AUDITING: A JOURNAL OF PRACTICE & THEORY American Accounting Association

Vol. 35, No. 4 DOI: 10.2308/ajpt-51452

November 2016
pp. 57–78

Evaluating the Intentionality of Identified Misstatements:

How Perspective Can Help Auditors in Distinguishing
Errors from Fraud
Erin L. Hamilton
University of Nevada, Las Vegas
SUMMARY: Although auditors are responsible for detecting misstatements arising from either error or fraud, the
auditing standards require very different audit responses when a misstatement is believed to be the result of an
intentional act (AS No. 14, PCAOB 2010a). Specifically, if auditors suspect intentional misstatement, then they
should perform additional audit procedures, reassess overall fraud risk and the integrity of management, and
communicate potential concerns to the audit committee. Thus, if auditors fail to recognize and respond to information
indicating a misstatement was caused intentionally, then audit quality may be impaired. The objective of this study is
to investigate whether auditors who consider the perspective of the manager responsible for a misstatement’s
occurrence are more sensitive to circumstances indicating the misstatement was intentional. Using an experiment
with audit managers and senior managers, I find that auditors who consider the client manager’s perspective assess
the misstatement as significantly more likely to be intentional when circumstances surrounding it indicate high, as
opposed to low, fraud risk. In contrast, auditors who do not consider the client manager’s perspective do not assess
intentionality any differently, regardless of the level of fraud risk.
Keywords: perspective taking; fraud risk; fraud detection; misstatement; audit quality.

INTRODUCTION

W
hen a misstatement is identified during the audit, auditors are responsible for evaluating whether it is indicative of
fraud (Public Company Accounting Oversight Board AS No. 14, PCAOB 2010a). The primary characteristic that
differentiates a fraudulent misstatement from an error is whether the underlying action that resulted in the
misstatement was intentional or unintentional (American Institute of Certified Public Accountants [AICPA] 2002). If the
classification of a misstatement as fraud or error hinges on the intent of the individual(s) who caused it, then it is important that
auditors consider fraud risk factors specifically related to the responsible individual(s). For example, a misstatement resulting
from a client manager’s actions should be particularly concerning if the manager directly benefited from the misstatement’s
occurrence. The objective of this study is to investigate whether auditors who consider the perspective of the client manager
responsible for a misstatement’s occurrence are more sensitive to circumstances indicating that the misstatement was
intentional.
Auditors are responsible for identifying material misstatements within the financial statements regardless of their cause;
however, differentiating between an unintentional error and an intentional act of deception has important implications for audit
quality. If auditors believe a misstatement may have been caused intentionally, then the auditing standards recommend very
different audit responses than in the case of an isolated error. According to AS No. 14, when auditors believe that a

This paper is based on my dissertation completed at the University of South Carolina. I thank my dissertation chair, Scott Vandervelde, for his time and
guidance, and my committee members, Marsha Keune, Brad Tuttle, and Douglas Wedell, for their valuable feedback. I am also very grateful to Joel
Owens for his assistance in coding data. I appreciate comments received from Ashley Austin, Jasmijn Bol, Billy Brink, Tina Carpenter, Mark Cecchini,
Jonathan Grenier, Ling Harris, Kathryn Kadous, Tim Keune, Al Leitch, Michael Majerczyk, Roger Martin, Bill Messier, Linda Quick, Aaron Saiewitz,
Kelli Saunders, Jason Smith, Carolyn Westfall, Richard White, Jennifer Winchel, and workshop participants at Miami University; Florida State University;
Portland State University; Tulane University; University of Central Florida; University of Nevada, Las Vegas; University of South Carolina; and
University of Virginia. Funding for the research project described in this article was provided by the Center for Audit Quality.
The views expressed in this article and its contents are those of the author alone and not those of the Center for Audit Quality.
Editor’s note: Accepted by Gregory M. Trompeter.
Submitted: June 2015
Accepted: March 2016
Published Online: March 2016
57
58 Hamilton

misstatement might be intentional, even if it does not clearly exceed quantitative materiality levels,1 they should reassess fraud
risk and the integrity of management, discuss concerns with client management and/or the audit committee, and gather
additional audit evidence to determine whether fraud has in fact occurred (PCAOB 2010a).2 Thus, audit quality may be
impaired if auditors mistakenly believe that a misstatement resulted from an unintentional error, when in fact it was caused
intentionally. In such a case, the auditors may simply correct the misstatement (or deem it immaterial) without reassessing fraud
risk or gathering additional audit evidence that could uncover a more pervasive issue with fraud.
Although studies suggest that auditors generally are able to diagnose high fraud risk during the planning phase of the audit
(e.g., PCAOB 2007; Hoffman and Zimbelman 2009), it is unclear how or how well auditors consider the intentionality of
identified misstatements during audit fieldwork. In a recent Practice Alert, the PCAOB warned auditors against dismissing
identified misstatements as immaterial without giving them adequate consideration (PCAOB 2012). This warning stems from
multiple instances in which regulators have reported that auditors have not been sufficiently skeptical of identified
misstatements in recent years. For instance, failing to further evaluate known misstatements is one of multiple auditor
deficiencies alleged by the SEC in fraud investigations taking place between 1998 and 2010 (Beasley, Carcello, Hermanson,
and Neal 2013). PCAOB inspection reports also have suggested that auditors do not always investigate whether identified
misstatements are indicative of fraud when existing circumstances suggest they should (PCAOB 2007). Thus, additional
improvements and additional research are needed with regard to this important aspect of fraud detection.
Prior research related to fraud detection has primarily focused on the planning phase of the audit (e.g., Asare and Wright
2004; Wilks and Zimbelman 2004a; Carpenter 2007; Hoffman and Zimbelman 2009; Simon 2012). However, evaluating the
intentionality of client management’s behavior is not possible when planning the audit. That is, it is not until auditors have
learned of a client’s actions that resulted in a misstatement that they may begin to consider whether the misstatement was
caused intentionally. Considering a misstatement’s intentionality should be based in large part on the circumstances
surrounding the identified misstatement (SEC 1999; PCAOB 2010a). Unfortunately, a well-established finding within the
psychology literature is that people, in general, often ignore the influential role of the situation (i.e., circumstances) in
influencing the behavior of others (e.g., Jones and Nisbett 1971; Galper 1976; Ross and Nisbett 1991; Gilbert 2002). Thus, in
an audit context, auditors also may fail to consider how existing circumstances might have influenced a client manager’s
actions, which in turn, can make it difficult for them to determine whether the manager intentionally caused a misstatement.
To overcome the challenges of evaluating a misstatement’s intentionality, auditors may benefit from considering the
perspective (e.g., thoughts, feelings, and perceptions) of the client manager responsible for the misstatement. According to
psychology theory, evaluating another’s actions from that person’s point of view leads to an increased understanding of how
existing circumstances might have influenced the individual’s behavior (e.g., Regan and Totten 1975; Vescio, Sechrist, and
Paolucci 2003; Eyal, Liberman, and Trope 2008). Therefore, I expect that auditors who consider the perspective of the client
manager responsible for a misstatement will recognize when existing circumstances may have provided the manager with the
incentive and opportunity to intentionally misstate.
Using an experiment, I provide 82 auditor participants at the manager level and above with a case describing an identified
misstatement that resulted from the actions of a client manager and ask them to assess the likelihood that it was caused
intentionally. I manipulate the circumstances surrounding the identified misstatement to be indicative of high fraud risk (i.e., the
manager had a direct incentive and opportunity to intentionally misstate) or low fraud risk (i.e., the manager had no such
incentive or opportunity). To evaluate whether auditors’ intentionality judgments are influenced by the consideration of
management’s perspective, I use a manipulation as well as a measure of perspective taking, consistent with approaches used in
prior research (e.g., Leith and Baumeister 1998; Galinsky and Moskowitz 2000; Mallett, Huntsinger, Sinclair, and Swim 2008).
The manipulation instructs auditors to take the perspective of the client manager. The measure of perspective taking is derived
from auditors’ open-ended responses and captures whether auditors spontaneously consider the perspective of the client
manager. I choose to use the measure of perspective taking, rather than the manipulation, because it has the ability to directly
capture those auditors who considered the client manager’s perspective by their own admission and differentiate them from
those who did not. While not all experienced auditors consider the perspective of client management when evaluating a
misstatement’s intentionality, those who do make judgments and decisions reflecting greater skepticism of the identified
misstatement when surrounding circumstances are indicative of high fraud risk.
Consistent with expectations, I find that auditors who considered the client manager’s perspective, compared to those who
did not, assessed the misstatement as significantly more likely to be intentional, but only when such increased skepticism was
warranted (i.e., only in the presence of high-fraud-risk information). Furthermore, while auditors who considered the client

1
An example of why a client manager may cause an intentional misstatement that is quantitatively small may be to increase earnings by a small amount
to obtain an earnings-based bonus.
2
Believing that a misstatement might be intentional should also influence auditors’ materiality judgments as intentional misstatements generally are
considered to be qualitatively material (Securities and Exchange Commission [SEC] 1999).

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 59

manager’s perspective assessed the misstatement’s intentionality significantly higher in the high-fraud-risk condition than in the
low-fraud-risk condition, I find that auditors who did not consider the manager’s perspective assessed the misstatement’s
intentionality the same regardless of whether the circumstances surrounding the misstatement were indicative of high or low
fraud risk. Thus, the ability to recognize when the circumstances surrounding a misstatement indicate that it was caused
intentionally is associated with whether or not auditors consider the perspective of the client manager responsible for its
occurrence. I also find that auditors who consider management’s perspective respond to the identified misstatement in more
appropriate ways. Specifically, auditors assessed the materiality of the misstatement to be higher and were more likely to
investigate the misstatement further and communicate their concerns to client management and the audit committee in the
presence of high, compared to low, fraud risk, but only if they had considered the perspective of the manager responsible for
the misstatement.
This study contributes to the audit literature and to audit practice in a number of ways. First, Trompeter, Carpenter, Desai,
Jones, and Riley (2013) have recently recommended that the consideration of fraud be expanded beyond the elements of the
fraud triangle (i.e., circumstances that exist prior to the perpetration of a fraudulent act) to include the ‘‘post-fraud state,’’ which
includes the fraudulent act itself, efforts to conceal the act, and the benefits that accrue to the perpetrator as a result of the act.
My study provides evidence regarding auditors’ evaluation of a potentially fraudulent act (i.e., an identified misstatement), thus
answering Trompeter et al.’s (2013) and Trompeter, Carpenter, Jones, and Riley’s (2014) call for research that includes the
consideration of the post-fraud state.3
Second, I provide evidence regarding a type of reasoning process that appears to help auditors assess the risk that a
misstatement identified during audit fieldwork was caused intentionally. While fraud is often thought of as being perpetrated by
companies, fraudulent acts are actually carried out by individual managers. Thus, in addition to considering fraud risks at the
company level, auditors also should consider fraud risks that are specific to an operating location or individual manager
(AICPA 2002). By considering a client manager’s perspective, auditors presumably gain insight into whether the manager
perceived misstating to be personally beneficial and reasonably easy to perpetrate and conceal, thus assisting in the evaluation
of the manager’s intentions.
Finally, my study responds to the PCAOB’s concerns that auditors are not sufficiently skeptical of identified misstatements
(PCAOB 2007, 2012). The results of my study suggest that auditors are more skeptical of an identified misstatement (i.e.,
believe it is more likely to be intentional) when surrounding circumstances are indicative of high versus low fraud risk, but only
when they consider the client manager’s perspective. One’s ability to successfully take the perspective of another is believed to
be improved through direct experience (Iannotti 1978; Chalmers and Townsend 1990). As such, audit firms may want to
consider integrating perspective taking into future firm trainings for tasks that require an understanding of management’s intent.
The remainder of this paper proceeds as follows. In the following section, I review the related literature and develop my
hypotheses. I then describe my methodology and report the study results in the third and fourth sections, respectively. Finally, I
discuss the conclusions and limitations of the study.

BACKGROUND AND HYPOTHESES DEVELOPMENT

Evaluating a Misstatement’s Intentionality and Attention to Fraud Risk Factors

Most of the fraud detection literature to date has focused on the planning phase of the audit (see Hammersley [2011] for an
extensive review of the fraud-related planning literature). The auditing standards require the performance of various fraud
planning procedures (e.g., fraud risk assessments, brainstorming sessions); thus, fraud should be a primary focus for auditors
during the planning phase of the audit (Braun 2000).4 However, the planning phase of the audit is not the only point at which
fraud should be considered. Auditors should be sensitive to fraud risks during all stages of the audit (AICPA 2002; PCAOB
2010a).
One point at which auditors should consider the potential for fraud is upon identifying a misstatement within the financial
statements (PCAOB 2010a). However, to date, there exists no direct evidence of auditors’ sensitivity to fraud risk factors when
assessing whether identified misstatements might be intentional. Assessing auditors’ sensitivity to fraud risk when evaluating
identified misstatements during audit fieldwork is important (and distinct from the consideration of fraud during the planning
phase) for a number of reasons. First, evidence suggests that auditors are not always sensitive to the presence of fraud during

3
In the High Risk condition of my study, case materials also include information suggesting that the client manager attempted to conceal the fraudulent
act and that he benefited from his actions. Thus, the High Risk version of the case includes all three elements of the post-fraud state discussed by
Trompeter et al. (2013) and Trompeter et al. (2014).
4
Hammersley (2011) concludes in her review of the fraud planning literature that auditors generally are sensitive to the presence of fraud risk factors
when planning the audit.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
60 Hamilton

the evidence evaluation (i.e., testing) phase of the audit (Phillips 1999; Hammersley, Bamber, and Carpenter 2010; Wilks and
Zimbelman 2004b). Braun (2000) reports that during the performance of testing procedures, auditors generally are not sensitive
to information indicative of fraud, particularly when they are under time pressure. Unlike the planning phase of the audit where
fraud is a primary focus, fraud detection becomes a secondary task during subsequent phases (Braun 2000). Anecdotal evidence
suggests that auditors anchor on their initial fraud judgments and, thus, are unlikely to change their beliefs after initially
assessing fraud risk as low (Wilks and Zimbelman 2004b).
Second, in recent years, regulators have emphasized the importance of remaining professionally skeptical and considering
the potential for fraud throughout the course of the audit (AICPA 2002; PCAOB 2010a, 2010b). Nonetheless, the PCAOB and
SEC have documented multiple instances in which auditors have failed to investigate whether departures from GAAP were
indicative of fraud, suggesting that auditors may not be applying sufficient levels of professional skepticism when
misstatements are identified (PCAOB 2007; Beasley et al. 2013). Because most misstatements that auditors encounter during
their careers are due to error rather than fraud, auditors are likely to believe that most identified misstatements are unintentional
(Libby 1985; Loebbecke, Eining, and Willingham 1989). If auditors are too quick to conclude that an identified misstatement is
unintentional, then they could fail to recognize when the circumstances surrounding a misstatement are indicative of high fraud
risk.
Finally, another reason to believe auditors may be insufficiently sensitive to fraud risk factors surrounding an identified
misstatement relates to a well-established finding within psychology concerning how people interpret the behavior of others.
Part of evaluating an identified misstatement involves obtaining an understanding of the client action(s) that contributed to the
misstatement and the circumstances under which it occurred. Many studies in psychology have reported that individuals tend to
ignore the role of the situation when evaluating the behavior of others (e.g., Jones and Nisbett 1971; Galper 1976; Gilbert 2002;
see also Ross and Nisbett [1991] for a review of this literature). Instead, these studies find that people tend to attribute the
actions of others to their dispositional traits. Thus, when a misstatement is identified, the psychology literature suggests that
auditors will fail to recognize how existing circumstances may have influenced a client manager to intentionally misstate. For
example, an auditor may attribute an identified misstatement to a client manager’s lack of competence (i.e., a disposition), thus
concluding it to be unintentional, while failing to recognize that the manager’s compensation or performance evaluation criteria
(i.e., circumstances) provided the manager with an incentive to misstate intentionally.
Failing to recognize the importance of the situation as a driving force in determining the behavior of others is considered
by many to be an error in reasoning (Jones and Nisbett 1971; Ross 1977; Gilbert and Jones 1986).5 The auditing standards are
clear about the importance of considering existing client circumstances when evaluating the potential for fraud. SAS No. 99
states that ‘‘even otherwise honest individuals can commit fraud in an environment that imposes sufficient pressure on them’’
(AICPA 2002, para. .07). If auditors fail to consider how existing circumstances may have influenced a client manager’s
actions, then they may not recognize when a misstatement is indicative of fraud.

Perspective Taking
Research in psychology suggests that a greater understanding of how existing circumstances might have influenced an
individual’s behavior is achieved by considering that individual’s point of view (i.e., perspective taking). Perspective taking is
the process of inferring another individual’s thoughts, feelings, concerns, etc. in a particular situation (Coutu 1951; Davis,
Conklin, Smith, and Luce 1996; Batson, Early, and Salvarani 1997; Epley and Caruso 2009). Perspective taking has been used
in a variety of contexts in psychology as a method for improving interpersonal understanding and is theorized to provide insight
into the thoughts, feelings, and intentions of others (e.g., Mendelsohn and Straker 1999; Galinsky and Moskowitz 2000). In
fact, Kozak, Marsh, and Wegner (2006) find that by instructing participants to imagine themselves (versus observe someone
else) enacting a series of behaviors, participants gave more consideration to the potential goals and purposes (i.e., intentions)
underlying the actions.
Prior studies in auditing have found that actively playing out the role of the client (i.e., role-playing) can lead to improved
negotiation outcomes and financial reporting quality, presumably because role playing provides insight into the client’s
perspective (Trotman, A. Wright, and S. Wright 2005; Church, Peytcheva, Yu, and Singtokul 2015). Trotman et al. (2005) find
that auditors who played the role of the client during a mock negotiation subsequently negotiated larger write-downs during the
actual negotiation compared to auditors receiving other negotiation interventions. Using a laboratory experiment, Church et al.
(2015) report that participants who were assigned the role of a client manager prior to being assigned an auditor role were more
accurate in discerning true earnings from a client manager’s reported earnings. Finally, Altiero, Kang, and Peecher (2013) find

5
This error in reasoning has been termed the ‘‘fundamental attribution error,’’ and describes the tendency to draw personal, dispositional inferences from
an individual’s actions without taking into account that the actions were likely influenced to a great extent by situational forces and constraints (Ross
1977).

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 61

that auditors prompted to think from the perspective of an investor assessed materiality to be significantly lower for a
qualitatively immaterial misstatement, compared to auditors who did not receive the prompt.
Consistent with the finding in psychology that individuals have a tendency to ignore the role of the situation in
influencing the behavior of others (Jones and Nisbett 1971; Gilbert 2002), I expect that auditors who use their natural, third-
person perspective to assess the risk that a misstatement was caused intentionally will be insensitive to the circumstances
surrounding the misstatement and their fraud implications. Given that perspective taking provides insight into how another
individual might perceive and respond to a set of circumstances (Jones and Nisbett 1971; Gilbert 2002), it is likely to be a
useful cognitive tool for recognizing when the circumstances surrounding an identified misstatement indicate that it was
caused intentionally. It is important to bear in mind, however, that the risk-based nature of auditing suggests it is ineffective
and inefficient for auditors to respond to all identified misstatements in a highly skeptical manner, as this behavior would
reduce the time and resources available for testing other high-risk audit areas. As such, enhanced skepticism is only
preferable when it occurs in the presence of high-fraud-risk (as opposed to low-fraud-risk) circumstances. Accordingly, I
expect that auditors who consider the manager’s perspective will assess an identified misstatement as more likely to be
intentional when the circumstances surrounding it are indicative of high fraud risk (i.e., the manager responsible for the
misstatement had an incentive and opportunity to misstate) as compared to low fraud risk.6 Stated formally, my first
hypothesis is as follows:
H1: Auditors who engage in perspective taking will assess the likelihood that an identified misstatement was caused
intentionally higher than auditors who do not engage in perspective taking, but only when the circumstances
surrounding the misstatement are indicative of high (as opposed to low) fraud risk.
When a client manager perceives fraud to be both personally beneficial (an incentive) and relatively easy to perpetrate and
conceal (an opportunity), the manager is more likely to form fraudulent intentions and engage in a fraudulent act (AICPA
2002).7 Considering the perspective of client management is expected to help auditors recognize whether a set of circumstances
might have been interpreted by the client as being ‘‘ripe’’ for fraud. As pointed out by Trompeter et al. (2013), a manager’s
assessment of whether an opportunity for fraud exists is less dependent on the actual circumstances that exist at the company
(e.g., the presence of strong internal controls) and more dependent on the manager’s perception of these circumstances. By
providing insight into how a client manager may have perceived a set of circumstances (e.g., Jones and Nisbett 1971; Ross and
Nisbett 1991; Epley and Caruso 2009), I propose that auditors who consider the client manager’s perspective will recognize the
extent to which the manager had an incentive and opportunity to intentionally misstate. Thus, in the presence of circumstances
indicative of high fraud risk, perspective taking should help auditors recognize the manager’s increased incentive and
opportunity for fraud.8
It is possible that perspective taking may not be equally effective in helping auditors evaluate a client manager’s incentive
and opportunity for fraud. Therefore, my final hypothesis is broken into two parts to determine which of these two elements of
the fraud triangle are made salient when taking the client manager’s perspective—the manager’s incentive to intentionally
misstate, the opportunity to misstate, or both. Stated formally, my final hypotheses are as follows:
H2a: Auditors who engage in perspective taking will assess the client manager’s incentive for fraud higher than auditors
who do not engage in perspective taking, but only when the circumstances surrounding an identified misstatement
are indicative of high (as opposed to low) fraud risk.
H2b: Auditors who engage in perspective taking will assess the client manager’s opportunity for fraud higher than
auditors who do not engage in perspective taking, but only when the circumstances surrounding an identified
misstatement are indicative of high (as opposed to low) fraud risk.

6
My first hypothesis specifies an ordinal interaction in which perspective taking is predicted to have an effect in the high-fraud-risk condition, but not in
the low-fraud-risk condition. The low-risk condition serves as a point of comparison for the judgments made by auditors in the high-risk condition and
is also used to rule out the possibility that perspective taking results in increased skepticism (i.e., increased assessments of intentionality) regardless of
the circumstances.
7
The ability to rationalize a fraudulent act is the third element of the fraud triangle (PCAOB 2010a). Because my theory relates to the use of perspective
taking to understand how the situation influences another’s behavior, I focus exclusively on incentives and opportunities, which are the elements most
related to client circumstances.
8
Similar to H1, the next two hypotheses predict an ordinal interaction whereby perspective taking is expected to have an effect in the high-fraud-risk
condition, but not in the low-fraud-risk condition. The low-fraud-risk condition ensures that perspective taking only results in increased skepticism of
an identified misstatement when such skepticism is warranted (i.e., in the high-fraud-risk condition, but not in the low-fraud-risk condition).

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
62 Hamilton

METHODOLOGY

Participants and Study Administration

Participants included 82 auditors from the Big 4 public accounting firms, consisting of 25 managers, 56 senior managers,
and one partner. Participants had a combined average of 9.3 years of audit experience.9 Through discussions with audit
practitioners, it was determined that evaluating whether identified misstatements might be intentional is typically performed by
auditors with this level of experience. A total of 63 participants accessed the experiment online using Qualtrics software and the
remaining 19 participants completed the study in paper-based format. There were no differences in responses based on firm or
medium of study administration.

Case Materials and Procedures

I created case materials in consultation with an audit partner from a Big 4 firm. The case materials were then reviewed by
three additional audit partners. Finally, the case was pilot-tested with audit managers and senior managers and was revised
based on partner feedback and pilot test results to increase the understandability of the case.
Participants were provided with case materials related to a hypothetical year-end audit engagement and were asked to
evaluate a misstatement identified by the audit team. All participants were provided with an excerpt from the audit program
instructing them to ‘‘consider whether each misstatement is or might be intentional,’’ consistent with the wording of AS No. 14
(PCAOB 2010a). After viewing the audit program instructions, participants received information related to the identified
misstatement, including the circumstances contributing to its occurrence and the manager responsible. The misstatement is
described as relating to a payment made to a supplier that was incorrectly recorded as a prepayment (i.e., an asset
classification). The payment should have been recorded as an expense as it related to a one-time penalty for failing to purchase
the level of product specified within a supply agreement. All participants received an explanation from client management that
the misstatement had resulted from an unintentional classification error. Included within the case information were the
manipulated fraud risk factors as well as information about the quantitative and qualitative materiality of the misstatement.
The materiality information in all conditions indicated that the misstatement was below all quantitative benchmarks, did
not allow the company to meet a bonus target or analyst earnings target, and did not mask a change in earnings or other trends.
The information concerning qualitative materiality was intended to suggest that the company (and management) as a whole did
not benefit from the misstatement. The manipulation of fraud risk, discussed in the next section, was designed to relate
specifically to the client manager responsible for the misstatement’s occurrence. Thus, if auditors only attend to information
concerning the impact of the misstatement on the company as a whole, then they may fail to recognize information suggesting
that the client manager who caused the misstatement had an incentive and opportunity to do so intentionally.
After participants finished reviewing the case information, they were asked to evaluate the likelihood that the misstatement
is or might be intentional. Participants then described the facts that they considered when evaluating whether the misstatement
was intentional, assessed the extent to which the client manager responsible had an incentive and opportunity for fraud,
indicated follow-up procedures, and answered demographic and manipulation check questions.

Independent Variables
The study employs two independent variables: Fraud Risk, which is manipulated between-participants at a high and low
level, and Perspective Taking, which is manipulated and measured. Both independent variables are described in greater detail in
this section.

Fraud Risk
Fraud Risk is manipulated by varying four pieces of information directly related to the client manager responsible for the
misstatement and his actions leading up to the misstatement.10 In this way, the likelihood that the client manager intentionally
caused the misstatement is varied, while information concerning the general level of fraud risk for the company remained
consistent between conditions. Two of the four fraud risk factors varied the extent to which the client manager personally
benefited from causing the misstatement (i.e., an incentive for the manager to intentionally misstate). Specifically, these two

9
One audit firm did not provide full demographic data; therefore, mean audit experience for my sample was calculated using a reduced sample of 58
participants.
10
Through a pretest, I obtained verification from an audit partner that each of the manipulated high (low) fraud risk factors did in fact increase (decrease)
the risk that the misstatement might be intentional.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 63

incentive-related factors varied (1) whether the client manager was responsible (or not) for setting unreachable purchasing
targets that resulted in a contractual penalty (i.e., a bad business decision that negatively impacted the company), and (2)
whether the client manager’s performance evaluation criteria indicated that the effects of the misstatement benefited (or
harmed) the manager’s performance. The remaining two fraud risk factors concerned the perceived ease or difficulty with
which the client manager could have perpetrated and concealed the misstatement (i.e., an opportunity for the manager to
intentionally misstate). The two opportunity-related factors varied (1) the likelihood that the misclassified transaction would be
reviewed by a superior, and (2) the presence (absence) of internal audit at the client manager’s division. See Appendix A for the
full manipulations.

Perspective Taking
Perspective taking has been analyzed in psychology studies by manipulating perspective taking via instructions (e.g.,
Davis et al. 1996; Galinsky and Moskowitz 2000), measuring individuals’ dispositional tendency toward perspective
taking via self-report (e.g., Davis 1983), or measuring participants’ consideration of another’s perspective as indicated
through open-ended responses (e.g., McPherson Frantz and Janoff-Bulman 2000; Mallett et al. 2008). Various methods
have been employed in prior research due to the difficulty of capturing the perspective-taking construct (e.g., Ickes,
Stinson, Bissonnette, and Garcia 1990). Therefore, I use a two-fold process to ensure that I am able to capture perspective
taking among my experienced auditor participants. I attempt to manipulate perspective taking by giving participants
instructions, and I use a measure based on participants’ open-ended responses. I describe both, in turn, and discuss why
the measure of perspective taking is the preferred method for analyzing the effects of perspective taking in the present
study.
Perspective-taking instructions. The first way I attempt to capture perspective taking is by manipulating the presence or
absence of Perspective Taking Instructions included within the audit program (PT Instructions-Present versus PT Instructions-
Absent). Participants in the PT Instructions-Present condition were told to evaluate the facts and circumstances related to the
misstatement using the following evaluation process: ‘‘Think from the perspective of the client-individual responsible for the
misstatement. Put yourself in the place of this client-individual and try to imagine what you would think and how you would
feel. . .’’11 In contrast, participants in the PT Instructions-Absent condition were told to evaluate the facts and circumstances ‘‘as
you normally would in practice’’ (see Appendix A for the full manipulations).
Although some studies in psychology have manipulated perspective taking via instructions, these studies have primarily
used undergraduate students as participants rather than highly experienced auditors (e.g., Regan and Totten 1975; Davis et al.
1996; Batson et al. 1997; Galinsky and Moskowitz 2000). A potential issue with manipulating perspective taking among
experienced auditors is that they likely have an established process in place for evaluating identified misstatements. As a result,
auditors may be less susceptible to simple instructions prompting them to consider the client manager’s perspective if that is not
part of their normal evaluation process.12 Alternatively, auditors who do not receive perspective-taking instructions (i.e., those
told to evaluate the misstatement as they ‘‘normally would in practice’’), still may consider the client manager’s perspective if
that is part of their ‘‘normal’’ evaluation process. That is, auditors may naturally (i.e., spontaneously) engage in perspective
taking, even when not explicitly told to do so (e.g., Leith and Baumeister 1998; Mallett et al. 2008). For instance, auditors who
encounter information indicative of high fraud risk may naturally consider how client management might perceive these high-
risk circumstances.
Perspective-taking measure. The second way I capture perspective taking is by constructing a measure from
participants’ open-ended responses, which is a method used in some psychology studies (e.g., Leith and Baumeister 1998;
McPherson Frantz and Janoff-Bulman 2000; Mallett et al. 2008). Because I am most interested in understanding the effects of
actively considering the perspective of a client manager, regardless of whether it is performed spontaneously or after receiving
instructions to do so, I construct (and ultimately use) a measure of Perspective Taking that is similar to measures used within
the psychology and audit literatures.
Perspective taking requires an individual to gather information from a given context and develop an understanding of
another by inferring the other’s thoughts and feelings under the circumstances (Gesn and Ickes 1999); the perspective-taking
measure is intended to capture these inferences. Prior studies in psychology that use a measure of perspective taking analyze

11
The wording of the Perspective Taking manipulation is consistent with perspective-taking studies in psychology, which instruct participants to imagine
themselves in another’s shoes and consider their own thoughts and feelings under the circumstances facing another (e.g., Davis et al. 1996; Galinsky
and Moskowitz 2000; Mallett et al. 2008).
12
Furthermore, studies in psychology suggest that perspective taking is an effortful process, which can be inhibited when there are high demands on one’s
cognitive resources (e.g., Davis et al. 1996). The present study includes substantially more information for participants to evaluate compared to the
typical psychology study. Thus, demands on participants’ cognitive resources are likely to be greater.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
64 Hamilton

whether participants document that they considered the other individual’s internal mental states—namely, the individual’s
thoughts, beliefs, feelings, wants, desires, and/or interpretations of events and circumstances. A similar coding of open-
ended responses has been used within the audit literature to capture auditors’ mental states, including problem
representations and mental simulations (e.g., Hammersley 2006; Trotman, Simnett, and Khalifa 2009). The strength of this
measured approach is that it relies on participants’ own words and admissions regarding their consideration of another’s
perspective.13
Based on this prior research, I construct the perspective-taking measure by analyzing responses to a question that asks
participants to explain the factors they considered while evaluating the misstatement’s intentionality. Two coders (the author
and a doctoral student with auditing experience) worked independently to code participant responses. Both coders were blind to
the participant’s experimental condition during the coding process, and the doctoral student coder also was blind to the
hypotheses. A dichotomous coding scheme was used so that each response was either coded as (1) including a consideration of
the client manager’s perspective (PT Present) or (2) not including a consideration of the client manager’s perspective (PT
Absent).14 Specifically, responses were coded as PT Present if they included one or more inferences regarding the specific
manager’s thoughts, beliefs, feelings, wants, desires, and/or interpretation of events or circumstances described in the case, or
coded PT Absent otherwise.
Figure 1 provides examples of responses coded as PT Present versus PT Absent. As illustrated by the examples, references
to the client’s perspective are independent of the case facts. That is, two participants can list the same case facts, but one may
consider these facts from the client’s perspective, while the other does not. The coding scheme only captures the perspective-
taking language. A total of 42 of the 82 participants (51.2 percent) were coded as PT Present, while the remaining 40
participants (48.8 percent) were coded as PT Absent. Inter-rater agreement is 89.0 percent. To test inter-rater reliability, I
compute Cohen’s Kappa, which is a common measure for nominal categorization that corrects for chance agreement (Cohen
1960). Results indicate that the categories have a Kappa value of 0.78, indicating substantial reliability (Landis and Koch 1977;
Fleiss 1981). All coding differences were mutually resolved.15

Dependent Variables
I collect three dependent measures to test my hypotheses: (1) misstatement intentionality assessments; (2) incentive
assessments; and (3) opportunity assessments. Participants’ assessments of misstatement intentionality (used to test H1) were
elicited by asking participants the likelihood that the identified misstatement is or might be intentional on an 11-point scale with
endpoints labeled ‘‘0 ¼ NOT at all Likely’’ and ‘‘10 ¼ VERY Likely.’’
To test H2a and H2b, I measure the extent to which auditors believe the client manager responsible for the misstatement
had an incentive (H2a) and opportunity (H2b) to intentionally misstate. For the incentive assessment, I ask participants to
indicate the extent to which the manager had an incentive for fraud (i.e., how beneficial was it to the manager to intentionally
misstate) on an 11-point scale ranging from ‘‘0 ¼ No Incentive (Not at all Beneficial)’’ to ‘‘10 ¼ Strong Incentive (Very
Beneficial).’’ For the opportunity assessment, I ask participants to indicate the extent to which the manager had an opportunity
for fraud (i.e., how difficult/easy was it for the manager to misstate) on an 11-point scale ranging from ‘‘0 ¼ No Opportunity
(Very Difficult)’’ to ‘‘10 ¼ Strong Opportunity (Very Easy).’’

13
Although I consider the measure’s reliance on the words and admissions of auditors primarily to be a strength of the measured variable approach, I also
acknowledge that it may be considered a weakness, as individuals sometimes lack insight into their own cognitive processes. To the extent that auditors
lack such self-insight or are unable to articulate their judgment process, the measure may fail to fully capture perspective taking.
14
A dichotomous measure was chosen, because I am most interested in whether or not auditors engaged in perspective taking and not necessarily the level
of perspective taking. A single statement regarding the thoughts, feelings, etc. of the client manager indicates that the auditor considered the manager’s
perspective. Additionally, using a dichotomous measure allows me to easily compare the perspective-taking measure to the manipulation of perspective
taking, which also has two levels (present versus absent).
15
A potential weakness in the measured variable is that it depends on participants’ documentation of the manager’s perspective. That is, some participants
may have considered the manager’s perspective, but failed to document their consideration, causing them to be misclassified as PT Absent. Therefore, I
constructed an alternative perspective-taking measure, which is less prone to this inherent weakness. Specifically, I used a question that asked
participants to self-report the extent to which they considered the manager’s perspective on a scale ranging from ‘‘0 ¼ I ONLY considered MY OWN
perspective’’ to ‘‘10 ¼ I ONLY considered [the manager’s] perspective,’’ with a midpoint of ‘‘5 ¼ I EQUALLY considered by own and [the manager’s]
perspectives.’’ Participants were then classified as PT Present if they either documented a consideration of the manager’s perspective (consistent with
the original measure) or indicated a high degree of perspective taking on the self-report measure (by selecting a value above the scale’s midpoint). By
supplementing the original measure with the self-report measure, I am able to capture participants who consciously considered, but failed to document,
the manager’s perspective. Results (untabulated) using this alternative measure yield identical inferences, providing comfort that any misclassification
resulting from the failure to document perspective-taking considerations is minor and does not significantly impact results.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 65

FIGURE 1
Perspective-Taking Measure Coding Scheme Examples

Figure 1 provides excerpts from participant responses to a question asking them to list the factors they considered when evaluating whether the
misstatement is or might be intentional. The table indicates whether the participant response was coded as PT Absent or PT Present. Italicized text within
the PT Present column points out the participant’s reference to the manager’s (i.e., Joe’s) perspective, interpretation, thoughts, feelings, and wants/desires
that were considered by the coders to be an admission that the participant had considered the manager’s perspective while evaluating the misstatement.

RESULTS

Capturing Perspective Taking

Recall that I employed two methods of capturing perspective taking that have been used in prior literature. First, I provided
half of my experienced auditor participants with perspective-taking instructions in an attempt to induce perspective taking.
Second, due to the potential difficulty with manipulating perspective taking with such experienced participants, I constructed a
measure to capture perspective taking by participants’ own admission.
To determine the effectiveness of the perspective-taking manipulation, I first analyze data collected in the post-
experimental questionnaire. Specifically, I analyze a question that asks participants to identify the case instructions they
received, where the two options were to evaluate the case information (1) ‘‘from the perspective of Joe Rogers, Purchasing
Manager (i.e., by imagining myself in the place of the client-individual responsible for the misstatement),’’ or (2) ‘‘as I normally
would in practice.’’ A total of 34 of the 42 participants who received perspective-taking instructions (81 percent) answered this
question correctly, and all 40 participants (100 percent) who were told to evaluate the misstatement as they normally would in
practice correctly recalled receiving these instructions. Although participants largely recalled the instructions they received, this
question cannot gauge whether the experienced auditor participants who received perspective-taking instructions actually
followed these instructions. Furthermore, it is possible that auditors instructed to evaluate the misstatement ‘‘as they normally
would in practice’’ may still have engaged in perspective taking if that is part of their ‘‘normal’’ evaluation process. Therefore, I
further analyze the effectiveness of the perspective-taking manipulation by looking at whether auditors receiving perspective-
taking instructions were more likely to document a consideration of the client manager’s perspective while evaluating the
misstatement compared to auditors who did not receive these instructions.
To perform this second analysis, I use the perspective-taking measure to determine whether participants who received
perspective-taking instructions were, by their own admission, more likely to consider the client manager’s perspective. I find
that of the 42 participants who received perspective-taking instructions, only 19 (45.2 percent) documented that they had

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
66 Hamilton

considered the client manager’s perspective (i.e., thoughts, feelings, etc.).16 Moreover, of the 40 participants who did not
receive perspective-taking instructions, 23 (57.5 percent) spontaneously considered the perspective of the client manager
without being prompted.
Since I am interested in understanding the benefits of perspective taking, regardless of whether it is performed
spontaneously or in response to instruction, I choose to use the method that more clearly divides those participants who
considered the client manager’s perspective from those who did not. Based on my analysis as a whole, I conclude that the
measured variable represents a more appropriate way of examining the effects of perspective taking in my study.17

Control Variables
Because I am using a measured independent variable, one could argue that any association between perspective taking and
auditors’ judgments could be due to additional effort exerted by participants who spontaneously engaged in perspective taking.
To rule out this alternative explanation, I include two control variables in all of my analyses as proxies for participant effort in
the experimental task: Total Facts and Time.18 Total Facts represents the total number of case facts listed by participants when
asked to explain the factors they considered while evaluating the misstatement,19 and Time represents the number of minutes
that participants spent on the experimental task.20 Neither Fraud Risk nor the measure of Perspective Taking is significantly
correlated with participant effort (i.e., Total Facts and Time) or participants’ level of experience (manager, senior manager, or
partner).21

Perspective-Taking Exploratory Analysis

Before analyzing the hypotheses regarding the effect of perspective taking on auditor judgments, I perform an exploratory
analysis to understand the circumstances under which auditors are more likely to consider the perspective of client
management. Specifically, I investigate whether auditors are more likely to consider client management’s perspective in the
presence of high-fraud-risk circumstances compared to low-risk circumstances. Because mindsets (i.e., cognitive states) can be
activated temporarily when individuals encounter certain situational variables (Hurtt 2010), I examine whether high-fraud-risk
circumstances, as compared to low-risk circumstances, activate a perspective-taking mindset in experienced auditors. When
tasked with evaluating whether a client manager intentionally caused a misstatement, auditors should be more likely to consider
that manager’s perspective (i.e., enter into a perspective-taking mindset) when their skepticism has been piqued by information
suggesting that the manager had an incentive and opportunity to cause the misstatement intentionally.

16
Using a measure of perspective taking similar to the one used in the present study, Leith and Baumeister (1998) find that 72 percent of their participants
actively engaged in perspective taking after receiving instructions to do so. There exist two plausible reasons for finding a lower percentage in my
study: (1) my auditor participants possess a greater level of experience than the undergraduate participants used in that study, and (2) my task requires
higher cognitive demands, which can make perspective taking particularly difficult (Davis et al. 1996).
17
When results are analyzed using the Perspective Taking Instructions variable, I find that regardless of perspective-taking condition, auditors’
assessments of the misstatement’s intentionality were higher in the high-risk, compared to the low-risk, condition (F1,78 ¼ 12.42, p ¼ 0.001,
untabulated). The finding of a main effect for Fraud Risk suggests that some participants were attuned to the heightened fraud risk indicated in the high-
risk condition. However, the absence of an interaction between Fraud Risk and the manipulation of Perspective Taking Instructions suggests that either
(1) perspective taking does not help auditors evaluate intentionality as theory suggests it should or (2) manipulating perspective taking is not an
effective method for separating those auditors who considered the manager’s perspective from those who did not. To rule out the former possibility, I
use the measure of perspective taking that more directly differentiates participants who engaged in perspective taking from those who did not. Finding
an interaction between fraud risk and the perspective-taking measure, consistent with my hypotheses, would provide evidence that perspective taking is
effective in evaluating the intentionality of misstatements, but that prompting experienced auditors to engage in perspective taking is not an effective
means of eliciting this mental process.
18
Three of the four firms provided additional demographic data. For these 58 participants, I conducted analyses with the following additional control
variables: years of audit experience, number of frauds encountered in the last three years, and a measure of how often participants are responsible for
evaluating the intentionality of misstatements on their audit teams (i.e., task experience), which was measured using an 11-point scale with endpoints
labeled ‘‘0 ¼ Never’’ and ‘‘10 ¼ Always.’’ Results (untabulated) with these control variables yield identical inferences.
19
I chose to use the number of case facts listed by participants, rather than the number of words, because perspective taking has been found to foster
information elaboration (Hoever, van Knippenberg, van Ginkel, and Barkema 2012). Thus, when auditors engage in perspective taking, I expect them
to elaborate to a greater extent about the manager’s thoughts, feelings, etc. Consistent with this expectation, I find that the number of words written by
participants is significantly correlated with the measure of perspective taking (r ¼ 0.28, p ¼ 0.012). However, when number of words is used as a control
variable in my analyses in place of Total Facts, results (untabulated) are unchanged.
20
Five participants were eliminated because their Time was in excess of 12 hours, which indicated that they had not completed the experiment in one
sitting (Time for all other participants was below two hours). Because some measures are memory-based, participants were instructed to complete the
study in one sitting. Results are qualitatively unchanged when these participants are included in the analyses.
21
To eliminate the possibility that participants’ level of experience impacts both perspective taking and their subsequent judgments, I ran all analyses with
Auditor Level as an additional independent variable with three levels (manager, senior manager, and partner). Auditor Level does not have a significant
effect in the ANCOVA model and is not significantly correlated with the measured Perspective Taking variable; therefore, I present pooled results.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 67

Using the measure of Perspective Taking, I analyze the proportion of auditors classified as PT Present (i.e., those indicating
that they considered some aspect of the client manager’s perspective while evaluating the misstatement) in the high-fraud-risk
condition compared to the low-fraud-risk condition. I find that auditors are significantly more likely to consider the perspective
of client management when they encounter circumstances indicative of high, compared to low, fraud risk (v21 ¼ 4.68, ptwo-tailed
¼ 0.031, untabulated). Specifically, I find that 25 of the 39 participants in the high-fraud-risk condition (64.1 percent)
considered the client manager’s perspective compared to 15 of the 38 participants (39.5 percent) in the low-fraud-risk
condition. It is interesting to note, however, that not all of my highly experienced auditor participants considered the
perspective of client management, even when faced with circumstances indicative of high fraud risk. As such, it is of interest to
examine whether auditors who did consider management’s perspective benefited from doing so, as predicted by my
hypotheses.22

Intentionality Judgments
H1 predicts that auditors who consider the manager’s perspective will assess the identified misstatement as more likely to be
intentional, compared to auditors who do not consider the manager’s perspective, but only when surrounding circumstances are
indicative of high fraud risk. Panel A of Table 1 provides descriptive statistics and Panel B provides the results of an ANCOVA
that includes Perspective Taking (PT) and Fraud Risk as the independent variables as well as the two control variables. Although I
report the default ANCOVA interaction in Panel B, H1 predicts a specific ordinal interaction. Therefore, I use a planned contrast
(with weights of þ3 for the PT Present/High Risk condition and 1 for all other conditions) as my primary test of H1. The planned
interaction contrast, presented in Panel C of Table 1, is significant (F1,71 ¼ 22.80, p , 0.001), providing support for H1.
Follow-up simple effects presented in Panel C of Table 1 show that the pattern of cell means is consistent with my
predictions, and Figure 2 depicts the resulting ordinal interaction. Specifically, I find that in the High Risk condition,
misstatement intentionality assessments are significantly higher for auditors who considered the client manager’s perspective
(PT Present) compared to those who did not (PT Absent) (F1,71 ¼ 12.38, p ¼ 0.001). Also consistent with expectations, I find no
effect for Perspective Taking in the Low Risk condition (F1,71 ¼ 0.11, p ¼ 0.736), suggesting that perspective taking only
increases auditor skepticism when such an increase is appropriate (i.e., in the presence of high-risk cues). Importantly, results
also indicate that auditors who considered the client manager’s perspective (PT Present) assess misstatement intentionality
significantly higher in the High versus Low Risk condition (F1,71 ¼ 11.32, p ¼ 0.001), whereas auditors who did not consider
the client manager’s perspective (PT Absent) do not assess misstatement intentionality differently between the Fraud Risk
conditions (F1,71 ¼ 0.02, p ¼ 0.893). Overall, these findings support the idea that without perspective taking, auditors have
difficulty recognizing when circumstances indicate a high risk of intentional misstatement.

Assessments of the Client Manager’s Incentive and Opportunity to Misstate

H2a and H2b predict that auditors who consider the client manager’s perspective will assess the manager’s incentive and
opportunity for fraud to be higher than auditors who do not consider the manager’s perspective, but only in the presence of
high-fraud-risk information. By testing H2a and H2b separately, I can determine whether perspective taking results in a greater
understanding of the client manager’s incentive to misstate (H2a), the opportunity to misstate (H2b), or a combination of both.
Panel A of Table 2 provides descriptive statistics for auditors’ incentive assessments and their opportunity assessments.
Panel B displays the ANCOVA models for each dependent variable. Similar to H1, H2a and H2b predict a specific ordinal
interaction. Therefore, I use a planned contrast (with weights of þ3 for the PT Present/High Risk condition and 1 for all other
conditions, consistent with the weights used to test H1) as my primary test of H2a and H2b. The planned interaction contrasts,
presented in Panel C of Table 2, are significant for both incentive assessments (F1,71 ¼ 38.19, p , 0.001) and opportunity
assessments (F1,71 ¼ 7.54, p ¼ 0.008), providing initial support for H2a and H2b.

22
The finding that more auditors engaged in perspective taking in the high-fraud-risk condition, compared to the low-risk condition, may raise concerns
regarding the endogeneity of perspective taking and fraud risk. That is, it may be unclear whether (1) perspective taking leads to the results reported in
my study, or (2) sensitivity to high fraud risk leads to perspective taking and the results reported in my study. The latter possibility is problematic in that
it would suggest that my results are explained by a heightened sensitivity to fraud risk and not by perspective taking. To rule out this possibility, it is
necessary to show that auditors who did not engage in perspective taking (PT Absent) in the high-risk condition were in fact sensitive to the presence of
high-risk cues (but failed to adopt a perspective taking mindset nonetheless). Therefore, I analyzed responses to a question that asked participants to
assess the level of fraud risk present at the company on a scale ranging from ‘‘0 ¼ Very LOW’’ to ‘‘10 ¼ Very HIGH.’’ I find that PT Absent auditors
assessed the level of fraud risk significantly higher in the high-risk condition (LS mean of 5.25) compared to the low-risk condition (LS mean of 3.73,
t35 ¼ 2.66, pone-tailed ¼ 0.029). As discussed later in the paper, I also find that PT Absent auditors assessed the manager’s incentive and opportunity for
fraud to be significantly higher in the high-risk, compared to the low-risk, condition. These findings provide evidence that the PT Absent auditors were
in fact sensitive to fraud risk, suggesting that sensitivity to fraud risk does not cause perspective taking and cannot explain differences in results
between perspective-taking conditions reported in my study.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
68 Hamilton

TABLE 1
H1
Effect of Perspective Taking and Fraud Risk on Misstatement Intentionality Assessments

Panel A: Least Squares Mean (Standard Deviation) of Misstatement Intentionality Assessmentsa

PT Present PT Absent
Low Risk High Risk Low Risk High Risk
(n ¼ 15) (n ¼ 25) (n ¼ 23) (n ¼ 14)
4.43 6.73 4.19 4.28
(2.20) (1.62) (2.32) (2.24)

Panel B: ANCOVA Table—Misstatement Intentionality Assessments

Source df SS F p-value
PT 1 32.14 7.48 0.008
Fraud Risk 1 25.94 6.04 0.016
PT  Fraud Risk 1 20.70 4.82 0.031
Total Facts (covariate) 1 8.51 1.98 0.164
Time (covariate) 1 3.23 0.75 0.389
Error 71 304.94

Panel C: Planned Contrast and Simple Effects

Comparison df F p-valuec
H1 Planned Contrastb 1 22.80 , 0.001
High Risk: PT Present . PT Absent 1 12.38 0.001
Low Risk: PT Present ¼ PT Absent 1 0.11 0.736
PT Present: High Risk . Low Risk 1 11.32 0.001
PT Absent: High Risk ¼ Low Risk 1 0.02 0.893
Table 1 reports descriptive statistics (Panel A) and results of an ANCOVA with misstatement intentionality assessments as the dependent variable,
Perspective Taking (PT) and Fraud Risk as independent variables, and the two effort control variables (Panel B). Panel C presents the planned contrast
used to test H1 and follow-up simple effects.
a
Assessments of misstatement intentionality were measured by asking participants to indicate the likelihood that an identified misstatement is or might be
intentional using an 11-point scale with endpoints labeled ‘‘0 ¼ Not at all Likely’’ and ‘‘10 ¼ Very Likely.’’
b
Contrast coefficients are þ3 for the PT Present/High Risk condition and 1 for the remaining three conditions.
c
Two-tailed test.

Follow-up simple effects are presented in Panel C of Table 2. The planned comparisons related to auditors’ incentive
assessments are consistent with the pattern of cell means predicted in H2a. Specifically, in the High Risk condition, auditors
who considered the client manager’s perspective (PT Present) assess the manager’s incentive for fraud to be significantly
higher than those who did not (PT Absent) (F1,71 ¼ 11.87, p ¼ 0.001). Similar to H1 and consistent with expectations, I do
not find a significant effect of PT in the Low Risk condition (F1,71 ¼ 0.80, p ¼ 0.374). I also find that in the PT Present
condition, auditors assess the client manager’s incentive to be significantly higher in the High Risk versus the Low Risk
condition (F1,71 ¼ 21.91, p , 0.001), although this contrast is also significant in the PT Absent condition (F1,71 ¼ 4.07, p ¼
0.048). Overall, these results provide support for H2a and suggest that when existing circumstances indicate a client manager
benefited from a misstatement, auditors who engage in perspective taking are more likely to recognize the manager’s
increased incentive for fraud.
With respect to auditors’ opportunity assessments, I find that the planned contrast used to test H2b is significant;
however, simple effects tests indicate that the form of the interaction is not consistent with the predicted pattern. Specifically,
simple effects tests presented in Panel C of Table 2 indicate that in the High Risk condition, auditors who considered the
client manager’s perspective (PT Present) did not assess the manager’s opportunity to be higher than auditors who did not
consider the manager’s perspective (PT Absent) (F1,71 ¼ 0.00, p ¼ 0.963). Rather, results suggest that the significance of the
interaction contrast is primarily attributable to a main effect of Fraud Risk, and that effect is slightly stronger in the PT

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 69

FIGURE 2
Results of H1
Interactive Effect of Perspective Taking and Fraud Risk on Misstatement Intentionality Assessments

Figure 2 depicts the results of the test of H1. The graph shows the interaction of Perspective Taking (Absent, Present) and Fraud Risk (Low Risk, High
Risk) on auditors’ misstatement intentionality assessments. Assessments of misstatement intentionality were measured by asking participants to indicate
the likelihood that an identified misstatement is or might be intentional using an 11-point scale with endpoints labeled ‘‘0 ¼ Not at all Likely’’ and ‘‘10 ¼
Very Likely.’’

Present condition (F1,71 ¼ 14.51, p , 0.001) than in the PT Absent condition (F1,71 ¼ 5.32, p ¼ 0.024). Thus, H2b is not
supported.23
Overall, these results suggest that while auditors are equally effective at evaluating a manager’s opportunity to
intentionally misstate regardless of perspective taking, those auditors who considered the manager’s perspective appear to have
a better understanding of how existing circumstances could provide a manager with an increased incentive for fraud.24 This
finding is consistent with the idea that perspective taking provides insight into the motives underlying another’s actions (e.g.,
Kozak et al. 2006).

Supplemental Analyses
Auditors’ Follow-Up Judgments and Procedures
According to the auditing standards, believing that a misstatement might be intentional should impact auditors’
assessments of fraud risk, materiality, and management’s integrity, as well as the performance of subsequent audit procedures
(SEC 1999; AICPA 2002; PCAOB 2010a). Therefore, I perform a supplemental analysis to determine whether auditors who
consider the perspective of client management not only assess the intentionality of an identified misstatement more in line with
the risk factors that are present, but also respond in appropriate ways. Table 3 presents the results of this analysis, which uses
the Bonferroni procedure to adjust p-values (all two-tailed) for multiple comparisons (t-values are not tabulated for

23
Recall that the two opportunity-related risk factors manipulated in the study both related to weaknesses that existed in the control environment. Because
auditors are accustomed to evaluating how a breakdown in controls might impact the risk of material misstatement, risk factors related to controls may
be particularly salient to auditors even in the absence of perspective taking, which may explain why I do not find support for H2b.
24
Although not formally hypothesized, participants were also asked to assess how easily the manager could have rationalized a fraudulent act on a scale
ranging from ‘‘0 ¼ Very Difficult to Rationalize’’ to ‘‘10 ¼ Very Easy to Rationalize.’’ When rationalization is used in place of the incentive and
opportunity dependent variables in my ANCOVA model, I do not find a significant main effect for Perspective Taking (ptwo-tailed ¼ 0.404) or Fraud
Risk (ptwo-tailed ¼ 0.172) or a significant interaction (ptwo-tailed ¼ 0.876).

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
70 Hamilton

TABLE 2
H2a and H2b
Effect of Perspective Taking and Fraud Risk on Incentive and Opportunity Assessments

Panel A: Least Squares Mean (Standard Deviation) of Incentive and Opportunity Assessmentsa
PT Present PT Absent
Low Risk High Risk Low Risk High Risk
(n ¼ 15) (n ¼ 25) (n ¼ 23) (n ¼ 14)
Incentive 3.87 7.21 3.21 4.71
(2.45) (1.58) (2.23) (2.72)
Opportunity 5.65 7.93 6.51 7.96
(1.80) (2.06) (1.56) (1.66)

Panel B: ANCOVA Tables—Incentive Assessment and Opportunity Assessment

Dependent Variable Source df SS F p-value
Incentive PT 1 44.46 9.52 0.003
Fraud Risk 1 105.83 22.67 , 0.001
PT  Fraud Risk 1 14.31 3.06 0.084
Total Facts (covariate) 1 17.84 3.82 0.055
Time (covariate) 1 1.87 0.40 0.529
Error 71 331.49
Opportunity PT 1 3.54 1.08 0.303
Fraud Risk 1 62.44 19.03 , 0.001
PT  Fraud Risk 1 2.97 0.90 0.345
Total Facts (covariate) 1 3.59 1.09 0.299
Time (covariate) 1 1.12 0.34 0.562
Error 71 232.98

Panel C: Planned Contrasts and Simple Effects

Dependent Variable Comparison df F p-valuec
Incentive H2a Planned Contrastb 1 38.19 , 0.001
High Risk: PT Present . PT Absent 1 11.87 0.001
Low Risk: PT Present ¼ PT Absent 1 0.80 0.374
PT Present: High Risk . Low Risk 1 21.91 , 0.001
PT Absent: High Risk ¼ Low Risk 1 4.07 0.048
Opportunity H2b Planned Contrastb 1 7.54 0.008
High Risk: PT Present . PT Absent 1 0.00 0.963
Low Risk: PT Present ¼ PT Absent 1 1.94 0.168
PT Present: High Risk . Low Risk 1 14.51 , 0.001
PT Absent: High Risk ¼ Low Risk 1 5.32 0.024
Table 2 reports descriptive statistics (Panel A) and results of ANCOVAs with incentive assessments and opportunity assessments as the dependent
variables, Perspective Taking (PT) and Fraud Risk as independent variables, and the two effort control variables (Panel B). Panel C of Table 2 presents the
planned contrasts used to test H2a and H2b and follow-up simple effects.
a
Assessments of the manager’s incentive were measured on an 11-point scale with endpoints labeled ‘‘0 ¼ No Incentive (Not at all Beneficial)’’ and ‘‘10 ¼
Strong Incentive (Very Beneficial).’’ Assessments of the manager’s opportunity were measured on an 11-point scale with endpoints labeled ‘‘0 ¼ No
Opportunity (Very Difficult)’’ and ‘‘10 ¼ Strong Opportunity (Very Easy).’’
b
Contrast coefficients are þ3 for the PT Present/High Risk condition and 1 for the remaining three conditions.
c
Two-tailed test.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 71

TABLE 3
Supplemental Analysis
Effect of Perspective Taking and Fraud Risk on Subsequent Audit Judgments and Procedures
PT Present PT Absent
Judgment/Procedure
Least Squares Mean Low Risk High Risk Low Risk High Risk
(Standard Deviation) (n ¼ 15) (n ¼ 25) (n ¼ 23) (n ¼ 14)
Fraud Risk Assessment 4.00a 5.91a 3.73c 5.25c
(1.65) (1.50) (1.45) (2.15)
Materiality Assessment 2.23a 4.49ab 3.56 2.39b
(1.49) (2.43) (2.11) (2.10)
Integrity Assessment 5.28 4.40 5.51 5.43
(1.58) (1.66) (1.41) (0.94)
Investigate Further for Presence of Fraud 5.34a 6.73a 5.49 6.39
(1.68) (0.46) (1.70) (0.65)
Increase Testing of Cash Disbursements 4.92a 6.44a 5.61 6.38
(2.13) (0.92) (1.64) (1.01)
Communicate Concerns to Client Management 3.30a 5.56a 3.89 4.65
(1.49) (1.45) (1.81) (1.82)
Communicate Concerns to Audit Committee 2.69a 4.88a 3.08 3.75
(1.54) (1.45) (1.62) (1.31)
Consult with Forensic Specialist 2.24a 4.13a 3.13 3.10
(1.08) (1.88) (1.63) (1.03)
Table 3 provides descriptive statistics regarding subsequent audit judgments and planned procedures indicated by participants after assessing misstatement
intentionality. Pairwise comparisons were performed between all means listed for each audit judgment/procedure. Fraud Risk (Materiality) Assessments
were collected on an 11-point scale with endpoints labeled ‘‘0 ¼ Very Low (Immaterial)’’ and ‘‘10 ¼ Very High (Material).’’ Integrity Assessment measures
participants’ assessment of the integrity/ethicality of the manager who caused the misstatement and was collected on an 11-point scale with endpoints
labeled ‘‘0 ¼ Not At All Ethical’’ and ‘‘10 ¼ Very Ethical.’’ The remaining subsequent audit procedures were collected on seven-point likelihood scales
with endpoints labeled ‘‘Very Unlikely’’ and ‘‘Very Likely.’’
a,b
For each audit judgment/procedure, two means with the same superscript (a or b) signifies that the means are different at p , 0.05 (two-tailed, adjusted
using the Bonferroni procedure). All analyses include the Total Facts and Time control variables.
c
Means are different at p , 0.10 (two-tailed, adjusted using the Bonferroni procedure).

presentational efficiency).25 With respect to fraud risk and materiality assessments, I find that PT Present auditors assess the
overall level of fraud risk significantly higher in the High Risk condition compared to the Low Risk condition (t38 ¼ 3.51, p ¼
0.005), and they assess the materiality of the misstatement to be significantly higher in the High Risk versus Low Risk
condition (t38 ¼ 3.18, p ¼ 0.013). Importantly, PT Present auditors also assess materiality significantly higher than PT Absent
auditors in the High Risk condition (t35 ¼ 2.92, p ¼ 0.028). In contrast, I do not find a significant difference between Fraud Risk
conditions for PT Absent auditors’ materiality assessments (t35 ¼ 1.58, p ¼ 0.710), but do find a marginally significant
difference for their fraud risk assessments (t38 ¼ 2.66, p ¼ 0.057). I find no differences between any of my experimental
conditions for auditors’ assessment of the manager’s integrity (all p-values . 0.460).26
Finally, to determine whether perspective taking is associated with auditor behavior, I analyze participants’ likelihood
of performing various audit procedures recommended within the auditing standards when a misstatement is believed to be
intentional (SEC 1999; PCAOB 2010a). I find that for each of these procedures, PT Present auditors are significantly more
likely to enact the procedure in the High Risk versus the Low Risk condition. These follow-up procedures include
investigating further for fraud (t38 ¼ 3.38, p ¼ 0.007), increasing testing procedures (t38 ¼ 3.10, p ¼ 0.017), communicating
concerns to client management (t38 ¼ 4.12, p ¼ 0.001), communicating concerns to the audit committee (t38 ¼ 4.48, p ,

25
Fraud Risk (Materiality) Assessments were collected on an 11-point scale with endpoints labeled ‘‘0 ¼ Very Low (Immaterial)’’ and ‘‘10 ¼ Very High
(Material).’’ Integrity Assessment measures participants’ assessment of the integrity/ethicality of the manager who caused the misstatement and was
collected on an 11-point scale with endpoints labeled ‘‘0 ¼ Not At All Ethical’’ and ‘‘10 ¼ Very Ethical.’’
26
The finding that auditors’ integrity assessments do not differ by condition may be because auditors typically evaluate the integrity of management as a
whole. Thus, when asked to evaluate the integrity of an individual manager, participants may have anchored on the case’s description of company
management as ‘‘generally appear(ing) to have high integrity.’’ Nonetheless, I do find a significant, negative correlation between auditors’ intentionality
assessments and their integrity assessments (r ¼0.59, p , 0.001, untabulated), suggesting that auditors believe a manager’s integrity is compromised
when he is suspected of intentionally misstating the financials.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
72 Hamilton

0.001), and consulting with a forensic specialist (t38 ¼ 3.79, p ¼ 0.002). For PT Absent auditors, I find no significant
differences in the likelihood of performing these procedures between the High Risk and Low Risk conditions (all p-values
. 0.200). These results suggest that auditors who consider the perspective of management are more likely to investigate an
identified misstatement further for evidence of fraud when existing circumstances indicate that it is appropriate to do so
(i.e., in the High Risk condition).27 The fact that PT Absent auditors’ responses were unchanged, regardless of the level of
fraud risk, suggests that they may over-audit in the presence of low fraud risk or under-audit in the presence of high fraud
risk (or both).

Alternative Explanation for the Exploratory Findings

Earlier in the paper I provided evidence, based on an exploratory analysis, that a perspective-taking mindset is activated
more frequently in experienced auditors when they encounter circumstances indicative of high fraud risk as compared to low
fraud risk. However, with the measured variable approach, it is possible that the measure of Perspective Taking is actually
capturing a greater propensity of auditors to document their consideration of the manager’s perspective in the high-risk
condition. That is, an alternative explanation for the greater number of auditors classified as PT Present in the high-risk
condition may be that auditors considered the manager’s perspective equally in both fraud risk conditions, but felt no need to
document this consideration in the presence of low risk, resulting in a greater number of auditors being classified as PT Present
in the high-risk condition.
To test this alternative explanation, it is necessary to separate engagement in perspective taking from documentation of
perspective taking. For the alternative explanation to be true, individuals engaged in perspective taking would have to exhibit a
lower propensity to document their consideration of the manager’s perspective in the low-risk condition compared to the high-
risk condition. To test this assertion, I perform a follow-up experiment with students, consisting of 32 masters of accounting
students and ten undergraduate students who had completed an accounting internship. Students were enrolled in either a
graduate level auditing course or a fraud course at two major state universities. I chose to use student participants because they
are more likely to follow simple instructions telling them to consider the manager’s perspective than are experienced auditors
(i.e., they are more likely to engage in perspective taking when prompted to do so), and engaging in perspective taking is a
critical component of this follow-up experiment. I provided the students with the same basic case facts that were previously
given to auditor participants, and I manipulated fraud risk as either high or low. To elicit perspective taking in all participants,
everyone was provided with perspective-taking instructions. To further ensure that the student participants followed the
instructions and engaged in perspective taking, regardless of their fraud risk condition, I used a very strong perspective-taking
manipulation.28 These departures from my original experiment (the use of student participants and a stronger perspective-taking
manipulation) should help ensure that all participants engage in perspective taking, thus allowing me to test whether the
propensity to document perspective taking differs between the high- and low-risk conditions once all participants have been
placed in a perspective-taking mindset.29
To provide evidence that participants were considering the manager’s perspective evenly between fraud risk conditions, I
asked participants the extent to which they evaluated the misstatement from the manager’s perspective on an 11-point scale that
ranged from ‘‘0 ¼ I only considered my own perspective’’ to ‘‘10 ¼ I only considered (the manager’s) perspective.’’ The overall
mean and median response was 7.0, which is significantly above the scale midpoint (t41 ¼ 6.72, ptwo-tailed , 0.001, untabulated).
Most importantly, this measure did not vary between the high-fraud-risk (mean of 7.0) and low-fraud-risk conditions (mean of
6.9, t40 ¼ 0.25, p ¼ 0.805, untabulated). Therefore, the consideration of management’s perspective appears to be equally
balanced between fraud risk conditions.
Having established that engagement in perspective taking is the same in both fraud risk conditions, I can now analyze
whether there is a lower propensity to document perspective taking considerations in the low-risk condition. To do this, I
asked and analyzed the same open-ended response used to create the measure of perspective taking in the primary auditor

27
Although the subsequent audit actions presented here are not specific enough to determine whether auditors would ultimately detect a specific fraud,
they provide evidence regarding whether or not auditors would simply deem an identified misstatement to be immaterial without taking further action
(even when risk is high), as appears to be the case for those auditors classified as PT Absent.
28
Specifically, students were told to ‘‘imagine yourself in [the manager’s] shoes and imagine that you are considering the perpetration of a fraudulent act.
Specifically, imagine that you are considering whether you should classify a payment related to a contractual penalty as a prepaid asset, rather than
appropriately classifying it as an expense. Below is a list of circumstances that existed at the time the misstatement occurred. Read and evaluate each
piece of information as if you are in [the manager’s] shoes.’’ After participants read through the case facts, they were again reminded of the perspective-
taking instructions by being told, ‘‘Once you have obtained an understanding of how the misstatement occurred from the perspective of [the manager],
please proceed to the next page.’’
29
Although these departures from my original experiment help make it possible to rule out the alternative explanation, they also limit the generalizability
of my findings. Thus, it is possible that the results obtained in the follow-up experiment may not generalize to auditor participants.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 73

study. I find that there is no difference in the frequency with which participants discuss the manager’s perspective in the
high-fraud-risk condition (56.5 percent) compared to the low-fraud-risk condition (47.4 percent) (v21 ¼ 0.35, p ¼ 0.554,
untabulated).30 That is, when the actual consideration of management’s perspective is equivalent between fraud risk
conditions, I find no evidence of a lower propensity to document this consideration in the low-risk, compared to the high-
risk, condition. In sum, this supplemental experiment provides evidence that differences in the measure of perspective
taking between the high- and low-risk conditions stem from differences in auditors’ propensity to consider, rather than
simply discuss, management’s perspective.

Information Considered in the Absence of Perspective Taking

One of the more interesting findings in my study was that auditors who did not consider management’s perspective (PT
Absent auditors) assessed the intentionality of an identified misstatement the same regardless of whether the circumstances
surrounding the misstatement were indicative of high fraud risk or low fraud risk. Given that these auditors did not respond to
the red flags present in the high-fraud-risk condition, the question then becomes, what information did the PT Absent auditors
use when evaluating the identified misstatement, particularly in the high-fraud-risk condition? One possibility is that these
auditors focused their attention on information related to the impact of the misstatement on the company and its management as
a whole, while failing to consider information that indicated the client manager who caused the misstatement had an incentive
and opportunity to do so intentionally.31
To explore this possibility, I analyzed participants’ responses to a question asking them to list the factors they considered
when evaluating the intentionality of the identified misstatement. I restrict the analysis to the high-fraud-risk condition, as this is
the condition of interest (i.e., the condition in which PT Absent auditors failed to increase their intentionality assessments
despite the presence of high-risk cues). I find that in the high-fraud-risk condition, PT Absent auditors were significantly more
likely to discuss the fact that the misstatement did not benefit the company as a whole (i.e., did not allow the company to meet a
bonus target or analyst earnings target or mask a change in earnings) compared to PT Present auditors (64.3 percent of PT
Absent auditors discussed these factors compared to 12.0 percent of PT Present auditors, v21 ¼ 11.52, ptwo-tailed ¼ 0.001,
untabulated).32 The PT Absent auditors were also significantly less likely to discuss one or more of the four red flags seeded
into the high-risk condition that related to the incentives and opportunities of the manager responsible for the misstatement
(71.4 percent of PT Absent auditors discussed at least one of the red flags compared to 100 percent of PT Present auditors, v21 ¼
7.96, ptwo-tailed ¼ 0.005, untabulated). Furthermore, PT Absent auditors listed significantly fewer of the four seeded red flags
compared to PT Present auditors (means of 1.14 and 2.20, respectively, t37 ¼ 3.53, ptwo-tailed ¼ 0.001, untabulated). Based on
these findings, it would appear that in the high-risk condition, PT Absent auditors focused their evaluations more on factors that
indicated the misstatement did not benefit the company and its management as a whole, while overlooking more nuanced risk
factors, which suggested the manager responsible for the misstatement had an incentive and opportunity to intentionally
misstate.

DISCUSSION AND CONCLUSION

Fraud detection continues to be an important consideration for auditors. A failure to detect fraud has serious
consequences for both the audit firm and the general public (Beasley, Carcello, Hermanson, and Neal 2010). As such, it is
important that auditors continue to improve their fraud detection skills. In the present study, I examined whether
considering the perspective of the client manager responsible for a misstatement’s occurrence influences auditors’
sensitivity to the level of fraud risk surrounding the misstatement and, in turn, the belief that the misstatement was
intentional.
The results of this study suggest that by considering management’s perspective, auditors are more likely to recognize
when the circumstances surrounding an identified misstatement indicate that it was caused intentionally. When presented
with information indicating that the client manager benefited from the misstatement (i.e., high fraud risk), auditors who
considered the manager’s perspective assessed the manager’s incentive for fraud significantly higher compared to auditors

30
Results are qualitatively unchanged when I limit the analysis to include only the 32 graduate auditing students.
31
Specifically, participants may have focused on information indicating that the misstatement did not allow the company to meet a bonus target or analyst
earnings target, and did not mask a change in earnings or other trends (i.e., information about qualitatively materiality). As discussed previously in the
paper, this information was intended to suggest that the company (and management) as a whole did not benefit from the misstatement. While such
information certainly should be considered by auditors, it should not be the primary focus in the high-risk condition, given the presence of other
information, suggesting the misstatement had been caused intentionally.
32
It is important to note that PT Present auditors only appear to discount these qualitative materiality factors in the high-risk condition (i.e., when other
high-risk cues are present). In the low-risk condition, there is no difference in the frequency with which auditors list the qualitative materiality factors
(v21 ¼ 1.30, ptwo-tailed ¼ 0.254, untabulated).

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
74 Hamilton

who had not considered the manager’s perspective. Most importantly, I find that when the circumstances surrounding a
misstatement were indicative of high fraud risk, auditors who considered the perspective of client management assessed the
intentionality of the misstatement to be higher than those who did not use perspective taking. In contrast, auditors who did
not consider management’s perspective did not assess intentionality differently when factors surrounding a misstatement
were indicative of high or low fraud risk. These results suggest that auditors who fail to consider management’s perspective
may not recognize the level of fraud risk surrounding an identified misstatement and, in turn, may fail to investigate a
misstatement further when doing so might be warranted by the circumstances. Indeed, the supplemental analysis in this
study provides support for this potential implication. Specifically, I find that when circumstances surrounding the
misstatement were indicative of high versus low fraud risk, only the auditors who had considered the client manager’s
perspective were significantly more likely to perform various follow-up procedures, such as increasing testing procedures
and communicating concerns to client management and the audit committee.
It is interesting to note, however, that auditors who did not consider the perspective of the manager responsible for the
misstatement still recognized that this manager had a greater incentive and opportunity for fraud in the high-risk, compared to
the low-risk, condition; nonetheless, their assessments of intentionality and responses to the misstatement remained unchanged.
In contrast, auditors who did engage in perspective taking not only recognized the manager’s increased incentive and
opportunity in the high-risk condition, but also the implications of this (i.e., that the misstatement was more likely to be
intentional and further actions should be taken to investigate the misstatement further). In conclusion, my results suggest that
perspective taking helps auditors evaluate the intentionality of identified misstatements primarily by helping auditors
understand how a client manager’s intentions (and subsequent actions) might be impacted by the presence of high-fraud-risk
circumstances.
Although my study offers a number of important insights, certain limitations also exist. First, the use of perspective taking
relies on the evaluation of fraud risk factors that are observable to the auditor. To the extent that a client manager’s fraudulent
intentions are influenced by factors that are unobservable to the auditor (e.g., personal financial troubles), the utility of
perspective taking might be reduced. Second, although the study provides participants with a number of information items to
evaluate, participants received less information than what is normally available in the field. Future research can explore whether
a more cognitively demanding information environment inhibits perspective taking in experienced auditors. Finally, I rely on a
measure of perspective taking, rather than a manipulation, which makes it more difficult to draw causal inferences from my
results. As discussed previously, I include a number of controls in my analyses to diminish concerns that participants’ effort or
experience may be driving the results.
My study provides several avenues for future research. First, it remains unclear whether auditors can be trained to use
perspective taking in practice. Future research should investigate the efficacy of various instructional or training
mechanisms that can induce perspective taking in experienced auditors. Second, I find that the presence of high-fraud-risk
circumstances prompted many of my auditor participants to consider management’s perspective spontaneously.
Nonetheless, not all of the highly experienced auditors in my study considered the client manager’s perspective even
when faced with circumstances indicative of high fraud risk. Future research can explore the factors that facilitate
perspective taking in experienced auditors as well as the factors that may inhibit it. For instance, some auditors may possess
a predisposition toward perspective taking, making it more likely that they will spontaneously consider events and
circumstances from alternative points of view. Future research can determine whether auditors who possess such a
predisposition outperform auditors who do not and the types of tasks that may benefit from a perspective taking mindset.
Finally, research can explore how various characteristics of auditors (e.g., the extent to which they identify with the client),
audit tasks (e.g., complexity), and the audit environment (e.g., accountability) interact with perspective taking to influence
audit outcomes.

REFERENCES
Altiero, E., Y. J. Kang, and M. E. Peecher. 2013. The Investor Perspective and Its Influence on Auditor Materiality Judgments. Working
paper, University of Illinois.
American Institute of Certified Public Accountants (AICPA). 2002. Consideration of Fraud in a Financial Statement Audit. Statement on
Auditing Standards No. 99. New York, NY: AICPA.
Asare, S. K., and A. M. Wright. 2004. The effectiveness of alternative risk assessment and program planning tools in a fraud setting.
Contemporary Accounting Research 21 (2): 325–352.
Batson, C. D., S. Early, and G. Salvarani. 1997. Perspective taking: Imagining how another feels versus imagining how you would feel.
Personality and Social Psychology Bulletin 23: 751–758.
Beasley, M. S., J. V. Carcello, D. R. Hermanson, and T. L. Neal. 2010. Fraudulent Financial Reporting 1998–2007: An Analysis of U.S.
Public Companies. New York, NY: Committee of Sponsoring Organizations of the Treadway Commission.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 75

Beasley, M. S., J. V. Carcello, D. R. Hermanson, and T. L. Neal. 2013. An Analysis of Alleged Auditor Deficiencies in SEC Fraud
Investigations: 1998–2010. Available at: http://www.thecaq.org/analysis-alleged-auditor-deficiencies-sec-fraud-investigations-
1998%E2%80%932010
Braun, R. L. 2000. The effect of time pressure on auditor attention to qualitative aspects of misstatements indicative of potential fraudulent
financial reporting. Accounting, Organizations and Society 25: 243–259.
Carpenter, T. D. 2007. Audit team brainstorming, fraud risk identification, and fraud risk assessment: Implications of SAS No. 99. The
Accounting Review 82 (5): 1119–1140.
Chalmers, J. B., and M. A. Townsend. 1990. The effects of training in social perspective taking on socially maladjusted girls. Child
Development 61:178–190.
Church, B. K., M. Peytcheva, W. Yu, and O. Singtokul. 2015. Perspective taking in auditor-manager interactions: An experimental
investigation of auditor behavior. Accounting, Organizations and Society 45: 40–51.
Cohen, J. 1960. A coefficient of agreement for nominal scales. Educational and Psychological Measurement 20 (1): 37–46.
Coutu, W. 1951. Role-playing vs. role-taking: An appeal for clarification. American Sociological Review 16 (2): 180–187.
Davis, M. H. 1983. Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and
Social Psychology 44 (1): 113–126.
Davis, M. H., L. Conklin, A. Smith, and C. Luce. 1996. Effect of perspective taking on the cognitive representation of persons: A merging
of self and other. Journal of Personality and Social Psychology 70 (4): 713–726.
Epley, N., and E. M. Caruso. 2009. Perspective taking: Misstepping into others’ shoes. In The Handbook of Imagination and Mental
Simulation, edited by K. D. Markman, W. M. P. Klein, and J. A. Suhr, 295–309. New York, NY: Psychology Press.
Eyal, T., N. Liberman, and Y. Trope. 2008. Judging near and distant virtue and vice. Journal of Experimental Social Psychology 44:
1204–1209.
Fleiss, J. L. 1981. Statistical Methods for Rates and Proportions. 2nd edition. New York, NY: John Wiley & Sons.
Galinsky, A. D., and G. B. Moskowitz. 2000. Perspective-taking: Decreasing stereotype expression, stereotype accessibility, and in-group
favoritism. Journal of Personality and Social Psychology 78 (4): 708–724.
Galper, R. E. 1976. Turning observers into actors: Differential causal attributions as a function of ‘‘empathy.’’ Journal of Research in
Personality 10: 328–335.
Gesn, P. R., and W. Ickes. 1999. The development of meaning contexts for empathic accuracy: Channel and sequence effects. Journal of
Personality and Social Psychology 77 (4): 746–761.
Gilbert, D. T. 2002. Inferential correction. In Heuristics and Biases: The Psychology of Intuitive Judgment, edited by T. Gilovich, D.
Griffin, and D. Kahneman, 167184. New York, NY: Cambridge University Press.
Gilbert, D. T., and E. E. Jones. 1986. Perceiver-induced constraints: Interpretation of self-generated reality. Journal of Personality and
Social Psychology 50: 269–280.
Hammersley, J. S. 2006. Pattern identification and industry-specialist auditors. The Accounting Review 81 (2): 309–336.
Hammersley, J. S. 2011. A review and model of auditor judgments in fraud-related planning tasks. Auditing: A Journal of Practice &
Theory 30 (4): 101–128.
Hammersley, J. S., E. M. Bamber, and T. D. Carpenter. 2010. The influence of documentation specificity and priming on auditors’ fraud
risk assessments and evidence evaluation decisions. The Accounting Review 85 (2): 547–571.
Hoever, I. J., D. van Knippenberg, W. P. van Ginkel, and H. G. Barkema. 2012. Fostering team creativity: Perspective taking as key to
unlocking diversity’s potential. Journal of Applied Psychology 97 (5): 982–996.
Hoffman, V. B., and M. F. Zimbelman. 2009. Do strategic reasoning and brainstorming help auditors change their standard audit
procedures in response to fraud risk? The Accounting Review 84 (3): 811–837.
Hurtt, R. K. 2010. Development of a scale to measure professional skepticism. Auditing: A Journal of Practice & Theory 29 (1): 149–171.
Iannotti, R. J. 1978. Effect of role-taking experiences on role taking, empathy, altruism, and aggression. Developmental Psychology 14:
119–124.
Ickes, W., L. Stinson, V. Bissonnette, and S. Garcia. 1990. Naturalistic social cognition: Empathic accuracy in mixed-sex dyads. Journal
of Personality and Social Psychology 59 (4): 730–742.
Jones, E. E., and R. E. Nisbett. 1971. The Actor and the Observer: Divergent Perceptions of the Causes of Behavior. Morristown, NJ:
General Learning Press.
Kozak, M. N., A. A. Marsh, and D. M. Wegner. 2006. What do I think you’re doing? Action identification and mind attribution. Journal
of Personality and Social Psychology 90 (4): 543–555.
Landis, J. R., and G. G. Koch. 1977. The measurement of observer agreement for categorical data. Biometrics 33: 159–174.
Leith, K. P., and R. F. Baumeister. 1998. Empathy, shame, guilt, and narratives of interpersonal conflicts: Guilt-prone people are better at
perspective taking. Journal of Personality 66 (1): 1–37.
Libby, R. 1985. Availability and the generation of hypotheses in analytical review. Journal of Accounting Research 23 (2): 648–667.
Loebbecke, J. K., M. M. Eining, and J. J. Willingham. 1989. Auditors’ experience with material irregularities: Frequency, nature, and
detectability. Auditing: A Journal of Practice & Theory 9 (Fall): 1–28.
Mallett, R. K., J. R. Huntsinger, S. Sinclair, and J. K. Swim. 2008. Seeing through their eyes: When majority group members take
collective action on behalf of an outgroup. Group Processes & Intergroup Relations 11 (4): 451–470.

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
76 Hamilton

McPherson Frantz, C., and R. Janoff-Bulman. 2000. Considering both sides: The limits of perspective taking. Basic and Applied Social
Psychology 22 (1): 31–42.
Mendelsohn, M., and G. Straker. 1999. Social perspective taking and the use of discounting in children’s perceptions of others’ helping
behavior. The Journal of Genetic Psychology: Research and Theory on Human Development 160 (1): 69–83.
Phillips, F. 1999. Auditor attention to and judgments of aggressive financial reporting. Journal of Accounting Research 37 (1): 167–189.
Public Company Accounting Oversight Board (PCAOB). 2007. Observations on Auditors’ Implementation of PCAOB Standards
Relating to Auditors’ Responsibilities with Respect to Fraud. Available at: http://pcaobus.org/Inspections/Documents/2007_01-22_
Release_2007-001.pdf
Public Company Accounting Oversight Board (PCAOB). 2010a. Evaluating Audit Results. Auditing Standard No. 14. New York, NY:
PCAOB.
Public Company Accounting Oversight Board (PCAOB). 2010b. Identifying and Assessing Risks of Material Misstatement. Auditing
Standard No. 12. New York, NY: PCAOB.
Public Company Accounting Oversight Board (PCAOB). 2012. Maintaining and Applying Professional Skepticism in Audits. Staff Audit
Practice Alert No. 10. Available at: http://pcaobus.org/Standards/QandA/12-04-2012_SAPA_10.pdf
Regan, D. T., and J. Totten. 1975. Empathy and attribution: Turning observers into actors. Journal of Personality and Social Psychology
32 (5): 850–856.
Ross, L. 1977. The intuitive psychologist and his shortcomings: Distortions in the attribution process. Advances in Experimental Social
Psychology 10: 173–220.
Ross, L., and R. E. Nisbett. 1991. The Person and the Situation: Perspectives of Social Psychology. New York, NY: McGraw-Hill.
Securities and Exchange Commission (SEC). 1999. Materiality. SEC Staff Accounting Bulletin No. 99. (August 12). Washington, DC:
SEC.
Simon, C. A. 2012. Individual auditors’ identification of relevant fraud schemes. Auditing: A Journal of Practice & Theory 31 (1): 1–16.
Trompeter, G. M., T. D. Carpenter, K. L. Jones, and R. A. Riley, Jr. 2014. Insights for research and practice: What we learn about fraud
from other disciplines. Accounting Horizons 28 (4): 769–804.
Trompeter, G. M., T. D. Carpenter, N. Desai, K. L. Jones, and R. A. Riley, Jr. 2013. A synthesis of fraud-related research. Auditing: A
Journal of Practice & Theory 32 (Supplement): 287–321.
Trotman, K. T., A. Wright, and S. Wright. 2005. Auditor negotiations: An examination of the efficacy of intervention methods. The
Accounting Review 80 (1): 349–367.
Trotman, K., R. Simnett, and A. Khalifa. 2009. Impact of the type of audit team discussions on auditors’ generation of material frauds.
Contemporary Accounting Research 26 (4): 1115–1142.
Vescio, T. K., G. B. Sechrist, and M. P. Paolucci. 2003. Perspective taking and prejudice reduction: The mediational role of empathy
arousal and situational attributions. European Journal of Social Psychology 33: 455–472.
Wilks, T. J., and M. F. Zimbelman. 2004a. Decomposition of fraud-risk assessments and auditors’ sensitivity to fraud cues. Contemporary
Accounting Research 21 (3): 719–745.
Wilks, T. J., and M. F. Zimbelman. 2004b. Using game theory and strategic reasoning concepts to prevent and detect fraud. Accounting
Horizons 18 (3): 173–184.

APPENDIX A
Presented below are excerpts from my experimental instrument, including an overview of the identified misstatement,
followed by my manipulations.

Misstatement Overview
Joe Rogers is the primary individual responsible for purchasing decisions at Green Division (a division of C&P). The
identified misstatement was the result of inaccurate information provided by Joe Rogers in a Payment Request form. In the
form, he classified a payment being made to a supplier as a prepayment for future purchases. The payment actually related to a
one-time penalty triggered by Green Division’s failure to purchase the minimum amount of product specified within a Supply
Agreement. As a result of the inaccurate payment classification, the accounting department recorded the payment as a prepaid
asset rather than an expense. The misclassification was not identified during the review process, in part because Joe Rogers did
not include documentation to support the purpose of the payment. C&P management believes that, ‘‘The payment was
incorrectly classified as a prepayment by Joe Rogers because prepayments are a frequent occurrence with many of our
suppliers. However, we rarely incur a penalty for missing a minimum purchasing target. As such, in filling out the Payment
Request form, Joe Rogers accidentally selected a prepayment classification rather than selecting the appropriate expense
classification, and unfortunately this mistake was not caught during the review process.’’

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
Evaluating the Intentionality of Identified Misstatements: How Perspective Can Help in Distinguishing Errors from Fraud 77

Manipulation of Perspective Taking Instructions (Absent versus Present)

Note: All participants received an excerpt from the audit program. The perspective-taking manipulation appeared within the
third step of the audit program, directly following the instructions to ‘‘consider whether each misstatement is or might be
intentional.’’

Perspective Taking Absent

In considering whether a misstatement might be intentional, evaluate the facts and circumstances related to the
misstatement as you normally would in practice.

Perspective Taking Present

In considering whether a misstatement might be intentional, evaluate the facts and circumstances related to the
misstatement by using the following evaluation process required by your firm:
Think from the perspective of the client-individual responsible for the misstatement. Put yourself in the place of this client-
individual and try to imagine what you would think and how you would feel about the act of misstating under the
circumstances faced by the client.

Manipulation of Fraud Risk (High Risk versus Low Risk)

Note: The High Risk and Low Risk manipulations were seeded within additional information provided to participants. The first
two manipulations vary the extent to which Joe Rogers had an incentive to misstate and the final two vary the extent to which
he had an opportunity to misstate. The differences in wording between conditions are in bold font (they were not bolded in the
actual experimental instrument).

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016
78 Hamilton

Auditing: A Journal of Practice & Theory

Volume 35, Number 4, 2016