors from different local offices of a Big 4 audit firm who
participated in an annual professional training course. They were encouraged to participate in
the case study by a representative of the audit firm’s top management. However, participation
was on a voluntary basis. Therefore, it can be argued that primarily diligent and motivated
persons agreed to complete the case study.918
The experimental task used in the present study necessitates the participation of experienced
auditors because the case deals with going concern uncertainties with which less experienced
audit staff might have insufficient experience in the natural audit setting. 919 Overall, the de-
velopment of a realistic experimental case study and the choice of participants who meet the
requirements of the task aimed to enhance the external validity of the study, i.e., to ensure that
the results of the study were not crucially driven by, and solely limited to, either context or
participant population.920
In total, 181 subjects participated in the experiment. However, 14 participants provided in-
complete responses or failed the manipulation check questions and were therefore excluded
from the final dataset.921 Thus, the analyses were conducted based upon on a final sample of
167 participants. Table 5 provides an overview of the demographic characteristics of the par-
ticipants.
918
In consistency with the state of the art in experimental research in auditing, the current study did not make
use of explicit extrinsic incentives (e.g., monetary rewards) to enhance participation in the study. Rather,
herein intrinsic incentives were emphasized which are also very powerful motivation- and involvement-
increasing mechanisms. See, e.g., Fischhoff (1982): 435. Intrinsic incentives encompass “a personal interest
in performing the task, a desire to ‘help out’ the experimenter, and a desire on the part of the subject to see
‘how well can I do.’” Wright/Aboul-Ezz (1988): 143.
919
For such claims, see Cushing/Ahlawat (1996): 116.
920
See Aronson/Wilson/Brewer (1998): 130.
921
For a description of the manipulation check results and the sample rectification procedure, see Section 5.5.1.
164
Variable N Frequency Mean (SD) Range
Table 5 reveals that, on average, participants possessed 10.60 years of general audit experi-
ence, which exceeds the mean experience values reported in the majority of prior studies on
auditors’ belief revisions.922 Subjects’ self-assessed experience with audits of firms in the
steel industry was relatively low (mean: 2.44 on a scale from 1 to 7). Participants’ experience
with the going concern task was likewise below-average (mean: 3.70 on a scale from 1 to 7).
Nearly half of the participants included in the sample (76 subjects or 45.5%) were profession-
ally certified auditors (in German: “Wirtschaftsprüfer” or abbreviated “WPs”). Of the remain-
ing 91 participants, 30 persons were certified tax consultants (in German: “Steuerberater” or
abbreviated “StB”)923, 2 participants were CPAs, and 2 participants had ACCA certifica-
922
This statement is true both with regard to studies using a going concern experimental task (e.g.,
Ashton/Kennedy (2002): 226; Cushing/Ahlawat (1996): 116; Messier (1992): 146; Trotman/Wright (1996):
185) as well as studies employing other audit tasks (e.g., Ashton/Ashton (1988): 631; Bamber/Ramsay/Tubbs
(1997): 256; Monroe/Ng (2000): 159f.).
923
In Germany, tax law is one of the four major areas of the public auditor examination (§ 4 WiPrPrüfV). For
certified tax consultants (StB), examination in tax law is waived within the public auditor examination. That
is, for tax consultants an abridged form of the public auditor examination is possible. Given the immense
165
tion.924 As the experimental task involved an assessment of the going concern ability of the
firm, not the provision of a related report choice, it was deemed reasonable to include not yet
professionally certified participants (usually seniors) in the sample as these are typically the
staff members who gather and document evidence on which audit judgments and choices are
eventually based.925 Overall, the participants in the present study possessed the experience
and qualification required to complete the experimental task.
The inspection of Table 5 further reveals that 63.5% of the participants were males and 33.5%
were females.926 The obtained distribution reflects the male dominance characteristic for the
auditing profession.927
In addition, Table 5 shows that the mean skepticism score of the participants in the present
study was 134.82 (SD: 14.66). The obtained scores ranged from 71 to 165, thus covering
62.67% of the full theoretical range. The mean and standard deviation values obtained herein
are relatively consistent with the results reported in prior research.928 However, prior studies
using the Hurtt scale documented much smaller coverage rates (ca. 40%). The considerable
discrepancy between the results of the present study and the findings reported by previous
research can be primarily attributed to the scores of four participants in the present study
which were below 100 points, while in prior studies the minimal Hurtt scores were above
100.929 However, as the theoretical low-point of the Hurtt scale is 30, scores above 70 do not
appear questionable.
complexity and scope of the non-abridged professional auditor examination in Germany, WP candidates typ-
ically gain StB certification first on the way to the WP certification. For further details on the regulations
concerning auditor education, certification and training in Germany, refer to Köhler et al. (2008): 117-119.
924
These indications are not contained in Table 5.
925
See Ashton/Kennedy (2002): 226. It might be remarked that the participant utilization procedure found in
prior research is quite inconsistent. Specifically, Arnold et al. (2000), Asare (1992), Cushing/Ahlawat
(1996), and Favere-Marchesi (2006) used only very experienced, high-rank subjects (audit partners and
(senior) managers) to study subjects’ belief revisions in a going concern task. However, the former two stud-
ies utilized a task involving the choice of an audit report. Guiral/Esteo (2006), Kennedy (1993), Messier
(1992), and Trotman/Wright (1996), on the other hand, used staff auditors and/or accounting students in ad-
dition to more experienced subjects (audit managers) to examine experience effect. In contrast, Ash-
ton/Kennedy (2002) included only staff auditors in their sample.
926
Schwind (2011): 146 reported fairly similar results for his sample of German auditors (62.2% male and
37.8% female subjects).
927
In Germany, there are currently 14,390 certified public auditors. Only 15.34% of them (2,207 persons) are
females. The vast majority of professional auditors in Germany (84.66% or 12,183 persons) are males. See
WPK (2014): 2f. While at the entrant level, the proportion of males and females is relatively similar, at high-
er ranks in the firms, males clearly represent the majority. See, e.g., Doucet/Hooks (1999): 72f.; Schwind
(2011): 146f.
928
For instance, Hurtt (2010): 161 obtained a mean trait skepticism score of 138.6 (SD: 12.6);
Quadackers/Groot/Wright (2012): 24 reported an average score of 133.09 (SD: 10.84); and Popova (2013):
147 found a mean skepticism score of 131.5 (no indication of the standard deviation value).
929
In particular, Hurtt (2010): 161f. obtained responses ranging from 111 to 173, which implies a coverage of
the full range of 41.3%. Quadackers/Groot/Wright (2012): 24, 45 documented a response range from 103 to
158, i.e., coverage of about 36.6%. Popova (2013): 147 reported scores ranging from 101 to 157, i.e., cover-
age of 37.3%.
166
Cronbach’s alpha for the Hurtt scale was 0.872, which suggests a very good internal con-
sistency (validity) of the scale.930 The median Hurtt score was 136.931 As touched upon in
Section 5.2, the dichotomization of the trait skepticism variable was conducted on the basis of
a median split. Subjects scoring above 136 were classified as high in professional skepticism;
the remaining subjects were classified as low in professional skepticism. This post-
stratification of the sample resulted in 80 participants being classified as high in professional
skepticism and 87 subjects being categorized as low in professional skepticism. The random
assignment of subjects to treatment conditions described in Section 5.3.4 resulted in nearly
even distribution of participants across the two information order conditions. Specifically, 85
subjects were assigned to the െ െ treatment condition and 82 participants were assigned
to the െ െ treatment condition. The distribution of participants across the four experi-
mental cells can be summarized as follows:
െ െ , high professional skepticism: 39 subjects
െ െ , low professional skepticism: 46 subjects
െ െ, high professionals skepticism: 41 subjects
െ െ, low professional skepticism: 41 subjects
5.5 Results
As indicated in Section 5.3.3, the aim of the manipulation checks was to verify whether par-
ticipants apprehended the positive and negative pieces of additional information in consisten-
cy with the intended direction and whether a relatively similar importance was attached to the
930
This information is not contained in Table 5. The Cronbach alpha value obtained herein is comparable to the
coefficients reported in prior auditing research. See Section 4.4.4.
931
This information is also not contained in Table 5.
932
See Keppel (1973): 511f.
933
See Stevens (2009): 218.
934
See, e.g., Stiffler (2008): 57f.; Tenenbaum/Driscoll (2005): 84. Note that the recommendation of at least 30
participants per experimental group is rather conservative. Overall, the necessity to use groups of approxi-
mately equal and sufficiently large sizes relates to issues of normal distribution (see Section 5.5.4.1) and
power of the statistical procedure.
167
four cues. Following, the manipulation check results are discussed beginning with an analysis
of participants’ responses regarding the direction of the four information items followed by a
consideration of the results pertaining to the importance attached to the four cues.
As touched upon in the previous section, in total, 181 subjects participated in the study. Of the
724 total observations (181 participants ൈ 4 directional indications per participant), 68 obser-
vations (9.39%) entailed missing responses. The belief revisions of the 18 subjects who
missed to classify the direction of information were reviewed in order to verify whether par-
ticipants’ responses were in the intended direction. Overall, of the 18 subjects, five responded
in a manner inconsistent with the intended manipulation (i.e., revised beliefs upwards (down-
wards) after viewing negative (positive) evidence) and were thus excluded from the sample.
Of the 656 classifications obtained (724 total observations less 68 missing values), 13 were in
the incorrect direction (1.98%). A comparison with the relevant auditing literature reveals that
the misclassification rate reported in the present study is below the rates obtained in virtually
all prior studies.935 An analysis of the misclassifications revealed that they were conducted by
a total of eleven participants, with the majority of subjects (nine) having committed merely
one misclassification and only two subjects exhibiting two misclassifications. Of the eleven
“violating” participants, only three changed their beliefs in the incorrect direction. In con-
sistency with prior research, the three “violating” subjects were excluded from the sample as
they obviously misperceived one of the information items.936 As a result of the manipulation
check, a total of eight participants (4.42%) were excluded from the sample. In addition, six
participants (3.31%) failed to complete the Hurtt scale in full. Specifically, five subjects did
not provide a response to one item of the scale and one participant missed to respond to eight
items. All missing values concerned different items, so that no systematic pattern was identi-
fied. As the responses of these six participants were unusable for the statistical testing of the
research hypotheses, they were excluded from the sample as well.937 Consequently, the statis-
935
For instance, Messier (1992): 146 obtained a misclassification rate of 14.0%, Cushing/Ahlawat (1996): 117
reported erroneous evidence direction classifications in 3.7% of all observations, Monroe/Ng (2000): 161
documented miscoding of the direction of evidence in 3.57% of all observations, Asare (1992): 387 in
2.65% of all cases, Trotman/Wright (1996): 186 in 2.5%, and Ashton/Ashton (1988): 635 in 2.34% of all ob-
servations. The only study that reports lower misclassification rates than those obtained herein is the study
by Favere-Marchesi (2006) who obtained an error rate of 0.0% for a total of 180 participants and 720 obser-
vations. The other relevant studies on auditors’ belief revisions did not report misclassification rates at all or
only provided fairly unspecific result descriptions not allowing for a precise evaluation and derivation of vi-
olation rates.
936
The sample correction procedure used herein follows Anderson/Maletta (1999): 82, Ashton/Ashton (1990):
8, Brown (2009): 26, McMillan/White (1993): 451, and Trotman/Wright (1996): 186. Note, however, that a
substantial part of prior research has readily used the complete sample even in the cases of true misconcep-
tion of the evidence items obtained. See, e.g., Asare (1992): 387; Cushing/Ahlawat (1996): 117; Monroe/Ng
(2000): 161. Messier (1992): 146f., on the other hand, only excluded subjects who provided three or four in-
correct evidence direction assessments. While from a research efficiency viewpoint, the use of the full sam-
ple is plausible, it can introduce noise in the analysis and result interpretation because it is not clear whether
subjects’ responses were triggered by the presentation order of information or by their misperceptions of the
information presented. Hence, as argued by Ashton/Ashton (1990): 8, only the responses of participants who
understand the directional implications of the additional information presented are deemed truly usable.
937
This approach is in consistency with Peytcheva (2014): 37 who excluded four out of 89 audit participants
(4.49%) due to incompletely filled-in Hurtt scales. Other studies using the Hurtt scale do not provide infor-
mation regarding the completeness of subjects’ responses.
168
tical analyses described below were performed on the basis of a final sample of 167 sub-
jects.938
A summary of the misclassifications by cues is provided in Table 6. The table shows that the
most frequently misclassified information cue was client management’s announcement of a
planned cash capital increase until the end of the current year. In addition, this cue was most
frequently perceived as neutral in nature. These findings can be potentially attributed to the
fact that this information relates to the mere announcement rather than the actual implementa-
tion of a cash capital increase. However, the mean importance assessment of the cue does not
suggest that it was systematically underweighted. The second frequently misclassified item
was the rejection of state funding. Again, the mean strength assessment of this item does not
indicate that it was systematically underweighted. The information cue concerning the favor-
able external report and banks’ willingness to negotiate over the prolongation of the credit
lines, on the other hand, was not misclassified at all by the participants in the study. However,
it was relatively frequently viewed as neutral. This finding might reflect subjects’ insecurity
regarding the (successful) outcome of the negotiations with the banks. Finally, the bankruptcy
of a major supplier of the company was clearly perceived as an important negative infor-
mation as indicated by the lowest rate of misclassifications and classifications as neutral evi-
dence as well as the highest mean importance assessment. This result can be attributed to the
fact that the loss of a major supplier is explicitly stated in IDW PS 270.11 as a factor poten-
tially casting doubt on an entity’s ability to continue as a going concern.
Overall, the four additional information cues were rated as reasonably similar in importance,
with a slightly higher importance attached to the two negative items. In comparison, in prior
research participants assessed the additional items as considerably more heterogeneous in
strength.939
938
To control for self-selection biases (i.e., the possibility that the final sample consisted of more diligent and
thorough subjects who worked more concentratedly and exerted more effort on the case than did the exclud-
ed participants), it was tested whether the results of the analyses reported below changed when the full sam-
ple of 181 subjects was used. As this was not the case, it can be concluded that the results reported herein
were not driven by the sample delimitation procedure.
939
For instance, Tubbs/Messier/Knechel (1990): 458 obtained absolute evidence strength ratings of 78.6 and
71.64 for the positive items as well as 29.67 and 5.33 for the negative items. Similarly, Messier/Tubbs
(1994): 64 reported mean importance ratings of 78.54 and 39.92 for the positive evidence cues as well as
60.83 and 33.47 for the negative evidence cues. The other auditing studies on information order effects ei-
ther did not control for subjects’ evidence importance perceptions or did not report the corresponding re-
169
As a final analytical procedure for assessing the quality and plausibility of participants’ re-
sponses, the amount of belief adjustments in the wrong direction (upwards (downwards) after
obtaining negative (positive) evidence) of subjects who correctly classified the direction of
the information items in the subsequent manipulation checks was explored.940 In total, 18 out
of 665 belief revisions (167 subjects in the final sample ൈ 4 belief revisions (S1, S2, S3, S4)
each less 3 missing observations) (2.71%) involved error in direction. In comparison, Hogarth
and Einhorn (1992) documented belief revisions in the incorrect direction in 10.0% of all ob-
servations.941 Unfortunately, prior auditing research on information order effects either did
not conduct such an analysis or did not report the corresponding results, so that a benchmark
for comparison is lacking. However, the small percentage of belief adjustments in the wrong
direction provides a hint of the superior judgment performance of professional auditors in
comparison to non-professional subjects typically employed in psychological studies.
Collectively, the results of these analyses suggest that the experimental manipulations were
successful both in terms of the hypothesized sign and the overall mean importance of the evi-
dence utilized in the present study. Therefore, any potential differences in subjects’ mean be-
lief revisions can confidently be attributed to the different order in which the information
items were presented to the participants.
5.5.2 Demographics
As a prelude to the consideration of the hypothesis testing procedures and their results, some
relevant demographic aspects are discussed in the present section. At the outset, the distribu-
tion of participants among treatment conditions ( െ െvs. െ െ ) is considered in order
to explore whether randomization resulted in two approximately balanced groups with regard
to participants’ demographic characteristics. The results of this analysis are summarized in
Table 7.
sults. Similarly to the two preceedingly cited auditing studies, Hogarth/Einhorn (1992): 22 obtained overall
importance assessments of 65 for positive evidence and of 38 for negative evidence. Despite the significant
differences in subjects’ perceptions of the strength of the evidence, in all three studies cited above the ma-
nipulation checks were viewed as successful.
940
Note that the classification of subjects’ upward (downward) belief revisions in light of negative (positive)
information as wrong or inconsistent results from the assumption of evaluation-type cognitive processing
and additive evidence integration (see Section 3.2.1.3). However, to the extent to which subjects mentally
utilize estimation-type processing and averaging evidence integration, their belief revisions would not be
deemed counter-intuitive or erroneous.
941
See Hogarth/Einhorn (1992): 28.
170
Treatment Condition Mean Differ- p-value
Variable t-statistics
െെ െെ ence (two-tailed)
Gender is the proportion of female subjects. Auditor certification is the proportion of subjects that are legally
certified auditors (WPs). Position is subjects’ position in the firm hierarchy, where 1= assistant/other; 2 =
senior; 3 = manager; 4 = senior manager; 5 = partner.
* Significance level at 0.05
Table 7 shows that the two treatment conditions significantly differed in respect to partici-
pants’ age, task experience, and auditor certification. In addition, marginally significant942
differences were obtained with regard to subjects’ general audit experience and position in the
firm hierarchy. With other words, despite random assignment of subjects to treatment condi-
tions, differences between participants were not completely evened out among the two treat-
ment groups. On average, subjects in the െ െ condition were older, possessed more ex-
perience, were more frequently certified public auditors, and occupied higher positions in the
firm hierarchy compared to participants in the െ െ condition.943 This result is not sur-
prising. Specifically, as argued by Keppel (1973), it is practically impossible to perform an
experiment in which treatment groups only differ in respect of the experimental manipulation
(i.e., are perfectly equivalent in all other aspects) because real-world experiments are not con-
ducted using a sufficiently large number of participants.944
To further explore the interrelations among participants’ demographic characteristics, a corre-
lation analysis was performed. Table 8 provides a summary of the correlation coefficients
between the following demographic variables: age (AGE), gender (GEN), professional skepti-
cism as measured by the Hurtt scale (PS), general audit experience (GEXP), industry-specific
experience (IEXP), task-specific experience (TEXP), auditor certification (ACT), and position
in the firm hierarchy (POS). As not all of the aforementioned demographic variables are con-
942
The term “marginally significant” is used herein to refer to significance level at 0.1. On the marginal signifi-
cance controversy, see Aron/Coups/Aron (2013): 155f.
943
It might be remarked that Hurtt/Eining/Plumlee (2010): 18 also obtained differences in subjects’ experience
between randomly assigned treatment groups.
944
See Keppel (1973): 23, 26.
171
tinuous (or dichotomous) and approximately normally distributed, Spearman rank correlation
coefficients are reported in addition to Pearson product-moment correlation coefficients.945
GEN: dummy variable coded 1 if the participant is female, and 0 otherwise. ACT: dummy variable coded 1 if
the participant is legally certified auditor (WP), and 0 otherwise.
Pearson correlation coefficients in the lower triangle; Spearman correlation coefficients in the upper triangle
*, **, *** Significance levels at 0.05, 0.01 and 0.001, respectively (two-tailed tests)
Table 8 indicates that trait professional skepticism as measured by the Hurtt scale is not sig-
nificantly correlated with any other demographic measure. All other demographic variables
(age, gender, experience (general, industry- and task specific), auditor certification, and posi-
tion) are significantly correlated with each other in a plausible pattern which points to male
participants being older, more experienced, more frequently having auditor certification, and
occupying higher positions in the audit firm hierarchy compared to female participants.946
With regard to participants’ scores on the Hurtt scale, female auditors were found to score
higher on the trait skepticism scale (mean score: 136.98) than male auditors (mean score:
945
Note that the criteria of normal distribution and continuous (or dichotomous) level of variables must be met
in order to use Pearson product-moment correlation. See, e.g., Field (2009): 177, 179f.; Pallant (2010): 126,
128. As to the scale levels of the variables listed in Table 8, gender and auditor certification are dichotomous
(also referred to as nominal, discrete, or categorical) variables, position in the firm hierarchy is an ordinal
variable, and all other variables – age, professional skepticism score, general experience, industry-specific
experience, and task-specific experience – are continuous (also referred to as interval) variables. For a de-
tailed discussion of the different scale levels and their characteristics, see Bortz/Schuster (2010): 13-23. The
assumption of interval scale level for psychometric tests, rating scales, etc. is in conformity with the state of
the art in the empirical research practice. See, among others, Bortz/Schuster (2010): 14, 23 (with a list of fur-
ther references along these lines) and Tabachnick/Fidell (2007): 7. As to the normal distribution criterion,
scores on the age, general audit experience, and industry-specific experience variables are right-skewed and
do not appear approximately normally distributed.
946
A more detailed overview of the differences in demographic characteristics among the genders is provided
in Appendix 4.
172
133.42). This finding is in line with the results reported by Fullerton and Durtschi (2004).947
However, the obtained difference between the genders was not statistically significant (t (160)
= -1.466, p (two-tailed) = 0.145).
Finally, a review of the distribution of demographic data across the four experimental cells is
presented in Table 9. All of the values summarized in this table represent cell means.
Experimental Groups
Variable
െ െ , low PS െ െ , high PS െ െ, low PS െ െ, high PS
Gender is the proportion of female subjects. Auditor certification is the proportion of subjects that are legally
certified auditors (WPs). Position is subjects’ position in the firm hierarchy, where 1= assistant/other; 2 =
senior; 3 = manager; 4 = senior manager; 5 = partner.
Table 9 reveals that although randomization and post-hoc blocking did not result in perfect
equality, the four groups are reasonably similar with regard to participants’ demographics.
However, to account for the identified between-group differences in demographic characteris-
tics (especially with regard to the two treatment groups) and to avoid any potentially con-
founding effects, subjects’ experience is included as a control variable in the supplemental
regression analysis (Section 5.5.4.3).948 To avoid potential multicollinearity issues arising
from the high correlation between general experience, age, auditor certification, and position
in the firm hierarchy, only general experience is included in the regression model to represent
the set of correlated demographic variables. In addition, subjects’ industry- and task-specific
experience is also controlled for in the regression analysis.949 However, as pointed out in Sec-
947
See Fullerton/Durtschi (2004): 23.
948
Beyond the consideration of potential experience effects by the inclusion of subjects’ experience in years as
a control variable in the regression analysis reported in Section 5.5.4.3, an additional test (analysis of vari-
ance (ANOVA)) was conducted with experience as a post-hoc blocking variable. The two experience groups
were built on the basis of a rank split. Specifically, following Asare/Cianci/Tsakumis (2009): 229, staff audi-
tors and audit seniors were classified as low in experience, whereas managers, senior managers, and partners
were categorized as highly experienced. As the ANOVA with information order and experience as experi-
mental factors yielded results which are qualitatively consistent with the results of the regression analysis,
only the latter is reported in the present work (see Section 5.5.4.3).
949
Note in passing that in the present study, a regression analysis is preferred over an analysis of covariance
(ANCOVA) to statistically control for experience effects for three major reasons. First, as participants’ ex-
173
tion 3, information order effects operate at a subconscious cognitive level and thus are not
likely to be affected by knowledge and general audit experience. Hence, there are no compel-
ling theoretical grounds for expecting that the hypothesized recency effects in participants’
belief revisions will be mitigated by experience.
To set up the stage for the primary statistical tests, the results of some relevant preliminary
analyses are presented and discussed in the present section. To begin with, as argued by Ho-
garth and Einhorn (1992), the presence of recency effects at the aggregate level does not nec-
essarily imply the same result at the individual level.950 To account for this insight, prior to
aggregate level investigation, the belief revisions of each participant were analyzed individu-
ally in order to determine whether an order effect occurred and of which kind it was (primacy
vs. recency). The results of this analysis are summarized in Table 10.
Psychological
Observed Effect n N % Reference (%)
Recency effect 85 167 50.90 60.87
Primacy effect 40 167 23.95 34.78
No order effects 42 167 25.15 4.35
Table 10: Individual-Subject Level Order Effects
As expected, the majority of participants (85 subjects or 50.90%) exhibited responses indica-
tive of recency, i.e., they overweighed evidence presented later in the sequence. However, a
considerable number of the subjects (42 subjects or 25.15%) responded in a manner consistent
with no order effects in belief revision, i.e., the final beliefs of these subjects reflected their
initial positions, resulting in a total belief revision (S4–S0) of zero, which is the normatively
appropriate response in light of two positive and two negative evidence pieces of (approxi-
mately) equal strength.951 Nearly the same number of participants (40 subjects or 23.95%)
responded in consistency with a tendency toward primacy effects, i.e., these subjects anchored
on evidence provided early in the sequence. Unfortunately, prior auditing studies did not pro-
vide a differentiated analysis of the nature and portion of order effects in participants’ re-
perience is only very weakly related to their belief revisions (Pearson product-moment correlation coeffi-
cient r = 0.044, p (two-tailed) = 0.574), it is unlikely that experience will provide a viable counter-
explanation for any differences between experimental groups. Hence, virtually nothing is to be gained by us-
ing the ANCOVA as it will produce results very similar to the ANOVA results. See Cone/Foster (1998):
186; Kutner et al. (2005): 919. Second, the statistical model underlying the ANCOVA rests on a set of very
stringent assumptions (see Section 5.5.4.3) and is fairly sensitive to minor violations thereof. Hence, the
ANCOVA technique should be applied with cautiousness and only in very simple and clear cases. See
Keppel (1973): 514-516; Kerlinger/Lee (2000): 523f. Third, AN(C)OVA is essentially a special case of a re-
gression model. See, e.g., Field (2009): 349; Kerlinger/Lee (2000): 307; Slovic/Lichtenstein (1971): 660;
Tabachnick/Fidell (2007): 119.
950
See Hogarth/Einhorn (1992): 25.
951
See, e.g., Pinsker/Pennington/Schafer (2009): 95.
174
sponses. Hence, a comparison with the relevant auditing literature is not possible. The only
reference study available is the generic belief revision examination conducted by Hogarth and
Einhorn (1992) who found that 14 out of 23 individuals (60.87%) responded in consistency
with recency, eight participants (34.78%) exhibited primacy effects in their responses, and
only one subject (4.35%) did not display information order effects in his/her belief revi-
sions.952 Hence, in comparison with the inexperienced, non-professional participants in Ho-
garth and Einhorn’s experiment, the professional auditors in the present study demonstrated a
significantly greater proportion of normatively accurate responses (i.e., free of cognitive bias)
and relatively smaller proportions of each type of order effects. This is an encouraging finding
with respect to auditors’ judgment quality.
As a further preliminary analytical procedure, the sequence of subjects’ mean belief changes
(likelihood assessments) by treatment condition is considered. The results of this analysis are
presented in Table 11.
Table 11 reveals that participants’ responses are in line with the predictions of the belief-
adjustment model. Specifically, on average, subjects in the െ െ condition exhibited two
downward belief revisions followed by two upward belief revisions, resulting in a final as-
sessment of the going concern likelihood of the Premium Steel AG which was above the ini-
tial position. In the െ െcondition, subjects’ responses exhibited the contrary pattern: two
upward belief revisions were followed by two downward belief revisions, resulting in a final
belief which was below the initial one. The mean initial likelihood assessments (S0) among
both groups were slightly above the theoretical midpoint of 50%. The difference in initial be-
liefs between the two treatment conditions was not significant (t (165) = 0.260, p (two-tailed)
= 0.795). With other words, after starting with fairly similar initial assessments of the viability
of the Premium Steel AG and obtaining the very same information items, the two treatment
groups arrived at final beliefs (going concern likelihood assessments) that differed by 12.82
percentage points. When represented graphically, participants’ responses yield the prominent
“fishtail” pattern (Figure 7) hypothesized by Hogarth and Einhorn (1992) and corroborated
by prior auditing research.953
952
See Hogarth/Einhorn (1992): 25, Table 3, Experiment 3, SbS-column.
953
See, e.g., Arnold et al. (2000): 118; Favere-Marchesi (2006): 76; Hogarth/Einhorn (1992): 23f., 29.
175
80
70
Mean Likelihood Assessments
60
50
(++--)
(--++)
40
30
20
S0 S1 S2 S3 S4
Sequential Beliefs
Several interesting findings arise from the exploration of the mean belief revisions among the
four experimental groups. In the െ െ condition, participants with low levels of trait skep-
ticism started with higher initial likelihood assessments (S0 = 58.29) and exhibited greater
(more extreme) downward belief revisions (S4–S0 = -6.10) than participants with high levels
of trait skepticism (S0 = 50.98, S4–S0 = -2.68). However, both differences in initial beliefs (t
(80) = 1.572, p (two-tailed) = 0.120) and in total belief revisions (t (80) = -0.937, p (two-
176
tailed) = 0.352) between high- and low-skepticism individuals were not statistically signifi-
cant. In the െ െ condition, on the other hand, exactly the opposite results were obtained:
Subjects with high levels of trait skepticism started with higher initial beliefs in the going
concern ability of the Premium Steel AG (S0 = 57.69) and exhibited greater (more extreme)
upward belief revisions (S4–S0 = 7.95) than subjects with low levels of trait skepticism (S0 =
53.70, S4–S0 = 7.17). Again, the differences in initial beliefs (t (83) = -0.791, p (two-tailed) =
0.431) and in total belief revisions (t (83) = -0.201, p (two-tailed) = 0.841) between high- and
low-skepticism individuals were not significant.
The mean sequential beliefs (likelihood assessments) among the four experimental groups are
depicted in Figure 8. The figure reveals that auditors’ sequential belief revisions clearly fol-
low the classic fishtail pattern irrespective of participants’ level of trait professional skepti-
cism.
90
80
Mean Likelihood Assessments
70
60
(++--)/high PS
(++--)/low PS
50
(--++)/high PS
(--++)/low PS
40
30
20
S0 S1 S2 S3 S4
Sequential Beliefs
177
5.5.4 Hypothesis Testing
The method employed herein to test the hypotheses outlined in the previous sections is that of
an analysis of variance (ANOVA). ANOVA is a parametric technique which in its most
common form is essentially an application and extension of the method of least squares.954
The objective of the ANOVA technique is to determine whether there exist any significant
mean differences among different treatment conditions (groups) in an experimental study.955
The concept of ANOVA was developed by R. A. Fischer in the 1920s. The technique was
initially applied in psychological research by Hoffman, Slovic and Rorer (1968), Rorer et al.
(1967), and Slovic (1969). In the field of auditing, ANOVA was first used by Ashton
(1974).956 Ever since, it has been the standard analysis technique employed in experimental
research in auditing.
The application of ANOVA involves the following major requirements: (1) independence of
observations957, (2) normal distribution of the residuals, and (3) variance homogeneity among
experimental groups. While the first requirement concerns the experimental design of the
study and was satisfied herein through the use of randomization and experimental control, the
latter two conditions concern characteristics of the population under consideration and thus
are basically beyond the researcher’s control. The latter two ANOVA assumptions are critical
for the validity of inferences drawn using the F-statistic and therefore require a profound ex-
amination and analysis.958
Following the recommendations of Thode (2002) and Hair, Black and Babin (2010), the nor-
mality assumption was tested using both informal (i.e., visual and descriptive tests) and for-
mal (i.e., inferential tests) assessment techniques.959 The results of the visual inspection of the
histogram, the normal Q-Q plot, and the boxplot (not shown here) indicate that the residuals
are approximately normally distributed, with skewness of 0.229 (SE: 0.188) and kurtosis of
0.126 (SE: 0.374).960 This finding is reinforced by consideration of the standardized values (z-
scores) for skewness and kurtosis (0.218 and 0.337, respectively), which are safely below the
954
See Eisenhart (1947): 3f. For a detailed formal, graphical, and descriptive analysis of the logic underlying
the ANOVA technique, see Gravetter/Wallnau (2012): 356-362 and Keppel (1973): 35-71.
955
See Gravetter/Wallnau (2012): 428; Kerlinger/Lee (2000): 307.
956
See Trotman (1998): 119; Slovic/Lichtenstein (1971): 660.
957
This assumption requires that the dependent measures for each participant are completely uncorrelated with
the responses from other participants in the study. This assumption is crucial to the statistical validity of in-
ferences drawn using the ANOVA procedure. See Hair/Black/Babin (2010): 441.
958
See, e.g., Lix/Keselman/Keselman (1996): 579. For an in-depth formal consideration of the ANOVA as-
sumptions, consult Eisenhart (1947): 9-19.
959
See Hair/Black/Babin (2010): 74; Thode (2002): 16. The latter highlights the variety of methods and tech-
niques available for testing normality, indicating more than 40 formal procedures for examining the sample
distribution as well as a number of informal procedures, including plotting, outlier testing, general goodness
of fit methods, etc. See Thode (2002): 1.
960
A perfectly normal distribution has skewness and kurtosis values equal to zero. Harwell et al. (1992): 316
classify skewness values above 0.5 as mild deviations from normality. Under this criterion, the present dis-
tribution (skewness value: 0.229) can be classified as reasonably normal.
178
critical z-value of ±1.96 for rejecting the null hypothesis of normal distribution in the popula-
tion at a 0.05 significance level.961 Hence, the results of the informal normality tests uniform-
ly show that the residuals are approximately normally distributed. For the sake of complete-
ness, a one-sample Kolmogorov-Smirnov test was conducted as well. The results of the test
indicate that the data are not perfectly normally distributed (asymptotic p-value (two-tailed) <
0.0001). It should be remarked though that the Kolmogorov-Smirnov test is a rather conserva-
tive inferential test which reacts sensitively to even small deviations from normality and thus
involves a relatively high chance for yielding results in favor of rejecting the null hypothesis
that the sample is drawn from a normal distribution. This is particularly true for large samples
because of the higher power of the statistical test. As a result of the central limit theorem,
normality is not an issue for large samples.962 In addition, it should be noted that the F-
statistic underlying the ANOVA technique has been found to be fairly robust to minor devia-
tions from normality, particularly for symmetrical populations and experimental cells of ap-
proximately equal size and more than twelve subjects, as is the case in the present study.963
Hence, based on these arguments and the results of the informal normality checks, it can be
concluded that the normality assumption is tenable.
Finally, the application of ANOVA requires the homogeneity of variance among experimental
groups. This condition has been viewed as more critical than the normality requirement be-
cause if it is violated, the statistical inference will be invalid regardless of the sample size.964
Based on the results of Levene’s F test (F (3, 163) = 0.366, p-value (two-tailed) = 0.778), the
underlying null hypothesis of which is that the variances in the different groups are equal, it
can be concluded that the homogeneity of variance assumption is fulfilled.965
In summary, the assumptions of independence of observations, normal distribution of residu-
als, and homogeneity of variance are tenable. Hence, the application of the ANOVA tech-
nique is reasonable and justified. Following, the central ANOVA results are presented and
discussed.
961
See Hair/Black/Babin (2010): 72f.; Rovai/Baker/Ponton (2014): 192f. Although both sources provide alter-
native formulas for determining the z-score for skewness and kurtosis, both lead to the same results and in-
terpretations.
962
See, e.g., Field (2013): 184; Kirk (2013): 52; Pole/Bondy (2010): 933f.; Rovai/Baker/Ponton (2014): 192f.;
Tabachnick/Fidell (2007): 78; Wooldridge (2009): 172-174. Note that in the relevant statistical literature,
there is no universally accepted benchmark regarding the sample size required for the central limit theorem
to find application and relieve the strictness of the normality assumption. Wooldridge (2009): 174f. indicates
that some researchers consider a sample size of 30 respondents sufficient, which in his view is not tenable,
particularly when several independent variables are involved. Hence, he recommends that the degrees of
freedom are also taken in consideration instead of solely focusing on the sample size. Pole/Bondy (2010):
934 provide a more conservative guideline for sample size classifications. They consider samples with less
(more) than 50 observations as small (large) samples. Similarly, Kirk (2013): 52 considers a sample size of
50 to 100 as sufficient for yielding a nearly normal sampling distribution.
963
See, e.g., Cochran (1947): 24f.; David/Johnson (1951): 57; Glass/Peckham/Sanders (1972): 246, 273;
Harwell et al. (1992): 316, 333f.; Kirk (2013): 99; Lix/Keselman/Keselman (1996): 582; Pallant (2010):
206; Tiku (1971): 913-915.
964
See Eisenhart (1947): 13; Pole/Bondy (2010): 932; Wooldridge (2009): 175.
965
See, e.g., Field (2009): 150.
179
5.5.4.2 Summary of Results
Table 13 reports the overall ANOVA results (Panel A) as well as the mean belief revisions
(S4–S0) per experimental group (Panel B).
Panel A: ANOVA
Source of Variance df Mean Square F-statistic p-value η2
Table 13: ANOVA Results and Experimental Group Means for Total Belief Revision
10 10
8 8
Estimated Marginal Means of S4-S0
6 6
4 4
2 2
0 low PS 0 --++
high PS ++--
-2 -2
-4 -4
-6 -6
-8 -8
-10 -10
--++ ++-- low PS high PS
967
See, e.g., Terrell (2012): 282.
968
See Bortz/Schuster (2010): 244f.; Leigh/Kinnear (1980): 842.
181
5.5.4.3 Supplemental Tests
In consistency with prior research, an analysis of covariance (ANCOVA) with the final belief
(S4) as the dependent variable and the initial belief (S0) as a covariate was performed.969 Pre-
liminary analyses were undertaken to ensure no violation of the assumptions of independ-
ence970, normality971, homoscedasticity972, linearity973, homogeneity of regression slopes974,
and independence of the covariate of any treatment effects975.976 Table 14 reports the overall
ANCOVA results (Panel A) and the mean final beliefs (S4) per experimental group (Panel B).
Panel A: ANCOVA
Source of Variance df Mean Square F-statistic p-value η2
Table 14: ANCOVA Results and Experimental Group Means for Final Belief
969
See Cushing/Ahlawat (1996): 118; Favere-Marchesi (2006): 74, 77; Messier/Tubbs (1994): 66; Monroe/Ng
(2000): 161; Trotman/Wright (1996): 187.
970
The independence assumption was taken into account via the utilization of a randomized between-subjects
experimental design.
971
The normality testing procedure is analogue to the procedure described in Section 5.5.4.1.
972
As might be recalled from Section 5.5.4.1, homoscedasticity (i.e., homogeneity of variance) is tested using
Levene’s F test. The test yielded a not significant result (F (3, 163) = 0.026, p (two-tailed) = 0.994), thus in-
dicating the tenability of the assumption.
973
The linearity assumption was tested and corroborated via visual inspection of the scatterplots.
974
Following Field (2009): 413f., the homogeneity of regression slopes was verified on the basis of the ob-
tained not significant tree-way covariate by experimental factors interaction (initial belief ൈ information or-
der ൈ professional skepticism) (F (1, 160) = 0.089, p (two-tailed) = 0.766).
975
The independence of the covariate (S0) of any treatment effects was both logically derived and formally
confirmed. Specifically, as the information order treatment (i.e., the four pieces of additional information
presented in a varying order between the two treatment conditions) was presented subsequent to the for-
mation and indication of initial beliefs, S0 could not have been influenced by the different order of infor-
mation presentation. This conclusion was confirmed by the not significant covariate by treatment interaction
(initial belief ൈ information order) (F (1, 160) = 0.001, p (two-tailed) = 0.977).
976
For a summary of the ANCOVA assumptions, see, e.g., Keppel (1973): 484; Van Breukelen (2010): 21.
182
In consistency with the theoretical predictions condensed in H1 and in line with the pattern of
results reported in the previous section, participants who processed evidence in the െ െ
order exhibited significantly greater mean final beliefs (63.06) than respondents who pro-
cessed evidence in the reverse order (50.24) (F (1, 162) = 28.051, p (two-tailed) < 0.0001).
With other words, auditors who obtained positive information at the end of the sequence pro-
vided significantly more favorable final assessments of the going concern likelihood of the
Premium Steel AG than auditors who faced negative information at the end of the series. The
obtained information order effect size is large (η2 = 0.148).
The ANCOVA results further indicate that the difference in mean final beliefs between partic-
ipants with high and low levels of skepticism is not statistically significant (F (1, 162) =
0.401, p (two-tailed) = 0.528). In fact, participants’ final judgments are virtually identical in
both skepticism groups (high skepticism: 56.75 vs. low skepticism: 56.78). Thus, no support
is provided for H2. The interaction effect between information order and trait professional
skepticism is not significant either (F (1, 162) = 0.112, p (two-tailed) = 0.738). Hence, in con-
sistency with the ANOVA results reported in the previous section, the effect of information
order was not found to change significantly at different levels of professional skepticism.
Specifically, among auditors classified as low in professional skepticism exhibited those who
were in the െ െ condition greater mean final beliefs (60.86) than subjects in the െ െ
condition (52.20). Consistent results were obtained for auditors classified as high in profes-
sional skepticism (mean final beliefs of 65.64 in the െ െ condition vs. 48.29 in the
െ െ condition). Hence, at both skepticism levels, information order was found to affect
subjects’ final beliefs in a manner consistent with a recency effect. It may be remarked that
even though the overall effect of information order on auditors’ final beliefs appears at first
glance to be much more stronger among high skepticism auditors (mean difference: 17.35)
compared to low skepticism auditors (mean difference: 8.67), this seems to have been caused
by differences in participants’ initial beliefs since after controlling for the influence of S 0, the
interaction effect was not found to reach statistical significance (see above).
Finally, in consistency with the notion underlying the belief-adjustment model that the start-
ing position serves as an anchor which determines the nature and magnitude of subsequent
adjustments, auditors’ mean final beliefs (S4) were found to be highly significantly influenced
by their initial assessments (S0) (F (1, 162) = 139.911, p < 0.0001). The effect size of S0 is
very large (η2 = 0.463).
Overall, the ANCOVA results corroborate the ANOVA results reported in the previous sec-
tion with regard to the main effects of information order and professional skepticism as well
as the interaction effect of these factors on auditors’ judgments.
As a final supplemental test, a linear regression analysis was conducted. The main aim of this
analysis was to explore the relationship between a set of demographic, effort-, and evidence-
related variables of interest and auditors’ going concern assessments.
The regression model tested herein was specified in two alternatives ways: In the first model
specification, subjects’ total belief revisions (BELREV) (S4െS0) were employed as the de-
pendent variable. In the second specification, the dependent variable was auditors’ final be-
183
liefs (FINBEL) (S4), and their initial beliefs (INTBEL) (S0) were included as a further control
variable. All other variables included in the supplemental analysis were identical between the
two model specifications. In particular, the test variables utilized include information order
(INFORD), professional skepticism (PS), and the interaction of both (INFORD*PS).977 Fur-
ther variables of interest concern subjects’ perceptions of the sufficiency of information
(SUFINF), self-reported certainty in the own assessment (CERTAIN), concentration (CONC),
and intensity of thought (INTENS). The rationale underlying the inclusion of these variables
was to provide some preliminary insights into the nature and strength of the relationship be-
tween uncertainty- and effort-related factors and auditors’ judgments. More specifically, as
might be recalled from Section 3.4, Kennedy (1993) views recency as an effort-related bias
that can be mitigated using effort-enhancing mechanisms such as documentation and account-
ability requirements. In this context, the present study sought to explore whether, when not
specifically manipulated or induced, effort (operationalized through self-reported concentra-
tion and intensity of thought exerted during the completion of the case study) provides a
meaningful contribution in explaining subjects’ responses (belief revisions and final beliefs).
A further question of research interest is whether and how strong do the perceived infor-
mation (in)sufficiency and (un)certainty in the own assessment, which have been generally
recognized to influence information processing, relate to auditors’ belief revisions and final
beliefs. Finally, participants’ general experience (GEXP),978 industry-specific experience
(IEXP), task-specific experience (TEXP), gender (GEN) as well as the individual assessments
of the importance of the additional information cues concerning the rejection of state funding
(IMPC1), the insolvency of a major supplier (IMPC2), the favorable external report and
banks’ willingness to negotiate (IMPC3), and the announcement of a cash capital increase
(IMPC4) were included as control variables in the regression analysis.
Overall, the regression models tested can be summarized as follows:
Model specification 1:
977
Note that in contrast to the ANOVA procedure, where professional skepticism represented a blocking (post
hoc median dichotomized) factor, in the regression analysis, a continuous measure of professional skepti-
cism was used: the mean-centered Hurtt score. In consistency with Quadackers/Groot/Wright (2012) and
following the best-practice recommendation by Aguinis/Gottfredson/Culpepper (2013): 1507 and
Dalal/Zickar (2012): 351f., mean centering of the Hurtt scores was performed in order to avoid collinearity
issues and to improve the interpretability of the results with regard to the interaction effect between
INFORD and PS. Mean centering did not alter the results.
978
It might be recalled from Section 5.5.2 that age, general audit experience, position in the firm hierarchy, and
auditor certification are strongly correlated. Hence, in order to avoid potential multicollinearity issues, only
general experience was included in the regression analysis. An inclusion of auditors’ age, position in the
firm hierarchy, and auditor certification as additional control variables in the regression analysis did not
qualitatively change the results reported below. Overall, none of the three variables was found to significant-
ly relate to auditors’ belief revision or final beliefs.
184
Model specification 2:
Table 15 and Table 16 report the correlations between the variables included in the regression
analysis for model specification 1 and 2, respectively. As not all of the variables satisfy the
parametric assumption of normal distribution, Spearman rank correlations are reported in ad-
dition to the Pearson product-moment correlation coefficients.979
979
Specifically, the distribution of scores on the GEXP, IEXP, and SUFINF variables are right-skewed, while
the scores on the IMPC1, IMPC2, IMPC3, IMPC4, and CONC variables are left-skewed.
185
186
BELREV INFORD PS SUFINF CERTAIN CONC INTENS GEXP IEXP TEXP GEN IMPC1 IMPC2 IMPC3 IMPC4
BELREV 1.000 -0.325*** 0.061 -0.124 -0.085 -0.073 -0.083 0.085 0.116 0.075 0.042 0.046 -0.183* 0.117 0.140
INFORD -0.332*** 1.000 0.034 0.019 -0.127 0.009 -0.021 -0.134 -0.108 -0.179* 0.044 -0.155* 0.108 0.056 0.137
PS 0.058 0.038 1.000 -0.141 -0.088 0.203** 0.151 -0.019 -0.065 0.094 0.098 0.167* -0.025 0.063 0.011
SUFINF -0.097 0.018 -0.091 1.000 0.392*** 0.197* 0.085 -0.060 0.006 -0.024 0.026 0.100 0.100 0.219** 0.134
CERTAIN -0.072 -0.131 -0.036 0.408*** 1.000 0.271*** 0.211** 0.169* 0.229** 0.279*** -0.214** 0.055 -0.017 0.044 0.101
CONC -0.039 0.031 0.237** 0.200** 0.281*** 1.000 0.723*** -0.060 -0.011 0.010 0.034 0.121 0.083 0.145 0.063
INTENS -0.057 -0.010 0.173* 0.100 0.216** 0.740*** 1.000 -0.021 0.031 0.085 0.009 0.004 -0.012 0.023 0.022
GEXP 0.044 -0.143 -0.030 0.036 0.179* -0.081 -0.048 1.000 0.298*** 0.357*** -0.185* -0.031 -0.151 0.018 0.026
IEXP 0.077 -0.112 -0.095 0.022 0.205** 0.017 0.092 0.311*** 1.000 0.328*** -0.172* -0.175* -0.207** -0.061 0.141
TEXP 0.079 -0.178* 0.102 -0.035 0.270*** 0.004 0.085 0.364*** 0.301*** 1.000 -0.351*** -0.085 -0.155* -0.038 0.000
GEN 0.024 0.044 0.115 0.035 -0.198* 0.032 0.012 -0.201* -0.167* -0.349*** 1.000 0.094 0.172* 0.037 0.100
IMPC1 0.073 -0.155* 0.106 0.105 0.126 0.137 0.030 -0.014 -0.161* -0.077 0.080 1.000 0.295*** 0.096 0.111
IMPC2 -0.148 0.124 -0.035 0.087 0.019 0.043 0.001 -0.132 -0.175* -0.132 0.170* 0.252** 1.000 0.095 0.076
IMPC3 0.156* 0.074 0.075 0.199** 0.002 0.124 0.017 0.019 -0.105 -0.032 0.082 0.101 0.107 1.000 0.216**
IMPC4 0.147 0.144 -0.022 0.149 0.067 0.049 0.041 -0.013 0.131 -0.013 0.103 0.116 0.132 0.225** 1.000
INFORD: dummy variable coded 1 if the participant is in the െ െ condition, and 0 otherwise. GEN: dummy variable coded 1 if the participant is female, and 0 oth-
erwise.
Pearson product-moment correlation coefficients in the lower triangle; Spearman rank correlation coefficients in the upper triangle
*, **, *** Significance levels at 0.05, 0.01, and 0.001, respectively (two-tailed tests)
FINBEL 1.000 -0.325*** -0.070 0.173* 0.145 0.079 -0.006 0.639*** 0.038 0.109 -0.110 -0.034 -0.027 -0.173* 0.307*** 0.282***
INFORD -0.301*** 1.000 0.034 0.019 -0.127 0.009 -0.021 -0.025 -0.134 -0.108 -0.179* 0.044 -0.155* 0.108 0.056 0.137
PS -0.026 0.038 1.000 -0.141 -0.088 0.203** 0.151 -0.108 -0.019 -0.065 0.094 0.098 0.167* -0.025 0.063 0.011
SUFINF 0.162* 0.018 -0.091 1.000 0.392*** 0.197* 0.085 0.264*** -0.060 0.006 -0.024 0.026 0.100 0.100 0.219** 0.134
CERTAIN 0.105 -0.131 -0.036 0.408*** 1.000 0.271*** 0.211** 0.210** 0.169* 0.229** 0.279*** -0.214** 0.055 -0.017 0.044 0.101
CONC 0.089 0.031 0.237** 0.200** 0.281*** 1.000 0.723*** 0.113 -0.060 -0.011 0.010 0.034 0.121 0.083 0.145 0.063
INTENS -0.001 -0.010 0.173* 0.100 0.216** 0.740*** 1.000 0.039 -0.021 0.031 0.085 0.009 0.004 -0.012 0.023 0.022
INTBEL 0.658*** -0.020 -0.071 0.233** 0.158* 0.117 0.046 1.000 -0.026 0.021 -0.147 -0.062 -0.075 -0.037 0.210** 0.166*
GEXP 0.004 -0.143 -0.030 0.036 0.179* -0.081 -0.048 -0.031 1.000 0.298*** 0.357*** -0.185* -0.031 -0.151 0.018 0.026
IEXP 0.121 -0.112 -0.095 0.022 0.205** 0.017 0.092 0.054 0.311*** 1.000 0.328*** -0.172* -0.175* -0.207** -0.061 0.141
TEXP -0.097 -0.178* 0.102 -0.035 0.270*** 0.004 0.085 -0.158* 0.364*** 0.301*** 1.000 -0.351*** -0.085 -0.155* -0.038 0.000
GEN -0.026 0.044 0.115 0.035 -0.198* 0.032 0.012 -0.043 -0.201* -0.167* -0.349*** 1.000 0.094 0.172* 0.037 0.100
IMPC1 -0.029 -0.155* 0.106 0.105 0.126 0.137 0.030 -0.087 -0.014 -0.161* -0.077 0.080 1.000 0.295*** 0.096 0.111
IMPC2 -0.176* 0.124 -0.035 0.087 0.019 0.043 0.001 -0.049 -0.132 -0.175* -0.132 0.170* 0.252** 1.000 0.095 0.076
IMPC3 0.329*** 0.074 0.075 0.199** 0.002 0.124 0.017 0.191* 0.019 -0.105 -0.032 0.082 0.101 0.107 1.000 0.216**
IMPC4 0.317*** 0.144 -0.022 0.149 0.067 0.049 0.041 0.186* -0.013 0.131 -0.013 0.103 0.116 0.132 0.225** 1.000
INFORD: dummy variable coded 1 if the participant is in the െ െ condition, and 0 otherwise. GEN: dummy variable coded 1 if the participant is female, and 0 oth-
erwise.
Pearson product-moment correlation coefficients in the lower triangle; Spearman rank correlation coefficients in the upper triangle
*, **, *** Significance levels at 0.05, 0.0,1 and 0.001, respectively (two-tailed tests)
187
An inspection of the correlation coefficients summarized in Table 15 and Table 16 suggests
that multicollinearity was not an issue as indicated by the fact that most correlations are below
0.3, with the only major exception being the correlation between CONCEN and INTENS
which is slightly above 0.7, but yet below the critical threshold value of 0.8 suggested in the
relevant literature.980 The conclusion of no multicollinearity is reinforced by the fact that the
variance inflation factors of the variables included in both model specifications are uniformly
less than 2.8, thus falling clearly below the critical value of 10.981 In addition to the multicol-
linearity investigation, the residuals scatterplots were examined to ensure the tenability of the
assumptions of normality, linearity, and homoscedasticity.982 All assumptions were found to
hold.
Before proceeding to the discussion of the results of the multiple regression analysis, the most
notable findings emerging from the inspection of the correlation matrices depicted in Table 15
and Table 16 are briefly considered. As expected, information order (INFORD) is significant-
ly related to both subjects’ belief revisions (BELREV) (r983 = -0.332, p < .0001) and final be-
liefs (FINBEL) (r = -0.301, p < .0001) in the theoretically predicted direction (i.e., involving
greater downward belief revisions (smaller final beliefs) for subjects in the െ െ condi-
tion (coded 1) compared to subjects in the െ െ condition (coded 0)). In addition, partici-
pants’ belief revisions are significantly correlated with the importance attached to the addi-
tional piece of positive information concerning the favorable external report and banks’
willingness to negotiate (IMPC3) (r = 0.156, p = 0.045). Similar correlation coefficients are
obtained for the importance assigned to the information cues concerning the insolvency of a
major supplier (IMPC2) (r = -0.148, p = 0.056) and the announcement of a cash capital in-
crease (IMPC4) (r = 0.147, p = 0.057). Unexpectedly, the correlation between the importance
assigned to the rejection of state funding (IMPC1) and subjects’ belief revisions is positive (r
= 0.073, p = 0.350), indicating that the greater importance subjects attached to this cue, the
greater (more upward) were their total belief revisions. However, when subjects’ final beliefs
(FINBEL) are considered, the direction of the association is negative (r = -0.029, p = 0.708)
and thus consistent with the intention of the experimental manipulation. Nonetheless, the cor-
relation coefficients presented in Table 16 corroborate the finding that the cue concerning the
rejection of state funding is the piece of information which is most weakly associated with
both dependent variables. The three other pieces of additional information significantly corre-
late with auditors’ final beliefs (IMPC2: r = -0.176, p = 0.023; IMPC3: r = 0.329, p < 0.0001;
IMPC4: r = 0.317, p < 0.0001). Strikingly, even though the manipulation check results dis-
cussed in Section 5.5.1 reveal a somewhat higher mean importance assessment of the two
980
See, e.g., Field (2009): 224; Harlow (2014): 1035.
981
See, e.g., Kleinbaum et al. (2008): 315; Myers (1990): 369. The tolerance statistic and variance inflation
factor values are not included in the summary of regression results (Table 17) but were checked to ensure
that multicollinearity did not bias the results.
982
For a concise discussion of the assumptions underlying multiple regression analysis, see, e.g., Field (2009):
220f.
983
For ease of exposure, only the relevant Pearson product-moment correlation coefficients are stated subse-
quently. The findings do not change qualitatively when the Spearman rank correlation coefficients are con-
sidered. All p-values listed above stem from two-tailed tests.
188
negative items of additional information compared to the two positive items (Table 6), the
correlation analysis results reported in Table 15 and Table 16 suggest that the negative items
(IMPC1, IMPC2) are more weakly related to subjects’ responses than are the positive cues
(IMPC3, IMPC4). However, it is important to note that the ex post indicated importance as-
sessments of the four additional information cues do not necessarily coincide with the mental
assessments of these cues generated in the course of the experimental treatment. Nevertheless,
the finding reported above is in line with the results of a number of prior auditing studies
which indicate that auditors tend to (subconsciously) overweight positive information (infor-
mation that confirms their positively-framed hypotheses) in updating beliefs and arriving at
judgments.984
An analysis of Table 16 also reveals that, as expected, participants’ initial beliefs (INTBEL)
are very strongly related to their final beliefs (FINBEL) (r = 0.658, p < 0.0001).985 Hence, the
higher (more optimistic) the starting position, the higher (more favorable) an individual’s fi-
nal belief in the going concern ability of the Premium Steel AG.
In addition, the following statistically significant correlations are worth noting: First, the more
concentrated subjects were (or assessed themselves) (CONC), the more sufficient they per-
ceived the information provided in the case study (SUFINF) (r = 0.200, p = 0.010), and the
more certain they were in their assessments (CERTAIN) (r = 0.281, p < 0.0001). Perceived
sufficiency of information and certainty in the own judgment were found to significantly and
positively correlate (r = 0.408, p < 0.0001). Second, remarkably, the greater importance par-
ticipants attached to the information concerning the favorable external report and banks’ will-
ingness to negotiate over the prolongation of the credit lines (IMPC3), the greater was partici-
pants’ perception of the sufficiency of the information provided in the case study for a
definite going concern evaluation of the Premium Steel AG (SUFINF) (r = 0.199, p = 0.010).
In addition, the higher subjects’ initial belief in the going concern ability of the firm
(INTBEL), the more sufficient they perceived the information provided in the case (r = 0.233,
p = 0.002), and the more certain they were in their own judgments (r = 0.158, p = 0.043).
Third, the greater subjects’ experience with the going concern task (TEXP), the lower were
their initial beliefs in the going concern ability of the firm (r = -0.158, p = 0.041). Fourth,
participants’ certainty in their own judgment was found to be positively associated with their
general experience (GEXP) (r = 0.179, p = 0.023), industry-specific experience (IEXP) (r =
0.205, p = 0.008), task-specific experience (TEXP) (r = 0.270, p < 0.0001), and the intensity
of thought exerted (INTENS) (r = 0.216, p = 0.005). Furthermore, in consistency with prior
984
See, e.g., Ayers/Kaplan (1993): 126; Bamber/Ramsay/Tubbs (1997): 250, 263; Church (1991): 531f.;
Morton (2001): 115f.; Pei/Reed/Koch (1992): 180; Waller/Felix (1984): 399. For a list of relevant psycho-
logical studies reporting profound confirmation proneness, see Smith/Kida (1991): 483.
985
Following Cohen (1988): 79-81, correlation coefficients ranging from 0.1 to 0.29 are classified as small,
from 0.30 to 0.49 as medium, and above 0.5 as large. However, as indicated by Fahrmeir et al. (2007): 139,
the nature of the variables under scrutiny should be taken into account. Specifically, for relatively exact
measures such as age, this rough classification appears tenable. For “soft measured” variables (characteris-
tics) such as responses on attitude and other ranking scales, however, obtained correlations are typically not
that strong, with coefficients of approximately 0.5 rather representing the maximum.
189
research findings indicating that women are generally less confident than men,986 the female
participants in the present study were found to be significantly less certain in their own judg-
ment than male participants (r = -0.198, p = 0.012). Finally, in consistency with the conten-
tions set forth at the end of Section 5.5.2, auditors’ general, industry-specific, and task-
specific experience was not found to be significantly associated with their belief revisions
(GEXP: r = 0.044, p = 0.574; IEXP: r = 0.077, p = 0.320; TEXP: r = 0.079, p = 0.309) or final
beliefs (GEXP: r = 0.004, p = 0.957; IEXP: r = 0.121, p = 0.119; TEXP: r = -0.097, p =
0.210).
Subsequently, the results of the multiple regression analysis are considered (Table 17).
986
See, e.g., OECD (2013): 56f.; Sonnert/Holton (1995): 145.
190
As shown in Table 17, both model specifications are statistically significant, with F (15, 143)
= 2.876, p = 0.001, adjusted R2 = 0.151 for Model 1 and F (16, 142) = 16.841, p < 0.0001,
adjusted R2 = 0.616 for Model 2.987 As expected, the regression analysis generated the same
results regarding the three test variables as the results obtained using the AN(C)OVA proce-
dures. In particular, the regression results reveal that information order has the strongest
unique and statistically significant contribution to explaining both participants’ belief revi-
sions (BELREV, Model 1) (β = -0.355, p < 0.0001) and final beliefs (FINBEL, Model 2) (β = -
0.351, p < 0.0001), while professional skepticism provides virtually no contribution to ex-
plaining the variance in the dependent variables (β = -0.034, p = 0.740 and β = 0.016, p =
0.815 for BELREV and FINBEL, respectively). Similarly, skepticism was not found to interact
significantly with information order in explaining auditors’ total belief revisions (β = 0.120, p
= 0.234) or final beliefs (β = 0.021, p = 0.763).
With regard to the other variables of interest (SUFINF, CERTAIN, CONC, and INTENS),
none was found to be significantly associated with auditors’ belief revisions or final beliefs.
Likewise, the results do not reveal a significant relationship between subjects’ experience
(GEXP, IEXP, and TEXP) or gender (GEN) and the two dependent variables. The only signif-
icant relationships between control measures and the dependent variables concern the im-
portance assigned to the two positive cues in Model 1 (IMPC3: β = 0.185, p = 0.022; IMPC4:
β = 0.180, p = 0.026) as well as in addition the negative cue concerning the insolvency of a
major supplier (IMPC2) in Model 2 (IMPC2: β = -0.150, p = 0.006; IMPC3: β = 0.222, p <
0.0001; IMPC4: β = 0.242, p < 0.0001). Finally, as expected, the regression results show that
the INTBEL variable has a very strong explanatory power regarding auditors’ final beliefs (β
= 0.536, p < 0.0001).
Overall, it can be concluded that inferences concerning the hypotheses under consideration
are robust to different parametric tests (ANOVA, ANCOVA, and regression analysis) as well
as different specifications of the dependent variable (total belief revision vs. final belief and
control for initial belief). Collectively, the results of the statistical tests show that auditors
who evaluate positive information at the end of the sequence indeed exhibit significantly
greater upward belief revisions (and more favorable final beliefs) than auditors who evaluate
negative information at the end of the series. Hence, H1 can be supported. No significant dif-
ference was found in the belief revisions (and final beliefs) of auditors at different levels of
trait professional skepticism. Hence, there is no support for H2. Finally, the overall results
indicate that the interaction between information order and professional skepticism is not sta-
tistically significant. With other words, evidence of recency was obtained for both high and
low levels of trait professional skepticism, suggesting that information order does not influ-
987
It is instructive to note that behavioral studies normally account for less variance than do non-behavioral
(e.g., archival) studies because human behavior in the natural environment is influenced by many (frequently
latent) factors that are not under the researcher’s control. See, e.g., Tabachnick/Fidell (2007): 55. Therefore,
effect size and R2-values are not considered of central interest in behavioral research and are typically not
reported. One of the very few studies reporting R2-values is Quadackers/Groot/Wright (2012). They also
conducted a linear regression analysis and obtained adjusted R2-values ranging from 0.00 to 0.262 for their
individual models. See Quadackers/Groot/Wright (2012): 27.
191
ence auditors at different levels of skepticism differently. Hence, the results of the present
study suggest a negative answer to the research question.
Before proceeding to the discussion of these findings and their implications, the results of a
set of additional analyses are presented and briefly discussed in the subsequent section. These
analyses were conducted to explore (i.e., test without formally stated hypotheses) the associa-
tion between trait professional skepticism and auditors’ evidence-related attitudes, judgment
confidence, and cognitive effort in the context of a critical, complex, forward-looking, and
uncertainty-involving judgment task (going concern) in which no perfectly exhaustive infor-
mation was available.
As discussed in Section 4.4, three of the six constituent elements underlying Hurtt’s concept
of trait professional skepticism and the Hurtt scale – questioning mind, suspension of judg-
ment, and search for knowledge – are evidence-related and involve an auditor’s tendency to
search for and evaluate sufficient audit evidence before forming a final belief or judgment.
Prior auditing research using the Hurtt scale has typically utilized experimental settings in
which auditors can actively search for and obtain additional information and audit evidence.
In contrast to previous studies, in order to draw valid inferences on the emergence of recency
effects in auditors’ belief revisions, a fixed experimental setting was utilized herein, i.e., par-
ticipants were not able to obtain information beyond the data presented in the case materials.
Hence, the present investigation offers the opportunity for some relevant preliminary insights
into the link between trait professional skepticism and auditors’ responses to limitedly availa-
ble evidence.
Theoretically, auditors possessing higher degrees of trait professional skepticism are expected
to perceive the evidence provided in the current case study as less sufficient for a conclusive
evaluation of the going concern ability of the Premium Steel AG than their less skeptical
counterparts. This expectation is based on the questioning mind and search for knowledge
dimensions of trait skepticism.988
A correlation analysis was performed to test whether trait professional skepticism is inversely
related to auditors’ perceptions of information sufficiency. In consistency with the foregoing
theoretical considerations, a statistically significant negative association was found between
participants’ scores on the Hurtt scale and their perceptions of the sufficiency of information
(rs989 = -0.141, p (one-tailed) = 0.035). However, the correlation is small, with the amount of
988
For a general consideration of these dimension, see Section 4.4.2.1 and Section 4.4.2.3
989
Due to violations of the parametric assumption of normality by several of the relevant tested variables (suf-
ficiency of information, number of additional cues listed, general and industry-specific experience as well as
the importance assignments of all four additional information pieces), the nonparametric Spearman rank cor-
relation coefficients are reported in this section.
192
shared (i.e., explained) variation between both variables being below 2% (coefficient of de-
termination rs2 = 1.99%).
In addition, dispositional skepticism is expected to display a positive association with audi-
tors’ need for additional information for arriving at a definite going concern evaluation.
Again, this expectation arises from the questioning mind and search for knowledge compo-
nents of trait professional skepticism.
An analysis of the bivariate correlation between the Hurtt score and the number of additional
information cues listed by participants as necessary for a conclusive going concern assess-
ment990 reveals that there is a marginally significant positive association between the two var-
iables (rs = 0.110, p (one-tailed) = 0.078). However, it is conceivable that respondents’ indica-
tion of additional information cues required for a conclusive assessment of the going concern
likelihood of the entity also relates to individual knowledge (approximated via general, indus-
try-, and task-specific experience) and the amount of cognitive effort exerted (approximated
via self-reports on concentration and intensity of thought during the completion of the case
study). However, neither experience nor effort were found to significantly correlate with the
number of additional cues indicated as necessary for a definite going concern evaluation. 991
A further evidence-related explorative test concerns the inspection of whether subjects’ as-
sessments of the importance of the four additional information cues differ between the two
skepticism levels. Following Hogarth and Einhorn’s (1992) notion of skepticism, more skep-
tical subjects can be expected to attach substantially more weight to negative information than
to positive evidence. Table 18 provides a summary of the mean importance assessments by
skepticism level and t-test results of differences between the two skepticism groups.
Bankruptcy of a major supplier and lack of െ 5.43 5.33 0.10 0.555 0.580
alternatives
Favorable external report and banks willing 4.76 5.01 -0.25 -1.150 0.252
to negotiate
Announcement of a cash capital increase 4.87 4.91 -0.04 -0.178 0.859
990
The number of information items listed by each respondent was independently counted by the author and by
an audit manager who was unfamiliar with the rationale underlying the free formulation question. All inter-
rater differences were discussed and mutually resolved. On average, subjects were found to list 2.11 pieces
of additionally required information (SD: 1.93). The range of items was from 0 to 9.
991
Specifically, the obtained bivariate correlations between the number of listed cues and the relevant
knowledge- and effort-related variables are as follows: general experience: rs = 0.071, p (one-tailed) = 0.184;
industry experience: rs = -0.046, p (one-tailed) = 0.278; task experience: rs = 0.102, p (one-tailed) = 0.094;
concentration: rs = -0.034, p (one-tailed) = 0.332; and intensity of thought: rs = 0.069, p (one-tailed) = 0.187.
193
The results presented in Table 18 indicate that participants’ evidence assessments do not dif-
fer significantly between subjects in the high skepticism group and subjects in the low skepti-
cism group. This result holds for both positive and negative evidence cues.992
Overall, trait professional skepticism was found to be only weakly (negatively) related to the
perceived sufficiency of evidence. The number of cues listed as necessary for a definite going
concern evaluation as well as the importance ratings of the four additional cues were not
found to significantly differ between high and low skepticism auditors. With other words,
auditors’ evidence-related responses appear only very moderately associated with their dispo-
sitional skepticism. These findings can be viewed as supporting the notion that as a trait, pro-
fessional skepticism is potentially only subtly manifested in human behavior. This aspect will
be discussed in Section 6.1.
A further aspect that was exploratively tested within the current study concerns the relation
between trait skepticism and judgment confidence (certainty in the own judgment). Overall,
an inverse relation can be expected between dispositional skepticism and judgment confi-
dence because of a skeptic’s ongoing questioning which, as indicated in Section 4.4.2.1, is not
merely restricted to extraneous assertions and claims but also concerns the own judgments
and beliefs.993
The relationship between trait professional skepticism (as measured by the Hurtt score) and
participants’ self-reported certainty in the own judgment was tested using bivariate correla-
tion. The results obtained suggest that although higher levels of skepticism are tendentially
associated with lower levels of judgment confidence, the relationship between the two varia-
bles is not statistically significant (rs = -0.088, p (one-tailed) = 0.131).
Finally, as indicated in Section 4.5, innate skepticism can be expected to correlate with great-
er effort, attention, and thoroughness in information processing. This contention is based on
skeptics’ ongoing questioning and insatiable need for information and proof, which generally
calls for cognitive resources and deliberation.
In consistency with this notion, the results of the conducted correlational analysis indicate that
skepticism as measured by the Hurtt score and effort as measured by participants’ self-
reports994 on the concentration and intensity of thought exerted during the completion of the
992
Similar results were obtained when the bivariate correlation between the Hurtt skepticism score (continuous
variable) and the importance assigned to the additional information cues was considered.
993
The view that there is an inverse relationship between skepticism and confidence is also shared by, e.g.,
Fuchs (2001): ii and Stren (2001): 133.
994
Even though self-reports may bear reliability issues (see, e.g., Cone/Foster (1998): 245f.), they have been
frequently applied as measures of effort in prior research. See, e.g., Payne/Ramsay (2005): 324.
194
case study are significantly positively related (concentration: rs = 0.203, p (one-tailed) =
0.004; intensity of thought: rs = 0.151, p (one-tailed) = 0.026). However, the obtained correla-
tions are small, with the amount of shared variation (rs2) between skepticism and concentra-
tion as well as intensity of thought being 4.12% and 2.28%, respectively.
In addition, the time expended for completing the study was considered as a further proxy for
effort. The underlying notion is that auditors exerting higher effort will need more time to
complete the case materials. The correlation coefficients obtained for the relationship between
time and intensity of thought (rs = 0.327, p (one-tailed) < 0.0001) as well as time and concen-
tration (rs = 0.346, p (one-tailed) < 0.0001) appear to support this expectation. However, the
relationship between trait skepticism as measured by the Hurtt score and completion time was
found to be weak and only marginally significant (rs = 0.141, p (one-tailed) = 0.064).
In addition to cognitive effort, the relation between skepticism and cognitive load was ex-
plored. As previously indicated, trait professional skepticism has been theorized to involve
ongoing questioning as well as profound, critical, and elaborative information processing.
These cognitive operations require and consume substantial cognitive resources and can lead
to high cognitive load. The expected positive relationship between skepticism and cognitive
load is directionally supported by the correlation coefficient obtained between the Hurtt score
and the self-reported perceived difficulty of the case (rs = 0.106, p (one-tailed) = 0.088).
However, the obtained correlation is only marginally significant.
To summarize, the results of the foregoing analyses suggest that trait skepticism is indeed
positively related to cognitive effort. However, the obtained correlation coefficients indicate a
relatively weak link between the two constructs. In addition, the results reported above sug-
gest that with increasing degrees of innate skepticism, auditors are tendentially more likely to
perceive the task as more difficult, which is viewed as an indication of increased cognitive
load. This aspect needs to be more profoundly explored by future research, since it relates to a
potential downside of skepticism and thoroughness in information processing: the occupation
of considerable cognitive resources, substantial cognitive load, and high likelihood of revert-
ing to heuristic processing to relieve cognitive strain.
Collectively, the results of the explorative analyses concerning the behavioral manifestations
of trait professional skepticism in the specific context of the present study suggest that alt-
hough auditors’ evidence-, confidence-, and effort-related responses are at least tendentially in
line with the theoretical expectations, the relationship between trait skepticism and the other
variables of interest is relatively weak. This finding might be attributed to the dispositional
nature of skepticism foundational to the Hurtt concept and her scale as well as the fact that
traits have generally been found to more subtly manifest themselves in human behavior than
states.995 Support for the dispositional nature of skepticism as measured by the Hurtt scale is
provided by the finding that the Hurtt score is not correlated with the attendance of skepticism
trainings in the past (rs = 0.069, p (one-tailed) = 0.189). This result suggests that trait skepti-
995
See Robinson/Curtis/Robertson (2013): 7, 27; Ross/Nisbett (2011): 27f.
195
cism is a stable, enduring, and cross-situationally consistent personality characteristic rather
than an acquired skill.
Finally, the results of the debriefing questions are briefly discussed. As indicated in Section
5.3.3, at the end of the experimental study, participants were asked to rate the understandabil-
ity of the research instrument. The mean assessment of the understandability of the research
instrument was 5.48 (SD: 1.18), which is significantly above the theoretical mid-point of the
response scale (4.5) as indicated by the results of a one-sample t-test (t (162) = 10.674, p
(two-tailed) < 0.0001). Accordingly, it can be concluded that participants perceived the case
materials as reasonably comprehensible.
In addition, participants were asked to provide comments on the study. Only twelve partici-
pants (7.19%) provided comments and remarks, the most notable of which are briefly consid-
ered subsequently. Specifically, three participants expressed interest in the research project
and curiosity about the aim and the results of the study. This fact underscores the validity of
the experiment. It suggests that participants were not able to anticipate the objective of the
study, so that it can be expected that they provided genuine responses. In addition, partici-
pants’ interest in the study suggests that the striven experimental realism of the case was
largely achieved. Two participants stressed the fact that a case study is merely a simplification
of reality and can never provide the detailed information and mirror the true complexity faced
in audit practice. One participant complained about the tight time frame of the study. One
participant remarked that this is a classic case for an emphasis of matter paragraph in auditors’
report as the going concern ability of the firm appears more likely than not, but is still not
certain. Finally, three participants emphasized the uncertainty inherent in the additionally
provided information cues and the lack of detail on them.
196
6 Conclusion
In this concluding section, the experimental results reported above are synthesized and dis-
cussed, the key implications of the findings are considered, the limitations of the present work
are outlined, and potential avenues for future research are suggested.
The purpose of the present thesis was two-fold: first, to provide further evidence on the influ-
ence of information order effects on auditors’ belief revisions, and second, to explore whether
auditors exhibiting different degrees of trait professional skepticism adjust their beliefs in a
different manner and are to a different extent subject to information order effects.
These objectives were achieved through the conduct of an experimental investigation with
experienced professional auditors. The experiment involved auditors’ sequential assessments
of the likelihood of a firm’s going concern ability. The order of information presentation was
manipulated between subjects by randomly assigning participants to either a െ െ condi-
tion or a െ െ condition. Based on the individual scores on the Hurtt scale, subjects were
post-experimentally classified as high or low in trait skepticism based on a median split of the
sample.
Overall, the results of the present study support the pervasiveness of information order effects
(recency) in the context of the sequential processing of complex, mixed evidence. The exper-
imental results reported herein suggest that while auditors’ belief revisions and overall judg-
ments are significantly influenced by the order in which subject obtain and process infor-
mation, an individual’s trait professional skepticism does not mitigate this bias. With other
words, auditors exhibiting high degrees of dispositional skepticism and those exhibiting low
levels of dispositional skepticism likewise fall prey to recency effects in their belief revisions.
Consequently, recency appears to dominate trait professional skepticism in participants’ re-
sponses.
The results obtained in the present study are in line with the findings reported in prior psycho-
logical and auditing research which unequivocally indicate that recency is a strong, highly
persistent, hard-wired, and ubiquitous cognitive bias that operates at a subconscious level,
which renders its recognition and alleviation quite difficult. The results reported here suggest
that a high level of trait skepticism is not sufficient to debias recency. This finding contrasts
with the results reported by Chan (1995) who, as discussed in Section 3.4, found that the cog-
nitive style of field independence (i.e., a personality trait) mitigated recency effects in audi-
tors’ belief revisions.996 However, only 20 auditors participated in his study, which is clearly
insufficient for drawing any confident inference.997 As indicated in Section 3.4, other studies
996
See Chan (1995): 26.
997
Recall that in the relevant literature, it has been recommended to use a cell size of at least 30 subjects. In the
case of four experimental cells of equal size, the total sample should consist of at least 120 subjects, i.e., 100
(!) subjects more than actually used by Chan (1995).
197
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2_6, © Springer Fachmedien Wiesbaden 2015
exploring the interrelation between information order effects and personality traits are miss-
ing. Prior research has generally indicated that effort-fostering environmental factors like
documentation requirements and accountability can, when particularly emphasized, alleviate
recency. Taken together, the results of the current study and prior research suggest that envi-
ronmental factors might be more effective in moderating recency effects than personal dispo-
sitions. This contention is in line with the findings in both psychological and auditing research
which suggests that states, i.e., temporary conditions evoked by situational factors, typically
have a greater influence on human judgments and behaviors (and their distortions) than
traits.998
While these findings and conclusions do not in any way diminish the fundamental importance
of professional skepticism in auditing, it may be that trait skepticism is more effective as a
corrective mechanism in “induced” situations where inconsistencies and red flags are present
and a state of professional skepticism is evoked rather than in cases involving cognitive bias-
es. This contention is in line with Robinson (2011) who argues that traits and states might be
useful for explaining different types of human behavior.999
Subsequently, the implications of the findings of the present study for the auditing profession,
standard setters and regulators as well as academic research are discussed. From a scientific
perspective, the suggestions and recommendations outlined subsequently follow a normative
approach, with the ultimate aim being to improve the quality of auditor judgment and to ap-
proach the normative benchmark of unbiased information processing.1000
Auditing Profession
The finding that auditors are systematically affected by the order of information presentation
in forming beliefs and making judgments regardless of their degree of innate skepticism im-
plies that an increased awareness is required in order to eliminate or at least reduce the delete-
rious influence of the subconsciously operating information order effects. 1001 Given that the
greatest portion of audit evidence and information is obtained from the client who is in the
position to resume and present information in a way that induces the most favorable position
for the firm, and in light of the fact that “last-minute” positive information and good news are
998
See McGhee/Shields/Birnberg (1978): 692-695; Mischel (1968): 146; Robinson (2011): 89;
Robinson/Curtis/Robertson (2013): 7, 27; Ross/Nisbett (2011): 27f. Bonner (2008): 101 even goes one step
further and argues that “personality characteristics may be small potatoes to accounting JDM.” Buss (1989):
1380f., in contrast, argues that paradigms may be constructed and empirical evidence provided both in favor
of a trait or a state view of personality. However, he considers the comparison and attempts to identify the
superiority of one aspect over the other as not really purposeful since both traits and states matter in human
behavior and only their joint consideration provides a more complete picture of the antecedents of judgment
and decision making.
999
See Robinson (2011): 17.
1000
See Stahl (2012): 135.
1001
This view is also shared by Adam (2007): 263. In addition, it is in line with Kahneman’s (2011): 28 notion
that cognitive illusions are very difficult to overcome and the most effective strategy for alleviating them
appears to be to learn to identify circumstances in which cognitive fallacies are likely, i.e., to sharpen one’s
awareness of bias-inducing factors.
198
common in practice,1002 an increased sensibility to the biasing influence of information order
effects appears crucial. In this context, Nelson (2009) suggests that the awareness and antici-
pation of order effects can serve as an opportunity to enhance skepticism by mentally ordering
and processing information in such way that any potential recency effect favors rather than
inhibits professional skepticism.1003 On a more general level, a heightened degree of inward
skepticism, i.e., skepticism directed toward the own judgment processes and their susceptibil-
ity to cognitive traps, seems critical for improving judgment quality and forming sound, well-
founded, and justifiable beliefs.1004 This is particularly important in the context of complex
and ambiguous audit issues and tasks because high complexity implies great cognitive load
and is thus likely to force individuals toward mental shortcuts and cognitive biases in infor-
mation processing.1005
In a nutshell, awareness is the most effective mechanism for preventing judgment from being
distorted by cognitive flaws. By being aware of cognitive traps and understanding well their
causes, typical situations where they are likely to arise might be better recognized and the
biasing influence might be prevented.1006
To date, however, the issue of cognitive biases in the context of auditors’ judgment and deci-
sion making has received very rudimentary consideration and modest attention within the
auditing profession. To the best of the author’s knowledge, only one of the Big 4 audit firms
explicitly addresses the issue of cognitive biases and potential mitigating strategies in its in-
ternal materials and proceedings.1007 Overall, cognitive biases seem to be largely disregarded
in audit practice, and it is not clear whether the lack of practical consideration is due to insuf-
ficient knowledge and awareness or due to the perceived irrelevance of cognitive phenomena
by the auditing profession. In either case, the results of prior research and the present study
suggest that cognitive biases may considerably impair the quality of auditors’ judgments and
thus represent an issue of practical concern. Consequently, it might be beneficial for audit
firms to extend the scope of their training materials and programs beyond the provision of
technical knowledge and skills to include training and guidance on sound judgment processes.
Prompts and reminders integrated into audit procedures (e.g., reminders to be aware of con-
trast effects and tendencies to over- or underweight information depending on its serial posi-
tion in a sequence of inconsistent evidence, to consider counterarguments, etc.) can help sus-
taining auditors’ awareness of cognitive fallacies throughout the course of the audit, thereby
alleviating biases in auditors’ beliefs and judgments.1008 Further general strategies for enhanc-
ing judgment quality discussed in the literature include, among others, the use of judgment-
1002
See Asare (1992): 391; Schreiber (2000): 174.
1003
See Nelson (2009): 2.
1004
The effectiveness of inward skepticism in mitigating fallacies in auditors’ information processing and judg-
ment has been demonstrated by Grenier (2013): 24.
1005
See Salterio/Koonce (1997): 573.
1006
See Glover/Prawitt (2013): 12; Kahneman (2011): 28; KPMG (2011): 31. For a general review on debiasing
as a strategy for improving judgment quality, see Bazerman/Moore (2013): 216-219.
1007
See KPMG (2011): 5-37.
1008
See Glover/Prawitt (2013): 12.
199
and decision-analysis tools in very complex judgment domains, analogical reasoning, taking
an outsider’s view, and understanding biases in others.1009
Standard Setters
Auditing standards do not provide any guidance on how individual pieces of evidence shall be
weighted and integrated into an overall belief or judgment. With more specific and stringent
judgment process guidance, the potential for bias might be considerably reduced.
With regard to professional skepticism, standard setters and regulators around the globe inten-
sively address the fundamentality of this concept in auditing and univocally call for improve-
ments in the application of skepticism in audit practice but fail to provide a straightforward
description of what professional skepticism essentially is. This conceptual indefiniteness casts
ambiguity concerning the proper application of professional skepticism in auditing. Hence,
based on the results of the theoretical analysis conducted in Section 4 and in consistency with
Glover and Prawitt (2013), it is recommended that standard setters work toward the achieve-
ment of a coherent definition and understanding of professional skepticism and how it has to
be applied to ensure high audit quality.1010 In this context, insights provided by recent aca-
demic studies are crucial.1011
Academic Research
The present study provides further support for the descriptive validity of the belief-adjustment
model. It adds to the body of psychological and auditing research demonstrating the robust-
ness of recency effects in human judgment. In addition, it offers important insights into the
behavioral manifestations of trait professional skepticism in a context that does not involve
some skepticism “prompts”. Overall, the fact that the results presented herein regarding the
debiasing capacity of the personality trait of skepticism contradict the findings reported by
Chan (1995) highlights the necessity to conduct well-designed studies with adequate sample
sizes in order to draw valid inferences. The potential avenues for future research are discussed
in the subsequent section along with the limitations of the present study.
As with all empirical research, the present study is subject to several limitations that should be
taken into consideration when interpreting the study’s results and findings. External validity
issues typically associated with experimental research, e.g., a simplified judgment setting
which contains limitedly available information, no client-specific knowledge, no possibility to
discuss the issue in the audit team as well as potentially weaker cognitive involvement of the
1009
For a detailed discussion of these strategies and a consideration of additional mechanisms for improving
judgment quality, see Bazerman/Moore (2013): 206-229.
1010
See Glover/Prawitt (2013): 22.
1011
See Franzel (2013).
200
subjects as compared to real-world auditing tasks, apply to this study.1012 Hence, the results
reported herein might not be generalizable beyond the context of the present study. In addi-
tion, it should be taken into consideration that the results reported in the present work were
obtained using a sample which was not randomly drawn. Specifically, auditors from a single
Big 4 audit firm participated in the present study. Thus, the results are not statistically gener-
alizable to the auditor population in general. The use of a single company, however, ensures
constancy with regard to training practice, organizational structure, and firm culture, thereby
reducing the threat of confounding effects related to firm-specific factors.1013 In addition, as
the pattern of results documented in the present study fits well into the picture drawn by prior
research, the non-statistical generalizability of the results beyond the context and subject pool
employed herein appears reasonable.1014
A further potential caveat concerns the threat of relevant omitted variables that might have
affected the results of the study. Of particular relevance to the present study appear the per-
sonal dispositions of tolerance for ambiguity, need for cognitive closure, and locus of control
since they concern the manner in which individuals process information, form beliefs, and
arrive at judgments under conditions of uncertainty. Hence, these personality factors might
have influenced auditors’ belief revisions. In addition, as discussed in Section 4.4.2, these
dispositions are strongly related to trait professional skepticism and might have moderated its
effect on auditors’ beliefs and judgments. However, as the distinct scales measuring tolerance
for ambiguity, need for cognitive closure as well as locus of control are quite comprehensive,
and in light of the time constraints for the conduct of the experiment, the three scales were not
employed in the present study. Future research might address the interrelations between these
personality traits and information order effects in the context of auditors’ information integra-
tion and belief revision processes.
In addition, the present study is limited to the consideration of trait professional skepticism
and its interrelation with information order effects in the context of auditors’ belief revisions.
However, it does not allow for inferences to be made regarding the influence of variations in
professional skepticism resulting from environmental factors and changing incentives on au-
ditors’ belief revisions. With other words, it is unclear whether and how state professional
skepticism would affect auditors’ belief revisions in the context of sequentially presented
complex and inconsistent evidence. This issue might be addressed by future research.
An additional potential caveat of the study concerns the translation of the Hurtt scale. Specifi-
cally, although great effort and thoroughness were exerted to ensure the reasonableness and
precision of the German translation of the scale, it is still possible that there are some subtle
discrepancies in the nuances of the original items contained in the Hurtt scale and their Ger-
man translations which could have influenced the trait skepticism scores and thus the results
obtained and inferences drawn within this work.1015 However, the Hurtt scores obtained in the
1012
See Asare/Cianci/Tsakumis (2009): 233; Cushing/Ahlawat (1996): 111, 121.
1013
See Bierstaker/Wright (2001): 52; Shaub/Lawrence (1996): 154.
1014
See Keppel (1973): 28f.
1015
For such claims in the context of cross-cultural research in auditing, see Cohen/Pant/Sharp (1995): 59.
201
present study are in line with prior research using the scale. This can be interpreted as an indi-
cation of the substantial adequacy of the translation of the Hurtt scale.
Finally, as indicated in Section 5.5.2, in spite of random assignment of participants to treat-
ment conditions, the െ െ group and the െ െgroup differed on a number of demo-
graphic variables. However, the supplemental tests conducted to control for the influence of
these demographic factors on auditors’ belief revisions did not reveal significant effects.
Following the consideration of the potential limitations of the study, in the subsequent para-
graphs, a number of directions that auditing researchers can pursue in the future are outlined.
With regard to auditors’ belief updating, one potentially fruitful avenue for future research
concerns the effect of information order on the amount of audit evidence collected to support
the individual belief (position). Of particular interest here is the influence of the first piece of
information on auditors’ evidence search activities. Specifically, it is unclear whether auditors
who obtain a piece of positive (negative) evidence at the outset prematurely stop searching for
additional information and accept (reject) the client’s propositions. Empirical evidence on this
question would be very valuable, both theoretically and practically, because the extent and
intensity of auditors’ information search proceedings directly affects the efficiency and effec-
tiveness of the audit.1016 In addition, valuable novel insights into the processes and determi-
nants underlying information aggregation and belief updating may be gained through the ap-
plication of refined models of human belief adjustment that take into account the deficiencies
of the Hogarth and Einhorn model (e.g., the quantum inference model by Trueblood and
Busemeyer (2010)).
With regard to professional skepticism, several possible directions for future research can be
identified. To begin with, future research might investigate the interaction between trait pro-
fessional skepticism and some other, potentially less pervasive, cognitive fallacies on audi-
tors’ skepticism, thereby providing additional insight into the issue addressed by Nelson
(2009) regarding the deleterious influence of cognitive traps on auditors’ skepticism.
An additional promising area for future research concerns the further exploration of the be-
havioral manifestations of trait professional skepticism. Of particular interest here is evidence
on the strength of the link between trait skepticism and skeptical judgment and behavior, es-
pecially in contexts which do not involve irregularities, contradictions, fraud, or other envi-
ronmental factors theorized to induce a state of professional skepticism. Relatedly, future re-
search on the state dimension of professional skepticism as well as on the interplay between
trait and state skepticism appears of central importance. This contention is reinforced by the
fact that in consumer research, it has been demonstrated that dispositional (trait) and situa-
tional (state) skepticism have independent and additive effects on human judgment and be-
havior. Specifically, while all subjects are generally found to be affected by skepticism-
inducing situational factors, individuals with high trait skepticism seem to respond more sen-
sitively to the particular environmental stimuli than individuals with low trait skepticism.1017
1016
See Asare/Messier (1991): 90.
1017
See Forehand/Grier (2003): 355.
202
This interdisciplinary finding underlines the importance of obtaining a deeper comprehension
of the isolated and combined effects of trait and state skepticism in auditors’ judgment and
decision making. For achieving this goal, the development of a rigorous psychometric scale
for ex ante measuring the state dimension of skepticism is crucial.1018
A further promising avenue for future research concerns the exploration of the interrelation
between trait professional skepticism and other individual dispositional characteristics, e.g.,
risk attitudes, field dependence, and moral development. Specifically, it is worth exploring
whether and under which circumstances skepticism and the personality features outlined
above reinforce or counteract each other and how this potential interaction influences audi-
tors’ behavior.
An additional skepticism-related aspect which deserves to be addressed in future research is
the exploration of professional skepticism in a multi-person setting. As prior research has
consistently indicated that group judgments frequently differ from individual judgments,1019 it
will be intriguing to see whether these findings also apply to professional skepticism. In par-
ticular, are audit teams more skeptical in their overall judgments and decisions as compared to
individual judgment and decision making? If so, is this due to the communication, interaction,
idea exchange, synergy effects,1020 or some other potentially relevant effects? Or do the costs
of group work (e.g., desire to be in line with the other team members, even if they are not suf-
ficiently skeptical) within audit teams overweight the potential benefits? Hence, it is worth
investigating whether and under which circumstances audit teams have the potential to under-
pin an individual’s professional skepticism, to alleviate cognitive biases typically found in
individual judgment, and ultimately to improve judgment quality and overall performance.
Group-related research can be of great relevance and benefit to audit practice as it may pro-
vide suggestions for purposeful and more effective team composition.
1018
See Hurtt (2010): 166. Hurtt’s call for future research focusing on the conceptualization of a state skepticism
scale has recently been addressed by Robinson (2011) and Robinson/Curtis/Robertson (2013) who adapted
the Hurtt scale to derive a distinct instrument that captures state professional skepticism. In brief, they
adopted the procedure used by Kluemper/Little/DeGroot (2009): 218 in their study of optimism and modi-
fied some of the evidence-related items contained in the Hurtt scale by adding concretizing phrases like
“while completing this case” at the outset of each item. As a result, the authors arrived at a 12-item scale
with a Cronbach’s alpha value of 0.86. An instance of a state-skepticism item reads as follows: “While com-
pleting this case, I waited to make decisions until I could get more information.” The response scale utilized
by Robinson (2011) and Robinson/Curtis/Robertson (2013) ranges from 1 (“I don’t agree at all”) to 7 (“Fully
agree”), and hence it differs from the response measure employed by Hurtt (2010). However, the authors do
not provide arguments for this deviation. For a description of the rationale of the scale development and the
results of the validation procedures, see Robinson (2011): 43-48 and Robinson/Curtis/Robertson (2013): 20f.
The entire scale is available on p. 52f. of the latter paper. Overall, it can be argued that even though the at-
tempt to shed light on state skepticism and to develop an instrument to operationalize it clearly deserves re-
spect and encouragement, it should be noted that the understanding of professional skepticism can only be
effectively expanded through thorough and extensive theoretical and empirical analyses, not a haphazard and
eclectic approach.
1019
See Section 2.4.2.2.4 for a review.
1020
Synergy effects are a major benefit arising from the hierarchical audit team structure where each member of
the team brings in his/her distinct, unique experience, thus contributing to a broader knowledge and idea
pool on which current audit judgments and decisions can be drawn. See Carpenter (2007): 1124.
203
Overall, future research must adopt a broad and multidisciplinary perspective because in light
of the multifacetedness and complexity of auditor judgment and professional skepticism “no
less a perspective will do”1021. The experience of the past decades has shown that research
from a number of other disciplines, including social and cognitive psychology, philosophy,
organizational science, consumer behavior, and forensic economics, has proven extremely
helpful in informing auditing research about relevant aspects of human information pro-
cessing and the role of personal dispositions and characteristics in this context. In addition,
considerable scientific advance can be expected from current progress in neuroscience which
can help to specify the neural substrates for the elementary information processes of the hu-
man mind, thereby opening the “black box” of human cognition and thought.1022
Finally, it can be concluded that even though the current understanding of the nature, cogni-
tive foundations, and processes underlying auditor judgment and professional skepticism is
far away from being perfect or complete, serious academic progress has been achieved in
these areas in the last years and decades. This development shows that while the glass might
still be half empty with regard to the current state of knowledge and understanding of auditor
judgment and professional skepticism, it is as well half full.1023 Future research has the chal-
lenging and critical task to continue the search for the missing puzzle pieces until a more
complete picture of auditor judgment and professional skepticism is obtained, or – figurative-
ly stated – until the glass is full.
1021
Einhorn/Hogarth (1981): 27.
1022
See Simon (1980): 77. Instances of studies demonstrating the current progress in the neuroscientific explana-
tion of human cognition include, among many others, Atlas/Wager (2013); Christoff (2013); Eichenbaum
(2004); Rugg (2004); Shankar/McClure (2013). For a concise overview of the neuroscientific approach to
human information processing and a discussion of relevant studies along these lines, see Kennedy et al.
(2005): 91-115.
1023
Note that Kramer (1999): 594 uses a similar formulation with regard to trust research.
204
Appendix 1
205
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2, © Springer Fachmedien Wiesbaden 2015
I dislike having to make 1 2 3 4 5 6
decisions quickly.
I have confidence in myself. 1 2 3 4 5 6
I do not like to decide until I’ve 1 2 3 4 5 6
looked at all of the readily
available information.
I like searching for knowledge. 1 2 3 4 5 6
I frequently question things 1 2 3 4 5 6
that I see or hear.
It is easy for other people to 1 2 3 4 5 6
convince me.
I seldom consider why people 1 2 3 4 5 6
behave in a certain way.
I like to ensure that I’ve 1 2 3 4 5 6
considered most available
information before making a
decision.
I enjoy trying to determine if 1 2 3 4 5 6
what I read or hear is true.
I relish learning. 1 2 3 4 5 6
The actions people take and the 1 2 3 4 5 6
reasons for those actions are
fascinating.
206
Appendix 2
Pretest Materials
herzlichen Dank, dass Sie sich die Zeit nehmen, diese Fallstudie zu bearbeiten. Bitte lesen Sie
zunächst die nachfolgenden Instruktionen sorgfältig durch, bevor Sie mit der Bearbeitung
beginnen. Sie werden im Folgenden mit Informationen zur Premium Steel AG konfrontiert
und werden gebeten, auf deren Basis Ihre persönlichen Einschätzungen bezüglich bestimm-
ter ökonomischer Sachverhalte abzugeben sowie anschließend einige allgemeine Fragen zu
beantworten. Bitte beachten Sie, dass in dieser Fallstudie nicht das prüferische Vorgehen im
Rahmen der gesetzlichen Abschlussprüfung nachgebildet werden soll und dass Ihre Einschät-
zungen lediglich auf Basis der in der Fallstudie bereitgestellten Information zu treffen sind; es
besteht keine Möglichkeit zur Erlangung weiterer Informationen.
Insgesamt werden Sie zur Bearbeitung der Fallstudie ca. 30 Minuten benötigen. Bitte bear-
beiten Sie die Fallstudienmaterialien selbständig, vollständig und in der vorgesehenen Rei-
henfolge. Sollten Sie Fragen zur Fallbearbeitung haben, signalisieren Sie uns dies bitte durch
Handzeichen.
Wir möchten explizit betonen, dass diese Fallstudie ausschließlich der Durchführung einer
empirischen Untersuchung im Rahmen eines Forschungsprojekts dient und nicht auf eine
Leistungsbeurteilung abzielt! Ihre Angaben und Antworten werden streng vertraulich behan-
delt und ausschließlich zu wissenschaftlichen Zwecken verwendet. Die Auswertung der Be-
fragungsergebnisse erfolgt im Rahmen des Forschungsprojekts auf aggregierter Ebene und es
können keine Rückschlüsse auf Ihre Person gezogen werden.
207
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2, © Springer Fachmedien Wiesbaden 2015
Teil 1
Bitte lesen Sie die nachfolgenden Informationen sorgfältig durch und beantworten Sie
die anschließenden Fragen!
Die Premium Steel AG ist ein führender Hersteller und Händler von Qualitätsspezialstahlprodukten
für die Automobil- und Anlagenbauindustrie. Insgesamt sind rund 10.500 Mitarbeiter an mehreren
Standorten deutschlandweit für das Unternehmen tätig.
Nach dem sehr erfolgreichen Geschäftsjahr 2007 hatte die Premium Steel AG – wie die meisten
Unternehmen in der Stahlbranche auch – ab dem letzten Quartal des Jahres 2008 mit den Folgen der
globalen Wirtschaftskrise zu kämpfen. Der durch die internationale Krise ausgelöste konjunkturelle
Einbruch hat die Automobil- und Maschinenbauindustrie als zentrale Abnehmer von Stahlprodukten
besonders schwer getroffen, was nicht spurlos an der Premium Steel AG vorbeigegangen ist. Im Ge-
schäftsjahr 2009 hatte die Gesellschaft einen Bestellrückgang von mehr als 50 % zu verzeichnen. Eine
extreme Unterauslastung der Produktionskapazitäten, ein signifikanter Umsatzeinbruch (von über 50
%) sowie eine drastische Ergebnisverschlechterung waren die Folgen der stark rückläufigen Entwick-
lung der Weltkonjunktur und der Stahlnachfrage.
Als Reaktion auf die Krisensituation hat die Premium Steel AG mit Hilfe eines externen Beratungs-
unternehmens ein Restrukturierungsprogramm, das operative und strukturelle Maßnahmen zur Kos-
tensenkung, Ertragsverbesserung und Reduktion des Working Capital umfasst, entwickelt und einge-
leitet. Bereits im Geschäftsjahr 2009 konnten erste Erfolge dieses Programms verzeichnet werden –
durch den Abbau von Vorräten und die Verbesserung des Forderungsmanagements konnte das Um-
laufvermögen (ohne liquide Mittel) um rd. 370 Mio. EUR vermindert werden, was einen positiven
Einfluss auf die Liquiditätslage der Premium Steel AG hatte. Ferner wurde das Investitionsvolumen
gegenüber den Vorjahren um mehr als 150 Mio. EUR zurückgefahren, um die Liquidität des Unter-
nehmens zu schonen. In den Geschäftsjahren 2006 bis 2008 hat das Unternehmen in Erwartung sich
nachhaltig positiv entwickelnder Absatzmärkte für Spezialstahl mit mehr als 300 Mio. EUR erheblich
in Produktionsmodernisierungen und punktuelle Kapazitätserweiterungen investiert.
Des Weiteren hat die negative Ergebnissituation der Premium Steel AG dazu geführt, dass die von den
Banken festgelegten Anforderungen an die Finanzkennzahlen (Financial Covenants) für den laufenden
Kreditvertrag (Kreditrahmen in Höhe von 550 Mio. EUR) erstmals zum 30.06.2009 nicht erfüllt wer-
den konnten. Somit mussten die Verbindlichkeiten aus diesem Vertrag als kurzfristig ausgewiesen
werden, da den Konsortialbanken das Recht zur außerordentlichen Kündigung des Konsortialvertrages
zusteht. Die Premium Steel AG finanziert sich über ein Kontokorrent-Darlehen mit der Hausbank
(Kreditrahmen seit 2006 unverändert 60 Mio. EUR), ein Investitionstilgungsdarlehen aus 2006 mit
einer Laufzeit von 8 Jahren in Höhe von ursprünglich 90 Mio. EUR sowie einen in 2006 über 5 Jahre
abgeschlossenen Konsortialkreditvertrag mit einem Maximalkreditrahmen von 550 Mio. EUR. Zur
Sicherstellung der Refinanzierung verfolgt die Premium Steel AG zwei Strategien: Zum einen sucht
sie Gespräche mit den Banken, die eine Restrukturierung der Finanzierung sowie eine Anpassung der
208
Financial Covenants an die veränderte Marktsituation zum Ziel haben. Zum anderen hat die Gesell-
schaft im Rahmen des “Konjunkturpakets II“ eine Bürgschaft der Bundesrepublik Deutschland sowie
eine Beteiligung der staatlichen KfW-Bank durch einen Direktkredit beantragt.
Sie sind als Abschlussprüfer der Premium Steel AG für das Geschäftsjahr 2009 bestellt worden und
prüfen die Gesellschaft bereits im dritten Jahr. Hierbei wenden Sie einen risikoorientierten Prüfungs-
ansatz an.
Bilanz (HGB)
31.12.2009 31.12.2008 31.12.2007
Aktiva
TEUR TEUR TEUR
A. Anlagevermögen
I. Immaterielle Vermögensgegenstände 7.380 7.572 5.798
II. Sachanlagen 593.717 661.133 588.576
III. Finanzanlagen 221 84 17
601.318 668.789 594.391
B. Umlaufvermögen
I. Vorräte 524.456 790.960 796.672
II. Forderungen und sonstige Vermögensgegenstände 296.617 400.012 569.440
III. Kassenbestand und Guthaben bei Kreditinstituten 10.184 20.174 35.236
831.257 1.211.146 1.401.348
C. Rechnungsabgrenzungsposten 920 901 640
1.433.495 1.880.836 1.996.379
209
Gewinn- und Verlustrechnung (HGB)
31.12.2009 31.12.2008 31.12.2007
TEUR TEUR TEUR
1. Umsatzerlöse 1.825.642 4.177.913 4.228.203
2. Erhöhung oder Verminderung des Bestandes an fertigen
-262.861 15.094 166.327
und unfertigen Erzeugnissen
3. Andere aktivierte Eigenleistungen 2.613 2.427 787
4. Sonstige betriebliche Erträge 70.266 53.795 54.637
1.635.660 4.249.229 4.449.954
5. Materialaufwand -1.125.631 -3.119.248 -3.224.218
6. Personalaufwand -487.146 -581.430 -545.732
7. Abschreibungen auf immaterielle Vermögensgegen-
-134.827 -147.318 -139.636
stände des Anlagevermögens und Sachanlagen
8. Sonstige betriebliche Aufwendungen -145.389 -202.430 -213.477
-1.892.993 -4.050.426 -4.123.063
9. Finanzergebnis -50.419 -57.782 -23.942
10. Ergebnis der gewöhnlichen Geschäftstätigkeit -307.752 141.021 302.949
11. Steuern vom Einkommen und vom Ertrag 18.576 -54.264 -117.145
12. Sonstige Steuern -2.660 -3.005 -2.576
15.916 -57.269 -119.721
13. Jahresüberschuss/-fehlbetrag -291.836 83.752 183.228
210
******
Bitte beurteilen Sie, ob die nachfolgenden Informationen jeweils positiv, negativ oder neutral im
Hinblick auf die Fortführungsfähigkeit der Premium Steel AG sind. Geben Sie bitte ferner auf einer
Skala von 1 bis 7 die Bedeutung an, die Sie diesen Informationen in Bezug auf die Beurteilung der
Fortführungsfähigkeit der Gesellschaft beimessen würden, wobei 1 „gar keine Bedeutung“ und 7
„sehr große Bedeutung“ ist.
Bitte bewerten Sie hierbei jeden Sachverhalt für sich und sehen Sie diesen nicht im Kontext zu
den anderen aufgeführten Sachverhalten.
Aufgrund der spürbaren Erholung in der Automobilindustrie und der langsamen Verbesserung der
allgemeinen konjunkturellen Situation geht das Management der Premium Steel AG davon aus, dass
sich die positive Entwicklung der Bestelleingänge der letzten Monate auch im weiteren Verlauf des
Geschäftsjahres 2010 fortsetzen wird und es rechnet mit einer tiefgreifenden Erholung des Absatzes
und der Produktionskapazitätsauslastung. Spätestens in 2011 soll die Gesellschaft laut Management-
prognose in die Gewinnzone zurückkehren und positive operative Cash Flows generieren.
Dem Antrag der Premium Steel AG auf eine Bürgschaft der Bundesrepublik Deutschland im Rahmen
des „Konjunkturpakets II“ sowie auf eine Beteiligung der staatlichen KfW-Bank durch einen Direkt-
kredit wird nicht entsprochen.
211
Auf der ersten Vorstandssitzung im Januar 2010 wird angekündigt, dass eine der wichtigsten Füh-
rungskräfte im Bereich Vertrieb die Premium Steel AG verlassen wird, um zum Hauptkonkurrenten
der Gesellschaft zu wechseln. Damit geht nicht nur ein erheblicher Know-how-Verlust für die Pre-
mium Steel AG einher, sondern auch der Verlust der wichtigsten Verbindung zu zwei Hauptkunden der
Gesellschaft, mit denen insgesamt rd. 27 % der Umsätze realisiert werden und die von der Führungs-
kraft akquiriert und betreut wurden. Zusammen mit der ausscheidenden Führungskraft verlassen
wichtige Vertriebsmitarbeiter die Premium Steel AG.
Auf einer Pressekonferenz im April 2010 hat der Finanzvorstand der Premium Steel AG angekündigt,
dass bis Ende 2010 eine Barkapitalerhöhung von 200 Mio. EUR durchgeführt werden soll.
Im März 2010 macht ein Kunde der Premium Steel AG Ansprüche in Höhe von 2,6 Mio. EUR auf-
grund wesentlicher Qualitätsmängel der gelieferten Stahlprodukte geltend. Diese werden vollumfäng-
lich durch die Garantierückstellung gedeckt. Weitere Ansprüche wurden zum aktuellen Zeitpunkt nicht
geltend gemacht. Es ist jedoch unklar, ob es sich bei den aufgetretenen Mängeln um einen Einzelfall
handelt oder ob sich die Qualitätsprobleme auf die gesamte Produktreihe beziehen und weitere Scha-
densfälle hervorrufen können. Im letzteren Fall geht das Management der Premium Steel AG von ei-
nem potentiellen Risiko in Höhe von 230 Mio. EUR aus.
212
Im Frühjahr 2010 wird bekanntgegeben, dass Steelo GmbH & Co. KG, einer der Hauptzulieferer der
Premium Steel AG, Insolvenz angemeldet hat. Daraus ergibt sich für die Premium Steel ein erhebli-
cher Lieferengpass und damit einhergehend auch die Gefahr, dass Kunden ihre Aufträge bei der Ge-
sellschaft zurückziehen. Aktuell laufen noch Verhandlungen mit anderen Lieferanten, doch abschlie-
ßende Vereinbarungen stehen noch aus.
Im März 2010 kann das Unternehmen eine Stundung von Verbindlichkeiten aus Lieferung und Leis-
tung i.H.v. 84 Mio. EUR über einen Zeitraum von zwei Jahren erreichen.
Das Unternehmen verfügt über ein Grundstück, dessen Verkauf die gewöhnliche Geschäftstätigkeit
des Unternehmens nicht wesentlich beeinträchtigen würde. Der aktuelle Marktpreis des Grundstücks
beläuft sich auf ca. 61 Mio. EUR. Das Management zieht in Erwägung, dieses Grundstück in den
nächsten Monaten zu veräußern, da die Kosten für die Verlagerung der durch den Verkauf betroffenen
Funktionen im Verhältnis zum erwarteten Verkaufserlös unerheblich sind.
Im April 2010 werden der Premium Steel AG aufwendige Modernisierungsmaßnahmen wegen festge-
stellter massiver Umweltverstöße in zwei Stahlwerken auferlegt. Aufgrund der momentanen Unfähig-
keit der Gesellschaft, das erforderliche Investitionsvolumen aufzubringen, droht die Stilllegung der
betroffenen Betriebsteile, die insgesamt ca. 23 % des Unternehmens ausmachen.
213
Teil 2
Anleitung
Nachfolgend finden Sie eine Reihe von allgemeinen Aussagen. Bitte beurteilen Sie diese Aussagen
aus Ihrer persönlichen Perspektive auf einer Skala von 1 „stimme überhaupt nicht zu“ bis 6 „stimme
voll und ganz zu“.
Der Aussagenkatalog ist einem wissenschaftlich anerkannten und vielfach angewandten Befragungs-
instrument entnommen. Es ist wichtig, dass Sie alle Aussagen beurteilen, selbst wenn sie sich
manchmal ähneln. Es gibt keine richtigen oder falschen Antworten! Bitte antworten Sie deshalb so
präzise und ehrlich wie möglich. Die Beantwortung soll spontan erfolgen und nicht länger als 5 Mi-
nuten beanspruchen. Ihre Angaben bleiben völlig anonym und fließen ausschließlich in das vorlie-
gende Forschungsprojekt ein.
Aussagen
214
stimme überhaupt stimme voll
nicht zu und ganz zu
Ich mag es, den Grund für das Verhalten anderer 1 2 3 4 5 6
Menschen zu verstehen.
Ich finde, dass Lernen spannend ist. 1 2 3 4 5 6
Ich nehme üblicherweise das, was ich sehe, lese 1 2 3 4 5 6
oder höre, für bare Münze.
Ich bin meiner nicht sicher. 1 2 3 4 5 6
Üblicherweise bemerke ich Inkonsistenzen in Erklärungen. 1 2 3 4 5 6
Meistens stimme ich dem, was die anderen in meiner 1 2 3 4 5 6
Gruppe denken, zu.
Ich mag es nicht, Entscheidungen schnell treffen zu müssen. 1 2 3 4 5 6
Ich habe Selbstvertrauen. 1 2 3 4 5 6
Ich mag es nicht zu entscheiden, bis ich mir alle 1 2 3 4 5 6
verfügbaren Informationen angesehen habe.
Ich mag die Suche nach Wissen. 1 2 3 4 5 6
Ich hinterfrage oft das, was ich sehe und höre. 1 2 3 4 5 6
Es ist leicht für andere Menschen, mich zu überzeugen. 1 2 3 4 5 6
Ich denke selten darüber nach, warum Menschen sich 1 2 3 4 5 6
auf eine bestimmte Weise verhalten.
Ich möchte sicherstellen, dass ich die meisten 1 2 3 4 5 6
verfügbaren Informationen berücksichtigt habe,
bevor ich eine Entscheidung treffe.
Ich genieße es zu versuchen, festzustellen, ob 1 2 3 4 5 6
das, was ich lese oder höre, wahr ist.
Ich genieße das Lernen. 1 2 3 4 5 6
Die Handlungen der Menschen und die Gründe 1 2 3 4 5 6
dafür sind faszinierend.
215
Demographische und sonstige Angaben
D. Berufsexamina: □ WP □ CPA
□ StB □ Sonstige:
F. Wie würden Sie das Ausmaß Ihrer Erfahrung in der Stahlindustrie einschätzen? Bitte Zutreffendes
ankreuzen:
G. Wie würden Sie das Ausmaß Ihrer Erfahrung in der Prüfung von fortführungsgefährdeten
Unternehmen einschätzen? Bitte Zutreffendes ankreuzen:
H. Inwieweit sind die in der Fallstudie enthaltenen Informationen ausreichend, um die Fortführungsfä-
higkeit der Premium Steel AG (eindeutig) zu beurteilen? Bitte Zutreffendes ankreuzen:
216
I. Welche Informationen würden Sie noch benötigen, um die Fortführungsfähigkeit der Premium
Steel AG eindeutig beurteilen zu können?
J. Bitte beurteilen Sie die nachfolgenden Aussagen in Bezug auf Ihre Vorgehensweise bei der
Beurteilung der Art und der Bedeutung der auf S. 5 bis 7 präsentierten Informationen auf einer
Skala von 1 „stimme überhaupt nicht zu“ bis 7 „stimme voll und ganz zu“:
L. Wie würden Sie die Verständlichkeit der Fallstudienmaterialien einschätzen? Bitte Zutreffendes
ankreuzen:
217
M. Anmerkungen/Kommentare:
218
Appendix 3
herzlichen Dank, dass Sie sich die Zeit nehmen, diese Fallstudie zu bearbeiten. Bitte lesen Sie
zunächst die nachfolgenden Instruktionen sorgfältig durch, bevor Sie mit der Bearbeitung
beginnen. Sie werden im Folgenden mit Informationen zur Premium Steel AG konfrontiert
und werden gebeten, auf deren Basis Ihre persönlichen Einschätzungen bezüglich bestimm-
ter ökonomischer Sachverhalte abzugeben sowie anschließend einige allgemeine Fragen zu
beantworten. Bitte beachten Sie, dass in dieser Fallstudie nicht das prüferische Vorgehen im
Rahmen der gesetzlichen Abschlussprüfung nachgebildet werden soll und dass Ihre Einschät-
zungen lediglich auf Basis der in der Fallstudie bereitgestellten Information zu treffen sind; es
besteht keine Möglichkeit zur Erlangung weiterer Informationen.
Es ist von höchster Bedeutung, dass Sie die Fallstudie in der vorgesehenen Reihenfolge
bearbeiten und einmal abgegebene Antworten nicht nachträglich revidieren.
Insgesamt werden Sie zur Bearbeitung der Fallstudie ca. 30 Minuten benötigen. Bitte bear-
beiten Sie die Fallstudienmaterialien selbständig und vollständig. Sollten Sie Fragen zur Fall-
bearbeitung haben, signalisieren Sie dies bitte durch Handzeichen.
Wir möchten explizit betonen, dass diese Fallstudie ausschließlich der Durchführung einer
empirischen Untersuchung im Rahmen eines Forschungsprojekts dient und nicht auf eine
Leistungsbeurteilung abzielt! Ihre Angaben und Antworten werden streng vertraulich behan-
delt und ausschließlich zu wissenschaftlichen Zwecken verwendet. Die Auswertung der Be-
fragungsergebnisse erfolgt im Rahmen des Forschungsprojekts auf aggregierter Ebene und es
können keine Rückschlüsse auf Ihre Person gezogen werden.
Bitte lesen Sie die nachfolgenden Informationen sorgfältig durch und beantworten Sie
die anschließenden Fragen!
Die Premium Steel AG ist ein führender Hersteller und Händler von Qualitätsspezialstahlprodukten
für die Automobil- und Anlagenbauindustrie. Insgesamt sind rund 10.500 Mitarbeiter an mehreren
Standorten deutschlandweit für das Unternehmen tätig.
Nach dem sehr erfolgreichen Geschäftsjahr 2007 hatte die Premium Steel AG – wie die meisten
Unternehmen in der Stahlbranche auch – ab dem letzten Quartal des Jahres 2008 mit den Folgen der
globalen Wirtschaftskrise zu kämpfen. Der durch die internationale Krise ausgelöste konjunkturelle
Einbruch hat die Automobil- und Maschinenbauindustrie als zentrale Abnehmer von Stahlprodukten
besonders schwer getroffen, was nicht spurlos an der Premium Steel AG vorbeigegangen ist. Im Ge-
schäftsjahr 2009 hatte die Gesellschaft einen Bestellrückgang von mehr als 50 % zu verzeichnen. Eine
extreme Unterauslastung der Produktionskapazitäten, ein signifikanter Umsatzeinbruch (von über 50
%) sowie eine drastische Ergebnisverschlechterung waren die Folgen der stark rückläufigen Entwick-
lung der Weltkonjunktur und der Stahlnachfrage.
Als Reaktion auf die Krisensituation hat die Premium Steel AG mit Hilfe eines externen Beratungs-
unternehmens ein Restrukturierungsprogramm, das operative und strukturelle Maßnahmen zur Kos-
tensenkung, Ertragsverbesserung und Reduktion des Working Capital umfasst, entwickelt und einge-
leitet. Bereits im Geschäftsjahr 2009 konnten erste Erfolge dieses Programms verzeichnet werden –
durch den Abbau von Vorräten und die Verbesserung des Forderungsmanagements konnte das Um-
laufvermögen (ohne liquide Mittel) um rd. 370 Mio. EUR vermindert werden, was einen positiven
Einfluss auf die Liquiditätslage der Premium Steel AG hatte. Ferner wurde das Investitionsvolumen
gegenüber den Vorjahren um mehr als 150 Mio. EUR zurückgefahren, um die Liquidität des Unter-
nehmens zu schonen. In den Geschäftsjahren 2006 bis 2008 hat das Unternehmen in Erwartung sich
nachhaltig positiv entwickelnder Absatzmärkte für Spezialstahl mit mehr als 300 Mio. EUR erheblich
in Produktionsmodernisierungen und punktuelle Kapazitätserweiterungen investiert.
Des Weiteren hat die negative Ergebnissituation der Premium Steel AG dazu geführt, dass die von den
Banken festgelegten Anforderungen an die Finanzkennzahlen (Financial Covenants) für den laufenden
Kreditvertrag (Kreditrahmen in Höhe von 550 Mio. EUR) erstmals zum 30.06.2009 nicht erfüllt wer-
den konnten. Somit mussten die Verbindlichkeiten aus diesem Vertrag als kurzfristig ausgewiesen
werden, da den Konsortialbanken das Recht zur außerordentlichen Kündigung des Konsortialvertrages
zusteht. Die Premium Steel AG finanziert sich über ein Kontokorrent-Darlehen mit der Hausbank
(Kreditrahmen seit 2006 unverändert 60 Mio. EUR), ein Investitionstilgungsdarlehen aus 2006 mit
einer Laufzeit von 8 Jahren in Höhe von ursprünglich 90 Mio. EUR sowie einen in 2006 über 5 Jahre
abgeschlossenen Konsortialkreditvertrag mit einem Maximalkreditrahmen von 550 Mio. EUR. Zur
Sicherstellung der Refinanzierung verfolgt die Premium Steel AG zwei Strategien: Zum einen sucht
sie Gespräche mit den Banken, die eine Restrukturierung der Finanzierung sowie eine Anpassung der
220
Financial Covenants an die veränderte Marktsituation zum Ziel haben. Zum anderen hat die Gesell-
schaft im Rahmen des “Konjunkturpakets II“ eine Bürgschaft der Bundesrepublik Deutschland sowie
eine Beteiligung der staatlichen KfW-Bank durch einen Direktkredit beantragt.
Sie sind als Abschlussprüfer der Premium Steel AG für das Geschäftsjahr 2009 bestellt worden und
prüfen die Gesellschaft bereits im dritten Jahr. Hierbei wenden Sie einen risikoorientierten Prüfungs-
ansatz an.
Bilanz (HGB)
31.12.2009 31.12.2008 31.12.2007
Aktiva
TEUR TEUR TEUR
D. Anlagevermögen
J. Immaterielle Vermögensgegenstände 7.380 7.572 5.798
II. Sachanlagen 593.717 661.133 588.576
III. Finanzanlagen 221 84 17
601.318 668.789 594.391
E. Umlaufvermögen
J. Vorräte 524.456 790.960 796.672
II. Forderungen und sonstige Vermögensgegenstände 296.617 400.012 569.440
III. Kassenbestand und Guthaben bei Kreditinstituten 10.184 20.174 35.236
831.257 1.211.146 1.401.348
F. Rechnungsabgrenzungsposten 920 901 640
1.433.495 1.880.836 1.996.379
221
Gewinn- und Verlustrechnung (HGB)
31.12.2009 31.12.2008 31.12.2007
TEUR TEUR TEUR
6. Umsatzerlöse 1.825.642 4.177.913 4.228.203
7. Erhöhung oder Verminderung des Bestandes an fertigen
-262.861 15.094 166.327
und unfertigen Erzeugnissen
8. Andere aktivierte Eigenleistungen 2.613 2.427 787
9. Sonstige betriebliche Erträge 70.266 53.795 54.637
1.635.660 4.249.229 4.449.954
10. Materialaufwand -1.125.631 -3.119.248 -3.224.218
6. Personalaufwand -487.146 -581.430 -545.732
8. Abschreibungen auf immaterielle Vermögensgegen-
-134.827 -147.318 -139.636
stände des Anlagevermögens und Sachanlagen
8. Sonstige betriebliche Aufwendungen -145.389 -202.430 -213.477
-1.892.993 -4.050.426 -4.123.063
9. Finanzergebnis -50.419 -57.782 -23.942
10. Ergebnis der gewöhnlichen Geschäftstätigkeit -307.752 141.021 302.949
11. Steuern vom Einkommen und vom Ertrag 18.576 -54.264 -117.145
12. Sonstige Steuern -2.660 -3.005 -2.576
15.916 -57.269 -119.721
13. Jahresüberschuss/-fehlbetrag -291.836 83.752 183.228
222
Frage
Wie hoch würden Sie auf Basis der zur Verfügung gestellten Informationen die Wahr-
scheinlichkeit einschätzen, dass die Premium Steel AG in den folgenden zwölf Monaten ihre
Geschäftstätigkeit fortführt? Bitte verwenden Sie für Ihre Antwort die nachstehende Skala:
Geschäftstätig- Geschäfts-
keit wird nicht
□ □ □ □ □ □ □ □ □ □ □ tätigkeit wird
fortgeführt fortgeführt
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (=100%)
(=0%)
223
Bitte lesen Sie die nachfolgenden Informationen sorgfältig durch und bearbeiten Sie
die Fragen in der vorgesehenen Reihenfolge!
Dem Antrag der Premium Steel AG auf eine Bürgschaft der Bundesrepublik Deutschland im Rahmen
des „Konjunkturpakets II“ sowie auf eine Beteiligung der staatlichen KfW-Bank durch einen
Direktkredit wird nicht entsprochen.
Frage
Wie hoch würden Sie auf Basis dieser zusätzlichen Information die Wahrscheinlichkeit
einschätzen, dass die Premium Steel AG in den folgenden zwölf Monaten ihre Geschäftstätigkeit
fortführt? Bitte verwenden Sie für Ihre Antwort die nachstehende Skala:
Geschäftstätig- Geschäfts-
keit wird nicht
□ □ □ □ □ □ □ □ □ □ □ tätigkeit wird
fortgeführt fortgeführt
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (=100%)
(=0%)
224
*****
Im Rahmen der Durchführung der Jahresabschlussprüfung geht Ihnen ferner auch die
nachfolgende Information zu:
Im Frühjahr 2010 wird bekanntgegeben, dass Steelo GmbH & Co. KG, einer der Hauptzulieferer der
Premium Steel AG, Insolvenz angemeldet hat. Daraus ergibt sich für die Premium Steel ein
erheblicher Lieferengpass und damit einhergehend auch die Gefahr, dass Kunden ihre Aufträge bei
der Gesellschaft zurückziehen. Aktuell laufen noch Verhandlungen mit anderen Lieferanten, doch
abschließende Vereinbarungen stehen noch aus.
Frage
Wie hoch würden Sie auf Basis dieser zusätzlichen Information die Wahrscheinlichkeit
einschätzen, dass die Premium Steel AG in den folgenden zwölf Monaten ihre Geschäftstätigkeit
fortführt? Bitte verwenden Sie für Ihre Antwort die nachstehende Skala:
Geschäftstätig- Geschäfts-
keit wird nicht
□ □ □ □ □ □ □ □ □ □ □ tätigkeit wird
fortgeführt fortgeführt
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (=100%)
(=0%)
225
******
Im Rahmen der Durchführung der Jahresabschlussprüfung geht Ihnen ferner auch die
nachfolgende Information zu:
Das renommierte internationale Beratungsunternehmen, mit dessen Hilfe das Restrukturie-
rungsprogramm entwickelt wurde, hat im Frühjahr 2010 ein Gutachten über die konjunkturellen
Entwicklungsaussichten für die nächsten zwei Jahre erstellt. Darin wird von einer mäßigen
Wiederbelebung der Konjunktur in 2010 sowie von einer spürbaren, nachhaltigen Besserung des
Marktumfeldes ab 2011 ausgegangen. Auf Basis dieser Markteinschätzungen wird im Gutachten der
Beratungsexperten die Premium Steel AG als überlebensfähig eingestuft. Infolge des Gutachtens
signalisieren die Banken Verhandlungsbereitschaft zur Verlängerung der Kreditlinien.
Frage
Wie hoch würden Sie auf Basis dieser zusätzlichen Information die Wahrscheinlichkeit
einschätzen, dass die Premium Steel AG in den folgenden zwölf Monaten ihre Geschäftstätigkeit
fortführt? Bitte verwenden Sie für Ihre Antwort die nachstehende Skala:
Geschäftstätig- Geschäfts-
keit wird nicht
□ □ □ □ □ □ □ □ □ □ □ tätigkeit wird
fortgeführt fortgeführt
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (=100%)
(=0%)
226
******
Im Rahmen der Durchführung der Jahresabschlussprüfung geht Ihnen ferner auch die
nachfolgende Information zu:
Auf einer Pressekonferenz im April 2010 hat der Finanzvorstand der Premium Steel AG angekündigt,
dass bis Ende 2010 eine Barkapitalerhöhung von 200 Mio. EUR durchgeführt werden soll.
Frage
Wie hoch würden Sie auf Basis dieser zusätzlichen Information die Wahrscheinlichkeit
einschätzen, dass die Premium Steel AG in den folgenden zwölf Monaten ihre Geschäftstätigkeit
fortführt? Bitte verwenden Sie für Ihre Antwort die nachstehende Skala:
Geschäftstätig- Geschäfts-
keit wird nicht
□ □ □ □ □ □ □ □ □ □ □ tätigkeit wird
fortgeführt fortgeführt
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (=100%)
(=0%)
227
Teil 2
Anleitung
Nachfolgend finden Sie eine Reihe von allgemeinen Aussagen. Bitte beurteilen Sie diese Aussagen
aus Ihrer persönlichen Perspektive auf einer Skala von 1 „stimme überhaupt nicht zu“ bis 6 „stimme
voll und ganz zu“.
Der Aussagenkatalog ist einem wissenschaftlich anerkannten und vielfach angewandten Befragungs-
instrument entnommen. Es ist wichtig, dass Sie alle Aussagen beurteilen, selbst wenn sie sich manch-
mal ähneln. Es gibt keine richtigen oder falschen Antworten! Bitte antworten Sie deshalb so präzise
und ehrlich wie möglich. Die Beantwortung soll spontan erfolgen und nicht länger als 5 Minuten be-
anspruchen. Ihre Angaben bleiben völlig anonym und fließen ausschließlich in das vorliegende For-
schungsprojekt ein.
Aussagen
228
stimme überhaupt stimme voll
nicht zu und ganz zu
Ich mag es, den Grund für das Verhalten anderer 1 2 3 4 5 6
Menschen zu verstehen.
Ich finde, dass Lernen spannend ist. 1 2 3 4 5 6
Ich nehme üblicherweise das, was ich sehe, lese 1 2 3 4 5 6
oder höre, für bare Münze.
Ich bin meiner nicht sicher. 1 2 3 4 5 6
Üblicherweise bemerke ich Inkonsistenzen in Erklärungen. 1 2 3 4 5 6
Meistens stimme ich dem, was die anderen in meiner 1 2 3 4 5 6
Gruppe denken, zu.
Ich mag es nicht, Entscheidungen schnell treffen zu müssen. 1 2 3 4 5 6
Ich habe Selbstvertrauen. 1 2 3 4 5 6
Ich mag es nicht zu entscheiden, bis ich mir alle 1 2 3 4 5 6
verfügbaren Informationen angesehen habe.
Ich mag die Suche nach Wissen. 1 2 3 4 5 6
Ich hinterfrage oft das, was ich sehe und höre. 1 2 3 4 5 6
Es ist leicht für andere Menschen, mich zu überzeugen. 1 2 3 4 5 6
Ich denke selten darüber nach, warum Menschen sich 1 2 3 4 5 6
auf eine bestimmte Weise verhalten.
Ich möchte sicherstellen, dass ich die meisten 1 2 3 4 5 6
verfügbaren Informationen berücksichtigt habe,
bevor ich eine Entscheidung treffe.
Ich genieße es zu versuchen, festzustellen, ob 1 2 3 4 5 6
das, was ich lese oder höre, wahr ist.
Ich genieße das Lernen. 1 2 3 4 5 6
Die Handlungen der Menschen und die Gründe 1 2 3 4 5 6
dafür sind faszinierend.
229
Weitere Fragen
Bitte beurteilen Sie, ob die nachfolgenden Informationen jeweils positiv, negativ oder neutral im
Hinblick auf die Fortführungsfähigkeit der Premium Steel AG sind. Geben Sie bitte ferner auf
einer Skala von 1 bis 7 die Bedeutung an, die Sie diesen Informationen in Bezug auf die
Beurteilung der Fortführungsfähigkeit der Gesellschaft beimessen würden, wobei 1 „gar keine
Bedeutung“ und 7 „sehr große Bedeutung“ ist.
Dem Antrag der Premium Steel AG auf eine Bürgschaft der Bundesrepublik Deutschland im Rahmen
des „Konjunkturpakets II“ sowie auf eine Beteiligung der staatlichen KfW-Bank durch einen
Direktkredit wird nicht entsprochen.
Im Frühjahr 2010 wird bekanntgegeben, dass Steelo GmbH & Co. KG, einer der Hauptzulieferer der
Premium Steel AG, Insolvenz angemeldet hat. Daraus ergibt sich für die Premium Steel ein
erheblicher Lieferengpass und damit einhergehend auch die Gefahr, dass Kunden ihre Aufträge bei
der Gesellschaft zurückziehen. Aktuell laufen noch Verhandlungen mit anderen Lieferanten, doch
abschließende Vereinbarungen stehen noch aus.
230
Auf einer Pressekonferenz im April 2010 hat der Finanzvorstand der Premium Steel AG angekündigt,
dass bis Ende 2010 eine Barkapitalerhöhung von 200 Mio. EUR durchgeführt werden soll.
231
Demographische und sonstige Angaben
D. Berufsexamina: □ WP □ CPA
□ StB □ Sonstige:
F. Wie würden Sie das Ausmaß Ihrer Erfahrung in der Stahlindustrie einschätzen? Bitte Zutreffendes
ankreuzen:
G. Wie würden Sie das Ausmaß Ihrer Erfahrung in der Prüfung von fortführungsgefährdeten
Unternehmen einschätzen? Bitte Zutreffendes ankreuzen:
H. Inwieweit sind die in der Fallstudie enthaltenen Informationen ausreichend, um die Fortführungsfä-
higkeit der Premium Steel AG (eindeutig) zu beurteilen? Bitte Zutreffendes ankreuzen:
232
I. Welche Informationen würden Sie noch benötigen, um die Fortführungsfähigkeit der Premium
Steel AG eindeutig beurteilen zu können?
J. Wie sicher fühlen Sie sich in Ihren Einschätzungen bezüglich der Fortführungsfähigkeit der
Premium Steel AG? Bitte Zutreffendes ankreuzen:
K. Bitte beurteilen Sie die nachfolgenden Aussagen in Bezug auf Ihre Vorgehensweise bei der
Beurteilung der Art und der Bedeutung der auf S. 5 bis 7 präsentierten Informationen auf einer
Skala von 1 „stimme überhaupt nicht zu“ bis 7 „stimme voll und ganz zu“:
233
M. Wie würden Sie die Verständlichkeit der Fallstudienmaterialien einschätzen? Bitte Zutreffendes
ankreuzen:
N. Anmerkungen/Kommentare:
234
Appendix 4
Auditor certification is the proportion of subjects that are legally certified auditors (WPs) within each sub-
group. Position is subjects’ position in the firm hierarchy, where 1= assistant/other; 2 = senior; 3 = manager; 4
= senior manager; 5 = partner.
*, **, *** Significance levels at 0.05, 0.1. and 0.01, respectively (two-tailed test)
Males Females
1.8%
senior/other
senior/other 16.1%
20.8% manager
36.8%
manager 57.1%
24.5% 25.0% senior manager
17.9% senior manager
partner
partner
missing values
235
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2, © Springer Fachmedien Wiesbaden 2015
References
Abdellaoui, Mohammed (2004): Rational Choice under Uncertainty. In: Bourgine, P./Nadal, J.
P. (eds.): Cognitive Economics: An Interdisciplinary Approach. Berlin/Heidelberg:
Springer: 15-13.
Abdolmohammadi, Mohammad (1985): Bayesian Inference Research in Auditing: Some
Methodological Suggestions. In: Contemporary Accounting Research, 2 (1): 76-94.
Abdolmohammadi, Mohammad (1986): Efficiency of the Bayesian Approach in Compliance
Testing: Some Empirical Evidence. In: Auditing: A Journal of Practice & Theory, 5 (2):
1-16.
Abdolmohammadi, Mohammad (1987): Bayesian Inference in Auditing: Evidence on the
Most Appropriate Assessment Techniques. In: Accounting and Business Research, 17
(68): 291-300.
Abdolmohammadi, Mohammad/Sarens, Gerrit (2011): An Investigation of the Association
between Cultural Dimensions and Variations in Perceived Use of and Compliance with
Internal Auditing Standards in 19 Countries. In: The International Journal of Account-
ing, 46 (4): 365-389.
Abdolmohammadi, Mohammad/Wright, Arnold (1987): An Examination of the Effects of
Experience and Task Complexity on Audit Judgments. In: The Accounting Review, 62
(1): 1-13.
Abou-Seada, Magda/Abdel-Kader, Magdy (2003): Behavioural Aspects of Auditors’
Evidence Evaluation: A Belief Revision Perspectie. Aldershot: Ashgate.
ACRA (2012): Practice Monitoring Programme: Sixth Public Report. URL: https://www.
acra.gov.sg/uploadedFiles/Content/Publications/Reports/PMPReport2012170712finalcle
an.pdf, accessed on: 25.07.2013.
Adam, Silke (2007): Das Going-Concern-Prinzip in der Jahresabschlussprüfung. Wiesbaden:
Deutscher Universitäts-Verlag.
Adams, John W. (1978): The Report of the Special Committee to Examine the Role of the
Auditor. Toronto: Canadian Institute of Chartered Accountants.
Adelman, Leonard/Tolcott, Martin A./Bresnick, Terry A. (1993): Examining the Effect of
Information Order on Expert Judgment. In: Organizational Behavior and Human Deci-
sion Processes, 56: 348-369.
Adelman, Leonard et al. (1996): Research with Patriot Air Defense Officers: Examining
Information Order Effects. In: Human Factors, 38 (2): 250-261.
Adelman, Leonard et al. (1997): Demonstrating the Effect of Context on Order Effects for an
Army Air Defense Task Using the Patriot Simulator. In: Journal of Behavioral Decision
Making, 10 (4): 327-342.
237
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2, © Springer Fachmedien Wiesbaden 2015
Aguinis, Herman/Gottfredson, Ryan K./Culpepper, Steven Andrew (2013): Best-Practice
Recommendations for Estimating Cross-Level Interaction Effects Using Multilevel
Modeling. In: Journal of Management, 39 (6): 1490-1528.
Ahlawat, Sunita S. (1999): Order Effects and Memory for Evidence in Individual versus
Group Decision Making in Auditing. In: Journal of Behavioral Decision Making, 12: 71-
88.
AICPA (1955): A Case Study on the Extent of Audit Samples. New York: AICPA.
AICPA (2010): Response to the European Commission’s Green Paper on Audit Policy:
Lessons from the Crisis. URL: http://www.aicpa.org/Advocacy/International/Down
loadableDocuments/AICPAresponse-EUGreenpaper.pdf, accessed on: 01.11.2013.
Ajzen, Icek (2005): Attitudes, Personality and Behavior, 2nd edition. New York: Open
University Press.
Akers, Michael D./Eaton, Tim V. (2003): Underreporting Of Chargeable Time: The Impact Of
Gender And Characteristics Of Underreporters. In: Journal of Managerial Issues, 15 (1):
82-96.
Akers, Michael D./Horngren, Charles T./Eaton, Tim V. (1998): Underreporting Chargeable
Time: A Continuing Problem For Public Accounting Firms. In: Journal of Applied
Business Research, 15 (1): 13-20.
Alchourrón, Carlos E./Gärdenfors, Peter/Makinson, David (1985): On the Logic of Theory
Change: Partial Meet Contraction and Revision Functions. In: The Journal of Symbolic
Logic, 50 (2): 510-530.
Alderman, C. Wayne/Deitrick, James W. (1982): Auditors’ Perceptions of Time Budget
Pressures and Premature Sign-Offs: A Replication and Extension. In: Auditing: A
Journal of Practice & Theory, 1 (2): 54-68.
Aldridge, R. et al. (1995): Measurement of Scene-Dependent Quality Variations in Digitally
Coded Television Pictures. In: IEE Proceedings - Vision, Image and Signal Processing,
142 (3): 149-154.
Allais, Maurice (1953): Le comportement de l’homme rationnel devant le risque, critique des
postulats et axiomes de l’école américaine. In: Econometrica, 21: 503-546.
Alleyne, Philmore/Hudaib, Mohammad/Pike, Richard (2013): Towards a Conceptual Model
of Whistle-Blowing Intentions Among External Auditors. In: The British Accounting
Review, 45 (1): 10-23.
Allport, Gordon Willard (1935): Attitudes. In: Murchison, C. (ed.): A Handbook of Social
Psychology. Worcester: Clark University Press: 789-994.
Allport, Gordon Willard (1954): The Nature of Prejudice. Reading et al.: Addison-Wesley.
238
Ameen, Elsie C./Guffey, Daryl M./McMillan, Jeffrey J. (1996): Gender Differences in
Determining the Ethical Sensitivity of Future Accounting Professionals. In: Journal of
Business Ethics, 15 (5): 591-597.
Anderson, Brenda H./Maletta, Mario J. (1999): Primacy Effects and the Role of Risk in
Auditor Belief-Revision Processes. In: Auditing: A Journal of Practice & Theory, 18 (1):
75-89.
Anderson, Norman H. (1981): Foundations of Information Integration Theory. New York:
Academic Press.
Anderson, Urton/Koonce, Lisa/Marchant, Garry (1991): A Model of Audit Judgment:
Cognition in a Professional Context. In: Ponemon, L. A./Gabhart, D. R. L. (eds.):
Auditing: Advances in Behavioral Research. New York et al.: Springer-Verlag: 43-73.
Anshakov, Oleg M./Gergely, Tamás (2010): Cognitive Reasoning: A Formal Approach.
Heidelberg et al.: Springer.
Antle, Rick et al. (2006): The Joint Determination of Audit Fees, Non-Audit Fees, and
Abnormal Accruals. In: Review of Quantitative Finance and Accounting, 27 (3): 235-
266.
APAK (2011): Bericht der Abschlussprüferaufsichtskommission über die Ergebnisse der
anlassunabhängigen Sonderuntersuchungen nach § 62b WPO für die Jahre 2007-2010.
URL: http://www.apak-aoc.de/pdf/APAK-AOC--Neuigkeiten_Bericht_Sonderunter
suchungen_2007-2010.pdf, accessed on: 11.03.2013.
Arkes, Hal R./Hammond, Kenneth R. (1986): General Introduction. In: Arkes, H.
R./Hammond, K. R. (eds.): Judgment and Decision Making: An Interdisciplinary
Reader, 2nd edition. Cambridge et al.: Cambridge University Press: 1-10.
Arnold, Donald F. et al. (2007): The Effect of Country and Culture on Perceptions of Appro-
priate Ethical Actions Prescribed by Codes of Conduct: A Western European Perspec-
tive among Accountants. In: Journal of Business Ethics, 70: 327-340.
Arnold, Donald F., Sr./Ponemon, Lawrence A. (1991): Internal Auditors’ Perceptions of
Whistle-Blowing and the Influence of Moral Reasoning: An Experiment. In: Auditing: A
Journal of Practice & Theory, 10 (2): 1-15.
Arnold, Vicky et al. (2000): The Effect of Experience and Complexity on Order and Recency
Bias in Decision Making by Professional Accountants. In: Accounting and Finance, 40:
109-134.
Aron, Arthur/Coups, Elliot J./Aron, Elaine N. (2013): Statistics for Psychology, 6th edition.
Upper Saddle River: Pearson.
Aronson, Elliot/Wilson, Timothy D./Akert, Robin M. (1994): Social Psychology: The Heart
and the Mind. New York: HarperCollins.
239
Aronson, Elliot/Wilson, Timothy D./Brewer, Marilynn B. (1998): Experimentation in Social
Psychology. In: Gilbert, D. T./Fiske, S. T./Lindzey, G. (eds.): The Handbook of Social
Psychology, 4th edition. Boston: McGraw-Hill: 99-142.
Asare, Stephen K. (1992): The Auditor’s Going-Concern Decision: Interaction of Task
Variables and the Sequential Processing of Evidence. In: The Accounting Review, 67
(2): 379-393.
Asare, Stephen K./Messier, William F. (1991): A Review of Audit Research Using the Belief-
Adjustment Model. In: Ponemon, L. A./Gabhart, D. R. L. (eds.): Auditing: Advances in
Behavioral Research. New York et al.: Springer-Verlag: 75-92.
Asare, Stephen K./Cianci, Anna M. /Tsakumis, George T. (2009): The Impact of Competing
Goals, Experience, and Litigation Consciousness on Auditors’ Judgments. In: Interna-
tional Journal of Auditing, 13: 223-236.
Asare, Stephen K./Cohen, Jeffrey/Trompeter, Greg (2005): The Effect of Non-Audit Services
on Client Risk, Acceptance and Staffing Decisions. In: Journal of Accounting and Public
Policy, 24 (6): 489-520.
Asch, S. E. (1946): Forming Impressions of Personality. In: The Journal of Abnormal and
Social Psychology, 41 (3): 258-290.
Aschauer, Ewald et al. (2013): Familiarity Threat Argument Revisited: Confidence, Trust and
Professional Skepticism. Working Paper presented at the 7th EARNet Symposium in
Trier, Germany.
Ashbaugh, Hollis/LaFond, Ryan/Mayhew, Brian W. (2003): Do Nonaudit Services
Compromise Auditor Independence? Further Evidence. In: The Accounting Review, 78
(3): 611-639.
Ashton, Alison Hubbard (1985): Does Consensus Imply Accuracy in Accounting Studies of
Decision Making? In: The Accounting Review, 60 (2): 173-185.
Ashton, Alison Hubbard/Ashton, Robert H. (1988): Sequential Belief Revision in Auditing.
In: The Accounting Review, 63 (4): 623-641.
Ashton, Robert H. (1974): An Experimental Study of Internal Control Judgments. In: Journal
of Accounting Research, 12: 143-157.
Ashton, Robert H. (1990): Pressure and Performance in Accounting Decision Settings: Para-
doxical Effects of Incentives, Feedback, and Justification. In: Journal of Accounting
Research, 28: 148-180.
Ashton, Robert H./Ashton, Alison Hubbard (1990): Evidence-Responsiveness in Professional
Judgment: Effects of Positive versus Negative Evidence and Presentation Mode. In:
Organizational Behavior and Human Decision Processes, 46: 1-19.
Ashton, Robert H./Kennedy, Jane (2002): Eliminating Recency with Self-Review: The Case
of Auditors’ ‘Going Concern’ Judgments. In: Journal of Behavioral Decision Making,
15: 221-231.
240
ASIC (2012): Audit Inspection Program Report for 2011-12. URL: http://www.asic.gov.au/
asic/pdflib.nsf/LookupByFileName/rep317-published-4-December-2012.pdf/$file/rep3
17-published-4-December-2012.pdf, accessed on: 25.07.2013.
Atlas, Lauren Y./Wager, Tor D. (2013): Expectancies and Beliefs: Insights from Cognitive
Neuroscience. In: Ochsner, K./Kosslyn, S. M. (eds.): The Oxford Handbook of Cogni-
tive Neuroscience, Volume 2: The Cutting Edges. Oxford: Oxford University Press:
359-381.
AUASB (2012): Professional Scepticism in an Audit of a Financial Report. URL: http://
www.auasb.gov.au/admin/file/content102/c3/Aug12_AUASB_Bulletin_Professional_Sc
epticism_in_an_Audit_of_a_Financial_Report.pdf, accessed on: 19.07.2013.
Awasthi, Vidya/Pratt, Jamie (1990): The Effects of Monetary Incentives on Effort and
Decision Performance: The Role of Cognitive Characteristics. In: The Accounting
Review, 65 (4): 797-811.
Ayers, Susam/Kaplan, Steven E. (1993): An Examination of the Effect of Hypothesis Framing
on Auditors’ Information Choices in an Analytical Procedure Task. In: Abacus, 29 (2):
113-130.
Baars, Bernard J. (1986): The Cognitive Revolution in Psychology. New York: Guilford
Press.
Baddeley, Alan D. (1997): Human Memory: Theory and Practice. Hove: Psychology Press.
Bailey, Charles D./Daily, Cynthia M./Phillips, Thomas J. (2011): Auditors’ Levels of Dispo-
sitional Need for Closure and Effects on Hypothesis Generation and Confidence. In:
Behavioral Research in Accounting, 23 (2): 27-50.
Baird, Jane E./Zelin, Robert C. (2000): The Effects of Information Ordering on Investor
Perceptions: An Experiment Utilizing Presidents’ Letters. In: Journal of Financial and
Strategic Decisions, 13 (3): 71-80.
Ballantine, Joan/McCourt, Patricia (2011): The Impact of Ethical Orientation and Gender on
Final Year Undergraduate Auditing Students’ Ethical Judgments. In: Accounting
Education, 20 (2): 187-201.
Ballwieser, Wolfgang/Berger, Karl-Heinz (1985): „Information und Wirtschaftlichkeit“ -
Hintergrund, Erwartungen und Ergebnisse. In: Ballwieser, W./Berger, K.-H. (eds.):
Information und Wirtschaftlichkeit. Wiesbaden: Gabler: 11-36.
Bamber, E. Michael/Bylinski, Joseph H. (1987): The Effects of the Planning Memorandum,
Time Pressure and Individual Auditor Characteristics on Audit Managers’ Review Time
Judgments. In: Contemporary Accounting Research, 4 (1): 127-143.
Bamber, E. Michael/Iyer, Venkataraman M. (2007): Auditors’ Identification with Their
Clients and Its Effect on Auditors’ Objectivity. In: Auditing: A Journal of Practice &
Theory, 26 (2): 1-24.
241
Bamber, E. Michael/Ramsay, Robert J./Tubbs, Richard M. (1997): An Examination of the
Descriptive Validity of the Belief-Adjustment Model and Alternative Attitudes to
Evidence in Auditing. In: Accounting, Organizations and Society, 22 (3/4): 249-268.
Bargh, John A. (1989): Conditional Automaticity: Varieties of Automatic Influence in Social
Perception and Cognition. In: Uleman, J. S./Bargh, J. A. (eds.): Unintended Thought.
New York/London: The Guilford Press: 3-51.
Baron, Jonathan (2007): Normative Models of Judgment and Decision Making. In: Koehler,
D. J./Harvey, N. (eds.): Blackwell Handbook of Judgment and Decision Making.
Malden/Oxford/Carlton: Blackwell Publishing: 19-36.
Bazerman, Max H./Moore, Don A. (2013): Judgment in Managerial Decision Making. Hobo-
ken: John Wiley & Sons.
Bazerman, Max H./Tenbrunsel, Ann E. (2011a): Blind Spots: Why We Fail to Do What’s
Right and What to Do about It. Princeton: Princeton University Press.
Bazerman, Max H./Tenbrunsel, Ann E. (2011b): Good People Often Let Bad Things Happen.
Why?: Ethical Breakdowns. In: Harvard Business Review: 58-65.
Bazerman, Max H./Loewenstein, George/Moore, Don A. (2002): Why Good Accountants Do
Bad Audits. In: Harvard Business Review: 97-102.
Bazerman, Max H./Morgan, Kimberley P./Loewenstein, George F. (1997): The Impossibility
of Auditor Independence. In: Sloan Management Review, 38 (4): 89-94.
Bazerman, Max H. et al. (2006): Reports of Solving Conflicts of Interest in Auditing are
Highly Exaggerated: Reply. In: Academy of Management Review, 31 (1): 43-49.
Beach, Lee Roy/Mitchell, Terence R. (1978): A Contingency Model for the Selection of
Decision Strategies. In: Academy of Management Review, 3 (3): 439-449.
Beasley, Mark S./Carcello, Joseph V./Hermanson, Dana R. (2001): Top 10 Audit Deficien-
cies. In: Journal of Accountancy, 191 (4): 63-66.
Bédard, Jean (1989): Expertise in Auditing: Myth or Reality? In: Accounting, Organizations
and Society, 14 (1/2): 113-131.
Bedard, Jean C./Biggs, Stanley F. (1991a): Pattern Recognition, Hypotheses Generation, and
Auditor Performance in an Analytical Task. In: The Accounting Review, 66 (3): 622-
642.
Bedard, Jean C./Biggs, Stanley F. (1991b): The Effect of Domain-Specific Experience on
Evaluation of Management Representations in Analytical Procedures. In: Auditing: A
Journal of Practice & Theory, 10 (Supplement): 77-90.
Bédard, Jean/Chi, Michelene T. H. (1993): Expertise in Auditing. In: Auditing: A Journal of
Practice & Theory, 12 (2): 21-45.
Bedard, Jean C./Johnstone, Karla M. (2010): Audit Partner Tenure and Audit Planning and
Pricing. In: Auditing: A Journal of Practice & Theory, 29 (2): 45-70.
242
Bedard, Jean C./Biggs, Stanley F./Maroney, James J. (1998): Sources of Process Gain and
Loss From Group Interaction in Performance of Analytical Procedures. In: Behavioral
Research in Accounting, 10 (Supplement): 207-233.
Beeler, Jesse D./Hunton, James E. (2002): Contingent Economic Rents: Insidious Threats to
Audit Independence. In: Advances in Accounting Behavioral Research, 5: 21-50.
Behling, Orlando/Law, Kenneth S. (2000): Translating Questionnaires and Other Research
Instruments: Problems and Solutions. Thousand Oaks/London/New Delhi: Sage
Publications.
Belkaoui, Ahmed (1989): Human Information Processing in Accounting. New
York/Westport/London: Quorum Books.
Bell, Timothy B./Peecher, Mark E./Solomon, Ira (2005): The 21st Century Public Company
Audit: Conceptual Elements of KPMG’s Global Audit Methodology. URL: http://busi
ness.illinois.edu/kpmg-uiuccases/monograph2.pdf, accessed on: 06.02.2014.
Bell, Timothy B. et al. (1997): Auditing Organizations Through a Strategic-Systems Lens: The
KPMG Business Measurement Process. New York: KPMG LLP.
Belmore, Susan M. (1987): Determinants of Attention During Impression Formation. In: Jour-
nal of Experimental Psychology: Learning, Memory, and Cognition, 13 (3): 480-489.
Benett, Nathan/Herold, David M./Ashford, Susan J. (1990): The Effects of Tolerance for
Ambiguity on Feedback-Seeking Behaviour. In: Journal of Occupational Psychology,
63: 343-348.
Bennink, Carl D./Spoelstra, Thomas (1979): Individual Differences in Field Articulation as a
Factor in Language Comprehension. In: Journal of Research in Personality, 13 (4): 480-
489.
Benston, George J./Hartgraves, Al L. (2002): Enron: What Happened and What We Can
Learn from It. In: Journal of Accounting and Public Policy, 21 (2): 105-127.
Bernardi, Richard A. (1994): Fraud Detection: The Effect of Client Integrity and Competence
and Auditor Cognitive Style. In: Auditing: A Journal of Practice & Theory, 13
(Supplement): 68-84.
Bernardi, Richard A./Arnold, Donald F., Sr. (2004): Testing the “Inverted-U” Phenomenon in
Moral Development on Recently Promoted Senior Managers and Partners. In:
Contemporary Accounting Research, 21 (2): 353-367.
Bernstein, Leopold A. (1967): The Concept of Materiality. In: The Accounting Review, 42
(1): 86-95.
Bierstaker, James Lloyd/Wright, Sally (2001): A Research Note Concerning Practical Prob-
lem-Solving Ability as a Predictor of Performance in Auditing Tasks. In: Behavioral Re-
search in Accounting, 13: 49-62.
243
Biggs, Stanley F./Wild, John J. (1985): An Investigation of Auditor Judgment in Analytical
Review. In: The Accounting Review, 60 (4): 607-633.
Bigus, Jochen (2011): Auditor’s Liability with Overcompensation and Reputation Losses. In:
OR Spectrum, 33 (2): 287-307.
Bik, Olof P. G. (2010): The Behavior of Assurance Professionals: A Cross-cultural Perspec-
tive. Delfi: Eburon.
Birnberg, Jacob G. (1964): Bayesian Statistics: A Review. In: Journal of Accounting
Research, 2 (1): 108-116.
Birnberg, Jacob G. (2011): A Proposed Framework for Behavioral Accounting Research. In:
Behavioral Research in Accounting, 23 (1): 1-43.
Blair-Broeker, Charles T./Ernst, Randal M. (2007): Thinking About Psychology: The Science
of Mind and Behavior, 2nd edition. New York: Worth Publishers.
Blay, Allen D. (2005): Independence Threats, Litigation Risk, and the Auditor’s Decision
Process. In: Contemporary Accounting Research, 22 (4): 759-789.
Block, Richard A./Harper, David R. (1991): Overconfidence in Estimation: Testing the An-
choring-and-Adjustment Hypothesis. In: Organizational Behavior and Human Decision
Processes, 49: 188-207.
Boatsman, James R./Moeckel, Cindy/Pei, Buck K. W. (1997): The Effects of Decision Conse-
quences on Auditors’ Reliance on Decision Aids in Audit Planning. In: Organizational
Behavior and Human Decision Processes, 71 (2): 211-247.
Bobek, Donna D./Hageman, Amy M./Radtke, Robin R. (2013a): A Research Note on the
Relationship between Professional Skepticism and Client Advocacy. Working Paper.
Bobek, Donna D./Hageman, Amy /Radtke, Robin (2013b): Ethical Sensitivity and Ethical
Decision-Making of Accounting Professionals: The Role of Context and Individual
Attributes. Working Paper.
Bobek, Donna D./Hageman, Amy M./Radtke, Robin R. (2013c): The Effects of Professional
Role, Decision Context and Gender on the Ethical Decision Making of Public
Accounting Professionals. Working Paper.
Bogdan, Radu J. (1986): The Importance of Belief. In: Bogdan, R. J. (ed.): Belief: Form,
Content, and Function. Oxford: Clarendon Press: 1-16.
Bogle, John C. (2000): Public Accounting: Profession or Business? URL: http://www.van
guard.com/bogle_site/sp20001016.html, accessed on: 18.3.2014.
Bollom, William J. (1988): Ethics and Self-Regulation for CPAs in the U.S.A. In: Journal of
Business Ethics, 7 (1/2): 55-61.
Bonner, Sarah E. (1994): A Model of the Effects of Audit Task Complexity. In: Accounting,
Organizations and Society, 19 (3): 213-234.
244
Bonner, Sarah E. (1999): Judgment and Decision-Making Research in Accounting. In:
Accounting Horizons, 13 (4): 385-398.
Bonner, Sarah E. (2008): Judgment and Decision Making in Accounting. Upper Saddle River:
Pearson Prentice Hall.
Bonner, Sarah E./Lewis, Barry L. (1990): Determinants of Auditor Expertise. In: Journal of
Accounting Research, 28 (Supplement): 1-20.
Bonner, Sarah E./Pennington, Nancy (1991): Cognitive Processes and Knowledge as Deter-
minants of Auditor Expertise. In: Journal of Accounting Literature, 10: 1-50.
Bonner, Sarah E./Walker, Paul L. (1994): The Effects of Instruction and Experience on the
Acquisition of Auditing Knowledge. In: The Accounting Review, 69 (1): 157-178.
Bortz, Jürgen/Schuster, Christof (2010): Statistik für Human- und Sozialwissenschaftler, 7th
edition. Berlin/Heidelberg: Springer-Verlag.
Boush, David M./Friestad, Marian/Rose, Gregory M. (1994): Adolescent Skepticism toward
TV Advertising and Knowledge of Advertiser Tactics. In: Journal of Consumer Re-
search, 21: 165-175.
Bowlin, Kendall O./Hobson, Jessen L./Piercey, David M. (2014): The Effects of Auditor
Rotation, Professional Skepticism, and Interactions with Managers on Audit Quality.
Working Paper. URL: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1914557, ac-
cessed on: 29.07.2014.
Bowrin, Anthony R./King, James II (2010): Time Pressure, Task Complexity, and Audit
Effectiveness. In: Managerial Auditing Journal, 25 (2): 160-181.
Boylan, Scott J. (2008): A Classroom Excercise on Unconscious Bias in Financial Reporting
and Auditing. In: Issues in Accounting Education, 23 (2): 229-245.
Boynton, William C./Kell, Walter G. (1996): Modern Auditing, 6th edition. New York: John
Wiley & Sons.
Brabeck, Mary (1984): Ethical Characteristics of Whistle Blowers. In: Journal of Research in
Personality, 18 (1): 41-53.
Brandon, Duane M. et al. (2007): The Joint Influence of Client Attributes and Cognitive
Moral Development on Students’ Ethical Judgments. In: Journal of Accounting
Education, 25 (1–2): 59-73.
Braun, Karen Wilken (2001): The Disposition of Audit-Detected Misstatements: An
Examination of Risk and Reward Factors and Aggregation Effects. In: Contemporary
Accounting Research, 18 (1): 71-99.
Brazel, Joseph et al. (2013): Hindsight Bias and Auditor Skepticism: Does the End Justify the
Means? Working Paper presented at the 7th EARNet Symposium in Trier, Germany.
Breesch, Diane/Branson, Joël (2009): The Effect of Gender on Audit Quality. In: The ICFAI
University Journal of Accounting Research & Audit Practices, 8 (3-4): 78-107.
245
Bronner, Rolf (1973): Entscheidungen unter Zeitdruck – Eine Experimentaluntersuchung zur
empirischen Theorie der Unternehmnung. Tübingen: J.C.B.Mohr (Paul Siebeck).
Brown, Charles A. (2009): Order Effects and the Audit Materiality Revision Choice. In: The
Journal of Applied Business Research, 25 (1): 21-36.
Brown-Liburd, Helen/Cohen, Jeffrey R./Trompeter, Greg (2013): Effects of Earnings Fore-
casts and Heightened Professional Skepticism on the Outcomes of Client-Auditor Nego-
tiation. In: Journal of Business Ethics, 116 (2): 311-325.
Bruine de Bruin, Wändi/Keren, Gideon (2003): Order Effects in Sequentially Judged Options
due to the Direction of Comparison. In: Organizational Behavior and Human Decision
Processes, 92: 91-101.
Bruni, Luigino/Sugden, Robert (2007): The Road Not Taken: How Psychology Was Removed
From Economics and How It Might Be Brought Back. In: The Economic Journal, 117:
146-173.
Bryant, Stephanie/Murthy, Uday/Wheeler, Patrick (2009): The Effects of Cognitive Style and
Feedback Type on Performance in an Internal Control Task. In: Behavioral Research in
Accounting, 21 (1): 37-58.
Buchman, Thomas A./Tetlock, Philip E./Reed, Ronald O. (1996): Accountability and Audi-
tors’ Judgments about Contingent Events In: Journal of Business Finance & Accounting,
23 (3): 379-398.
Buckley, John W. (1978): An Exploration of Professional Identity. In: Loeb, S. E. (ed.): Ethics
in the Accounting Profession. Santa Barbara et al.: John Wiley & Sons: 63-84.
Budner, Stanley (1962): Intolerance of Ambiguity as a Personality Variable. In: Journal of
Personality, 30 (1): 29-50.
Bunge, Mario (1991): A Skeptic’s Beliefs and Disbeliefs. In: New Ideas in Psychology, 9 (2):
131-149.
Burnyeat, M. F. (1983): Can the Skeptic Live His Skepticism? In: Burnyeat, M. F. (ed.): The
Skeptical Tradition. Berkeley/Los Angeles/London: University of California Press: 117-
148.
Burton, John C. (1980): A Critical Look at Professionalism and Scope of Services. In: Journal
of Accountancy, 149 (4): 48-56.
Buss, Arnold H. (1989): Personality as Traits. In: American Psychologist, 44 (11): 1378-1388.
Butcharov, Panayot (1998): Skepticism About the External World. Oxford et al.: Oxford
University Press.
Butler, Stephen A. (1985): Application of a Decision Aid in the Judgmental Evaluation of
Substantive Test of Details Samples. In: Journal of Accounting Research, 23 (2): 513-
526.
246
Butt, Jane L./Campbell, Terry L. (1989): The Effects of Information Order and Hypothesis-
Testing Strategies on Auditors’ Judgments. In: Accounting, Organizations and Society,
14 (5/6): 471-479.
Canary, Daniel J./Emmers-Sommer, Tara M./Faulkner, Sandra (1997): Sex and Gender Dif-
ferences in Personal Relationships. New York: Guilford Press.
Caplan, Paula J. et al. (1997): Gender Differences in Human Cognition. Oxford et al.: Oxford
University Press.
CAQ (2010): Deterring and Detecting Financial Reporting Fraud: A Platform for Action.
URL: http://www.thecaq.org/docs/reports-and-publications/deterring-and-detecting-fin
ancial-reporting-fraud-a-platform-for-action.pdf?sfvrsn=0, accessed on: 16.5.2014.
CAQ (2011): In-Depth Guide to Public Company Auditing: The Financial Statement Audit.
URL: http://www.thecaq.org/docs/audit-committees/in-depth_guidetopubliccompanyau
diting.pdf?sfvrsn=0, accessed on: 14.05.2014.
Carcello, Joseph V./Nagy, Albert L. (2004): Audit Firm Tenure and Fraudulent Financial Re-
porting. In: Auditing: A Journal of Practice & Theory, 23 (2): 55-69.
Carcello, Joseph V./Neal, Terry L. (2000): Audit Committee Composition and Auditor
Reporting. In: The Accounting Review, 75 (4): 453-467.
Carcello, Joseph V./Hermanson, Roger H./McGrath, Neal T. (1992): Audit Quality Attrib-
utes: The Perceptions of Audit Partners, Preparers, and Financial Statement Users. In:
Auditing: A Journal of Practice & Theory, 11 (1): 1-15.
Carcello, Joseph V./Vanstraelen, Ann/Willenborg, Michael (2009): Rules Rather than Discre-
tion in Audit Standards: Going-Concern Opinions in Belgium. In: The Accounting
Review, 84 (5): 1395-1428.
Cardello, Armand V. (1996): The Role of the Human Senses in Food Acceptance. In:
Meiselman, H. L./MacFie, H. J. H. (eds.): Food Choice, Acceptance and Consumption.
London: Chapman & Hall: 1-82.
Carey, John L. (1956): Professional Ethics of Certified Public Accountants. New York:
American Institute of Accountants.
Carey, Peter/Simnett, Roger (2006): Audit Partner Tenure and Audit Quality. In: The
Accounting Review, 81 (3): 653-676.
Carmichael, Douglas R./Pany, Kurt (1993): Reporting on Uncertainties, Including Going
Concern. In: AICPA (ed.): The Expectation Gap Standards: Progress, Implementation
Issues, Research Opportunities. Jersey City: AICPA: 35-58.
Carnota, Raul/Rodrigues, Ricardo (2010): AGM Theory and Artificial Intelligence. In:
Olsson, E. J./Enqvist, S. (eds.): Belief Revision meets Philosophy of Science. Dordrecht
et al.: Springer.
247
Carpenter, Tina D. (2007): Audit Team Brainstorming, Fraud Risk Identification and Fraud
Risk Assessment: Implications of SAS No. 99. In: The Accounting Review, 82 (5):
1119-1140.
Carpenter, Tina D./Reimers, Jane L. (2013): Professional Skepticism: The Effects of a
Partner’s Influence and the Level of Fraud Indicators on Auditors’ Fraud Judgments and
Actions. In: Behavioral Research in Accounting, 25 (2): 45-69.
Carpenter, Tina D./Durtschi, Cindy/Gaynor, Lisa Milici (2011): The Incremental Benefits of
a Forensic Accounting Course on Skepticism and Fraud-Related Judgments. In: Issues in
Accounting Education, 26 (1): 1-21.
Carruthers, Peter (2003): Moderately Massive Modularity. In: O’Hear, A. (ed.): Minds and
Persons. Cambridge: Cambridge University Press: 67-90.
Carruthers, Peter (2006): The Architecture of the Mind. Oxford: Oxford University Press.
Cash, Richard M. (2011): Advancing Differentiation: Thinking and Learning for the 21st
Century. Minneapolis: Free Spirit Publishing.
Chambers Universal Learners’ Dictionary (1994): New Delhi: Allied Chambers.
Chang, C. Janie/Hwang, Nen-Chen Richard (2003): The Impact of Retention Incentives and
Client Business Risks on Auditors’ Decisions Involving Aggressive Reporting Practices.
In: Auditing: A Journal of Practice & Theory, 22 (2): 207-218.
Chapman, Gretchen B./Johnson, Eric J. (2002): Incorporating the Irrelevant: Anchros in
Judgments of Belief and Value. In: Gilovich, T./Griffin, D./Kahneman, D. (eds.): Heu-
ristics and Biases: The Psychology of Intuitive Judgment. Cambridge et al.: Cambridge
University Press: 120-138.
Chapman, Gretchen B./Bergus, George R./Elstein, Arthur S. (1996): Order of Information
Affects Clinical Judgment. In: Journal of Behavioral Decision Making, 9 (3): 201-211.
Chavas, Jean-Paul (2004): Risk Analysis in Theory and Practice. San Diego/London:
Elsevier.
Chazen, Charles/Solomon, Kenneth I. (1975): The Art of Defensive Auditing. In: Journal of
Accountancy, 140 (4): 66-71.
Chekushina, Tatiana/Loth, Kai Sofia (2014): Die kritische Grundhaltung bei der Abschluss-
prüfung. In: Neue Zeitschrift für Gesellschaftsrecht, 17 (3): 85-92.
Chen, Serena/Chaiken, Shelly (1999): The Heuristic-Systematic Model in Its Broader Con-
text. In: Chaiken, S./Trope, Y. (eds.): Dual-Process Theories in Social Psychology. New
York: Guilford: 73-96.
Chen, Shimin/Sun, Sunny Y. J./Wu, Donghui (2010): Client Importance, Institutional Improve-
ments, and Audit Quality in China: An Office and Individual Auditor Level Analysis. In:
The Accounting Review, 85 (1): 127-158.
248
Chih-Ying, Chen/Chan-Jane, Lin/Yu-Chen, Lin (2008): Audit Partner Tenure, Audit Firm
Tenure, and Discretionary Accruals: Does Long Auditor Tenure Impair Earnings
Quality? In: Contemporary Accounting Research, 25 (2): 415-445.
Child, D. J. (1981): Culture, Contingency and Capitalism in Cross-National Study of Organi-
zations. In: Cummings, L. L./Staw, B. M. (eds.): Research in Organizational Behavior.
Greenwich: JAI Press: 303–356.
Chmielewicz, Klaus (1994): Forschungskonzeptionen der Wirtschaftswissenschaft, 3rd edi-
tion. Stuttgart: Shäffer-Poeschel.
Choo, Freddie (1986): Job Stress, Job Performance, and Auditor Personality Characteristics.
In: Auditing: A Journal of Practice & Theory, 5 (2): 17-34.
Choo, Freddie (1995): Auditors’ Judgment Performance Under Stress: A Test of the Predicted
Relationship by Three Theoretical Models. In: Journal of Accounting, Auditing & Fi-
nance, 10 (3): 611-641.
Choo, Freddie (1996): Auditors’ Knowledge Content and Judgment Performance: A Cogni-
tive Script Approach. In: Accounting, Organizations and Society, 21 (4): 339-359.
Choo, Freddie/Tan, Kim (2000): Instruction, Skepticism, And Accounting Students’ Ability
To Detect Frauds In Auditing. In: Journal of Business Education, (1): 72-87.
Chow, Chee W./McNamee, Alan H./Plumlee, R. David (1987): Practitioners’ Perceptions of
Audit Step Difficulty and Criticalness: Implications for Audit Research. In: Auditing: A
Journal of Practice & Theory, 6 (2): 123-133.
Christ, Mary York (1993): Evidence on the Nature of Audit Planning Problem Representa-
tions: An Examination of Auditor Free Recalls. In: The Accounting Review, 68 (2): 304-
322.
Christoff, Kalina (2013): Thinking. In: Ochsner, K./Kosslyn, S. M. (eds.): The Oxford
Handbook of Cognitive Neuroscience, Volume 2: The Cutting Edges. Oxford: Oxford
University Press: 318-333.
Chung, Janne/Monroe, Gary S. (1998): Gender Differences in Information Processing: An
Empirical Test of the Hypothesis-Confirming Strategy in an Audit Context. In:
Accounting & Finance, 38 (2): 265-279.
Chung, Janne/Monroe, Gary S. (2001): A Research Note on the Effects of Gender and Task
Complexity on an Audit Judgment. In: Behavioral Research in Accounting, 13: 111-125.
Church, A. Timothy (2000): Culture and Personality: Toward an Integrated Cultural Trait
Psychology. In: Journal of Personality, 68 (4): 651-703.
Church, Bryan K. (1990): Auditors’ Use of Confirmation Processes. In: Journal of
Accounting Literature, 9: 81-112.
249
Church, Bryan K. (1991): An Examination of the Effect that Commitment to a Hypothesis has
on Auditors’ Evaluations of Confirming and Disconfirming Evidence. In: Contemporary
Accounting Research, 7 (2): 513-534.
Churchill, Gilbert A., Jr. (1979): A Paradigm for Developing Better Measures of Marketing
Constructs. In: Journal of Marketing Research, 16 (1): 64-73.
CICA (1995): Professional Judgment and the Auditor. Toronto: CICA.
Cochran, Kathryn F./Davis, J. Kent (1987): Individual Differences in Inference Processes. In:
Journal of Research in Personality, 21 (2): 197-210.
Cochran, W. G. (1947): Some Consequences When the Assumptions for the Analysis of
Variance are not Satisfied. In: Biometrics, 3 (1): 22-38.
Cohen, Jacob (1988): Statistical Power Analysis for the Behavioral Sciences, 2nd edition.
Hillsdale: Lawrence Erlbaum Associates.
Cohen, Jeffrey R./Dalton, Derek W./Harp, Nancy (2014): The Effect of Professional
Skepticism on Job Attitudes and Turnover Intentions within the Audit Profession. URL:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2410547, accessed on: 29.04.2014.
Cohen, Jeffrey R./Pant, Laurie W./Sharp, David J. (1992): Cultural and Socioeconomic
Constraints on International Codes of Ethics: Lessons from Accounting. In: Journal of
Business Ethics, 11 (9): 687-700.
Cohen, Jeffrey R./Pant, Laurie W./Sharp, David J. (1993): Culture-Based Ethical Conflicts
Confronting Multinational Accounting Firms. In: Accounting Horizons, 7 (3): 1-13.
Cohen, Jeffrey R./Pant, Laurie W./Sharp, David J. (1995): An Exploratory Examination of
International Differences in Auditors’ Ethical Perceptions. In: Behavioral Research in
Accounting, 7: 37-64.
Cohen, Jeffrey R./Pant, Laurie W./Sharp, David J. (2001): An Examination of Differences in
Ethical Decision-Making Between Canadian Business Students and Accounting Profes-
sionals. In: Journal of Business Ethics, 30 (4): 319-336.
Cohen, LeoNora M./Ambrose, Don (1999): Adaptation and Creativity. In: Runco, M.
A./Pritzker, S. R. (eds.): Encyclopedia of Creativity. San Diego: Academic Press: 9-22.
Colbert, Janet L. (1989): The Effect of Experience on Auditors’ Judgments. In: Journal of
Accounting Literature, 8: 137-149.
Colby, Anne/Kohlberg, Lawrence (1987): The Measurement of Moral Judgment, Volume 1.
Cambridge: Cambridge University Press.
Colman, Andrew M. (2009): Oxford Dictionary of Psychology, 3rd edition. Oxford/New
York: Oxford University Press.
Cone, John D./Foster, Sharon L. (1998): Dissertations and Theses from Start to Finish:
Psychology and Related Fields, 7th printing. Washington: American Psychological
Association.
250
Cook, John W./Winkle, Gary M. (1988): Auditing, 4th edition. Boston: Houghton Mifflin.
Coolican, Hugh (2013): Research Methods and Statistics in Psychology, 5th edition. New
York: Routledge.
Copeland, Christopher (1996): Professional Skepticism. In: Business Credit, 98 (8): 37-38.
Corless, John C. (1972): Assessing Prior Distributions for Applying Bayesian Statistics in
Auditing. In: The Accounting Review, 47 (3): 556-566.
Cornman, James Welton /Lehrer, Keith Edward/Pappas, George Sotiros (1992): Philoso-
phical Problems and Arguments: An Introduction, 4th edition. Indianapolis: Hackett
Publishing.
Costa, Paul T., Jr./McCrae, Robert R. (1992): Four Ways Five Factors are Basic. In:
Personality and Individual Differences, 13 (6): 653-665.
Costabile, Kristi A./Klein, Stanley B. (2005): Finishing Strong: Recency Effects in Juror Judg-
ments. In: Basic & Applied Social Psychology, 27 (1): 47-58.
Cowperthwaite, Philip (2010): Culture Matters: How Our Culture Affects the Audit. In:
Accounting Perspectives, 9 (3): 175-215.
Cozzens, Michael D./Contractor, Noshir S. (1987): The Effect of Conflicting Information on
Media Skepticism. In: Communication Research, 14 (4): 437-451.
CPAB (2012): Meeting the Challenge: “A Call to Action”: 2011 Public Report. URL:
http://www.cpab-ccrc.ca/Documents/Topics/Public%20Reports/2011%20Public_Repor
t_EN_FNL_tag.pdf, accessed on: 06.08.2014.
Craft, Jana L. (2013): A Review of the Empirical Ethical Decision-Making Literature: 2004–
2011. In: Journal of Business Ethics, 117 (2): 221-259.
Craig, Edward (ed.) (2005): The Shorter Routledge Encyclopedia of Philosophy. London et
al.: Routledge.
Cronbach, Lee J. (1951): Coefficient Alpha and the Internal Structure of Test. In:
Psychometrika, 16 (3): 297-334.
Cropley, Arthur (2006): In Praise of Convergent Thinking. In: Creativity Research Journal, 18
(3): 391-404.
Crosby, Michael A. (1981): Bayesian Statistics in Auditing: A Comparison of Probability
Elicitation Techniques. In: The Accounting Review, 56 (2): 355.
Crowne, Douglas P./Marlowe, David (1960): A New Scale of Social Desirability Independent
of Psychopathology. In: Journal of Consulting Psychology, 24 (4): 349-354.
Cuccia, Andrew D./Mc Gill, Gary A. (2000): The Role of Decision Strategies in Understand-
ing Professionals’ Susceptibility to Judgment Biases. In: Journal of Accounting
Research, 38 (2): 419-435.
251
Cuccia, Andrew D./Hackenbrack, Karl/Nelson, Mark W. (1995): The Ability of Professional
Standards to Mitigate Aggressive Reporting. In: The Accounting Review, 70 (2): 227-
248.
Curtis, Mary B./Taylor, Eileen Z. (2009): Whistleblowing in Public Accounting: The
Influence of Identity Disclosure, Situational Context and Personal Characteristics In:
Accounting and the Public Interest, 9: 191-220.
Cushing, Barry E. (2000): Economic Analysis of Skepticism in an Audit Setting. In:
Solomon, I./Peecher, M. E. (eds.): Proceedings of the 14th Symposium on Auditing
Research at the University of Illinois at Urbana-Champaign: 1-3.
Cushing, Barry E./Ahlawat, Sunita S. (1996): Mitigation of Recency Bias in Audit Judgment:
The Effect of Documentation. In: Auditing: A Journal of Practice & Theory, 15 (2): 110-
122.
Dalal, Dev K./Zickar, Michael J. (2012): Some Common Myths About Centering Predictor
Variables in Moderated Multiple Regression and Polynomial Regression. In: Organiza-
tional Research Methods, 15 (3): 339-362.
Damien, Paulet al. (2013): Bayesian Theory and Applications. Oxford: Oxford University
Press.
Dani, D. N. (1989): Scientific Attitude and Cognitive Styles. New Delhi: Northern Book
Centre.
Daniel, Shirley J. (1988): Some Empirical Evidence about the Assessment of Audit Risk in
Practice. In: Auditing: A Journal of Practice & Theory, 7 (2): 174-181.
David, F. N./Johnson, N. L. (1951): The Effect of Non-Normality on the Power Function of
the F-Test in the Analysis of Variance. In: Biometrika, 38: 43-57.
Davis, Jon S./Solomon, Ira (1989): Experience, Expertise, and Expert-Performance Research
in Public Accounting. In: Journal of Accounting Literature, 8: 150-164.
Davis, Larry R./Simon, Daniel T. (1992): The Impact of SEC Disciplinary Actions on Audit
Fees. In: Auditing: A Journal of Practice & Theory, 11 (1): 58-68.
DeAngelo, Linda Elizabeth (1981a): Auditor Independence, ‘Low Balling’, and Disclosure
Regulation. In: Journal of Accounting and Economics, 3 (2): 113-127.
DeAngelo, Linda Elizabeth (1981b): Auditor Size and Audit Quality. In: Journal of Account-
ing and Economics, 3: 183-199.
Deckers, Marc/Hermann, Christian (2013): Die kritische Grundhaltung des Abschlussprüfers
(professional scepticism) – Verhaltensanforderung und Charaktereigenschaft im Fokus
aktueller Regulierungsbestrebungen. In: Der Betrieb, 66 (41): 2315-2321.
Deis, Donald R., Jr./Giroux, Gary A. (1992): Determinants of Audit Quality in the Public
Sector. In: The Accounting Review, 67 (3): 462-479.
252
Dennis, Ian (2013): The Nature of Professional Judgment in Auditing. Working Paper
presented at the 7th EARNet Symposium in Trier, Germany.
Dennis, Sean/Johnstone, Karla M. (2014): Audit Partner Leadership Tone and Professional
Skepticism in Fraud Brainstorming. Working Paper. URL: http://www3.nd.edu/~
carecob/Workshops/13-14Workshops/Johnstone%20Paper.pdf, accessed on: 29.04.2014.
Dermer, Jerry D. (1973): Cognitive Characteristics and the Perceived Importance of
Information. In: The Accounting Review, 48 (3): 511-519.
de Ruyter, Ko/Wetzels, Martin (1999): Commitment in Auditor–Client Relationships: Ante-
cedents and Consequences. In: Accounting, Organizations and Society, 24 (1): 57-75.
Deutsch, Morton (1958): Trust and Suspicion. In: Journal of Conflict Resolution, 2: 265-279.
DeVellis, Robert F. (2003): Scale Development: Theory and Applications, 2nd edition.
Thousand Oaks/London/New Delhi: Sage Publications.
Devolder, P. A. (1993): A Scale of Practical Problem-Solving Ability. In: Experimental Aging
Research, 19 (2): 129-146.
DeZoort, F. Todd/Lord, Alan T. (1994): An Investigation of Obedience Pressure Effects on
Auditors’ Judgments. In: Behavioral Research in Accounting, 6 (Supplement): 1-30.
DeZoort, F. Todd/Lord, Alan T. (1997): A Review and Synthesis of Pressure Effects Research
in Accounting. In: Journal of Accouting Literature, 16: 28-85.
Digman, John M. (1996): The Curious History of the Five-Factor Model. In: Wiggins, J. S.
(ed.): The Five-Factor Model of Personality: Theoretical Perspectives. New York/Lon-
don: The Guilford Press: 1-20.
Dilla, William N. et al. (1991): Predictive Bankruptcy Judgments by Auditors: A Probabilistic
Approach. In: Ponemon, L. A./Gabhart, D. R. L. (eds.): Auditing: Advances in Behav-
ioral Research. New York et al.: Springer-Verlag: 113-129.
Dillard, Jesse F./Kauffman, N. Leroy/Spires, Eric E. (1991): Evidence Order and Belief
Revision in Management Accounting Decisions. In: Accounting, Organizations and
Society, 16 (7): 619-633.
Dobbins, Gregory H./Lane, Irving M./Steiner, Dirk D. (1988): A Note on the Role of
Laboratory Methodologies in Applied Behavioural Research: Don’t Throw Out the Baby
with the Bath Water. In: Journal of Organizational Behavior, 9 (3): 281-286.
Donnelly, David P./Quirin, Jeffrey J./O’Bryan, David (2003): Attitudes Toward
Dysfunctional Audit Behavior: The Effects Of Locus Of Control, Organizational Com-
mitment, and Position. In: The Journal of Applied Business Research, 19 (1): 95-107.
Dopuch, Nicholas/King, Ronald R./Schwartz, Rachel (2001): An Experimental Investigation
of Retention and Rotation Requirements. In: Journal of Accounting Research, 39 (1): 93-
117.
253
Doucet, Mary S./Doucet, Thomas A (1996): Commentary on Ethics, Experience and
Professional Skepticism: A Situational Analysis. In: Behavioral Research in Accounting,
8 (Supplement): 158-168.
Doucet, Mary S./Hooks, Karen L. (1999): Toward an Equal Future. In: Journal of Account-
ancy, 187 (6): 71-76.
Douglas, Patricia Casey/Davidson, Ronald A./Schwartz, Bill N. (2001): The Effect of Organi-
zational Culture and Ethical Orientation on Accountants’ Ethical Judgments. In: Journal
of Business Ethics, 34 (2): 101-121.
Duska, Ronald (2005): The Good Auditor - Skeptic or Wealth Accumulator? Ethical Lessons
Learned from the Arthur Andersen Debacle. In: Journal of Business Ethics, 57 (1): 17-
29.
Dworkin, Gerald (1988): The Theory and Practice of Autonomy. Cambridge: Cambridge
University Press.
Eagly, Alice H. (1987): Sex Differences in Social Behavior: A Social-Role Interpretation.
Hillsdale: Lawrence Erlbaum Associates.
Eagly, Alice H./Chaiken, Shelly (1993): The Psychology of Attitudes. Belmont: Wadsworth,
Cengage Learning.
Earley, Christine E. (2001): Knowledge Acquisition in Auditing: Training Novice Auditors to
Recognize Cue Relationships in Real Estate Valuation. In: The Accounting Review, 76
(1): 81-97.
Earley, Christine E. (2003): A Note on Self-Explanation as a Training Tool for Novice
Auditors: The Effects of Outcome Feedback Timing and Level of Reasoning on Perfor-
mance. In: Behavioral Research in Accounting, 15: 111-124.
Earley, Christine E./Hoffman, Vicky B./Joe, Jennifer R. (2008): Reducing Management’s
Influence on Auditors’ Judgements: An Experimental Investigation of SOX 404
Assessments. In: The Accounting Review, 83 (6): 1461-1485.
Earley, P. Christopher et al. (1990): Impact of Process and Outcome Feedback on the Rela-
tion of Goal Setting to Task Performance. In: Academy of Management Journal, 33 (1):
87-105.
Edelmann, Walter/Wittmann, Simone (2012): Lernpsychologie, 7th edition. Weinheim: Beltz
Verlag.
Eichenbaum, Howard (2004): An Information Processing Framework for Memory Represen-
tation by the Hippocampus. In: Gazzaniga, M. S. (ed.): The Cognitive Neurosciences.
Cambridge: MIT Press: 679-690.
Eid, Michael/Luhmann, Maike (2012): Models of Measurement of Persons in Situations. In:
Seel, N. M. (ed.): Encyclopedia of the Sciences of Learning. New York et al.: Springer:
2323-2325.
254
Einhorn, Hillel J. (1976): A Synthesis: Accounting and Behavioral Science. In: Journal of
Accounting Research, 14 (Supplement): 196-206.
Einhorn, Hillel J./Hogarth, Robin M. (1978): Confidence in Judgment: Persistence of the
Illusion of Validity. In: Psychological Review, 85 (5): 395-416.
Einhorn, Hillel J./Hogarth, Robin M. (1981): Behavioral Decision Theory: Processes of Judg-
ment and Choice. In: Journal of Accounting Research, 19 (1): 1-31.
Einhorn, Hillel J./Hogarth, Robin M. (1985): A Contrast/Surprise Model for Updating
Beliefs. Unpublished Manuscript. URL: http://oai.dtic.mil/oai/oai?verb=getRecord&me
tadataPrefix=html&identifier=ADA153937, accessed on: 21.03.2014.
Einhorn, Hillel J./Hogarth, Robin M. (1987): Adaptation and Inertia in Belief Updating: The
Contrast-Inertia Model. Unpublished Manuscript.
Eining, Martha M./Jones, Donald R./Loebbecke, James K. (1997): Reliance on Decision Aids:
An Examination of Auditors’ Assessment of Management Fraud. In: Auditing: A
Journal of Practice & Theory, 16 (2): 1-19.
Eisenhart, Churchill (1947): The Assumptions Underlying the Analysis of Variance. In:
Biometrics, 3 (1): 1-21.
Elias, Rafik Z. (2002): Determinants of Earnings Management Ethics Among Accountants. In:
Journal of Business Ethics, 40 (1): 33-45.
Elliott, Robert K. (2002): Twenty-First Century Assurance. In: Auditing: A Journal of
Practice & Theory, 21 (1): 139-146.
Ellsberg, Daniel (1961): Risk, Ambiguity, and the Savage Axioms. In: Quarterly Journal of
Economics, 75: 643-649.
El-Masry, El-Hussein E./Hansen, Kathryn A. (2008): Factors Affecting Auditors’ Utilization
of Evidential Cues: Taxonomy and Future Research Directions. In: Managerial Auditing
Journal, 23 (1): 26-50.
Emby, Craig (1994): Framing and Presentation Mode Effects in Professional Judgment:
Auditors’ Internal Control Judgments and Substantive Testing Decisions. In: Auditing:
A Journal of Practice & Theory, 13: 102-115.
Endrawes, Medhat /Monroe, Gary S. (2012): Professional Scepticism of Auditors: A Cross-
Cultural Experiment. Working Paper. URL: http://www.unisa.edu.au/Global/business/
centres/cags/docs/seminars/Medhat%20G%20Monroe%20paper%20June%2027%20201
2.pdf, accessed on: 20.05.2014.
Engler, Barbara (2014): Personality Theories, 9th edition. Belmont: Wadsworth.
Epley, Nicholas/Gilovich, Thomas (2005): When Effortful Thinking Influences Judgmental
Anchoring: Differential Effects of Forewarning and Incentives on Self-Generated and
Externally Provided Anchors. In: Journal of Behavioral Decision Making, 18 (3): 199-
212.
255
Epley, Nicholas/Gilovich, Thomas (2006): The Anchoring-and-Adjustment Heuristic. In:
Psychological Science, 17 (4): 311-318.
European Commission (2009): Impact of the Economic Crisis on Key Sectors of the EU –
The Case of the Manufacturing and Construction Industries - October 2009 Update.
URL: http://ec.europa.eu/enterprise/newsroom/cf/_getdocument.cfm?doc_id=5509, ac-
cessed on: 29.04.2014.
European Commission (2010a): Green Paper - Audit Policy: Lessons from the Crisis. URL:
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2010:0561:FIN:EN:PDF,
accessed on: 13.11.2013.
European Commission (2010b): Impact of the Economic Crisis on Key Sectors of the EU –
The Case of the Manufacturing and Construction Industries. URL: http://ec.europa.eu/
enterprise/newsroom/cf/_getdocument.cfm?doc_id=5633, accessed on: 29.04.2014.
Evans, Jonathan St. B. T. (2008): Dual-Processing Accounts of Reasoning, Judgment, and
Social Cognition. In: Annual Review of Psychology, 59: 255-278.
Evans, Jonathan St. B. T. (2013a): Dual-Process Theories of Deductive Reasoning: Facts and
Fallacies. In: Holyoak, K. J./Morrison, R. G. (eds.): The Oxford Handbook of Thinking
and Reasoning. Oxford et al.: Oxford University Press: 115-133.
Evans, Jonathan St. B. T. (2013b): Reasoning, Rationality and Dual Processes: Selected
Works of Jonathan St B.T. Evans. Hove/New York: Psychology Press.
Eynon, Gail et al. (1996): An International Comparison of Ethical Reasoning Abilities:
Accounting Students from Ireland and the United States. In: Journal of Accounting
Education, 14 (4): 477-492.
Eysenck, Michael W./Keane, Mark T. (2010): Cognitive Psychology: A Students’ Handbook,
6th edition. Hove/New York: Psychology Press.
Fahrer, Marion (2012): Gathering the Evidence. In: Charter: Journal of the Institute of
Chartered Accountants in Australia, 83 (8): 30-32.
Fahrmeir, Ludwig et al. (2007): Statistik: Der Weg zur Datenanalyse, 6th edition. Berlin/Hei-
delberg: Springer-Verlag.
Falk, Haimet al. (1999): Auditor Independence, Self-Interested Behavior and Ethics: Some
Experimental Evidence. In: Journal of Accounting and Public Policy, 18 (4–5): 395-428.
Farag, Magdy S./Elias, Rafik Z. (2012): The Impact of Accounting Students’ Professional
Skepticism on Their Ethical Perception of Earnings Management. In: Research on
Professional Responsibility and Ethics in Accounting, 16: 185-200.
Farmer, Timothy A./Rittenberg, Larry E./Trompeter, Gregory M. (1987): An Investigation of
the Impact of Economic and Organizational Factors on Auditor Independence. In:
Auditing: A Journal of Practice & Theory, 7 (1): 1-14.
256
Favere-Marchesi, Michael (2006): “Order Effects” Revisited: The Importance of Chronology.
In: Auditing: A Journal of Practice & Theory, 25 (1): 69-83.
Favere-Marchesi, Michael/Emby, Craig E. N. (2005): The Impact of Continuity on Con-
curring Partner Reviews: An Exploratory Study. In: Accounting Horizons, 19 (1): 1-10.
Fein, Steven/Hilton, James L. (1994): Judging Others in the Shadow of Suspicion. In: Motiva-
tion and Emotion, 18 (2): 167-198.
Feist, Gregory J. (2006): The Psychology of Science and the Origins of the Scientific Mind.
New Haven/London: Yale University Press.
Felix, William L. (1974): A Decision Theory View of Auditing. In: Stettler, H. (ed.): Contem-
porary Auditing Problems: Proceedins of the Touche Ross/University of Kansas Sympo-
sium on Auditing Problems. Kansas: University of Kansas: 63-71.
Felix, William L./Kinney, William R. (1982): Research in the Auditor’s Opinion Formulation
Process: State of the Art. In: The Accounting Review, 57 (2): 245-271.
Fernberger, Samuel W. (1920): Interdependence of Judgments within the Series for the
Method of Constant Stimuli. In: Journal of Experimental Psychology, 3 (2): 126-150.
Field, Andy (2009): Discovering Statistics Using SPSS, 3rd edition. London et al.: Sage
Publications.
Field, Andy (2013): Discovering Statistics using IBM SPSS Statistics, 4th edition. London et
al.: Sage Publications.
Findley, Maureen J./Cooper, Harris M. (1983): Locus of Conrol and Academic Achievement:
A Literature Review. In: Journal of Personality and Social Psychology, 44 (2): 419-427.
Fink, Arlene (1995): How to Analyze Survey Data. Thousand Oaks/London/New Delhi: Sage
Publications.
Finucane, Melissa L. et al. (2000): The Affect Heuristic in Judgments of Risks and Benefits.
In: Journal of Behavioral Decision Making, 13: 1-17.
Firth, Michael (1990): Auditor Reputation: The Impact of Critical Reports Issued by Govern-
ment Inspectors. In: Journal of Economics, 21 (3): 374-387.
Fischhoff, Baruch (1975): Hindsight ≠ Foresight: The Effect of Outcome Knowledge on Judg-
ment under Uncertainty. In: Journal of Experimental Psychology: Human Perceptions
and Performance, 1: 288-299.
Fischhoff, Baruch (1982): Debiasing. In: Kahneman, D./Slovic, P./Tversky, A. (eds.): Judg-
ment under Uncertainty: Heuristics and Biases. Cambridge et al.: Cambridge University
Press: 335-351.
Fischhoff, Baruch/Slovic, Paul/Lichtenstein, Sarah (1977): Knowing with Certainty: The
Appropriateness of Extreme Confidence. In: Journal of Experimental Psychology, 3 (4):
552-564.
257
Fiske, Susan T. (1980): Attention and Weight in Person Perception: The Impact of Negative
and Extreme Behavior. In: Journal of Personality and Social Psychology, 38 (6): 889-
906.
Fiske, Susan T./Taylor, Shelly E. (2008): Social Cognition: From Brains to Culture. New
York: McGraw-Hill.
Fleming, James S./Courtney, Barbara E. (1984): The Dimensionality of Self-Esteem II: Hier-
archical Facet Model for Revised Measurement Scales. In: Journal of Personality and
Social Psychology, 46 (2): 404–421.
Fodor, Jerry A. (1983): The Modularity of Mind, 2nd print. Cambridge/London: MIT Press.
Fogelin, Robert J. (1994): Pyrrhonian Reflections on Knowledge and Justification. New
York: Oxford University Press.
Ford, Gary T./Smith, Darlene B./Swasy, John L. (1990): Consumer Skepticism of Advertising
Claims: Testing Hypotheses from Economics of Information. In: Journal of Consumer
Research, 16 (4): 433-441.
Forehand, Mark R./Grier, Sonya (2003): When Is Honesty the Best Policy? The Effect of
Stated Company Intent on Consumer Skepticism. In: Journal of Consumer Psychology,
13 (3): 349.
Fornelli, Cindy/Desmond, Angela (2011): The Art of Skepticism. URL: http://www.directo
rship.com/the-art-of-skepticism/, accessed on: 16.9.2013.
Forsman, Lennart/Johnson, Maarit (1996): Dimensionality and Validity of Two Scales
Measuring Different Aspects of Self-Esteem. In: Scandinavian Journal of Psychology,
37: 1–15.
Forsyth, Donelson R. (1980): A Taxonomy of Ethical Ideologies. In: Journal of Personality
and Social Psychology, 39 (1): 175-184.
Francis, Jere R. (2011a): A Framework for Understanding and Researching Audit Quality. In:
Auditing: A Journal of Practice & Theory, 30 (2): 125-152.
Francis, Jere R. (2011b): Auditing without Borders. In: Accounting, Organizations and Socie-
ty, 36 (4): 318-323.
Frankel, Richard M./Johnson, Marilyn F./Nelson, Karen K. (2002): The Relation between
Auditors’ Fees for Nonaudit Services and Earnings Management. In: The Accounting
Review, 77 (4): 71-105.
Franzel, Jeanette M. (2013): Auditor Objectivity and Skepticism – What’s Next? URL:
http://pcaobus.org/News/Speech/Pages/08052013_AAA.aspx, accessed on: 21.9.2013.
FRC (2010): Auditor Scepticism: Raising the Bar. URL: http://www.frc.org.uk/getattachment/
2a1e0146-a92c-4b7e-bf33-305b3b10fcd2/Discussion-Paper-Auditor-Scepticism-Raisin
g-the-Ba.aspx, accessed on: 06.02.2013.
258
FRC (2012): Professional Scepticism: Establishing a Common Understanding and
Reaffirming its Central Role in Delivering Audit Quality. URL: http://www.frc.org.
uk/Our-Work/Publications/APB/Briefing-Paper-Professional-Scepticism.aspx, accessed
on 06.03.2014.
FRC (2013): Audit Quality Inspections: Annual Report 2012/13. URL: http://www.frc.org.uk/
Our-Work/Publications/Audit-Quality-Review/Audit-Quality-Inspections-Annual-Repo
rt-2012-13.aspx, accessed on: 25.07.2013.
Frederick, David M. (1991): Auditors’ Representation and Retrieval of Internal Control
Knowledge. In: The Accounting Review, 66 (2): 240-258.
Frisch, Deborah (1993): Reasons for Framing Effects In: Organizational Behavior and Hu-
man Decision Processes, 54: 399-429.
Fromlet, Hubert (2001): Behavioral Finance - Theory and Practical Application. In: Business
Economics, 36 (3): 63-69.
Fuchs, Stephan (2001): Against Essentialism: A Theory of Culture and Society. Cambridge:
Harvard University Press.
Fukukawa, Hironori/Mock, Theodore J. (2011): Does Assertion Framing Affect Professional
Skepticism? Working Paper presented at the 2011 International Symposium on Audit
Research in Quebec, Canada.
Fülbier, Rolf Uwe (2004): Wissenschaftstheorie und Betriebswirtschaftslehre In: Wirtschafts-
wissenschaftliches Studium, 33 (5): 266-271.
Fullerton, Rosemary R./Durtschi, Cindy (2004): The Effect of Professional Skepticism on the
Fraud Detection Skills of Internal Auditors. Working Paper. URL: http://papers.ssrn.co
m/sol3/papers.cfm?abstract_id=617062, accessed on: 06.02.2014.
Gaertner, James F./Ruhe, John A. (1981): Job-Related Stress in Public Accounting. In:
Journal of Accountancy, 151 (6): 68-74.
Galotti, Kathleen M. et al. (2009): Cognitive Psychology: In and Out of the Laboratory.
Toronto: Nelson Education.
Garfield, Jay L. (1991): Introduction: Carving the Mind at Its Joints. In: Garfield, J. L. (ed.):
Modularity in Knowledge and Representation and Natural-Language Understanding.
Cambridge: MIT Press: 1-16.
Gavious, Ilanit (2007): Alternative Perspectives to Deal with Auditors’ Agency Problem. In:
Critical Perspectives on Accounting, 18 (4): 451-467.
Gazzaniga, Michael S. (1985): The Social Brain: Discovering the Networks of the Mind. New
York: Basic Books.
Geiger, Marshall A./Raghunandan, K. (2002): Auditor Tenure and Audit Reporting Failures.
In: Auditing: A Journal of Practice & Theory, 21 (1): 67-78.
259
Gelardi, Alexander M. G. (2010): Information Quantity and Order in Students’ Tax Research
Judgments. In: Asian Journal of Finance & Accounting, 2 (1): 25-45.
Gemünden, Hans Georg (1993): Informationsverhalten. In: Hauschildt, J./Grün, O. (eds.):
Ergebnisse empirischer betriebswirtschaftlicher Forschung - Zu einer Realtheorie der
Unternehmung. Stuttgart: Schäffer-Poeschel Verlag: 839-877.
Ghauri, Pervez N./Grønhaug, Kjell (2005): Research Methods in Business Studies: A
Practical Guide, 3rd edition. Harlow: Financial Times Prentice Hall.
Gibbins, Michael (1984): Propositions about the Psychology of Professional Judgment in
Public Accounting. In: Journal of Accounting Research, 22 (1): 103-125.
Gibbins, Michael/Newton, James D. (1994): An Empirical Exploration of Complex Account-
ability in Public Accounting. In: Journal of Accounting Research, 32 (2): 165-186.
Gigerenzer, Gerd (2007): Fast and Frugal Heuristics: The Tools of Bounded Rationality. In:
Koehler, D. J./Harvey, N. (eds.): Blackwell Handbook of Judgment and Decision
Making. Malden/Oxford/Carlton: Blackwell Publishing: 62-88.
Gilbert, Daniel T. (1991): How Mental Systems Believe. In: American Psychologist, 46 (2):
107-119.
Gilbert, Daniel T. (1999): What the Mind’s Not. In: Chaiken, S./Trope, Y. (eds.): Dual-
Process Theories in Social Psychology. New York: Guilford Press: 3-11.
Glass, Gene V./Peckham, Percy D./Sanders, James R. (1972): Consequences of Failure to
Meet Assumptions Underlying the Fixed Effects Analyses of Variance and Covariance.
In: Review of Educational Research, 42 (3): 237-288.
Glover, Steven M. (1997): The Influence of Time Pressure and Accountability on Auditors’
Processing of Nondiagnostic Information. In: Journal of Accounting Research, 35 (2):
213-226.
Glover, Steven M./Prawitt, Douglas F. (2013): Enhancing Auditor Professional Skepticism.
URL:http://www.thecaq.org/docs/research/skepticismreport.pdf,accessed on 24.05.2014.
Glynn, John J. (1985): Value for Money Auditing in the Public Sector. Englewood Cliffs:
Prentice-Hall International.
Goffman, Erving (1959): The Presentation of Self in Everyday Life. Garden City: Doubleday
Anchor Books.
Gold, Anna/Hunton, James E./Gomaa, Mohamed I. (2009): The Impact of Client and Auditor
Gender on Auditors’ Judgments. In: Accounting Horizons, 23 (1): 1-18.
Goldman, Arieh/Barlev, Benzion (1974): The Auditor-Firm Conflict of Interests: Its
Implications for Independence. In: The Accounting Review, 49 (4): 707-718.
Goodwin, Jenny (1999): The Effects of Source Integrity and Consistency of Evidence on
Auditors’ Judgments. In: Auditing: A Journal of Practice & Theory, 18 (2): 1-16.
260
Gramling, Audrey A. (1999): External Auditors’ Reliance on Work Performed by Internal
Auditors: The Influence of Fee Pressure on This Reliance Decision. In: Auditing: A
Journal of Practice & Theory, 18 (Supplement): 117-135.
Gravetter, Frederick J/Wallnau, Larry B. (2012): Essentials of Statistics for the Behavioral
Sciences, 8th edition. Belmont: Wadswoth, Cengage Learning.
Greco, John (2008): Introduction. In: Greco, J. (ed.): The Oxford Handbook of Skepticism.
Oxford et al.: Oxford University Press: 3-6.
Green, Sharon Weiner/Wolf, Ira K. (2011): Barron’s New GRE Graduate Record
Examination, 19th edition. Hauppauge: Barron Educational Service.
Green, Wendy (2008): Are Industry Specialists More Efficient and Effective in Performing
Analytical Procedures? A Multi-stage Analysis. In: International Journal of Auditing, 12
(3): 243-260.
Grenier, Jonathan H. (2013): Encouraging Professional Skepticism in the Industry Special-
ization Era. Working Paper. URL: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1
533622, accessed on: 19.12.2013.
Griffin, Dale/Tversky, Amos (1992): The Weighing of Evidence and the Determinants of
Confidence. In: Cognitive Psychology, 24: 411-435.
Gross, Richard (2009): Introduction to Basic Psychology. In: Puri, B./Treasaden, I. (eds.):
Psychiatry: An Evidence-Based Text. Boca Raton: CRC Press: 127-136.
Gross, Richard (2012): Psychology: The Science of Mind and Behaviour, 6th edition.
Abingdon: Hodder Education.
Grumet, Louis (2003): Through the Eyes of an Auditor: Trust and Verify. In: CPA Journal, 73
(1): 9-9.
Guilford, J. P. (1950): Creativity. In: American Psychologist, 5 (9): 444-454.
Guilford, J. P. (1959): Personality. New York: McGraw-Hill.
Guilford, J. P. (1967): The Nature of Human Intelligence. New York: McGraw-Hill.
Guiral, Andrés/Esteo, Francisco (2006): Are Spanish Auditors Skeptical in Going Concern
Evaluations? In: Managerial Auditing Journal, 21 (6): 598-620.
Guiral-Contreras, Andrés/Gonzalo-Angulo, Jose A./Rodgers, Waymond (2007): Information
Content and Recency Effect of the Audit Report in Loan Rating Decisions. In: Account-
ing and Finance, 47: 285-304.
Gul, Ferdinand A. (1983): A Note on the Relationship between Age, Experience, Cognitive
Styles and Accountants’ Decision Confidence. In: Accounting and Business Research,
14 (53): 85-88.
Gul, Ferdinand A. (1984): The Joint and Moderating Role of Personality and Cognitive Style
on Decision Making. In: The Accounting Review, 59 (2): 264-277.
261
Gul, Ferdinand A. (1986): Tolerance for Ambiguity, Auditors’ Opinions and Their Effects on
Decision Making. In: Accounting and Business Research, 16 (62): 99-105.
Gul, Ferdinand A./Zaid, Omer (1981): Field Dependence and Accountants’ Confidence in
Decisions. In: Psychological Reports, 49: 949-950.
Habib, Ahsan (2012): Non-Audit Service Fees and Financial Reporting Quality: A Meta-
Analysis. In: Abacus, 48 (2): 214-248.
Hackenbrack, Karl (1992): Implications of Seemingly Irrelevant Evidence in Audit Judgment.
In: Journal of Accounting Research, 30 (1): 126-136.
Hackenbrack, Karl/Nelson, Mark W. (1996): Auditors’ Incentives and Their Application of
Financial Accounting Standards. In: The Accounting Review, 71 (1): 43-59.
Hair, Joseph F./Black, William C./Babin, Barry J. (2010): Multivariate Data Analysis: A
Global Perspective, 7th edition. Upper Saddle River: Pearson Education.
Hallie, Philip P. (ed.) (1985): Sextus Empiricus: Selections from the Major Writings on
Scepticism, Man & God. Indianapolis: Hackett Publishing.
Halpern, Diane F. (2013): Sex Differences in Cognitive Abilities, 4th edition. New
York/Hove: Taylor & Francis.
Hammar, Äsa (2012): Convergent Thinking and Learning. In: Seel, N. M. (ed.): Encyclopedia
of the Sciences of Learning. New York et al.: Springer: 811-812.
Hammersley, Jacqueline S. (2006): Pattern Identification and Industry-Specialist Auditors. In:
The Accounting Review, 81 (2): 309-336.
Hammersley, Jacqueline S. (2011): A Review and Model of Auditor Judgments in Fraud-
Related Planning Tasks. In: Auditing: A Journal of Practice & Theory, 30 (4): 101-128.
Hands, David S./Avons, S. E. (2001): Recency and Duration Neglect in Subjective Assess-
ment of Television Picture Quality. In: Applied Cognitive Psychology, 15 (6): 639-657.
Handzic, Meliha (2004): Knowledge Management: Through the Technology Glass. Singa-
pore: World Scientific Publishing.
Hardies, Kris/Breesch, Diane/Branson, Joel (2014): Do (Fe)Male Auditors Impair Audit
Quality? Evidence from Going Concern Opinions. Working Paper. URL:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1583206, accessed on: 05.08.2014.
Harding, Noel/Trotman, Ken T. (2011): Enhancing Professional Skepticism via the Fraud
Brainstorming Discussion Outcomes. Working Paper.
Harlow, Lisa L. (2014): The Essence of Multivariate Thinking: Basic Themes and Methods,
2nd edition. New York/Hove: Routledge.
Harper, William L. (1976): Rational Conceptual Change. In: PSA (ed.): Proceedings of the
Biennial Meeting of the Philosophy of Science Association: 462-494.
262
Hartmann, Frank (2005): The Effects of Tolerance for Ambiguity and Uncertainty on the
Appropriateness of Accounting Performance Measures. In: Abacus, 41 (3): 241-264.
Harwell, Michael R. et al. (1992): Summarizing Monte Carlo Results in Methodological
Research: The One- and Two-Factor Fixed Effects ANOVA Cases. In: Journal of
Educational and Behavioral Statistics, 17 (4): 315-339.
Haskins, Mark E./Baglioni, A. J., Jr./Cooper, Cary L. (1990): An Investigation of the
Sources, Moderators, and Psychological Symptoms of Stress Among Audit Seniors. In:
Contemporary Accounting Research, 6 (2): 361-385.
Hastie, Reid (1980): Memory for Behavioral Information that Confirms or Contradicts a
Personality Impression. In: Hastie, R. et al. (eds.): Person Memory : The Cognitive Basis
of Social Perception. Hillsdale: Erlbaum: 155-177.
Hawkins, Scott A./Hastie, Reid (1990): Hindsight: Biased Judgments of Past Events After the
Outcomes Are Known. In: Psychological Bulletin, 107: 311-327.
Hayes, John/Allinson, Christopher W. (1998): Cognitive Style and the Theory and Practice of
Individual and Collective Learning in Organizations. In: Human Relations, 51 (7): 847-
871.
Haynes, Christine M./Jenkins, J. Gregory/Nutt, Stacey R. (1998): The Relationship Between
Client Advocacy and Audit Experience: An Exploratory Analysis. In: Auditing: A
Journal of Practice & Theory, 17 (2): 88-104.
Heintz, James A./White, Gwendolen B. (1989): Auditor Judgment in Analytical Review: Some
Further Evidence. In: Auditing: A Journal of Practice & Theory, 8 (2): 22-39.
Helmreich, Robert/Stapp, Joy (1974): Short Forms of the Texas Social Behavior Inventory
(TSBI): An Objective Measure of Self-Esteem. In: Bulletin of the Psychonomic Society,
4 (5-A): 473-475.
Heninger, William G. (2001): The Association between Auditor Litigation and Abnormal
Accruals. In: The Accounting Review, 76 (1): 111-126.
Hicks, Ernest L. (1974): Standards for the Attest Function. In: Journal of Accountancy, 138
(2): 39-45.
Highhouse, Scott/Gallo, Andrew (1997): Order Effects in Personnel Decision Making. In:
Human Performance, 10 (1): 31-46.
Higson, Andrew (1997): Developments in Audit Approaches: From Audit Efficiency to Audit
Effectiveness? In: Sherer, M./Turley, S. (eds.): Current Issues in Auditing, 3rd edition.
London: Paul Chapman Publishing: 198-215.
Higson, Andrew (2003): Corporate Financial Reporting: Theory and Practice. London/Thou-
sand Oaks/New Delhi: Sage Publications.
Hill, Gayle W. (1982): Group Versus Individual Performance: Are N + 1 Heads Better Than
One? In: Psychological Bulletin, 91 (3): 517-539.
263
Hilton, James L./Fein, Steven /Miller, Dale T. (1993): Suspicion and Dispositional Inference.
In: Personality and Social Psychology Bulletin, 19 (5): 501-512.
Hirschey, Mark (2009): Managerial Economics, 12th edition. Mason: South-Western, Cen-
gage Learning.
Hite, Peggy A./Stock, Tobby (1996): Taxpayer Belief Revisions and the Importance of
Controlling for Prior Beliefs. In: Advances in Taxation, 8: 81-99.
Ho, Joanna L./Chang, Chengyee Janie (1994): Does National Culture or Professional Knowl-
edge Affect Auditors’ Probabilistic Conjunction Judgments? A Study of the United
States versus Taiwan. In: International Journal of Accounting, 29: 189-205.
Hoffman, Paul J./Slovic, Paul/Rorer, Leonard G. (1968): An Analysis-of-Variance Model for
the Assessment of Configural Cue Utilization in Clinical Judgment. In: Psychological
Bulletin, 69 (5): 338-349.
Hoffman, Vicky B./Patton, James M. (1997): Accountability, the Dilution Effect, and Con-
servatism in Auditors’ Fraud Judgments. In: Journal of Accounting Research, 35 (2):
227-237.
Hoffrage, Ulrich (2004): Overconfidence. In: Pohl, R. F. (ed.): Cognitive Illusions: A Hand-
book on Fallacies and Biases in Thinking, Judgment and Memory. Hove/New York:
Psychology Press: 235-254.
Hofstede, Geert H. (1980): Culture’s Consequences: International Differences of Work
Related Values. Beverly Hills: Sage Publications.
Hofstede, Geert H. (2001): Culture’s Consequences: Comparing Values, Behaviors, Institu-
tions and Organizations Across Nations, 2nd edition. Thousand Oaks: Sage Publications.
Hogan, Chris E. et al. (2008): Financial Statement Fraud: Insights from the Academic Litera-
ture. In: Auditing: A Journal of Practice & Theory, 27 (2): 231-252.
Hogarth, Robin M./Einhorn, Hillel J. (1992): Order Effects in Belief Updating: The Belief-
Adjustment Model. In: Cognitive Psychology, 24: 1-55.
Hogarth, Robin M./Reder, Melvin W. (1986): Introduction: Perspectives from Economics and
Psychology. In: Hogarth, R. M./Reder, M. W. (eds.): Rational Choice – The Contrast
Between Economics and Psychology. Chicago/London: University of Chicago Press: 1–
23.
Holt, Charles A./Smith, Angela M. (2009): An Update on Bayesian Updating. In: Journal of
Economic Behavior & Organization, 69: 125-134.
Hookway, Christoper (1990): Scepticism. London/New York: Routledge.
Hosmer, Larue Tone (1994): Strategic Planning as if Ethics Mattered. In: Strategic Manage-
ment Journal, 15: 17-34.
Houston, Richard W. (1999): The Effects of Fee Pressure and Client Risk on Audit Seniors’
Time Budget Decisions. In: Auditing: A Journal of Practice & Theory, 18 (2): 70-86.
264
Hronsky, Jane J. F./Houghton, Keith A. (2001): The Meaning of a Defined Accounting
Concept: Regulatory Changes and the Effect on Auditor Decision Making. In: Account-
ing, Organizations and Society, 26 (2): 123-139.
Huemer, Michael (2001): Skepticism and the Veil of Perception. Lanham: Rowman & Little-
field.
Hughes, Susan B. et al. (2009): The Impact of Cultural Environment on Entry-level Auditors’
Abilities to Perform Analytical Procedures. In: Journal of International Accounting,
Auditing and Taxation, 18: 29-43.
Hume, David (1748): An Enquiry Concerning Human Understanding, reprint edition 1999
(edited by Beauchamp, T. L.). Oxford et al.: Oxford University Press.
Hurtt, R. Kathy (2010): Development of a Scale to Measure Professional Skepticism. In:
Auditing: A Journal of Practice & Theory, 29 (1): 149-171.
Hurtt, R. Kathy/Eining, Martha/Plumlee, R. David (2010): Linking Professional Skepticism to
Auditors’ Behaviors. Working Paper. URL: http://www.google.de/url?sa=t&rct=j&
q=&esrc=s&source=web&cd=1&ved=0CCcQFjAA&url=https%3A%2F%2F1.800.gay%3A443%2Fhttp%2Fwww.profb
ailey.com%2Fseminar%2FReadings%2FSkepticism%2FLinking%2520Professional%25
20Skepticism.doc&ei=GZbjU-GFO-XRywPKzICwAw&usg=AFQjCNGg7IuGDb5G8
Wwn68moO6s58jGjsQ&sig2=f-pY1UYr81r8FJ7azd3fGg&bvm=bv.72676100,d.bGE,
accessed on: 08.02.2014.
Hurtt, R. Kathy et al. (2013): Research on Auditor Professional Skepticism: Literature
Synthesis and Opportunities for Future Research. In: Auditing: A Journal of Practice &
Theory, 32 (Supplement 1): 45-97.
Hyatt, Troy A./Prawitt, Douglas F. (2001): Does Congruence between Audit Structure and
Auditors’ Locus of Control Affect Job Performance? In: The Accounting Review, 76
(2): 263-274.
IAASB (2012): Professional Skepticism in an Audit of Financial Statements - Staff Questions
and Answers. URL: http://www.ifac.org/sites/default/files/publications/files/IAASB%2
0Professional%20Skepticism%20QandA-final.pdf, accessed on: 06.02.2013.
IAASB (2013): Handbook of International Quality Control, Auditing Review, Other
Assurance, and Related Services Pronouncements - Volume I. New York: IFAC.
IFIAR (2013): Annual Report and Accounts 2012. URL: https://www.ifiar.org/IFIAR/media/
Documents/General/About%20Us/IFIAR%20Member%20Updates%202013/IFIAR-An
nual-Report-2012.pdf, accessed on: 29.07.2013.
James, William (1890): The Principles of Psychology, Volume 2, reprint edition 2010. New
York: Digireads.com Publishing.
James, William (1912): The Will to Believe: And Other Essays in Popular Philosophy, reprint
edition 2010. Portland: Floating Press.
265
Janvrin, Diane/Bierstaker, James/Lowe, D. Jordan (2008): An Examination of Audit Infor-
mation Technology Use and Perceived Importance. In: Accounting Horizons, 22 (1): 1-
21.
Jeffrey, Cynthia/Dilla, William/Weatherholt, Nancy (2004): The Impact of Ethical Develop-
ment and Cultural Constructions on Auditor Judgments: A Study of Auditors in Taiwan.
In: Business Ethics Quarterly, 14 (3): 553-579.
Jennings, Daniel E./Zeithaml, Carl P. (1983): Locus of Control: A Review and Directions for
Entrepreneurial Research. Academy of Management Proceedings: 417-421.
Jiambalvo, James/Waller, William (1984): Decomposition and Assessments of Audit Risk.
In: Auditing: A Journal of Practice & Theory, 3 (2): 80-88.
Johnson, Eric N. (1995): Effects of Information Order, Group Assistance, and Experience on
Auditors’ Sequential Belief Revision. In: Journal of Economic Psychology, 16: 137-160.
Johnson, Oliver A. (1978): Skepticism and Cognitivism: A Study in the Foundations of
Knowledge. Berkeley/Los Angeles/London: University of California Press.
Johnson, Van E./Kaplan, Steven E. (1991): Experimental Evidence on the Effects of Account-
ability on Auditor Judgments. In: Auditing: A Journal of Practice & Theory, 10 (Supple-
ment): 96-107.
Johnson, Van E./Khurana, Inder K./Reynolds, J. Kenneth (2002): Audit-Firm Tenure and the
Quality of Financial Reports. In: Contemporary Accounting Research, 19 (4): 637-660.
Johnstone, Karla M./Sutton, Michael H./Warfield, Terry D. (2001): Antecedents and Conse-
quences on Independence Risk: Framework for Analysis. In: Accounting Horizons, 15
(1): 1-18.
Jones, Joanne/Massey, Dawn W./Thorne, Linda (2003): Auditors’ Ethical Reasoning: Insights
from Past Research and Implications for the Future. In: Journal of Accounting Litera-
ture, 22: 45-103.
Jonnergård, Karin/Stafsudd, Anna/Elg, Ulf (2010): Performance Evaluations as Gender Bar-
riers in Professional Organizations: A Study of Auditing Firms. In: Gender, Work &
Organization, 17 (6): 721-747.
Joyce, Edward J. (1976): Expert Judgment in Audit Program Planning. In: Journal of
Accounting Research, 14 (Supplement): 29-60.
Joyce, Edward J./Biddle, Gary C. (1981): Anchoring and Adjustment in Probabilistic
Inference in Auditing. In: Journal of Accounting Research, 19 (1): 120-145.
Junaidi/Miharjo, Setiyono/Hartadi, Bambang (2012): Does Auditor Tenure Reduce Audit
Quality? In: Gadjah Mada International Journal of Business, 14 (3): 303-315.
Kachelmeier, Steven J./Messier, William F., Jr. (1990): An Investigation of the Influence of a
Nonstatistical Decision Aid on Auditor Sample Size Decisions. In: The Accounting
Review, 65 (1): 209-226.
266
Kahle, Jennifer/Pinsker, Robert/Pennington, Robin (2005): Belief Revision in Accounting: A
Literature Review of the Belief-Adjustment Model. In: Advances in Accounting Behav-
ioral Research, 8: 1-40.
Kahneman, Daniel (1973): Attention and Effort. Englewood Cliffs: Prentice-Hall.
Kahneman, Daniel (2003a): A Psychological Perspective on Economics. In: The American
Economic Review, 93 (2): 162-168.
Kahneman, Daniel (2003b): Maps of Bounded Rationality: Psychology for Behavioral Eco-
nomics. In: The American Economic Review, 93 (5): 1449-1475.
Kahneman, Daniel (2011): Thinking, Fast and Slow. London: Penguin Books.
Kahneman, Daniel/Frederick, Shane (2002): Representativeness Revisited: Attribute Substi-
tution in Intuitive Judgment. In: Gilovich, T./Griffin, D./Kahneman, D. (eds.): Heuristics
and Biases: The Psychology of Intuitive Judgment. Cambridge et al.: Cambridge Univer-
sity Press: 49-81.
Kahneman, Daniel/Frederick, Shane (2005): A Model of Heuristic Judgment. In: Holyoak, K.
J./Morrison, R. G. (eds.): Cambridge Handbook of Thinking and Reasoning. New York:
Cambridge University Press: 267-293.
Kalat, James W. (2012): Biological Psychology, 11th edition. Belmont: Wadsworth, Cengage
Learning.
Kao, Jennifer L./Li, Yan/Zhang, Wenjun (2013): Has SOX Enhanced Non-Big 4 Auditors’
Ability to Deal with Client Pressure? In: Accounting Perspectives, 12 (2): 141-164.
Kaplan, Steven E./Reckers, Philip M. J. (1984): An Empirical Examination of Auditors’
Initial Planning Processes. In: Auditing: A Journal of Practice & Theory, 4 (1): 1-19.
Karni, Edi (2013): Axiomatic Foundations of Expected Utility and Subjective Probability. In:
Machina, M./Viscusi, W. K. (eds.): Handbook of the Economics of Risk and Uncertain-
ty. Oxford/Amsterdam: North-Holland.
Kaufman, Bruce E. (1999): Expanding the Behavioral Foundations of Labor Economics. In:
Industrial and Labor Relations Review, 52 (3): 361-392.
Kayadelen, Engin (2007): Zur Durchführung von analytischen Prüfungshandlungen im
Rahmen der Jahresabschlussprüfung. Dissertation, Technische Universität Darmstadt.
URL: http://tuprints.ulb.tu-darmstadt.de/epda/000971/DissertationKayadelen.pdf, ac-
cessed on: 21.04.2014.
Kee, Herbert W./Knox, Robert E. (1970): Conceptual and Methodological Considerations in
the Study of Trust and Suspicion. In: The Journal of Conflict Resolution, 14 (3): 357-
366.
Kelley, Tim/Margheim, Loren (1990): The Impact of Time Budget Pressure, Personality, and
Leadership Variables on Dysfunctional Auditor Behavior. In: Auditing: A Journal of
Practice & Theory, 9 (2): 21-42.
267
Keltz, Ilean K./Adelman, Leonard (2008): Testing Information Order Effects in a Long Series
of Evidence. In: Engineering Management Journal, 20 (2): 22-29.
Kennedy, Jane (1993): Debiasing Audit Judgment with Accountability: A Framework and
Experimental Results. In: Journal of Accounting Research, 31 (2): 231-245.
Kennedy, Jane (1995): Debiasing the Curse of Knowledge in Audit Judgment. In: The
Accounting Review, 70 (2): 249-273.
Kennedy, Jane/Peecher, Mark E. (1997): Judging Auditors’ Technical Knowledge. In: Journal
of Accounting Research, 35 (2): 279-293.
Kennedy, Robert S. et al. (2005): Quantifying Human Information Processing (QHIP): Can
Practice Effects Alleviate Bottlenecks? In: McBride, D. K./Schmorrow, D. (eds.): Quan-
tifying Human Information Processing. Lanham/Oxford: Lexington Books: 91-115.
Keppel, Geoffrey (1973): Design and Analysis: A Researcher’s Handbook. Englewood Cliffs:
Prentice-Hall.
Keren, Gideon (2011): On the Definition and Possible Underpinnings of Framing Effects: A
Brief Review and a Ciritical Evaluation. In: Keren, G. (ed.): Perspectives on Framing.
New York/Hove: Psychology Press: 3-34.
Keren, Gideon/Schul, Yaacov (2009): Two Is Not Always Better Than One: A Critical Evalu-
ation of Two-System Theories. In: Perspectives on Psychological Science, 4 (6): 533-
550.
Keren, Gideon/Teigen, Karl H. (2007): Yet Another Look at the Heuristics and Biases
Approach. In: Koehler, D. J./Harvey, N. (eds.): Blackwell Handbook of Judgment and
Decision Making. Malden/Oxford/Carlton: Blackwell Publishing: 89-109.
Kerler, William A./Killough, Larry N. (2009): The Effects of Satisfaction with a Client’s
Management During a Prior Audit Engagement, Trust, and Moral Reasoning on
Auditors’ Perceived Risk of Management Fraud. In: Journal of Business Ethics, 85 (2):
109-136.
Kerlinger, Fred N./Lee, Howard B. (2000): Foundations of Behavioral Research, 4th edition.
Singapore et al.: Wadsworth, Thomson Learning.
Kerr, David S./Ward, D. Dewey (1994): The Effects of Audit Task on Evidence Integration
and Belief Revision. In: Behavioral Research in Accounting, 6: 21-42.
Kerstholt, José H./Jackson, Janet L. (1998): Judicial Decision Making: Order of Evidence
Presentation and Availability of Background Information. In: Applied Cognitive
Psychology, 12 (5): 445-454.
Kida, Thomas (1984): The Impact of Hypothesis-Testing Strategies on Auditors’ Use of
Judgment Data. In: Journal of Accounting Research, 22 (1): 332-340.
268
King, Ronald (1991): Using Experimental Economics in Auditing Research. In: Ponemon, L.
A./Gabhart, D. R. L. (eds.): Auditing: Advances in Behavioral Research. New York et
al.: Springer-Verlag: 93-112.
Kinney, William R. (1975): A Decision-Theory Approach to the Sampling Problem in Audit-
ing. In: Journal of Accounting Research, 13 (1): 117-132.
Kinney, William R. (1994): Audit Litigation Research: Professional Help is Needed. In:
Accounting Horizons, 8 (2): 80-86.
Kinney, William R., Jr./Uecker, Wilfred C. (1982): Mitigation of the Consequences of
Anchoring in Auditor Judgments. In: The Accounting Review, 57 (1): 55-69.
Kinney, William R., Jr./Palmrose, Zoe-Vonna/Scholz, Susan (2004): Auditor Independence,
Non-Audit Services, and Restatements: Was the U.S. Government Right? In: Journal of
Accounting Research, 42 (3): 561-588.
Kirk, Roger E. (2013): Experimental Design: Procedures for the Behavioral Sciences, 4th
edition. Thousand Oaks: Sage Publications.
Kleinbaum, David et al. (2008): Applied Regression Analysis and Other Multivariable
Methods, 4th edition. Belmont: Thomson.
Kleinmuntz, Don N. (1990): Decomposition and the Control of Error in Decision-Analytic
Models. In: Hogarth, R. M. (ed.): Insights in Decision Making: A Tribute to Hillel J.
Einhorn. Chicago/London: The University of Chicago Press: 107-126.
Kluemper, Donald H./Little, Laura M./DeGroot, Timothy (2009): State or Trait: Effects of
State Optimism on Job-Related Outcomes. In: Journal of Organizational Behavior, 30
(2): 209-231.
Kluger, Avraham N./DeNisi, Angelo (1996): The Effects of Feedback Interventions on Perfor-
mance: A Historical Review, a Meta Analysis, and a Preliminary Feedback Intervention
Theory. In: Psychological Bulletin, 119 (2): 254-284.
Kluger, Avraham N./DeNisi, Angelo (1998): Feedback Interventions: Toward the Understand-
ing of a Double-Edged Sword. In: Current Directions in Psychological Science, 7 (3):
67-72.
Knapp, Carol A./Knapp, Michael C. (2001): The Effects of Experience and Explicit Fraud
Risk Assessment in Detecting Fraud with Analytical Procedures. In: Accounting,
Organizations and Society, 26 (1): 25-37.
Knapp, Michael C. (1985): Audit Conflict: An Empirical Study of the Perceived Ability of
Auditors to Resist Management Pressure. In: The Accounting Review, 60 (2): 202-211.
Knechel, W. Robert/Vanstraelen, Ann (2007): The Relationship between Auditor Tenure and
Audit Quality Implied by Going Concern Opinions. In: Auditing: A Journal of Practice
& Theory, 26 (1): 113-131.
269
Knechel, W. Robert et al. (2013): Audit Quality: Insights from the Academic Literature. In:
Auditing: A Journal of Practice & Theory, 32: 385-421.
Knoblett, James A. (1970): The Applicability of Bayesian Statistics in Auditing. In: Decision
Sciences, 1: 423-440.
Koch, Christopher/Schunk, Daniel (2013): Limiting Liability? - Risk and Ambiguity Attitudes
under Real Losses. In: sbr, 65 (1): 54-75.
Koch, Christopher/Wüstemann, Jens (2009): Review of Bias Research in Auditing: Oppor-
tunities for Combining Psychological and Economic Research. Working Paper. URL:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1032961, accessed on: 13.06.2014.
Kochetova-Kozloski, Natalia/Messier, William F., Jr./Eilifsen, Aasmund (2011): Improving
Auditors’ Fraud Judgments Using a Frequency Response Mode. In: Contemporary Ac-
counting Research, 28 (3): 837-858.
Kohlberg, Lawrence (1958): The Development of Modes of Moral Thinking and Choice in
the Years 10 to 16. Chicago: University of Chicago.
Kohlberg, Lawrence (1984): Essays on Moral Development, Volume 2: Psychology of Moral
Development. San Francisco: Harper & Row.
Köhler, Annette G. et al. (2008): Audit Regulation in Germany: Improvements Driven by
Internationalization. In: Quick, R./Turley, S./Willekens, M. (eds.): Auditing, Trust and
Governance: Regulation in Europe. London/New York: Routledge: 111-143.
Koonce, Lisa/Anderson, Urton/Marchant, Garry (1995): Justification of Decisions in Audit-
ing. In: Journal of Accounting Research, 33 (2): 369-384.
Kornmeier, Martin (2007): Wissenschaftstheorie und wissenschaftliches Arbeiten – Eine Ein-
führung für Wirtschaftswissenschaftler. Heidelberg: Physica-Verlag.
Koslow, Scott (2000): Can the Truth Hurt? How Honest and Persuasive Advertising Can
Unintentially Lead to Increased Consumer Skepticism. In: Journal of Consumer Affairs,
34 (2): 245-268.
Kotchetova, Natalia/Salterio, Steven (2007): Judgment and Decision-Making Accounting
Research: A Quest to Improve the Production, Certification, and Use of Accounting
Information. In: Koehler, D. J./Harvey, N. (eds.): Blackwell Handbook of Judgment and
Decision Making. Malden/Oxford/Carlton: Blackwell Publishing: 547-566.
Kowalczyk, Tamara K./Wolfe, Christopher J. (1998): Anchoring Effects Associated with Rec-
ommendations from Expert Decision Aids: An Experimental Analysis In: Behavioral
Research in Accounting, 10 (Supplement): 147-169.
Koziol, Leonard F./Budding, Deborah Ely (2009): Subcortical Structures and Cognition:
Implications for Neuropsychological Assessment. New York: Springer.
KPMG (2011): Elevating Professional Judgment in Auditing and Accounting: The KPMG
Professional Judgment Framework. Delaware: KPMG LLP.
270
Kramer, Roderick M. (1999): Trus and Distrust in Organizations: Emerging Perspectives,
Enduring Questions. In: Annual Review of Psychology, 50: 569-598.
Krech, David/Crutchfield, Richard S. (1948): Theory and Problems of Social Psychology.
New York: McGraw-Hill.
Krishnamoorthy, Ganesh/Mock, Theodore J./Washington, Mary T. (1999): A Comparative
Evaluation of Belief Revision Models in Auditing. In: Auditing: A Journal of Practice &
Theory, 18 (2): 105-127.
Krommes, Werner (2011): Handbuch Jahresabschlussprüfung: Ziele Technik Nachweise -
Wegweiser zum sicheren Prüfungsurteil, 3rd edition. Wiesbaden: Gabler.
Krosnick, Jon A./Fabrigar, Leandre R. (1997): Designing Rating Scales for Effective Meas-
urement in Surveys. In: Lyberg, L. et al. (eds.): Survey Measurement and Process
Quality. New York: John Wiley & Sons: 141-164.
Kruglanski, Arie W. (1989): Lay Epistemics and Human Knowledge: Cognitive and Motiva-
tional Bases. New York/London: Plenum Press.
Kruglanski, Arie W. (1990): Motivations for Judging and Knowing: Implications for Causal
Attribution. In: Higgins, E. T./Sorrentino, R. M. (eds.): The Handbook of Motivation
and Cognition. Foundation of Social Behavior. New York: Guilford Press: 333-368.
Krull, George/Reckers, Philip M. J./Wong-On-Wing, Bernard (1993): The Effect of Experi-
ence, Fraudulent Signals and Information Presentation Order on Audtiors’ Beliefs. In:
Auditing: A Journal of Practice & Theory, 12 (2): 143-153.
Kühberger, Anton (1998): The Influence of Framing on Risky Decisions: A Meta-analysis. In:
Organizational Behavior and Human Decision Processes, 75 (1): 23-55.
Kung, Fan-Hua /Huang, Cheng Li (2013): Auditors’ Moral Philosophies and Ethical Beliefs.
In: Management Decision, 51 (3): 479-500.
Kurtz, Paul (1992): The New Skepticism: Inquiry and Reliable Knowledge. Buffalo: Prome-
theus Books.
Kurtz, Paul (2001): Skepticism and Humanism: The New Paradigm. New Brunswick/London:
Transaction Publishers.
Kutner, Michael H. et al. (2005): Applied Linear Statistical Models, 5th edition. Boston et al.:
McGraw-Hill.
Lagerlund, Henrik (2010): A History of Skepticism in the Middle Ages. In: Lagerlund, H.
(ed.): Rethinking the History of Skepticism: The Missing Medieval Background. Leiden:
Brill: 1-28.
Lagnado, David (2011): Thinking about Evidence. In: Dawid, P./Twining, W./Vasaliki, M.
(eds.): Evidence, Inference and Enquiry. Oxford et al.: Oxford University Press: 183-
223.
271
Lammenranta, Markus (2008): The Pyrrhonian Problematic. In: Greco, J. (ed.): The Oxford
Handbook of Skepticism. Oxford et al.: Oxford University Press: 9-33.
LaSalle, Randall E. (1997): Presentation Order Effects on Accounting Students’ Ethical
Judgments. In: Journal of Accounting Education, 15 (1): 19-38.
Latané, Bibb/Williams, Kipling/Harkins, Stephen (1979): Many Hands Make Light the Work:
The Causes and Consequences of Social Loafing. In: Journal of Personality and Social
Psychology, 37 (6): 823-832.
Latham, Claire Kamm/Linville, Mark (1998): A Review of the Literature in Audit Litigation.
In: Journal of Accounting Literature, 17: 175-213.
Laws, D. J./O’Hagan, A. (2002): A Hierarchical Bayes Model for Multilocation Auditing. In:
Journal of the Royal Statistical Society: Series D (The Statistician), 51 (4): 431-450.
Lee, Chih-Chen/Welker, Robert B./Wang, Te-Wei (2013): An Experimental Investigation of
Professional Skepticism in Audit Interviews. In: International Journal of Auditing, 17
(2): 213-226.
Lee, Chi-Wen Jevons/Gu, Zhaoyang (1998): Low Balling, Legal Liability and Auditor
Independence. In: The Accounting Review, 73 (4): 533-555.
Leigh, James H./Kinnear, Thomas C. (1980): On Interaction Classification. In: Educational
and Psychological Measurement, 40 (4): 841-843.
Lerner, Jennifer S./Tetlock, Philip E. (1999): Accounting for the Effect of Accountability. In:
Psychological Bulletin, 125 (2): 255-275.
Leung, Patrick W./Trotman, Ken T. (2005): The Effects of Feedback Type on Auditor
Judgment Performance for Configural and Non-Configural Tasks. In: Accounting, Orga-
nizations and Society, 30 (6): 537-553.
Levi, Isaac (1977): Subjunctives, Dispositions and Chances. In: Synthese, 34 (4): 423-455.
Levi, Isaac (1991): The Fixation of Belief and Its Undoing: Changing Beliefs Through In-
quiry. Cambridge/New York/Melbourne: Cambridge University Press.
Lewicki, Roy J./Bunker, Barbara Benedict (1996): Developing and Maintaining Trust in Work
Relationships. In: Kramer, R. M./Tyler, T. R. (eds.): Trust in Organizations: Frontiers of
Theory and Research. Thousand Oaks: Sage Publications: 114-139.
Li, Chan (2009): Does Client Importance Affect Auditor Independence at the Office Level?
Empirical Evidence from Going-Concern Opinions. In: Contemporary Accounting
Research, 26 (1): 201-230.
Libby, Robert (1981): Accounting and Human Information Processing: Theory and Applica-
tions. Englewood Cliffs: Prentice-Hall.
Libby, Robert (1985): Availability and the Generation of Hypotheses in Analytical Review.
In: Journal of Accounting Research, 23 (2): 646-665.
272
Libby, Robert (1991): Experimental Research and the Distinctive Features of Accounting
Settings. In: Ponemon, L. A./Gabhart, D. R. L. (eds.): Auditing: Advances in Behavioral
Research. New York et al.: Springer-Verlag: 1-20.
Libby, Robert/Frederick, David M. (1990): Experience and the Ability to Explain Audit
Findings. In: Journal of Accounting Research, 28 (2): 348-367.
Libby, Robert/Kinney, William R., Jr. (2000): Does Mandated Audit Communication Reduce
Opportunistic Corrections to Manage Earnings to Forecasts? In: The Accounting Re-
view, 75 (4): 383-404.
Libby, Robert/Lewis, Barry L. (1977): Human Information Processing Research in Account-
ing: The State of the Art. In: Accounting, Organizations and Society, 2 (3): 245-268.
Libby, Robert/Lewis, Barry L. (1982): Human Information Processing Research in Account-
ing: The State of the Art in 1982. In: Accounting, Organizations and Society, 7 (3): 231-
285.
Libby, Robert/Libby, Patricia A. (1989): Expert Measurement and Mechanical Combination
in Control Reliance Decisions. In: The Accounting Review, 64 (4): 729-747.
Libby, Robert/Lipe, Marlys Gascho (1992): Incentives, Effort, and the Cognitive Processes
Involved in Accounting-Related Judgments. In: Journal of Accounting Research, 30 (2):
249-273.
Libby, Robert/Luft, Joan (1993): Determinants of Judgment Performance in Accounting
Settings: Ability, Knowledge, Motivation, and Environment. In: Accounting,
Organizations and Society, 18 (5): 425-450.
Libby, Robert/Tan, Hun-Tong (1994): Modeling the Determinants of Audit Expertise. In:
Accounting, Organizations and Society, 19 (8): 701-716.
Libby, Robert/Bloomfield, Robert/Nelson, Mark W. (2002): Experimental Research in Finan-
cial Accounting. In: Accounting, Organizations and Society, 27: 775-810.
Lightner, Sharon M./Adams, Steven J./Lightner, Kevin M. (1982): The Influence of Situa-
tional, Ethical, and Expectancy Theory Variables on Accountants’ Underreporting
Behavior. In: Auditing: A Journal of Practice & Theory, 2 (1-12): 1.
Lightner, Sharon M./Leisenring, James J./Winters, Alan J. (1983): Underreporting Chargeable
Time. In: Journal of Accountancy, 155 (1): 52-57.
Lim, Chee-Yeow/Tan, Hun-Tong (2010): Does Auditor Tenure Improve Audit Quality?
Moderating Effects of Industry Specialization and Fee Dependence. In: Contemporary
Accounting Research, 27 (3): 923-957.
Lipe, Marlys Gascho (2008): Discussion of “Judging Audit Quality in Light of Adverse
Outcomes: Evidence of Outcome Bias and Reverse Outcome Bias”. In: Contemporary
Accounting Research, 25 (1): 275-282.
273
Litman, Jordan A. (2005): Curiosity and the Pleasures of Learning: Wanting and Liking New
Information. In: Cognition & Emotion, 19 (6): 793-814.
Litman, Jordan A./Silvia, Paul J. (2006): The Latent Structure of Trait Curiosity: Evidence for
Interest and Deprivation Curiosity Dimensions. In: Journal of Personality Assessment,
86 (3): 318-328.
Litwin, Mark S. (1995): How to Measure Survey Reliability and Validity. Thousand Oaks:
Sage Publications.
Livesley, W. John/Jang, Kerry L./Vernon, Philip A. (2003): Genetic Basis of Personality
Structure. In: Millon, T./Lerner, M. J. (eds.): Handbook of Psychology, Volume 5:
Personality and Social Psychology. Hoboken: John Wiley & Sons: 59-83.
Lix, Lisa M./Keselman, Joanne C./Keselman, H. J. (1996): Consequences of Assumption
Violations Revisited: A Quantitative Review of Alternatives to the One-Way Analysis of
Variance “F” Test. In: Review of Educational Research, 66 (4): 579-619.
Loebbecke, James K./Eining, Martha M./Willingham, John J. (1989): Auditors’ Experience
with Material Irregularities: Frequency, Nature, and Detectability. In: Auditing: A
Journal of Practice & Theory, 9 (1): 1-28.
Loewenstein, George (1999): Experimental Economics from the Vantage Point of Behav-
ioural Economics. In: The Economic Journal, 109: F25-F34.
Loftus, Geoffrey R./Loftus, Elizabeth F. (1976): Human Memory: The Processing of Informa-
tion. Hillsdale: Lawrence Erlbaum Associates.
Lom, Petr (2001): The Limits of Doubt: The Moral and Political Implications of Skepticism.
Albany: State University of New York.
Lord, Alan T. (1992): Pressure: A Methodological Consideration for Behavioral Research in
Auditing. In: Auditing: A Journal of Practice & Theory, 11 (2): 89-108.
Lord, Alan T./DeZoort, F. Todd (2001): The Impact of Commitment and Moral Reasoning on
Auditors’ Responses to Social Influence Pressure. In: Accounting, Organizations and
Society, 26 (3): 215-235.
Love, Vincent J./Lawson, Celia (2009): Auditing in Turbulent Economic Times: Plan Ahead
and Apply Skepticism and Judgment. In: The CPA Journal, (5): 30-34.
Low, Kin-Yew (2004): The Effects of Industry Specialization on Audit Risk Assessments and
Audit-Planning Decisions. In: The Accounting Review, 79 (1): 201-219.
Luckett, Peter F./Eggleton, Ian R. (1991): Feedback and Management Accounting: A Review
of Research into Behavioural Consequences. In: Accounting, Organizations and Society,
16 (4): 371-394.
Lundholm, Helge (1936): The Psychology of Belief. Durham: Duke University Press.
MacDonald, A. P. J. (1970): Revised Scale for Ambiguity Tolerance: Reliability and Validity
In: Psychological Reports, 26: 791–798.
274
Magee, Robert P./Tseng, Mei-Chiun (1990): Audit Pricing and Independence. In: The
Accounting Review, 65 (2): 315-336.
Mahajan, Jayashree (1992): The Overconfidence Effect in Marketing Management Predic-
tions. In: Journal of Marketing Research, 29: 329-342.
Majid, Abdul/Pragasam, John (1997): Interactions of Intolerance of Ambiguity and of
Contingent Liability on Auditors’ Avoidance of Litigation. In: Psychological Reports, 81
(3): 935-944.
Majors, Tracie M./Shefchik, Lori B./Vitalis, Adam (2014): The Interactive Effect of Ego
Depletion and Trait Skepticism on Auditor Performance. Working Paper. URL: http://pa
pers.ssrn.com/sol3/papers.cfm?abstract_id=2408227, accessed on: 29.04.2014.
Makkawi, Bilal/Abdolmohammadi, Mohammad J. (2004): Determinants of the Timing of In-
terim and Majority of Audit Work. In: International Journal of Auditing, 8 (2): 139-151.
Makkawi, Bilal/Rutledge, Robert W. (2000): Evaluating Audit Risk: The Effects of Industry
Characteristics, Experience and Tolerance-for-Ambiguity. In: Advances in Behavioral
Accounting Research, 3: 63–82.
Mangleburg, Tamara F./Bristol, Terry (1998): Socialization and Adolescents’ Skepticism
toward Advertising. In: Journal of Advertising, 27 (3): 11-21.
Manis, Melvin (1967): Context Effects in Communication. In: Journal of Personality and
Social Psychology, 5 (3): 326–334.
Manry, David L./Mock, Theodore J./Turner, Jerry L. (2008): Does Increased Audit Partner
Tenure Reduce Audit Quality? In: Journal of Accounting, Auditing & Finance, 23 (4):
553-572.
March, James G./Simon, Herbert A. (1958): Organizations. New York/London/Sydney: John
Wiley & Sons.
Margheim, Loren/Kelley, Tim/Pattison, Diane (2005): An Empirical Analysis Of The Effects
Of Auditor Time Budget Pressure And Time Deadline Pressure. In: The Journal of Ap-
plied Business Research, 21 (1): 21-35.
Marques, Pedro Augusto/Azevedo-Pereira, José (2009): Ethical Ideology and Ethical Judg-
ments in the Portuguese Accounting Profession. In: Journal of Business Ethics, 86 (2):
227-242.
Marris, Robin (1992): Implications for Economics. In: Simon, H. A. et al. (eds.): Economics,
Bounded Rationality and the Cognitive Revolution. Aldershot/Brookfield: Edward
Elger: 194-223.
Marschark, Marc (1997): Psychological Development of Deaf Children. Oxford et al.: Oxford
University Press.
275
Marten, Kai-Uwe/Quick, Reiner/Ruhnke, Klaus (2011): Wirtschaftsprüfung: Grundlagen des
betriebswirtschaftlichen Prüfungswesens nach nationalen und internationalen Normen,
4th edition. Stuttgart: Schäffer-Poeschel.
Mason, Alister K./Gibbins, Michael (1991): Judgment and U.S. Accounting Standards. In:
Accounting Horizons, 5 (2): 14-24.
Massey, Dawn W./Thorne, Linda (2006): The Impact of Task Information Feedback on
Ethical Reasoning. In: Behavioral Research in Accounting, 18: 103-116.
Matthews, Gerald/Deary, Ian J./Whiteman, Martha C. (2003): Personality Traits, 2nd edition.
Cambridge et al.: Cambridge University Press.
Maurer, Todd J./Alexander, Ralph A. (1991): Contrast Effects in Behavioral Measurement:
An Investigation of Alternative Process Explanations. In: Journal of Applied Psycholo-
gy, 76 (1): 3-10.
Mautz, Robert Kuhn/Sharaf, Hussein Amer (1961): The Philosophy of Auditing. Sarasota:
American Accounting Association.
May, Mike (2009): Sensation and Perception. New York: Infobase Publishing.
Mayhew, Brian W./Schatzberg, Jeffrey W./Sevcik, Galen R. (2001): The Effect of Accounting
Uncertainty and Auditor Reputation on Auditor Objectivity. In: Auditing: A Journal of
Practice & Theory, 20 (2): 49-70.
Mayper, Alan G./Anderson, Urton/Kilpatrick, Bob G. (1999): Tax Workpaper Review: The
Impact of Staff Conclusions and Argument Presentation. In: Advances in Taxation, 11:
145-167.
McCracken, Susan A. (2003): Auditors’ Strategies to Protect Their Litigation Reputation: A
Research Note. In: Auditing: A Journal of Practice & Theory, 22 (1): 165-179.
McCrae, Robert R./Costa, Paul T., Jr. (1996): Toward a New Generation of Personality
Theories: Theoretical Contexts for the Five-Factor Model. In: Wiggins, J. S. (ed.): The
Five-Factor Model of Personality: Theoretical Perspectives. New York/London: The
Guilford Press: 51-87.
McCrae, Robert R./John, Oliver P. (1992): An Introduction to the Five-Factor Model and Its
Applications. In: Journal of Personality, 60: 175-215.
McDaniel, Linda S. (1990): The Effects of Time Pressure and Audit Program Structure on
Audit Performance. In: Journal of Accounting Research, 28 (2): 267-285.
McGhee, Walter/Shields, Michael D./Birnberg, Jacob G. (1978): The Effects of Personality
on a Subject’s Information Processing. In: The Accounting Review, 53 (3): 681-697.
McGinn, Marie (1989): Sense and Certainty: A Dissolution of Scepticism. Oxford et al.: Basil
Blackwell.
McKenna, Eugene F. (2000): Business Psychology and Organisational Behaviour: A Stu-
dent’s Handbook, 3rd edition. Hove: Psychology Press.
276
McKinnon, Jill L. (1984): Cultural Constraints on Audit Independence in Japan. In: The
International Journal of Accounting, 19: 17–44.
McManus, Lisa/Subramaniam, Nava (2009): Ethical Evaluations and Behavioural Intentions
of Early Career Accountants: The Impact of Mentors, Peers and Individual Attributes.
In: Accounting & Finance, 49 (3): 619-643.
McMillan, Jeffrey J./White, Richard A. (1993): Auditors’ Belief Revisions and Evidence
Search: The Effect of Hypothesis Frame, Confirmation Bias, and Professional Skepti-
cism. In: The Accounting Review, 68 (3): 443-465.
Meeden, Glen/Sargent, Dawn (2007): Some Bayesian Methods for Two Auditing Problems.
In: Communications in Statistics: Theory & Methods, 36 (15): 2727-2740.
Meidan, Abraham (2004): Skepticism Is True. Boca Raton: Universal Publishers.
Melnick, Edward L./Everitt, Brian S. (eds.) (2008): Encyclopedia of Quantitative Risk
Analysis and Assessment, Volume 1 A-C. Chichester: John Wiley & Sons.
Messick, Samuel (1976): Personality Consistencies in Cognition and Creativity. In: Messick,
S. et al. (eds.): Individuality in Learning: Implications of Cognitive Style and Creativity
for Human Development. San Francisco/Washington/London: Jossey-Bass.
Messier, William F. (1992): The Sequencing of Audit Evidence: Its Impact on the Extent of
Audit Testing and Report Formulation. In: Accounting and Business Research, 22 (86):
143-150.
Messier, William F./Quilliam, William C. (1992): The Effect of Accountability on Judgment:
Development of Hypotheses for Auditing. In: Auditing: A Journal of Practice & Theory,
11: 123-138.
Messier, William F./Tubbs, Richard M. (1994): Recency Effects in Belief Revision: The Im-
pact of Audit Experience and the Review Process. In: Auditing: A Journal of Practice &
Theory, 13 (1): 57-72.
Messier, William F./Kozloski, Thomas M./Kochetova-Kozloski, Natalia (2010): An Analysis
of SEC and PCAOB Enforcement Actions against Engagement Quality Reviewers. In:
Auditing: A Journal of Practice & Theory, 29 (2): 233-252.
Miller, George A. (2003): The Cognitive Revolution: A Historical Perspective. In: Trends in
Cognitive Sciences, 7 (3): 141-144.
Mischel, Walter (1968): Personality and Assessment. New York: John Wiley & Sons.
Mock, Theodore J. et al. (1997): Auditors’ Uncertainty Representation and Evidence Aggre-
gation. In: Behavioral Research in Accounting, 9 (Supplement): 123-147.
Moeckel, Cindy (1990): The Effect of Experience on Auditors’ Memory Errors. In: Journal of
Accounting Research, 28 (2): 368-387.
Moeckel, Cindy/Plumlee, R. David (1989): Auditors’ Confidence in Recognition of Audit
Evidence. In: The Accounting Review, 64 (4): 653-666.
277
Möller, Manuela/Sigillo, Francesco (2010): Der Enron-Andersen-Skandal und dessen Ein-
fluss auf das Reputationskapital der Institution Wirtschaftsprüfung: Eine empirische Stu-
die zu Kursreaktionen als Folge des Enron-Andersen-Skandals am Schweizer Kapital-
markt. In: Gesellschafts- und Kapitalmarktrecht: 1-31.
Monroe, Gary S./Ng, Juliana (2000): An Examination of Order Effects in Auditors’ Inherent
Risk Assessments. In: Accounting and Finance, 40: 153-168.
Montgomery, Daniel D. et al. (2002): Auditors’ New Procedures for Detecting Fraud. In:
Journal of Accountancy, 193 (5): 63-66.
Mook, Douglas G. (1983): In Defense of External Invalidity. In: American Psychologist, 38:
379-387.
Moore, Don A. et al. (2006): Conflicts of Interest and the Case of Auditor Independence:
Moral Seduction and Strategic Issue Cycling. In: Academy of Management Review, 31
(1): 10-29.
Moors, Agnes/De Houwer, Jan (2007): What is Automaticity? An Analysis of Its Component
Features and Their Interrelations. In: Bargh, J. A. (ed.): Social Psychology and the Un-
conscious: The Automaticity of Higher Mental Processes. New York/Hove: Psychology
Press: 11-50.
Morales-Sánchez, Rafael/Cabello-Medina, Carmen (2013): The Role of Four Universal Moral
Competencies in Ethical Decision-Making. In: Journal of Business Ethics, 116 (4): 717-
734.
Morrow, Patricia Bockelman/Fiore, Stephen M. (2013): Team Cognition: Coordination across
Individuals and Machines. In: Lee, J. D./Kirlik, A./Dainoff, J. (eds.): The Oxford Hand-
book of Cognitive Engineering. Oxford et al.: Oxford University Press: 200-215.
Morton, Jane E. (2001): Confidence as a Determinant of the Direction of Order Effects:
Theory and a Test in an Internal Control Judgment Setting. In: Academy of Accounting
and Financial Studies Journal, 5 (1): 95-123.
Moser, Paul K. (2008): Religious Skepticism. In: Greco, J. (ed.): The Oxford Handbook of
Skepticism. Oxford et al.: Oxford University Press: 200-224.
Moskowitz, Gordon B./Skurnik, Ian/Galinsky, Adam D. (1999): The History of Dual-Process
Notions and the Future of Preconscious Control. In: Chaiken, S./Trope, Y. (eds.): Dual-
Process Theories in Social Psychology. New York: Guilford Press: 12-36.
Mueller, Jennifer M./Anderson, John C. (2002): Decision Aids for Generating Analytical Re-
view Alternatives: The Impact of Goal Framing and Audit-Risk Level. In: Behavioral
Research in Accounting, 14: 157-177.
Mumpower, Jeryl L./Stewart, Thomas R. (1996): Expert Judgement and Expert Disagreement.
In: Thinking & Reasoning, 2 (2/3): 191-212.
Murray, Zowie (2012): Professional Scepticism. In: Chartered Accountants Journal, 91 (11):
36-38.
278
Mutchler, Jane F. (1986): Empirical Evidence Regarding the Auditor’s Going-Concern Opin-
ion Decision. In: Auditing: A Journal of Practice & Theory, 6 (1): 148-163.
Myers, James N./Myers, Linda A./Omer, Thomas C. (2003): Exploring the Term of the
Auditor-Client Relationship and the Quality of Earnings: A Case for Mandatory Auditor
Rotation? In: The Accounting Review, 78 (3): 779-799.
Myers, Raymond H. (1990): Classical and Modern Regression with Applications, 2nd edition.
Belmont: Thomson.
Naess, Arne (1968): Scepticism. London: Routledge & Kegan Paul.
Nasution, Damai/Östermark, Ralf (2012): The Impact of Social Pressures, Locus of Control,
and Professional Commitment on Auditors’ Judgment: Indonesian Evidence. In: Asian
Review of Accounting, 20 (2): 163-178.
Naylor, F. D. (1981): A State-Trait Curiosity Inventory. In: Australian Psychologist, 16 (2):
172-183.
Neisser, Ulric (1967): Cognitive Psychology. New York: Appleton-Century-Corfts.
Neisser, Ulric (1976): Cognition and Reality: Principles and Implications of Cognitive Psy-
chology. New York: W.H. Freeman and Company.
Nelson, Mark W. (2009): A Model and Literature Review of Professional Skepticism in
Auditing. In: Auditing: A Journal of Practice & Theory, 28 (2): 1-34.
Nelson, Mark W./Kinney, William R., Jr. (1997): The Effect of Ambiguity on Loss Contin-
gency Reporting Judgements. In: The Accounting Review, 72 (2): 257-274.
Nelson, Mark W./Tan, Hun-Tong (2005): Judgment and Decision Making Research in
Auditing: A Task, Person, and Interpersonal Interaction Perspective. In: Auditing: A
Journal of Practice & Theory, 24 (Supplement): 41-71.
Nelson, Mark W./Elliott, John A./TarpIey, Robin L. (2002): Evidence from Auditors about
Managers’ and Auditors’ Earnings Management Decisions. In: The Accounting Review,
77 (4): 175-202.
Nichols, Donald R./Price, Kenneth H. (1976): The Auditor-Firm Conflict: An Analysis Using
Concepts of Exchange Theory. In: The Accounting Review, 51 (2): 335-346.
Nisbett, Richard/Ross, Lee (1980): Human Inference: Strategies and Shortcomings of Social
Judgment. Engelwood Cliffs: Prentice-Hall.
Nisbett, Richard E./Zukier, Henry/Lemley, Ronald E. (1981): The Dilution Effect: Nondiag-
nostic Information Weakens the Implications of Diagnostic Information. In: Cognitive
Psychology, 13 (2): 248-277.
Nolder, Christine J. (2012): The Role of Professional Skepticism, Attitudes and Emotions on
Auditor’s Judgments. Dissertation, Bentley University. Ann Arbor: ProQuest.
279
Northrop, Alona/Arsneault, Shelly (2007): Handbook of Research Methods in Public Admin-
istration. In: Yang, K./Miller, G. J. (eds.): Public Administration and Public Policy, 2nd
edition. Boca Raton: Taylor & Francis: 213-239.
Norton, Robert W. (1975): Measurement of Ambiguity Tolerance. In: Journal of Personality
Assessment, 39 (6): 607-619.
Noyes, Jan/Garland, Kate/Bruneau, Daniel (2004): Humans: Skills, Capabilities and Limita-
tions. In: Sandom, C./Harvey, R. S. (eds.): Human Factors for Engineers. Sevenage:
Institution of Engineering and Technology: 35-56.
NZAuASB (2013): Explanatory Guide Au7: References to Practical Guidance on Professional
Scepticism in an Audit of Financial Statements. URL: http://www.xrb.govt.nz/include
s/download.aspx?ID=126424, accessed on: 17.10.2013.
O’Bryan, David/Quirin, Jeffrey J./Donnelly, David P. (2005): Locus Of Control And Dys-
functional Audit Behavior. In: Journal of Business & Economics Research, 3 (10): 9-14.
O’Clock, Priscilla/Devine, Kevin (1995): An Investigation of Framing and Firm Size on the
Auditor’s Going Concern Decision. In: Accounting and Business Research, 25 (99):
197-207.
O’Donnell, Ed/Prather-Kinsey, Jenice (2010): Nationality and Differences in Auditor Risk
Assessment: A Research Note with Experimental Evidence. In: Accounting, Organiza-
tions and Society, 35: 558-564.
O’Malley, Shaun F. (2000): The Panel of Audit Effectiveness: Report and Recommendations.
Stamford: Public Oversight Board.
Obermiller, Carl/Spangenberg, Eric R. (1998): Development of a Scale to Measure Consumer
Skepticism Toward Advertising. In: Journal of Consumer Psychology, 7 (2): 159-186.
Obermiller, Carl/Spangenberg, Eric R. (2000): On the Origin and Distinctness of Skepticism
Toward Advertising. In: Marketing Letters, 11 (4): 311.
OECD (2013): Women and Financial Literacy: OECD/INFE Evidence, Survey and Policy
Responses. URL: http://www.oecd.org/daf/fin/financial-education/TrustFund2013_OEC
D_INFE_Women_and_Fin_Lit.pdf, accessed on: 13.05.2014.
Oswald, Margit E./Grosjean, Stefan (2004): Confirmation Bias. In: Pohl, R. F. (ed.): Cogni-
tive Illusions: A Handbook on Fallacies and Biases in Thinking, Judgment and Memory.
Hove/New York: Psychology Press: 79-96.
Over, David (2007): Rationality and the Normative/Descriptive Distinction. In: Koehler, D.
J./Harvey, N. (eds.): Blackwell Handbook of Judgment and Decision Making.
Malden/Oxford/Carlton: Blackwell Publishing: 3-18.
Owhoso, Vincent/Weickgenannt, Andrea (2009): Auditors’ Self-Perceived Abilities in Con-
ducting Domain Audits. In: Critical Perspectives on Accounting, 20 (1): 3-21.
280
Pallant, Julie (2010): SPSS Survival Manual: A Step by Step Guide to Data Analysis Using
SPSS, 4th edition. Maidenhead: McGraw-Hill.
Palmrose, Zoe-Vonna (1988): An Analysis of Auditor Litigation and Audit Service Quality.
In: The Accounting Review, 63 (1): 55-73.
Parret, Herman (1983): Introduction: Beliefs and Believing: The Web and the Spinning. In:
Parret, H. (ed.): On Believing: Epistemological and Semiotic Approaches. Berlin/New
York: de Gruyter: 1-15.
Passer, Michael W./Smith, Ronald Edward (2010): Psychology: The Science of Mind and
Behavior. Boston et al.: McGraw-Hill.
Patel, Chris (2003): Some Cross-Cultural Evidence on Whistle-Blowing as an Internal Con-
trol Mechanism. In: Journal of International Accounting Research, 2: 69-96.
Patel, Chris/Psaros, Jim (2000): Perceptions of External Auditors’ Independence: Some
Cross-Cultural Evidence. In: The British Accounting Review, 32 (3): 311-338.
Patel, Chris/Harrison, Graeme L./McKinnon, Jill L. (2002): Cultural Influences on Judg-
ments of Professional Accountants in Auditor–Client Conflict Resolution. In: Journal of
International Financial Management & Accounting, 13 (1): 1-31.
Patterson, Denise M. (2001): Causal Effects of Regulatory, Organizational and Personal
Factors on Ethical Sensitivity. In: Journal of Business Ethics, 30 (2): 123-159.
Payne, Elizabeth A./Ramsay, Robert J. (2005): Fraud Risk Assessments and Auditors’ Profes-
sional Skepticism. In: Managerial Auditing Journal, 20 (3): 321-330.
Payne, John W. (1976): Task Complexity and Contingent Processing in Decision Making: An
Information Search and Protocol Analysis. In: Organizational Behavior and Human
Performance, 16 (2): 366-387.
Payne, John W. (1982): Contingent Decision Behavior. In: Psychological Bulletin, 92 (2):
382-402.
Payne, John W./Bettman, James R./Johnson, Eric J. (1993): The Adaptive Decision Maker.
Cambridge: Cambridge University Press.
PCAOB (2012): Maintaining and Applying Professional Skepticism in Audits. URL: http://
pcaobus.org/Standards/QandA/12-04-2012_SAPA_10.pdf, accessed on: 07.10.2013.
Peecher, Mark E. (1996): The Influence of Auditors’ Justification Processes on Their Deci-
sions: A Cognitive Model and Experimental Evidence. In: Journal of Accounting Re-
search, 34 (1): 125-140.
Peecher, Mark E./Solomon, Ira (2001): Theory and Experimentation in Studies of Audit Judg-
ments and Decisions: Avoiding Common Research Traps. In: International Journal of
Auditing, 5: 193-203.
281
Peecher, Mark E./Solomon, Ira/Trotman, Ken T. (2013): An Accountability Framework for
Financial Statement Auditors and Related Research Questions. In: Accounting, Organi-
zations and Society, 38: 596-620.
Peecher, Mark E. et al. (2010): The Effects of a Supervisor’s Active Intervention in Subordi-
nates’ Judgements, Directional Goals, and Perceived Technical Knowledge Advantage
on Audit Team Judgements. In: The Accounting Review, 85 (5): 1763-1786.
Pei, Buck K. W./Reckers, Philip M. J./Wyndelts, Robert W. (1990): The Influence of Informa-
tion Presentation Order on Professional Tax Judgment. In: Journal of Economic Psychol-
ogy, 11: 119-146.
Pei, Buck K. W./Reckers, Philip M. J./Wyndelts, Robert W. (1992): Tax Professionals Belief
Revision: The Effects of Information Presentation Sequence, Client Preference, and
Domain Experience. In: Decision Sciences, 23: 175-199.
Pei, Buck K. W./Reed, Sarah A./Koch, Bruce S. (1992): Auditor Belief Revisions in a Perfor-
mance Auditing Setting: An Application of the Belief-Adjustment Model. In: Account-
ing, Organizations and Society, 17 (2): 169-183.
Pennington, Nancy/Hastie, Reid (1986): Evidence Evaluation in Complex Decision Making.
In: Journal of Personality and Social Psychology, 51 (2): 242-258.
Perera, M. H. B. (1989): Towards a Framework to Analyze the Impact of Culture on Ac-
counting. In: International Journal of Accounting, 24: 42-56.
Peterson, John (2008): Aquinas: A New Introduction. Lanham et al.: University Press of
America.
Peytcheva, Marietta (2014): Professional Skepticism and Auditor Cognitive Performance in a
Hypothesis-Testing Task. In: Managerial Auditing Journal, 29 (1): 27-49.
Pincus, Karen V. (1990): Auditor Individual Differences and Fairness of Presentation Judg-
ments. In: Auditing: A Journal of Practice & Theory, 9 (3): 150-166.
Pincus, Karen V. (1991): Audit Judgment Confidence. In: Behavioral Research in Account-
ing, 3: 39-65.
Pinker, Steven (1994): The Language Instinct. New York: W. Morrow and Co.
Pinker, Steven (1997): How the Mind Works. New York: W.W. Norton.
Pinsker, Robert (2007): Long Series of Information and Nonprofessional Investors’ Belief
Revision. In: Behavioral Research in Accounting, 19: 197-214.
Pinsker, Robert (2011): Primacy or Recency? A Study of Order Effects When Nonprofes-
sional Investors are Provided a Long Series of Disclosures. In: Behavioral Research in
Accounting, 23 (1): 161-183.
Pinsker, Robert/Pennington, Robin/Schafer, Jennifer Kahle (2009): The Influence of Roles,
Advocacy, and Adaption to the Accounting Decision Environment. In: Behavioral Re-
search in Accounting, 21 (2): 91-111.
282
Plotnik, Rod/Kouyoumdjian, Haig (2010): Introduction to Psychology, 9th edition. Belmont:
Wadsworth, Cengage Learning.
Plotnik, Rod/Kouyoumdjian, Haig (2013): Introduction to Psychology, 10th edition. Belmont:
Cengage Learning.
Plous, Scott (1993): The Psychology of Judgment and Decision Making. New York:
McGraw-Hill.
Plumlee, David/Rixom, Brett A./Rosman, Andrew J. (2012): Training Auditors to Think
Skeptically. Working Paper. URL: http://web.ku.edu/~audsymp/myssi/_pdf/Plumlee%
20et%20al.%202012%20Training%20Auditors%20to%20Think%20Skeptically%20-
%20April%202012%20KU%20Symposium.pdf, accessed on: 06.11.2013.
Pohl, Rüdiger F. (2004): Hindsight Bias. In: Pohl, R. F. (ed.): Cognitive Illusions: A
Handbook on Fallacies and Biases in Thinking, Judgment and Memory. Hove/New
York: Psychology Press: 363-378.
Pole, Jason D./Bondy, Susan J. (2010): Normality Assumption. In: Salkind, N. J. (ed.): Ency-
clopedia of Research Design, Volume 2. Thousand Oaks et al.: Sage Publications: 932-
935.
Ponemon, Lawrence A. (1990): Ethical Judgments in Accounting: A Cognitive-Developmen-
tal Perspective. In: Critical Perspectives on Accounting, 1 (2): 191-215.
Ponemon, Lawrence A. (1992a): Auditor Underreporting of Time and Moral Reasoning: An
Experimental Lab Study. In: Contemporary Accounting Research, 9 (1): 171-189.
Ponemon, Lawrence A. (1992b): Ethical Reasoning and Selection-Socialization in Account-
ing. In: Accounting, Organizations and Society, 17 (3/4): 239-258.
Ponemon, Lawrence A. (1993): The Influence of Ethical Reasoning on Auditors’ Perceptions
of Management Competence and Integrity. In: Advances in Accounting, 11: 1-29.
Ponemon, Lawrence A./Gabhart, David R. L. (1990): Auditor Independence Judgments: A
Cognitive-Developmental Model and Experimental Evidence. In: Contemporary Ac-
counting Research, 7 (1): 227-251.
Ponemon, Lawrence A./Gabhart, David R. L. (1994): Ethical Reasoning Research in the
Accounting and Auditing Professions. In: Rest, J. R./Narváez, D. (eds.): Moral
Development in the Professions: Psychology and Applied Ethics. New Jersey: Hillsdale:
101-119.
Popkin, Richard H. (1979): The History of Scepticism from Erasmus to Spinoza. Berkeley:
University of California Press.
Popkin, Richard H./Stroll, Avrum (2002): Skeptical Philosophy for Everyone. New York:
Prometheus Books.
Popova, Velina (2013): Exploration of Skepticism, Client-Specific Experiences, and Audit
Judgments. In: Managerial Auditing Journal, 28 (2): 140-160.
283
Pound, Roscoe (1953): The Lawyer from Antiquity to Modern Times: With Particular Refer-
ence to the Development of Bar Associations in the United States. St. Paul: West
Publishing Company.
Pratt, Jamie (1980): The Effects of Personality on a Subject’s Information Processing: A
Comment. In: The Accounting Review, 55 (3): 501-506.
Preston, Malcolm G. (1936a): Contrast Effects and the Psychometric Functions. In: The
American Journal of Psychology, 48 (4): 625-631.
Preston, Malcolm G. (1936b): Contrast Effects and the Psychophysical Judgments. In: The
American Journal of Psychology, 48 (3): 389-402.
Prevignano, Carlo/Thibault, Paul J. (2003): Continuing the Discussion with John J.
Gumperz. In: Eerdmans, S./Prevignano, C./Thibault, P. J. (eds.): Language and Interac-
tion: Discussions with John J. Gumperz. Amsterdam/Philadelphia: John Benjamins: 149-
162.
Pritchard, Robert D. et al. (2013): Long-Term Effects of Goal Setting on Performance with
the Productivity Measurement and Enhancement System (ProMES). In: Locke, E.
A./Latham, G. P. (eds.): New Developments in Goal Setting and Task Performance.
New York/Hove: Routledge: 233-245.
Qandil, Johanna Souad (2014): Wahrnehmung der Qualität der Abschlussprüfung: Eine
theoretische und empirische Analyse für den deutschen Kapitalmarkt. Wiesbaden:
Springer Gabler.
Quadackers, Luc (2009): A Study of Auditors’ Skeptical Characteristics and Their
Relationship to Skeptical Judgments and Decisions. Dissertation, Vrije Universiteit
Amsterdam. URL: http://dare.ubvu.vu.nl/bitstream/handle/1871/13280/8931.pdf;jsession
id=D89295F46D9861C1DED98CDF637A2EE4?sequence=5, accessed on: 15.07.2013.
Quadackers, Luc/Groot, Tom/Wright, Arnold (2012): Auditors’ Professional Skepticism:
Neutrality versus Presumptive Doubt. Working Paper. URL: http://papers.ssrn.com/sol3/
papers.cfm?abstract_id=2162947, accessed on: 30.05.2013.
Quick, Reiner/Warming-Rasmussen, Bent (2007): Unabhängigkeit des Abschlussprüfers –
Zum Einfluss von Beratungsleistungen auf Unabhängigkeitswahrnehmungen von
Aktionären. In: Zeitschrift für Betriebswirtschaft, 77 (10): 1007–1033.
Quine, Williard van Orman /Ullian, Joseph S. (1978): The Web of Belief, 2nd edition. New
York: Random House.
Radtke, Robin R. (2000): The Effects of Gender and Setting on Accountants’ Ethically
Sensitive Decisions. In: Journal of Business Ethics, 24 (4): 299-312.
Raghunathan, Bhanu (1991): Premature Signing-Off of Audit Procedures: An Analysis. In:
Accounting Horizons, 5 (2): 71-79.
Randall, Donna M./Fernandes, Maria F. (1991): The Social Desirability Response Bias in
Ethics Research. In: Journal of Business Ethics, 10 (11): 805-817.
284
Rasso, Jason (2013): The Influence of Mindsets on Professional Skepticism in Auditors.
Working Paper. URL: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2357676,
accessed on 04.12.2013.
Ratzinger-Sakel, Nicole V. S. (2013): Auditor Fees and Auditor Independence: Evidence from
Going Concern Reporting Decisions in Germany. In: Auditing: A Journal of Practice &
Theory, 32 (4): 129-168.
Razumnikova, Olga M. (2012): Divergent Thinking and Learning. In: Seel, N. M. (ed.):
Encyclopedia of the Sciences of Learning. New York et al.: Springer: 1028-1031.
Reckers, Philip M. J./Schultz, Joseph J. (1993): The Effects of Fraud Signals, Evidence Order,
and Group-Assisted Counsel on Independent Auditor Judgment. In: Behavioral Research
in Accounting, 5: 124-143.
Reheul, Anne-Mie et al. (2013): The Relationship Between Personal Auditor Characteristics
and Audit Opinions. Working Paper presented at the 7th EARNet Symposium in Trier,
Germany.
Reidy, Mari/Theobald, Jonathan (2011): Financial Reporting Fraud: Prevention Starts at the
Top. URL: http://www.financialexecutives.org/KenticoCMS/getattachment/News---Pub
lications/news/FEI-Fraud-Literacy-Quiz/FEI-Mag-Article-on-Fraud---Nov--2011.pdf.as
px, accessed on: 16.08.2013.
Rempel, John K. /Holmes, John G./Zanna, Mark P. (1985): Trust in Close Relationships. In:
Journal of Personality and Social Psychology, 49 (1): 95-112.
Rennie, Morina D./Kopp, Lori S./Lemon, W. Morley (2010): Exploring Trust and the Auditor-
Client Relationship: Factors Influencing the Auditor’s Trust of a Client Representative.
In: Auditing: A Journal of Practice & Theory, 29 (1): 279-293.
Rest, James R. (1979): Development in Judging Moral Issues. Minneapolis: University of
Minnesota Press.
Rest, James R. et al. (1999): Postconventional Moral Thinking: A Neo-Kohlbergian Ap-
proach. Mahwah: Lawrence Erlbaum Associates.
Reynolds, Paul Davidson (1971): A Primer in Theory Construction. Boston: Allyn and Bacon.
Rezaee, Zabihollah/Riley, Richard (2010): Financial Statement Fraud: Prevention and Detec-
tion, 2nd edition. Hoboken: John Wiley & Sons.
Rich, Jay S./Solomon, Ira/Trotman, Ken T. (1997a): Multi-Auditor Judgment/Decision
Making Research: A Decade Later. In: Journal of Accounting Literature, 16: 86-126.
Rich, Jay S./Solomon, I./Trotman, K. T. (1997b): The Audit Review Process: A
Characterization from the Persuasion Perspective. In: Accounting, Organizations and
Society, 22 (5): 481-505.
Roberts, Michael L. (2010): Independence, Impartiality, and Advocacy in Client Conflicts. In:
Research in Accounting Regulation, 22 (1): 29-39.
285
Robinson, John A./Bennink, Carl D. (1978): Field Articulation and Working Memory. In:
Journal of Research in Personality, 12 (4): 439-449.
Robinson, Shani N. (2011): An Experimental Examination of the Effects of Goal Framing and
Time Pressure on Audiotrs’ Professional Skepticism. Dissertation, University of North
Texas.
Robinson, Shani N./Curtis, Mary B./Robertson, Jesse C. (2013): A Person-Situation Approach
to the Examination of Professional Skepticism: Consideration of Time Pressure and
Goal Framing. Working Paper. URL: http://papers.ssrn.com/sol3/papers.cfm?abstract_
id=2276478, accessed on: 04.12.2013.
Rohrmann, Bernd (2005): Risk Attitude Scales: Concepts, Questionnaires, Utilizations.
Project Report. URL: http://www.rohrmannresearch.net/pdfs/rohrmann-racreport.pdf,
accessed on: 11.03.2013.
Rorer, Leonard G. et al. (1967): Configural Judgments Revealed. In: Proceedings of the 75th
Annual Convention of the American Psychological Association, 2: 195-196.
Rose, Anna M./Rose, Jacob M. (2003): The Effects of Fraud Risk Assessments and a Risk
Analysis Decision Aid on Auditors’ Evaluation of Evidence and Judgment. In:
Accounting Forum, 27 (3): 312-338.
Rose, Jacob M. (2007): Attention to Evidence of Aggressive Financial Reporting and
Intentional Misstatement Judgments: Effects of Experience and Trust. In: Behavioral
Research in Accounting, 19: 215-229.
Rosenberg, Morris (1965): Society and the Adolescent Self-Image. Princeton: Princeton
University Press.
Rosenthal, Robert/Rosnow, Ralph L./Rubin, Donald B. (2000): Contrasts and Effect Sizes in
Behavioral Research: A Correlational Approach. Cambridge: Cambridge University
Press.
Ross, Lee/Nisbett, Richard E. (2011): The Person and the Situation: Perspectives of Social
Psychology. London: Pinter & Martin.
Roth, Alvin E. (1988): Laboratory Experimentation in Economics: A Methodological
Overview. In: The Economic Journal, 98: 974-1031.
Rott, Hans (2008): Belief Revision. In: Adler, J. E./Rips, L. J. (eds.): Reasoning: Studies of
Human Inference and Its Foundations. Cambridge et al.: Cambridge University Press:
514-534.
Rott, Hans/Williams, Mary-Anne (2001): Introduction. In: Williams, M./Rott, H. (eds.):
Frontiers in Belief Revision. Dordrecht: Kluwer Academic Publishers: 1-6.
Rotter, Julian B. (1966): Generalized Expectancies for Internal versus External Control of
Reinforcement. In: Psychological Monographs: General and Applied, 80 (1): 1-28.
286
Rotter, Julian B. (1967): A New Scale for the Measurement of Interpersonal Trust. In: Journal
of Personality, 35 (4): 651-665.
Rotter, Julian B. (1980): Interpersonal Trust, Trustworthiness, and Gullibility. In: American
Psychologist, 35 (1): 1-7.
Rovai, Alfred P./Baker, Jason D./Ponton, Michael K. (2014): Social Science Research Design
and Statistics: A Practitioner’s Guide to Research Methods and IBM SPSS, 2nd edition.
Chesapeake: Watertree Press.
Rugg, Michael D. (2004): Retrieval Processing in Human Memory: Electrophysiological and
fMRI Evidence. In: Gazzaniga, M. S. (ed.): The Cognitive Neurosciences. Cambridge:
MIT Press: 727-737.
Runco, Mark A. (2009): Divergent Thinking. In: Kerr, B. (ed.): Encyclopedia of Giftedness,
Creativity, and Talent, Volume 1. Thousand Oaks: Sage Publications: 252-253.
Russell, Bertrand (1921): The Analysis of Mind. New York: McMillan.
Rutledge, Robert W. (1995): The Ability to Moderate Recency Effects through Framing of
Management Accounting Information. In: Journal of Managerial Issues, 7 (1): 27-40.
Salterio, S./Koonce, L. (1997): The Persuasiveness of Audit Evidence: The Case of Account-
ing Policy Decisions. In: Accounting, Organizations and Society, 22 (6): 573-587.
Satava, David/Caldwell, Cam/Richards, Linda (2006): Ethics and the Auditing Culture:
Rethinking the Foundation of Accounting and Auditing. In: Journal of Business Ethics,
64 (3): 271-284.
Schandl, Charles W. (1978): Theory of Auditing: Evaluation, Investigation, and Judgement.
Houston: Scholars Book Company.
Schatzberg, Jeffrey W. et al. (2005): A Reexamination of Behavior in Experimental Audit
Markets: The Effects of Moral Reasoning and Economic Incentives on Auditor Re-
porting and Fees. In: Contemporary Accounting Research, 22 (1): 229-264.
Schellenberg, J. L. (2007): The Wisdom to Doubt: A Justification of Religious Skepticism.
Ithaca: Cornell University Press.
Schmitt, Manfred J./Steyer, Rolf (1993): A Latent State-Trait Model (Not Only) for Social
Desirability. In: Personality and Individual Differences, 14 (4): 519-529.
Schneider, Walter/Shiffrin, Richard M. (1977): Controlled and Automatic Human Information
Processing: I. Detection, Search, and Attention. In: Psychological Review, 84 (1): 1-66.
Schreiber, Stefan M. (2000): Das Informationsverhalten von Wirtschaftsprüfern - Eine
Prozessanalyse aus verhaltenswissenschaftlicher Perspektive. Wiesbaden: Deutscher
Universitäts-Verlag.
Schul, Yaacov/Burnstein, Eugene/Bardi, Anat (1996): Dealing with Deceptions that are Diffi-
cult to Detect: Encoding and Judgment as a Function of Preparing to Receive Invalid
Information. In: Journal of Experimental Social Psychology, 32 (3): 228-253.
287
Schultz, Joseph J., Jr./Reckers, Philip M. J. (1981): The Impact of Group Processing on
Selected Audit Disclosure Decisions. In: Journal of Accounting Research, 19 (2): 482-
501.
Schwind, Jochen (2011): Die Informationsverarbeitung von Wirtschaftsprüfern bei der
Prüfung geschätzter Werte: Eine verhaltenswissenschaftliche und empirische Analyse.
Wiesbaden: Gabler Verlag.
Scott, William R. (1973): A Bayesian Approach to Asset Valuation and Audit Size. In: Journal
of Accounting Research, 11 (2): 304-330.
Seow, Jean-Lin (2009): Cue Usage in Financial Statement Fraud Risk Assessments: Effects of
Technical Knowledge and Decision Aid Use. In: Accounting & Finance, 49 (1): 183-
205.
Shadish, William R./Cook, Thomas D./Campbell, Donald T. (2002): Experimental and Quasi-
Experimental Designs for Generalized Causal Inference. Belmont: Wadsworth, Cengage
Learning.
Shankar, Maya U./McClure, Samuel M. (2013): Decision Neuroscience. In: Ochsner,
K./Kosslyn, S. M. (eds.): The Oxford Handbook of Cognitive Neuroscience, Volume 2:
The Cutting Edges. Oxford: Oxford University Press: 334-341.
Shaub, Michael K. (1994): An Analysis of the Association of Traditional Demographic
Variables with the Moral Reasoning of Auditing Students and Auditors. In: Journal of
Accounting Education, 12 (1): 1-26.
Shaub, Michael K. (1996): Trust and Suspicion: The Effects of Situational and Dispositional
Factors on Auditors’ Trust of Clients. In: Behavioral Research in Accounting, 8: 154-
174.
Shaub, Michael K. (2004): Trust as a Threat to Independence: Emotional Trust, Auditor-
Client Interdependence, and Their Impact on Professional Skepticism. In: Research on
Professional Responsibility and Ethics in Accounting, 9: 169-188.
Shaub, Michael K./Braun, Robert L. (2014): Call of Duty: A Framework for Auditors’ Ethical
Decisions. In: Mintz, S. (ed.): Accounting for the Public Interest: Perspectives on Ac-
countability, Professionalism and Role in Society. New York et al.: Springer: 3-25.
Shaub, Michael K./Lawrence, Janice E. (1996): Ethics, Experience and Professional
Skepticism: A Situational Analysis. In: Behavioral Research in Accounting, 8
(Supplement): 124-157.
Shaub, Michael K./Lawrence, Janice E. (1999): Differences in Auditors’ Professional
Skepticism across Career Levels in the Firm. In: Advances in Accounting Behavioral
Research, 2: 61-83.
Shaub, Michael K./Lawrence, Janice E. (2002): A Taxonomy of Auditors’ Professional
Skepticism. In: Research on Accounting Ethics, 8: 167-194.
288
Shaub, Michael K./Finn, Don W./Munter, Paul (1993): The Effects of Auditors’ Ethical
Orientation on Commitment and Ethical Sensitivity. In: Behavioral Research in Ac-
counting, 5: 145.
Shaw, Tamsin (2010): Nietzsche’s Political Skepticism. Princeton: Princeton University Press.
Shelton, Sandra Waller (1999): The Effect of Experience on the Use of Irrelevant Evidence in
Auditor Judgment. In: The Accounting Review, 74 (2): 217-224.
Shields, Michael D./Solomon, Ira/Waller, William S. (1987): Effects of Alternative Sample
Space Representations on the Accuracy of Auditors’ Uncertainty Judgments. In:
Accounting, Organizations and Society, 12 (4): 375-385.
Silvoso, J. A. et al. (1973): A Statement of Basic Auditing Concepts. Sarasota: American
Accounting Association.
Sim, Marc (2010): National Culture Effects on Groups Evaluating Internal Control. In:
Managerial Auditing Journal, 25 (1): 53-78.
Simnett, Roger (1996): The Effect of Information Selection, Information Processing, and Task
Complexity on Predictive Accuracy of Auditors. In: Accounting, Organizations and
Society, 21 (7/8): 699-719.
Simon, Herbert A. (1955): A Behavioral Model of Rational Choice. In: Quarterly Journal of
Economics, 69: 99-118.
Simon, Herbert A. (1957a): Administrative Behavior: A Study of Decision-Making Processes
in Administrative Organizations, 2nd edition. New York: McMillan.
Simon, Herbert A. (1957b): Models of Man: Social and Rational. New York: John Wiley &
Sons.
Simon, Herbert A. (1959): Theories of Decision-Making in Economics and Behavioral Sci-
ence. In: The American Economic Review, 49 (3): 253-283.
Simon, Herbert A. (1980): The Behavioral and Social Sciences. In: Science, 209: 72-78.
Simon, Herbert A. (1992): Introductory Comment. In: Simon, H. A. et al. (eds.): Economics,
Bounded Rationality and the Cognitive Revolution. Aldershot/Brookfield: Edward
Elger: 3-7.
Simon, Herbert A. (1997): Models of Bounded Rationality: Empirically Grounded Economic
Reason. Cambridge: MIT Press.
Simon, Herbert A. (2013): Administrative Behavior: A Study of Decision-Making Processes
in Administrative Organizations, 4th edition. New York: The Free Press.
Simpson, David D./Ostrom, Thomas M. (1976): Contrast Effects in Impression Formation. In:
Journal of Personality and Social Psychology, 34 (4): 625-629.
Sinnott-Armstrong, Walter (2007): Moral Skepticism. Oxford et al.: Oxford University Press.
289
Skinner, Douglas J./Srinivasan, Suraj (2012): Audit Quality and Auditor Reputation:
Evidence from Japan. In: The Accounting Review, 87 (5): 1737-1767.
Slovic, Paul (1969): Analyzing the Expert Judge: A Descriptive Study of a Stockbroker’s
Decision Processes In: Journal of Applied Psychology, 53 (4): 255-263.
Slovic, Paul/Lichtenstein, Sarah (1971): Comparison of Bayesian and Regression Approaches
to the Study of Information Processing in Judgment. In: Organizational Behavior and
Human Performance, 6 (6): 649-744.
Smith, James F./Kida, Thomas (1991): Heuristics and Biases: Expertise and Task Realism in
Auditing. In: Psychological Bulletin, 109 (3): 472-489.
Smith, Neil (2003): Dissociation and Modularity: Reflections on Language and Mind. In:
Banich, M. T./Mack, M. (eds.): Mind, Brain, and Language: Multidisciplinary Perspec-
tives. Mahwah: Lawrence Erlbaum Associates: 87-111.
Solomon, Ira (1982): Probability of Assessment by Individual Auditors and Audit Teams: An
Empirical Investigation. In: Journal of Accounting Research, 20 (2): 689-710.
Solomon, Ira (1987): Multi-Auditors Judgment/Decision Making Research. In: Journal of
Accounting Literature, 6: 1-25.
Solomon, Ira/Shields, Michael D. (1995): Judgment and Decision-Making Research in Audit-
ing. In: Ashton, R. H./Ashton, A. H. (eds.): Judgment and Decision-Making Research in
Accounting and Auditing. Cambridge et al.: Cambridge University Press: 137-175.
Solomon, Ira/Trotman, Ken T. (2003): Experimental Judgment and Decision Research in
Auditing: The First 25 Years of AOS. In: Accounting, Organizations and Society, 28:
395-412.
Solomon, Ira/Shields, Michael D./Whittington, O. Ray (1999): What Do Industry-Specialist
Auditors Know? In: Journal of Accounting Research, 37 (1): 191-208.
Solso, Robert L./MacLin, Otto H./MacLin, M. Kimberly (2008): Cognitive Psychology, 8th
edition. Boston et al.: Pearson.
Soltani, Bahram (2007): Auditing: An International Approach. Harlow: Pearson Education.
Sonnert, Gerhard/Holton, Gerald (1995): Who Succeeds in Science?: The Gender Dimension.
New Brunswick: Rutgers University Press.
Sontheimer, Kevin (2006): Behavioral Versus Neoclassical Economics: Paradigm Shift or
Generalization? In: Altman, M. (ed.): Handbook of Contemporary Behavioral Econom-
ics: Foundations and Developments. New York/London: M.E. Sharpe: 237-256.
Sperber, Dan (1996): Explaining Culture: A Naturalistic Approach. Oxford/Cambridge:
Blackwell.
Spielberger, Charles D. (1995): State-Trait Personality Inventory (STPI). Redwood City:
Mind Garden.
290
Spilker, Brian C./Prawitt, Douglas F. (1997): Adaptive Responses to Time Pressure: The
Effects of Experience on Tax Information Search Behavior. In: Behavioral Research in
Accounting, 9: 172-198.
Srivastava, Rajendra P./Mock, Theodore J./Turner, Jerry L. (2009): Bayesian and Belief-
Functions Formulas for Auditor Independence Risk Assessment. In: International
Journal of Auditing, 13: 163-183.
Stahl, Silvana (2012): Der Urteilsbildungsprozess von Abschlussprüfern: Eine experimentelle
Untersuchung zur Verarbeitung von Prüfungsnachweisen. Frankfurt am Main et al.:
Peter Lang.
Stanga, Keith G./Turpen, Richard A. (1991): Ethical Judgments on Selected Accounting
Issues: An Empirical Study. In: Journal of Business Ethics, 10 (10): 739-747.
Stanovich, Keith E. (1999): Who Is Rational?: Studies of Individual Differences in Reasoning.
Mahwah: Lawrence Erlbaum Associates.
Stanovich, Keith E. (2004): The Robot’s Rebellion: Finding Meaning in the Age of Darwin.
Chicago/London: University of Chicago Press.
Stanovich, Keith E. (2009): What Intelligence Tests Miss: The Psychology of Rational
Thought. New Haven: Yale University Press.
Stanovich, Keith E. (2011): Rationality and the Reflective Mind. Oxford et al.: Oxford
University Press.
Stanovich, Keith E./West, Richard F. (2000): Individual Differences in Reasoning:
Implications for the Rationality Debate? In: Behavioral and Brain Sciences, 23: 645-726.
Sternberg, Robert J. (1988): The Triarchic Mind: A New Theory of Human Intelligence. New
York: Viking.
Sternberg, Robert J. (2009): Cognitive Psychology, 5th edition. Belmont: Wadsworth.
Sternberg, Robert J./Ben-Zeev, Talia (2001): Complex Cognition: The Psychology of Human
Thought. Oxford: Oxford University Press.
Sternberg, Robert J./Kaufman, James C. (1998): Human Abilities. In: Annual Review of
Psychology, 49 (1): 479-502.
Stevens, James P. (2009): Applied Multivariate Statistics for the Social Sciences, 5th edition.
New York/London: Routledge.
Stewart, Trevor R./Kinney, William R., Jr. (2013): Group Audits, Group-Level Controls, and
Component Materiality: How Much Auditing Is Enough? In: The Accounting Review,
88 (2): 707-737.
Steyer, Rolf/Schmitt, Manfred/Eid, Michael (1999): Latent State–Trait Theory and Research in
Personality and Individual Differences. In: European Journal of Personality, 13 (5): 389-
408.
291
Stiffler, Kirsten L. (2008): Adolescents and Anger: An Investigation of Variables that
Influence the Expression of Anger. Dissertation, Indiana University of Pennsylvania.
Ann Arbor: ProQuest.
Stoner, James A. F. (1968): Risky and Cautious Shifts in Group Decisions: The Influence of
Widely Held Values. In: Journal of Experimental Social Psychology, 4 (4): 442-459.
Stough, Charlotte L. (1969): Greek Skepticism: A Study in Epistemology. Berkeley/Los
Angeles: University of California Press.
Stout, G. F. (1896): Analytic Psychology, Volume 1, reprint edition 2013. London:
Routledge.
Strawson, P. F. (1985): Skepticism and Naturalism: Some Varieties. New York: Columbia
University Press.
Stren, Richard (2001): Knowledge Networks and New Approaches to ‘Development’. In:
Stein, J. G. et al. (ed.): Networks of Knowledge: Collaborative Innovation in Internation-
al Learning. Toronto/Buffalo/London: University of Toronto Press: 133-150.
Suedfeld, Peter/Tetlock, Philip E. (2001): Individual Differences in Information Processing
In: Tesser, A./Schwarz, N. (eds.): Blackwell Handbook of Social Psychology: Intraindi-
vidual Processes. Malden/Oxford: Blackwell Publishers: 284-304.
Superson, Anita M. (2009): The Moral Skeptic. Oxford et al.: Oxford University Press.
Sweeney, John T./Roberts, Robin W. (1997): Cognitive Moral Development and Auditor
Independence. In: Accounting, Organizations and Society, 22 (3/4): 337-352.
Sweller, John/Ayres, Paul/Kalyuga, Slava (2011): Cognitive Load Theory. New York et al.:
Springer.
Swieringa, Robert J./Weick, Karl E. (1982): An Assessment of Laboratory Experiments in
Accounting. In: Journal of Accounting Research, 20: 56-101.
Tabachnick, Barbara G./Fidell, Linda S. (2007): Using Multivariate Statistics, 5th edition.
Boston et al.: Pearson.
Tan, Hun-Tong (1995): Effects of Expectations, Prior Involvement, and Review Awareness
on Memory for Audit Evidence and Judgment. In: Journal of Accounting Research, 33
(1): 113-135.
Tan, Hun-Tong/Kao, Alison (1999): Accountability Effects on Auditors’ Performance: The
Influence of Knowledge, Problem-Solving, and Task Complexity. In: Journal of
Accounting Research, 37 (1): 209-223.
Tan, Hun-Tong/Libby, Robert (1997): Tacit Managerial versus Technical Knowledge as
Determinants of Audit Expertise in the Field. In: Journal of Accounting Research, 35
(1): 97-113.
292
Tan, Hun-Tong/Ng, Terence Bu-Peow/Mak, Bobby Wai-Yeong (2002): The Effects of Task
Complexity on Auditors Performance: The Impact of Accountability and Knowledge. In:
Auditing: A Journal of Practice & Theory, 21 (2): 81-95.
Tan, Soo Jiuan (2002): Can Consumers’ Scepticism be Mitigated by Claim Objectivity and
Claim Extremity? In: Journal of Marketing Communications, 8 (1): 45-64.
Taylor, Mark H. (2000): The Effects of Industry Specialization on Auditors’ Inherent Risk
Assessments and Confidence Judgements. In: Contemporary Accounting Research, 17
(4): 693-712.
Tellegen, Auke et al. (1988): Personality Similarity in Twins Reared Apart and Together. In:
Journal of Personality and Social Psychology, 54 (6): 1031-1039.
Tenenbaum, Gershon/Driscoll, Marcy Perkins (2005): Methods of Research in Sport Sci-
ences: Quantitative and Qualitative Approaches. Aachen et al.: Meyer & Meyer Sport.
Terrell, Steven R. (2012): Statistics Translated: A Step-by-Step Guide to Analyzing and
Interpreting Data. New York: The Guilford Press.
Tetlock, Philip E. (1983): Accountability and Complexity of Thought. In: Journal of Person-
ality and Social Psychology, 45 (1): 74-83.
Tetlock, Philip E./Boettger, Richard (1989): Accountability: A Social Magnifier of the Dilu-
tion Effect. In: Journal of Personality and Social Psychology, 57 (3): 388-398.
Tetlock, Philip E./Skitka, Linda/Boettger, Richard (1989): Social and Cognitive Strategies for
Coping With Accountability: Conformity, Complexity, and Bolstering. In: Journal of
Personality and Social Psychology, 57 (4): 632-640.
Theis, Jochen Christopher/Yankova, Kristina/Eulerich, Marc (2012): Information Order
Effects in the Context of Management Commentary: Initial Experimental Evidence. In:
Journal of Management Control, 23 (2): 133-150.
Thode, Henry C., Jr. (2002): Testing For Normality. New York/Basel: Marcel Dekker.
Thoma, Stephen J./Rest, James/Barnett, Robert (1986): Moral Judgment, Behavior, Decision
Making, and Attitudes In: Rest, J. (ed.): Moral Development: Advances in Theory and
Research. New York/Westport/London: Praeger: 133-175.
Thomas, William C./Ward, Bart H./Henke, Emerson O. (1991): Auditing: Theory and Prac-
tice, 3rd edition. Boston: PWS-Kent.
Thorne, Linda (2000): The Development of Two Measures to Assess Accountants’ Prescrip-
tive and Deliberative Moral Reasoning. In: Behavioral Research in Accounting, 12: 139-
169.
Thorne, Linda/Massey, Dawn W./Magnan, Michel (2003): Institutional Context and Auditors’
Moral Reasoning: A Canada-U.S. Comparison. In: Journal of Business Ethics, 43 (4):
305-321.
293
Thurstone, Louis Leon (1938): Primary Mental Abilities. Chicago: University of Chicago
Press.
Tiku, M. L. (1971): Power Function of the F-Test Under Non-Normal Situations. In: Journal
of the American Statistical Association, 66 (336): 913-916.
Tisdell, Clem (1975): Concepts of Rationality in Economics. In: Philosophy of the Social
Sciences, 5 (3): 259-272.
Toba, Yoshihide (2011): Toward a Conceptual Framework of Professional Skepticism in
Auditing. In: Waseda Business & Economic Studies, (47): 83-116.
Tooby, John/Cosmides, Leda (1992): The Psychological Foundations of Culture. In: Barkow,
J./Cosmides, L./Tooby, J. (eds.): The Adapted Mind: Evolutionary Psychology and the
Generation of Culture. New York: Oxford University Press: 19-136.
Trevino, Linda Klebe (1986): Ethical Decision Making in Organizations: A Person-Situation
Interactionist Model. In: The Academy of Management Review, 11 (3): 601-617.
Trevino, Linda Klebe/Youngblood, Stuart A. (1990): Bad Apples in Bad Barrels: A Causal
Analysis of Ethical Decision-Making Behavior. In: Journal of Applied Psychology, 75
(4): 378-385.
Trompeter, Greg (1994): The Effect of Partner Compensation Schemes and Generally
Accepted Accounting Principles on Audit Partner Judgment. In: Auditing: A Journal of
Practice & Theory, 13 (2): 56-68.
Trompeter, Gregory M. et al. (2013): A Synthesis of Fraud-Related Research. In: Auditing,
32: 287-321.
Trotman, Ken T. (1985): The Review Process and the Accuracy of Auditor Judgments. In:
Journal of Accounting Research, 23 (2): 740-752.
Trotman, Ken T. (1998): Audit Judgment Research - Issues Addressed, Research Methods and
Future Directions. In: Accounting and Finance, 38: 115-156.
Trotman, Ken T. (1999): Perspectives on JDM Research in Auditing over 25 Years. In:
Advances in Accounting Behavioral Research, 2 (March): 15-31.
Trotman, Ken T. (2005): Discussion of Judgment and Decision Making Research in Auditing:
A Task, Person, and Interpersonal Interaction Perspective. In: Auditing: A Journal of
Practice & Theory, 24 (Supplement): 73-87.
Trotman, Ken T. (2011): A Different Personal Perspective through the Behavioral Accounting
Literature. In: Behavioral Research in Accounting, 23 (1): 203-208.
Trotman, Ken T./Wright, Arnold (1996): Recency Effects: Task Complexity, Decision Mode,
and Task-Specific Experience. In: Behavioral Research in Accounting, 8: 175-192.
Trotman, Ken T./Wright, Arnold (2000): Order Effects and Recency: Where Do We Go From
Here? In: Accounting and Finance, 40: 169-182.
294
Trotman, Ken T./Yetton, Philip W. (1985): The Effect of the Review Process on Auditor
Judgments. In: Journal of Accounting Research, 23 (1): 256-267.
Trotman, Ken T./Simnett, Roger/Khalifa, Amna (2009): Impact of the Type of Audit Team
Discussions on Auditors’ Generation of Material Frauds. In: Contemporary Accounting
Research, 26 (4): 1115-1142.
Trotman, Ken T./Tan, Hwee C./Ang, Nicole (2011): Fifty-Year Overview of Judgment and
Decision-Making Research in Accounting. In: Accounting & Finance, 51 (1): 278-360.
Trotman, Ken T./Yetton, Philip W./Zimmer, Ian R. (1983): Individual and Group Judgments of
Internal Control Systems. In: Journal of Accounting Research, 21 (1): 286-292.
Trueblood, Jennifer S./Busemeyer, Jerome R. (2010): A Comparison of the Belief-Adjustment
Model and the Quantum Inference Model as Explanations of Order Effects in Human
Inference. In: Ohlsson, S./Catrambone, R. (eds.): Proceedings of the 32nd Annual Con-
ference of the Cognitive Science Society. Austin: Cognitive Science Society: 1166-1171.
Trueblood, Jennifer S./Busemeyer, Jerome R. (2011): A Quantum Probability Account of
Order Effects in Inference. In: Cognitive Science, 35: 1518-1552.
Tsakumis, George T. (2007): The Influence of Culture on Accountants’ Application of
Financial Reporting Rules. In: Abacus, 43 (1): 27-48.
Tsui, Judy S. L. (1996): Auditors’ Ethical Reasoning: Some Audit Conflict and Cross Cultural
Evidence. In: The International Journal of Accounting, 31 (1): 121-133.
Tsui, Judy S. L./Gul, Ferdinand A. (1996): Auditors’ Behaviour in an Audit Conflict
Situation: A Research Note on the Role of Locus of Control and Ethical Reasoning. In:
Accounting, Organizations and Society, 21 (1): 41-51.
Tsui, Judy S. L./Windsor, Carolyn (2001): Some Cross-Cultural Evidence on Ethical
Reasoning. In: Journal of Business Ethics, 31 (2): 143-150.
Tubbs, Richard M./Messier, William F./Knechel, W. Robert (1990): Recency Effects in the
Auditor’s Belief-Revision Process. In: The Accounting Review, 65 (2): 452-460.
Tubbs, Richard M. et al. (1993): Order Effects in Belief Updating with Consistent and Incon-
sistent Evidence. In: Journal of Behavioral Decision Making, 6: 257-269.
Turner, Cynthia Williams (2001): Accountability Demands and the Auditor’s Evidence
Search Strategy: The Influence of Reviewer Preferences and the Nature of the Response
(Belief vs. Action). In: Journal of Accounting Research, 39 (3): 683-706.
Tuttle, Brad/Coller, Maribeth/Burton, F. Greg (1997): An Examination of Market Efficiency:
Information Order Effects in a Laboratory Market. In: Accounting, Organizations and
Society, 22 (1): 89-103.
Tversky, Amos/Kahneman, Daniel (1974): Judgment under Uncertainty: Heuristics and
Biases. In: Science, 185 (4157): 1124-1131.
295
Tversky, Amos/Kahneman, Daniel (1981): The Framing of Decisions and the Psychology of
Choice. In: Science, 211 (4481): 453-458.
Tversky, Amos/Kahneman, Daniel (1986): Rational Choice and the Framing of Decisions. In:
The Journal of Business, 59 (4): S251-S278.
Van Boven, Leaf et al. (2013): Judgment and Decision-Making. In: Carlston, D. E. (ed.): The
Oxford Handbook of Social Cognition. Oxford et al.: Oxford University Press: 375-401.
Van Breukelen, Gerard J. P. (2010): Analysis of Covariance (ANCOVA), Volume 1. In:
Salkind, N. J. (ed.): Encyclopedia of Research Design. Thousand Oaks et al.: Sage
Publications: 20-26.
Van Hiel, Alain/Mervielde, Ivan (2003): The Need for Closure and the Spontaneous Use of
Complex and Simple Cognitive Structures. In: Journal of Social Psychology, 143 (5):
559-568.
Venkataraman, Ramgopal/Weber, Joseph P./Willenborg, Michael (2008): Litigation Risk,
Audit Quality, and Audit Fees: Evidence from Initial Public Offerings. In: The
Accounting Review, 83 (5): 1315-1345.
Villoro, Luis (1998): Belief, Personal, and Propositional Knowledge. Amsterdam/Atlanta:
Rodopi.
Visala, Aku (2011): Naturalism, Theism, and the Cognitive Study of Religion: Religion
Explained? Farnham/Burlington: Ashgate.
von Neumann, John/Morgenstern, Oscar (1944): Theory of Games and Economic Behavior.
Princeton: Princeton University Press.
Wahlen, James M./Baginski, Stephen P./Bradshaw, Mark (2010): Financial Reporting,
Financial Statement Analysis and Valuation: A Strategic Perspective, 7th edition.
Mason: South-Western, Cengage Learning.
Waller, William S./Felix, William I., Jr. (1984): The Auditor and Learning from Experience:
Some Conjectures. In: Accounting, Organizations and Society, 9 (3/4): 383-406.
Walters, Glenn D. (2002): Criminal Belief Systems: An Integrated-Interactive Theory of
Lifestyles. Westport: Praeger Publishers.
Wang, Karl J./Tuttle, Brad M. (2009): The Impact of Auditor Rotation on Auditor–Client
Negotiation. In: Accounting, Organizations and Society, 34 (2): 222-243.
Ward, James (1933): Psychological Principles, reprint edition 2013. Cambridge: Cambridge
University Press.
Wärneryd, Karl-Erik (1999): The Psychology of Saving: A Study on Economic Psychology.
Cheltenham: Edward Elgar Publishing.
Webb, William M./Worchel, Philip (1986): Trust and Distrust. In: Worchel, S. (ed.):
Psychology of Intergroup Relationships, 2nd edition. Chicago: Nelson-Hall Publishers:
213-228.
296
Wedemeyer, Phil D. (2010): A Discussion of Auditor Judgment as the Critical Component in
Audit Quality: A Practitioner’s Perspective. In: International Journal of Disclosure and
Governance, 7 (4): 320-333.
Wee, Chow-Hou/Lim, Wei-Shi/Lee, Roger (1994): Enterpreneurship: A Review with
Implications for Future Research. In: Journal of Small Business and Entrepreneurship,
11 (4): 25-49.
Weeks, William A. et al. (1999): The Effects of Gender and Career Stage on Ethical Judgment.
In: Journal of Business Ethics, 20 (4): 301-313.
Weld, H. P./Roff, Merrill (1938): A Study in the Formation of Opinion Based upon Legal
Evidence. In: The American Journal of Psychology, 51 (4): 609-628.
Weldon, Rebecca B./Corbin, Jonathan C./Reyna, Valerie F. (2013): Gist Processing in
Judgment and Decision Making: Developmental Reversals Predicted by Fuzzy-Trace
Theory. In: Markovits, H. (ed.): The Developmental Psychology of Reasoning and
Decision-Making. Hove: Psychology Press: 36-62.
Wells, Joseph T. (2003): The Fraud Examiners. In: Journal of Accountancy, 196 (4): 76-80.
Wentzel, Kathryn R. (2005): Peer Relationships, Motivation, and Academic Performance at
School. In: Elliot, A. J./Dweck, C. S. (eds.): Handbook of Competence and Motivation.
New York: Guilford: 279-296.
Westra, Laura S. (1986): Whose “Loyal Agent”? Towards an Ethic of Accounting. In: Journal
of Business Ethics, 5 (2): 119-128.
White, B. Jack (1964): Availability of Categories and Contrast-Effects in Judgment. In: The
American Journal of Psychology, 77 (2): 231-239.
Wickelgren, Wayne A. (1979): Cognitive Psychology. Engelwood Cliffs: Prentice-Hall.
Wilks, T. Jeffrey (2002): Predecisional Distortion of Evidence as a Consequence of Real-Time
Audit Review. In: The Accounting Review, 77 (1): 51-71.
Williams, Michael (2008): Hume’s Skepticism. In: Greco, J. (ed.): The Oxford Handbook of
Skepticism. Oxford et al.: Oxford University Press: 80-107.
Willoughby, Michael/Carmona, Pedro/Momparler, Alexandre (2012): The Effects of the
Provision of Consulting Services on Audit Reporting Quality. In: Service Industries
Journal, 32 (3): 411-429.
Wilson, Jonathan (2014): Essentials of Business Research: A Guide to Doing Your Research
Project, 2nd edition. Thousand Oaks: Sage Publications.
Wilson, Timothy D. (2002): Strangers to Ourselves: Discovering the Adaptive Unconscious.
Cambridge/London: The Belknap Press of Harvard University Press.
Windsor, Carolyn A./Ashkanasy, Neal M. (1995): The Effect of Client Management
Bargaining Power, Moral Reasoning Development, and Belief in a Just World on
Auditor Independence. In: Accounting, Organizations and Society, 20 (7/8): 701-720.
297
Witkin, H. A. et al. (1977): Field-Dependent and Field-Independent Cognitive Styles and
Their Educational Implications. In: Review of Educational Research, 47 (1): 1-64.
Witkin, Herman A. (1976): Cognitive Style in Academic Performance and in Teacher-Student
Relations. In: Messick, S. et al. (ed.): Individuality in Learining: Implications of
Cognitive Style and Creativity for Human Development. San
Francisco/Washington/London: Jossey-Bass: 38-72.
Wittmann, Waldemar (1959): Unternehmung und unvollkommene Information:
Unternehmerische Voraussicht, Ungewißheit und Planung. Köln/Opladen:
Westdeutscher Verlag.
Wong-On-Wing, Bernard/Lui, Gladie (2013): Beyond Cultural Values: An Implicit Theory
Approach to Cross-Cultural Research in Accounting Ethics. In: Behavioral Research in
Accounting, 25 (1): 15-36.
Woodhouse, Mark B. (2011): A Preface to Philosophy, 9th edition. Boston: Wadsworth.
Wooldridge, Jeffrey M. (2009): Introductory Econometrics: A Modern Approach, 4th edition.
Mason: South-Western, Cengage Learning.
WPK (2014): Mitgliederstatistik der WPK. URL: http://www.wpk.de/fileadmin/documents/
WPK/Organisation/WPK-Statistiken_Januar_2014.pdf, accessed on: 24.06.2014.
Wright, Arnold (1988a): The Impact of Prior Working Papers on Auditor Evidential Planning
Judgments. In: Accounting, Organizations and Society, 13 (6): 595-605.
Wright, Sally/Wright, Arnold M. (1997): The Effect of Industry Experience on Hypothesis
Generation and Audit Planning Decisions. In: Behavioral Research in Accounting, 9:
273-294.
Wright, William F. (1988b): Audit Judgment Consensus and Experience. In: Ferris, K. R.
(ed.): Behavioral Accounting Research: A Critical Analysis. Columbus: Century VII
Publishing Company: 305-328.
Wright, William F. (1995): Superior Loan Collectibility Judgments Given Graphical Displays.
In: Auditing: A Journal of Practice & Theory, 14 (2): 144-154.
Wright, William F./Aboul-Ezz, Mohamed E. (1988): Effects of Extrinsic Incentives on the
Quality of Frequency Assessments. In: Organizational Behavior and Human Decision
Processes, 41: 143-152.
Wrightsman, Lawrence S. (1974): Assumptions about Human Nature: A Social-Psychological
Approach. Monterey: Brooks/Cole Publishing.
Yamamura, Jeanne H. et al. (1996): A Comparison of Japanese and U.S. Auditor Decision-
Making in Behavior. In: The International Journal of Accounting, 31 (3): 347-363.
Ye, Ping/Carson, Elizabeth/Simnett, Roger (2011): Threats to Auditor Independence: The
Impact of Relationship and Economic Bonds. In: Auditing: A Journal of Practice &
Theory, 30 (1): 121-148.
298
Ylinen, Mika/Gullkvist, Benita (2012): The Effects of Tolerance for Ambiguity and Task
Uncertainty on the Balanced and Combined Use of Project Controls. In: European
Accounting Journal, 21 (2): 395-415.
Zalabardo, José L. (2012): Scepticism & Reliable Belief. Oxford: Oxford University Press.
Zhao, Yifan/Harding, Noel (2013): Improving the Interpretation of Complex Audit Evidence:
The Beneficial Role of Order Effects. In: Abacus, 49 (4): 476-505.
Zimbelman, Mark F. (1997): The Effects of SAS No. 82 on Auditors’ Attention to Fraud Risk
Factors and Audit Planning Decisions. In: Journal of Accounting Research, 35 (3): 75-
97.
Zouboulakis, Michel S. (2014): The Varieties of Economic Rationality: From Adam Smith to
Contemporary Behavioural and Evolutionary Economics. Abingdon/New York:
Routledge.
Zuckerman, Marvin (2003): Biological Bases of Personality In: Millon, T./Lerner, M. J.
(eds.): Handbook of Psychology, Volume 5: Personality and Social Psychology.
Hoboken: John Wiley & Sons: 85-116.
299
List of Normative Rules and Regulations
AICPA ET 101 (1988): American Institute of Certified Public Accountants Code of
Professional Conduct ET Section 101 – Independence. New York: AICPA.
IDW PS 200 (2000): IDW Prüfungsstandard 200: Ziele und allgemeine Grundsätze der
Durchführung von Abschlussprüfungen. State: 28.06.2000. In: Die Wirtschaftsprüfung,
53: 706.
IDW PS 210 (2012): IDW Prüfungsstandard: Zur Aufdeckung von Unregelmäßigkeiten im
Rahmen der Abschlussprüfung. State: 12.12.2012. In: Die Wirtschaftsprüfung, 66: 7.
IDW PS 270 (2010): IDW Prüfungsstandard: Die Beurteilung der Fortführung der
Unternehmenstätigkeit im Rahmen der Abschlussprüfung. State: 09.09.2010. In: Die
Wirtschaftsprüfung, 63 (Supplement): 1.
IES 3 (2004): International Education Standard 3: Professional Skills and General Education.
URL: http://www.ifac.org/sites/default/files/publications/files/ies-3-professional-skills-
1.pdf.
IESBA Code (2013): Handbook of the Code of Ethics for Professional Accountants. URL:
http://www.ifac.org/sites/default/files/publications/files/2013-IESBA-Handbook.pdf.
ISA 200 (2009): International Standard on Auditing 200: Overall Objectives of the
Independent Auditor and the Conduct of an Audit in Accordance with International
Standards on Auditing. URL: http://www.ifac.org/sites/default/files/downloads/a008-
2010-iaasb-handbook-isa-200.pdf.
ISA 570 (2009): International Standard on Auditing 2010: Going Concern. URL:
http://www.ifac.org/sites/default/files/downloads/a031-2010-iaasb-handbook-isa-
570.pdf.
SAS 1 (2003): References in Auditors’ Reports to the Standards of the Public Company
Accounting Oversight Board. URL:
http://pcaobus.org/Standards/Auditing/Pages/Auditing_ Standard_1.aspx.
SAS 99, AU Section 316 (2002): Consideration of Fraud in a Financial Statement Audit. URL:
http://www.aicpa.org/Research/Standards/AuditAttest/DownloadableDocuments/AU-
00316. pdf.
SAS 109, AU Section 314 (2006): Understanding the Entity and Its Environment and
Assessing the Risks of Material Misstatement. URL: http://www.aicpa.org/Research/
Standards/AuditAttest/DownloadableDocuments/AU-00314.pdf.
SAS 122, AU-C Section 200 (2012): Overall Objectives of the Independent Auditor and the
Conduct of an Audit in Accordance With Generally Accepted Auditing Standards. URL:
http://www.aicpa.org/Research/Standards/AuditAttest/DownloadableDocuments/AU-C-
0020 0.pdf.
301
K. Yankova, The Influence of Information Order Effects and Trait Professional
Skepticism on Auditors’ Belief Revisions, Auditing and Accounting Studies,
DOI 10.1007/978-3-658-08871-2, © Springer Fachmedien Wiesbaden 2015
SAS 126, AU-C Section 570 (2012): The Auditor’s Consideration of an Entity’s Ability to
Continue as a Going Concern (Redrafted). URL: http://www.aicpa.org/Research/Stand
ards/AuditAttest/DownloadableDocuments/AU-C-00570.pdf.
WiPrPrüfV (2007): Prüfungsverordnung für Wirtschaftsprüfer nach §§ 14 und 131l der
Wirtschaftsprüferordnung (Wirtschaftsprüferprüfungsverordnung – WiPrPrüfV) of
20.07.2004 (BGBl. I p. 1707), last amended by Article 3 of the Act of 03.09.2007
(BGBl. I p. 2178).
302