audit, as reflected in professional auditing standards (e.g., IFAC 2012; ... While there is no universally accepted definition of professional skepticism (see, e.g.,.
Auditors’ Professional Skepticism: Neutrality versus Presumptive Doubt* LUC QUADACKERS, VU University Amsterdam TOM GROOT, VU University Amsterdam ARNOLD WRIGHT, Northeastern University
1. Introduction Professional skepticism is considered to be an essential element of the financial statement audit, as reflected in professional auditing standards (e.g., IFAC 2012; PCAOB 2008) and the audit methodologies of international audit firms.1 The academic and professional auditing literatures also emphasize the importance of the use of professional skepticism (see, e.g., Kadous 2000; Mautz and Sharaf 1961; Nelson 2009; Financial Reporting Council 2010). In addition, analyses of fraud-related U.S. Securities and Exchange Commission (SEC) cases conclude that a lack of sufficient professional skepticism is often cited as the reason that auditors fail to detect material misstatements (e.g., Beasley, Carcello, and Hermanson 2001; Public Oversight Board [POB] 2000; Benston and Hartgraves 2002). While there is no universally accepted definition of professional skepticism (see, e.g., Hurtt 2010; Nelson 2009; Doucet and Doucet 1996), two perspectives have emerged in the current literature and auditing standards: neutrality and presumptive doubt (see, e.g., Nelson 2009). Neutrality refers to a perspective in which the auditor assumes no bias in management’s representations, ex ante (e.g., Nelson 2009). Presumptive doubt represents an auditor’s attitude in which some level of dishonesty or bias by management is assumed, unless evidence indicates otherwise (POB 2000; Bell, Peecher, and Solomon 2005). Importantly, there is a lack of consensus on which of these two perspectives of skepticism is most appropriate for audit practice. The purpose of this study is to examine the relationship between auditors’ skeptical perspective (neutrality and presumptive doubt) and auditor skeptical judgments and decisions across client risk settings (higher versus lower control environment risk). Neutrality is measured by the Hurtt Professional Skepticism Scale (HPSS; Hurtt 2010), and presumptive doubt is measured by the inverse of the Rotter Interpersonal Trust Scale (RIT; Rotter *
1.
Accepted by Hun-Tong Tan. We are indebted to the Big 4 auditing firm that participated in this study. Furthermore, we are grateful for the comments received at the ARNN Accounting Symposium 2007 in Leuven, the EAA Annual Congress 2008 in Rotterdam, the International Symposium on Audit Research 2008 in Pasadena, the AAA Annual Meeting in 2008, and the IAAER Conference 2012 in Amsterdam. In addition we would like to thank the participants at the research seminars at Bentley College, Northeastern University and Maastricht University, as well as the anonymous reviewers and associate editor, for their valuable comments and suggestions. Finally, we are grateful to those who helped in developing the research materials, in coding, and in analyzing the results. The detailed genesis of the concept of “professional skepticism” in the auditing standards is not determinable. The technical managers of the Audit and Attest Standards of the American Institute of Certified Public Accountants (AICPA) checked the archives but were unable to find information on this issue. However, it appears to have been first used in professional standards in SAS No. 16 (1977, the predecessor standard to SAS No. 53). However, prior to this, skepticism is considered in the academic literature. For instance, Mautz and Sharaf (1961) discuss the importance of skepticism in auditing.
Contemporary Accounting Research Vol. 31 No. 3 (Fall 2014) pp. 639–657 © CAAA doi:10.1111/1911-3846.12052
640
Contemporary Accounting Research
1967). Since RIT measures an individual’s level of trust, we utilize the inverse to reflect skepticism or a lack of trust (inversed RIT). These measures reflect two auditor personality traits with respect to skepticism. In his model of professional skepticism, Nelson (2009) identifies such individual traits as an important determinant of auditor behavior. We hypothesize that both perspectives may drive auditor skeptical judgments and decisions, but in the higher-risk situation presumptive doubt will be more predictive of auditor behaviors. This expectation implies presumptive doubt is more closely related to auditor skeptical behaviors than neutrality, especially in higher-risk settings where professional standards prescribe greater skeptical judgments and actions. In sum, in this study auditor skepticism is examined based on the relationships between two widely recognized trait measures of skepticism (i.e., HPSS and [inversed] RIT), representing two skeptical perspectives, a situational risk factor (control environment risk), and auditors’ planning judgments and decisions. This study presents the first empirical evidence, to our knowledge, of the link between the two predominant perspectives of skepticism in auditing (neutrality and presumptive doubt) and auditors’ professional skeptical judgments and decisions. Knowing how the two main skeptical perspectives are related to auditor judgments and decisions across risk situations is important for considering which perspective is optimal in addressing client risks. Ninety-six auditors participated in an experiment, in which they performed an analytical procedures task and indicated how they would respond to an unexpected fluctuation in the financial statements, in terms of risk assessments and audit program planning. In addition, they completed instruments measuring RIT and HPSS. As hypothesized, the findings show that inversed RIT, reflecting presumptive doubt, is more predictive than HPSS of auditors’ skeptical judgments and decisions across risk scenarios, but particularly in the higher-risk setting. This result suggests that the presumptive doubt perspective of professional skepticism is more predictive than neutrality of auditor skeptical judgments and decisions in higher-risk situations. The remainder of the paper is organized as follows. In section 2, the theory, prior literature, and the hypothesis are described. The research method is discussed in section 3, and the results are presented in section 4. The final section provides a discussion of the findings and their implications for future research and practice. 2. Theory, literature and hypothesis Neutrality and presumptive doubt have emerged as the predominant perspectives of auditor professional skepticism. In support of a focus on neutrality, Cushing (2000) states that auditors should attempt to be unbiased in forming their beliefs; there should be no bias in either a positive (“trusting”) or negative (“suspicious”) direction. Nelson (2009) argues that this is the main focus of current auditing standards. Under this perspective, evidence needs to be sought and evaluated by the auditor to confirm management’s assertions but also to rule out alternative explanations. This focus essentially fits the “trust but verify” principle. In contrast, presumptive doubt takes a different view (e.g., POB 2000; Bell et al. 2005). Bell et al. (2005) assert that an auditor with presumptive doubt assumes some level of dishonesty or bias by management, unless evidence indicates otherwise. This view appears consistent with that of forensic auditors (POB 2000). The Panel on Audit Effectiveness recommends that auditors adopt this perspective (POB 2000) and proposes a “forensic-type fieldwork” phase of the audit. From this perspective, skeptical auditors will focus more on error-related evidence than on nonerrors, such as changes in business conditions (cf. McMillan and White 1993; Smith and Kida 1991). The presumptive doubt perspective is most visible in the auditing standards concerning fraud (e.g., IFAC: ISA CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
641
240 2012), since these standards focus on the possibility of an intentional material misstatement.2 Both presumptive doubt and neutrality measure an auditor’s trait or tendency to exhibit skeptical behavior. Prior research in psychology has found that dispositional characteristics, or traits, influence judgments and decisions (e.g., Eagly and Chaiken 1993, 2005; Ajzen 2005). In the setting of professional skepticism, Nelson (2009) identifies an auditor’s skeptical traits as an important factor. Hence, theory implies that auditors with a more skeptical disposition will exhibit more skeptical judgments and decisions (e.g., suspend judgment and engage in more substantive testing) than auditors with a less skeptical disposition. Thus, under either the neutrality or presumptive doubt perspective, an auditor who has a trait toward greater skepticism would generally want more evidence to be persuaded sufficiently than one who has less. However, with greater presumptive doubt, the auditor would focus on evidence of material misstatements or fraud. In contrast, with greater neutrality, the auditor would seek a balance of greater amounts of evidence supporting as well as refuting the financial statement assertions. Whereas neutrality appears to be the basic perspective underlying most of the auditing standards (e.g., Nelson 2009), Bell et al. (2005) argue that societal expectations of auditors have moved from the neutral stance more toward the presumptive doubt perspective of professional skepticism, due to economic downturns and major business improprieties. This change is reflected in the Nelson (2009, 1) definition of professional skepticism (PS): “indicated by auditor judgments and decisions that reflect a heightened assessment of the risk that an assertion is incorrect, conditional on the information available to the auditor.” He states that “this definition reflects more of a “presumptive doubt” than a “neutral” view of PS, implying that auditors who exhibit high PS need relatively more persuasive evidence (in terms of quality and/or quantity) to be convinced that an assertion is correct.” According to Nelson’s (2009) model of professional skepticism, traits constitute an important set of determinants of skeptical judgments and actions, along with incentives, knowledge, and audit experience and training. Nelson defines traits as auditor’s nonknowledge attributes that can influence professional skepticism. Consistent with Libby and Luft (1993), Nelson views traits as individual characteristics that are stable by the time an auditor commences audit training and practice. Nelson defines three categories of traits related to professional skepticism: problemsolving ability, ethics/moral reasoning, and skepticism. In this paper we concentrate on two widely cited scales that measure skepticism: (the inverse of) Rotter’s Interpersonal Trust Scale (RIT) and the Hurtt Professional Skepticism Scale (HPSS). In general, dispositional skeptical characteristics (i.e., skeptical traits) of auditors are expected to be predictive of auditors’ skeptical judgments and decisions (e.g., Nelson 2009). For example, if an auditor in general has a suspicious nature, this is likely to lead to more skeptical judgments and decisions (e.g., Shaub 1996). The presumptive doubt perspective and interpersonal trust This study is the first one in auditing to use the Rotter Interpersonal Trust Scale (Rotter 1967) to capture auditor skepticism. This scale is widely accepted (see, e.g., Hoell 2004; Johnson-George and Swap 1982; Stack 1978; Webb and Worchel 1986). Interpersonal 2.
According to the IAASB document “Staff Questions & Answers—Professional Skepticism in an Audit of Financial Statements” (2012), the ISAs in some cases “require the auditor to make presumptions about risks of fraud, the assessment of risks of material misstatement, or specify procedures that are required to be performed. They do so recognizing the importance of professional skepticism in areas where it has been demonstrated that they are more susceptible to misstatement, including misstatement due to fraud.” For example, paragraphs 26 and 27 of ISA 240 discuss the required presumption that there are risks of fraud in revenue recognition.
CAR Vol. 31 No. 3 (Fall 2014)
642
Contemporary Accounting Research
trust can be defined as “a generalized expectancy held by an individual or a group that the word, promise, verbal or written statement of another individual or group can be relied upon” (Rotter 1967, 651). The basic thought is that if an auditor has a lower level of interpersonal trust, he or she is assumed to be more skeptical (e.g., Shaub 1996; Hurtt 2010). Thus, the inverse of interpersonal trust (i.e., lack of trust) is directly related to the presumptive doubt perspective of professional skepticism (e.g., Nelson 2009). In nonauditing studies, Rotter’s Interpersonal Trust Scale has been associated more strongly with actual behaviors than other interpersonal trust scales, such as the trustworthiness part of Wrightsman’s Philosophies of Human Nature Scale (Stack 1978; Rotter 1980). Accordingly, this study examines the influence of inversed RIT, as a measure of presumptive doubt, on auditors’ professional judgments and decisions. Previous auditing studies using trust scales (other than RIT) have reported mixed results in explaining skeptical judgments and decisions. The trustworthiness and independence parts of the Wrightsman Philosophies of Human Nature Scale (Wrightsman 1964, 1974) do not significantly relate to the auditor decision to trust a client in a study by Shaub (1996). But the trustworthiness part is reported by Rose (2007) to be related significantly to skeptical judgments. In particular, less trusting auditors pay more attention to evidence of aggressive reporting and increase the belief that intentional misstatement has occurred. Hence, the results of using the Wrightsman subscale are mixed. Furthermore, Shaub’s Client Trust Scale only shows significant results in two of the 18 regressions tested. In those two instances, there was an incentive present to overstate sales. Choo and Tan (2000) used a modified version of the Rempel, Holmes, and Zanna Trust Scale (1985), which showed some significant results: classroom instruction interacted with skeptical attitude (as measured by the trust scale) in affecting the ability to detect frauds. In all, the trust scales used in previous auditing studies do not yield consistent results. However, as noted, these studies do not employ the Rotter Interpersonal Trust Scale. Neutrality and the Hurtt Professional Skepticism Scale Several authors have stressed the need for development of a specific professional skepticism scale for auditing (e.g., Choo and Tan 2000; Hurtt 2010). In response, Hurtt (2010) developed such an instrument (HPSS). Hurtt (2010) argues that HPSS focuses more on having and pursuing doubt than on a particular direction of doubt, which is consistent with the neutrality perspective of professional skepticism. There is limited and mixed empirical evidence that the scores on the Hurtt Professional Skepticism Scale are related to auditors’ skeptical behavior (e.g., Hurtt, Eining, and Plumlee 2012; Fullerton and Durtschi 2004). Hurtt et al. (2012) find that scores on this scale are significantly related to contradictions detected in the working papers and that auditors with higher levels of professional skepticism, as measured by HPSS, generate a moderately higher number of alternatives. However, auditors with higher scores on the scale detect fewer mechanical errors. Fullerton and Durtschi (2004) find that internal auditors with a higher score on HPSS require greater evidence search, in the presence of fraud symptoms. Fraud training appears to reduce marginally the differences between high and low skeptical auditors, indicating that training can influence professional skepticism to some extent. To corroborate and extend prior research, we examine how HPSS relates to RIT in terms of explaining auditors’ skeptical judgments and decisions across different risk settings. The influence of situational factors: Client control environment risk In high client risk situations, more skeptical judgments and decisions are necessary, since such situations expose the individual auditor and the firm to increased reputation and CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
643
other risks (e.g., litigation).3 Investors also face a greater likelihood of losses due to the presence of material errors and fraud. In Nelson’s model on professional skepticism, this is recognized in the link between incentives and skeptical judgments and actions (Nelson 2009). Hurtt (2010) describes this as “state skepticism.” Professional standards dictate that engagements with a higher risk of material misstatement should be audited with increased professional skepticism (IFAC: ISA 240 2012). Furthermore, the auditor should corroborate clients’ explanations of unexpected fluctuations more fully, if the risk related to the areas of explanation is high (e.g., Hirst and Koonce 1996). One of the most pervasive client risks is the client’s control environment. It is well recognized in the literature that the control environment has a significant impact on the likelihood of material errors and fraud (e.g., Haskins 1987; Bernardi 1994; O’Leary, Iselin, and Sharma 2006). Financial reporting problems of companies have been found to be more pervasive when there is a weaker control environment (see, e.g., Beasley 1996; COSO 1992). Further, a common feature of many well-publicized major frauds is the weakness of the control environment, such as for Enron and WorldCom. Prior studies document that auditors are well aware of the importance of the control environment; and this risk is taken into consideration when, for instance, auditors assess risks, plan audit testing, and decide whether or not to accept an engagement (e.g., Asare and Knechel 1995; Cohen and Hanno 2000). The importance and impact of the control environment on auditors’ judgments is well established and is not a focus of the current study. Rather, we examine how auditor skepticism affects planning judgments under different risk settings, as will be discussed in the following section. Variations in control environment risk provide a context for us to investigate this issue. The interactive effects of skeptical perspective and control environment risk on auditor judgments and decisions The focus of this study is to examine the relationship between the two skeptical perspectives (neutrality and presumptive doubt) and auditors’ professional judgments and decisions across risk settings. It is likely that auditor judgments and decisions will be most strongly dependent on skeptical characteristics in higher-risk settings, with auditors possessing the trait of a high level of skepticism reacting to the situation by taking highly cautious actions (cf. Das and Teng 2004). We posit that the interaction is especially likely for auditors having a high level of presumptive doubt, since this perspective is one of distrust. Distrust coupled with signs of a higher risk of misstatement is posited to place the auditor at a high level of alert. In contrast, auditors with a neutrality focus do not explicitly assume the presence of material misstatements and, thus, when confronted with indications of greater risk, will be more cautious, but not to the same extent as those inclined toward distrust (e.g., Nelson 2009; Bell et al. 2005). Therefore, we hypothesize that in a higher-risk setting (such as a weaker control environment), auditors exhibiting greater presumptive doubt (i.e., higher scores on inversed RIT) will especially be on alert and, as a result, exhibit greater skeptical judgments than those with a pronounced neutrality focus (i.e., higher scores on HPSS). The nature of the expected interactions is presented in Figures 1 and 2. As shown, for both HPSS and inversed RIT, we expect auditors to exercise greater skeptical judgments as risks increase. However, the differences between the levels of skeptical judgments in the lower- versus higher-risk settings are posited to be significantly 3.
Also the psychological literature suggests that judgments and decisions will be related to situational characteristics (see, e.g., Eagly and Chaiken 1993, 2005; Ajzen 2005; Kee and Knox 1970; Bhattacharya, Devinney, and Pillutla 1998).
CAR Vol. 31 No. 3 (Fall 2014)
644
Contemporary Accounting Research
Figure 1 A depiction of the hypothesized interaction between risk setting and level of neutral skepticism (HPSS)
Figure 2 A depiction of the hypothesized interaction between risk setting and level of presumptive doubt (inversed RIT)
greater for increased levels of presumptive doubt (inversed RIT) than neutrality (HPSS).4 Stated formally our hypothesis is as follows: HYPOTHESIS. Presumptive doubt will be more strongly associated with auditors’ skeptical judgments and decisions than neutrality, particularly when control environment risk is higher. 3. Method Research setting The study utilizes an experimental case, adapted from Peecher (1996), which is embedded in a planning stage analytical procedures setting. Such a setting has important ramifications 4.
At very high levels of risk, auditors, regardless of skeptical disposition, may be prompted to react strongly (e.g., by either greatly extending testing or withdrawing from the engagement). In this study we examine the more typical levels of higher and lower risk.
CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
645
on audit efficiency and effectiveness (see, e.g., Cohen and Kida 1989; Hirst and Koonce 1996; Peecher 1996; Koonce, Anderson, and Marchant 1995). The case contains an unexpected material increase in a company’s (“MAEdic N.V.”) gross margin. Since the audit client is the most common source of explanations concerning unexpected fluctuations while conducting analytical procedures (see, e.g., Hirst and Koonce 1996; Trompeter and Wright 2010), a client explanation is provided. The CFO gives a nonerror explanation, stating that the increase in gross margin is caused by a change in the sales mix. In view of the fact that management may lack independence, auditors should evaluate client explanations with professional skepticism (e.g., Bedard and Biggs 1991; Glover, Jiambalvo, and Kennedy 2000; IFAC: ISA 240 2012). In the experiment, control environment risk is manipulated as higher or lower by using two vignettes based on Cohen and Hanno (2000). Skeptical disposition is measured by both the RIT and HPSS scales. Dependent variables The fundamental issue addressed in this study is to examine the relationship between auditors’ skeptical perspective (neutrality and presumptive doubt) and auditor skeptical judgments and decisions, across client risk settings. There is no universally accepted judgment or decision that is deemed to best reflect auditor skeptical behavior. Therefore, based on a review of prior research, we examine six proxies for auditors’ skeptical judgments and decisions. We posit the same general relationship for each of these dependent variables; that is, higher auditor skeptical disposition will lead to greater skeptical behavior, particularly in a higher-risk setting. An auditor should carefully reflect on the information that is provided by a client. Particularly for a skeptical auditor, it is common to ponder over the incentives a client might have in furnishing information, for example an explanation for an unexpected fluctuation. Auditors concerned about management veracity are directed by auditing standards to exercise more skeptical judgments and decisions (e.g., SAS 99 2002). The variables studied in this respect in this study are the likelihood that management’s explanation (a change in sales mix) accounts for substantially all of the increase in gross margin (>85 percent) (the likelihood that management explanation is right), as well as the likelihood of fraud (cf. Peecher 1996; Shaub 1996; Shaub and Lawrence 1996; Payne and Ramsay 2005; Choo and Tan 2000; Knapp and Knapp 2001). The assumption is that a higher score on the likelihood that management explanation is right implies less professional skepticism, and a higher score on the likelihood of fraud indicates higher professional skepticism. Furthermore, skeptical auditors are expected to build explanations, hypotheses, or scenarios that can function as alternative interpretations for the information that they examine. Auditors exhibit more skeptical judgments and decisions when they (1) are able to generate a greater number of plausible alternative explanations; (2) provide more errorexplanations (since these are counter-explanations to that provided by the client and entail greater risks to users and the auditor); and (3) assess a higher probability of the accuracy of error explanations than of nonerror explanations. As a result, the variables studied in this respect are the number of alternative explanations, the number of total error explanations, and the likelihood that the error explanations account for substantially all of the increase in gross margin (called the weight of the error explanations) (cf. Peecher 1996; McMillan and White 1993).5 The assumption is that the more alternative explanations 5.
Because of significant expectations for auditors to detect fraud (e.g., ISA 240 2012), further analyses were done on the two variables that constitute the variable the number of total error explanations: number of nonintentional error explanations, and number of intentional error explanations. The overall pattern of the results (untabulated) remains similar for these two variables.
CAR Vol. 31 No. 3 (Fall 2014)
646
Contemporary Accounting Research
generated, the more error explanations generated, and the higher the likelihood of the error explanations, the more professional skepticism is exhibited. Finally, an indication of skepticism is the extent to which auditors want to perform further testing. The most common variable of extent of testing in prior studies is the number of budgeted hours (e.g., Shaub 1996; Shaub and Lawrence 1996; Hurtt et al. 2012). A reference point of 100 hours for last year was provided to minimize variance in responses, by providing a benchmark, and because it is common in an audit setting to have a reference point (e.g., prior year hours). The assumption is that the more hours budgeted, the more professional skepticism is shown. Independent variables: Two scales measuring skeptical characteristics As mentioned, this study uses two scales to measure skeptical characteristics: Rotter’s Interpersonal Trust Scale or RIT (1967), and the Hurtt Professional Skepticism Scale or HPSS (2010).6 The primary difference in the perspective of RIT and HPSS is evident when one considers the nature of the questions. The RIT scale focuses on an individual’s basic level of trust or distrust in other individuals or organizations; that is, if one has a tendency of distrust (skepticism), there is a presumption of dishonesty unless evidence indicates otherwise. For instance, consider the following three questions in the RIT scale:
• • •
In dealing with strangers one is better off to be cautious until they have provided evidence that they are trustworthy. Using the honor system of not having a teacher present during exams would probably result in increased cheating. It is safe to believe that in spite of what people say most people are primarily interested in their own welfare.
In contrast, the focus of the HPSS is the extent to which an individual pursues extensive evidence (both positive and negative) in reaching a decision, which is reflective of neutrality. The following three questions taken from the HPSS scale illustrate this:
• • •
I wait to decide on issues until I can get more information. I like to understand the reason for other peoples’ behavior. I like to ensure that I’ve considered most available information before making a decision.
RIT is widely used and is found to have high construct validity and reliability (see, e.g., Rotter 1967). Based on prior literature, as discussed previously, the inversed RIT is used as a proxy for the presumptive doubt perspective of professional skepticism in this study. A higher score on the inversed RIT means lower trust and thus higher professional skepticism. This recoding enables a straightforward comparison with HPSS, since higher scores on either scale then both reflect greater skepticism.7 HPSS is a more recent measure of skepticism. Analyses employing students and professional subjects indicate that the scale has adequate inter-item consistency and test–retest reliability (e.g., Hurtt 2010). HPSS is used as a proxy for the neutrality perspective of professional skepticism. A pilot test was performed to test the construct validity of HPSS in measuring neutrality and (inversed) RIT in measuring presumptive doubt during a “train-the-trainers” 6. 7.
Please see supporting information, “S1: Items in Rotter Interpersonal Trust Scale and Hurtt Professional Skepticism Scale” as an addition to the online article for the questions underlying each scale. For inversed RIT, the original item scores were inversed (1 became 5, 2 became 4, et cetera), which makes the resulting total score directly interpretable as the degree of presumptive doubt. The Pearson correlation (r) between RIT and inversed RIT is 1.00. Using the nonreversed RIT item scores results in identical regression coefficients, except the sign for the RIT coefficient is negative rather than positive.
CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
647
session for a course on professional skepticism. Thirty-two (32) knowledgeable individuals were asked to assess a random sample of questions from each scale on a scale ranging from neutrality (value 1) to presumptive doubt (value 6).8 Beforehand, each of the constructs was defined. The average assessment is significantly different at p < 0.001 for the questions concerning RIT (mean = 4.62) from that for the HPSS (mean = 3.01) questions, in the direction expected (paired samples t-test; t = 7.048). These results indicate that (inversed) RIT is perceived to measure presumptive doubt significantly more than HPSS does, which provides support for the construct validity of the two measures. Independent variables: Manipulated and control variables Control environment risk (CER) is manipulated as lower (coded as 0) or higher (coded as 1). The manipulation of the control environment is shown in the Exhibit and is based on the case materials of Cohen and Hanno 2000, which are used with permission. Research instrument validation and manipulation checks The experiment was conducted during three regular audit firm annual summer training courses.9 The research instrument contains two parts: (1) the case description and the tasks (i.e., the dependent variables); and (2) the scales to measure skeptical characteristics, demographic information and debriefing questions. The only exception is that one of the dependent variables (i.e., the likelihood of fraud) was asked in the second part of the questionnaire to prevent the participants from becoming sensitized toward the possibility of fraud. Two questions on a 9-point scale were used as manipulation checks concerning the control environment risk (i.e., lower versus higher): one question on control environment effectiveness and another on overall control risk (e.g., Cohen and Hanno 2000). The question regarding control environment effectiveness was: “How effective do you consider the control environment of MAEdic N.V. to be?” The question concerning overall control risk was: “According to you, what would be the level of overall control risk (control risk at the level of the organization) of MAEdic N.V., given the information available?” The results of the manipulation check for both variables are significant in the expected direction (independent samples t-test; p < 0.01 for the question on control environment effectiveness and p < 0.05 for the question on overall control risk).10 Sample All of the participants come from one Big 4 firm, on an availability basis. There were 96 participants in the experiment: 25 partners, 41 managers, and 27 seniors (3 participants did not provide staff-level information). On average, the auditors had 15.36 years (8.95 standard deviation) of general auditing experience and 14.75 years (9.36) of experience with conducting analytical procedures. Both general experience and task-specific experience 8. 9.
10.
The items were selected at random and consisted of 40 percent of the items of each scale (RIT consists of 25 items and 10 items were selected at random; HPSS consists of 30 items and 12 items were selected). The experiment was done in the official language that was used during the firms’ sessions, which was Dutch. Therefore, the case-materials were translated into Dutch. Translations were developed for the Interpersonal Trust Scale and for the Hurtt Professional Skepticism Scale. For that purpose a combination of the Parallel Blind Technique and Translation/Back-Translation methods was used (Behling and Law 2000). Two experienced auditors compared the English and the Dutch versions of the case description and evaluated the contents of the case materials. In addition, four professors provided remarks on the research instrument. Two pilot tests were conducted with 19 staff-level auditors and minor modifications were made to the instrument on the basis of the comments received. Two questions were asked about case realism and understandability using a 9-point scale. T-tests show that the mean scores for both were significantly above the middle point of the scales (p-values are smaller than 0.001), suggesting participants believed the case information was realistic and clear.
CAR Vol. 31 No. 3 (Fall 2014)
648
Contemporary Accounting Research
suggest that the participants possess the requisite task knowledge. As a result of random assignment, the number of participants was relatively balanced across the two control environment risk conditions: 47 participants in the higher control environment risk condition and 49 participants in the lower control environment risk condition. All participants completed the two skeptical characteristics measurement scales in part two of the experiment. To mitigate the possibility that the order of the skeptical disposition scales may influence participants’ behaviors, the scales were randomly ordered. Oneway ANOVAs show that there are no significant order effects in the administration of the two scales for skeptical disposition. Administration of the experiment One of the authors attended all three experimental sessions and provided a brief introduction before the start of the experiment. The single remark concerning the topic of the study was that it comprised a case on conducting preliminary analytical procedures. After the introduction, the participants received an envelope containing the research materials. A printed instruction regarding completion of the instrument was attached to the envelope. Participants were randomly assigned to the two control environment risk conditions. A large envelope included two smaller envelopes with part one (risk assessment and audit planning tasks) and part two (skeptical dispositions and debriefing questions) of the research instrument. The two parts of the instruments were identified with identical numbers in order to enable ex post matching. The participants were asked to write down their name on the envelope in order to induce a feeling of accountability, as is present in practice (e.g., Asare, Trompeter, and Wright 2000). At the end of part two, participants were informed that if they would like a summary of the findings they should note their email address (57 out of the 96, or 59 percent, did, indicating a substantial level of interest). Coding As noted, in the case description the chief financial officer (CFO) provided a nonerror explanation for the increase in the gross margin percentage. The participants were requested to think about possible alternative explanations for the increase in the gross margin percentage. These explanations were independently coded by one of the authors and an experienced audit manager, blind to experimental condition. The coding encompassed an assessment of the type and plausibility of the explanation (is the explanation logical and in the right direction to explain the fluctuation?). Explanations were classified into the following categories: nonerror explanations; unintentional error explanations; intentional error explanations; and ambiguous unintentional/intentional error explanations. Cohen’s Kappa Coefficient regarding the coding of the plausibility of the explanation was 0.868 (p = 0.001) and Cohen’s Kappa Coefficient for coding the type of explanation was 0.862 (p = 0.001). These levels of agreement are strong (e.g., Landis and Koch 1977) and indicate high inter-rater reliability. Differences were discussed by the two coders and mutually resolved. 4. Results Descriptive statistics regarding the measures of the two skeptical characteristics are shown in Table 1. The mean score of RIT (73.95) is very close to the theoretical midpoint of the scale (75.00). However, the mean score of HPSS (133.09) is considerably higher than the theoretical midpoint of the scale (105.00). Furthermore, the actual range for HPSS encompasses about 35 percent of the theoretical range—which resembles the findings of Hurtt CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
649
TABLE 1 Descriptive statistics for the scales measuring the two skeptical perspectives
Measurement scale Rotter Interpersonal Trust Scale Hurtt Professional Skepticism Scale
Mean score
Standard deviation
Theoretical range
Actual range
Theoretical midpoint
Cronbach alpha
73.95
8.22
25–125
45–94
75
0.764
133.09
10.84
30–180
103–158
105
0.834
(2010)—while the actual range of RIT covers about 50 percent of the theoretical range. These differences may be caused by self-selection of individuals who enter and/or are promoted in the field of auditing. However, both scales exhibit a large variance. Importantly, the Pearson correlation (r = .084) between RIT and HPSS is not significant (p = .417), indicating they measure different skeptical perspectives. As reported in Table 1, the Cronbach alpha value for RIT is 0.764 and the alpha for HPSS is 0.834, which are acceptable (see, e.g., Nunnally 1978). One-sample Kolmogorov Smirnov tests and an examination of the histograms indicate that the measurement scales are approximately normally distributed. Descriptive statistics for the dependent variables are shown in Table 2, panel A, reflecting a wide range in auditor judgments and decisions. Relationship between skeptical characteristics and auditor judgments and decisions To provide initial descriptive evidence, in Table 2, panel B, the correlations (i.e., strength of association) between skeptical disposition (inversed RIT and HPSS) and the level of skeptical judgments and decisions are presented for the lower- and higher-risk settings. In the lower-risk setting (stronger control environment), only two of the correlations between inversed RIT and the dependent variables are significant (the likelihood that management explanation is right), or marginally significant (the likelihood of fraud). Three of the correlations between HPSS and the dependent variables are significant (the likelihood of fraud and the number of total error explanations), or marginally significant (the number of budgeted hours). These results suggest that in the low-risk setting both inversed RIT and HPSS are about equally predictive of auditor skeptical judgments and decisions. Most importantly, though, in the higher-risk setting (weaker control environment), all of the correlations between inversed RIT and the dependent variables are significant, except for the likelihood that management explanation is right, which is marginally significant. In contrast, only one correlation (the number of budgeted hours) is marginally significant for HPSS. These results indicate that presumptive doubt has a greater effect on auditors’ skeptical judgments when client risks are higher than does neutrality, consistent with our hypothesis. In particular, auditors with higher levels of presumptive doubt exhibit pronounced skeptical judgments and decisions in the higher control environment risk setting.11 Two sets of linear regressions were conducted in order to test the hypothesis regarding the association between the skeptical characteristics and individual judgments and decisions at varying risk levels.12 The first set (the “presumptive doubt set”) uses control 11. 12.
The use of nonparametric correlations also leads to similar conclusions as the parametric Pearson correlations reported. The scores on the skeptical characteristics were mean centered. Although there is debate about whether or not mean centering is beneficial (see, e.g., Kromrey and Foster-Johnson 1998), most sources believe it aids in improving the interpretation of the results (see, e.g., Aguinis and Gottfredson 2010; Dalal and Zickar 2011). The findings concerning the hypothesis remain identical, whether mean centered or not, since interactions are not affected by mean centering.
CAR Vol. 31 No. 3 (Fall 2014)
650
Contemporary Accounting Research
TABLE 2 Descriptive statistics and correlations Panel A: Descriptive statistics for the dependent variables measuring skeptical judgments and decisions
Dependent variable The likelihood that management explanation is right The likelihood of fraud The number of alternative explanations The number of total error explanations The weight of total error explanations The number of budgeted hours
Mean score (n)
Standard deviation
Theoretical range
Actual range
35.64 (n = 92)
21.96
0–100
0–100
22.19 1.73 1.43 28.80 39.15
0–100 0–∞ 0–∞ 0–100 0–∞
2–90 0–9 0–5 0–95 80–300
30.31 3.22 1.63 31.26 142.73
(n (n (n (n (n
= = = = =
93) 95) 95) 93) 88)
Panel B: Pearson correlations between skeptical judgments and decisions and the two skeptical perspectives (significance between parentheses) Inversed RIT Lower control environment risk The likelihood that management explanation is right The likelihood of fraud The number of alternative explanations The number of total error explanations The weight of the error explanations The number of budgeted hours Higher control environment risk The likelihood that management explanation is right The likelihood of fraud The number of alternative explanations The number of total error explanations The weight of the error explanations The number of budgeted hours
HPSS
.387 .234 .104 .025 .087 .020
(.003)*** (.056)* (.241) (.432) (.278) (.447)
.065 .336 .036 .259 .110 .214
(.329) (.010)*** (.405) (.037)** (.229) (.077)*
.199 .309 .312 .388 .408 .447
(.097)* (.018)** (.016)** (.004)*** (.003)*** (.002)***
.180 .095 .039 .094 .163 .238
(.121) (.266) (.398) (.265) (.142) (.065)*
Notes: The table shows the Pearson correlations (one-sided p-values) for the dependent variables and the two skeptical trait measures (inversed Rotter’s Interpersonal Trust Scale (inversed RIT) and the Hurtt Professional Skepticism Scale (HPSS)). The significant correlations are indicated by * for p < .10, ** for p < .05, and *** for p < .01.
environment risk (CER), inversed RIT and the interaction between CER and inversed RIT as independent variables, and the six skeptical judgments and decisions alternately as dependent variables. The second set of regressions (the “neutrality set”) uses control environment risk (CER), HPSS, and the interaction between CER and HPSS as independent variables and the six skeptical judgments and decisions alternately as dependent variables.13
13.
Given potential relationships between the dependent variables, a MANOVA was also conducted. All six of the dependent variables discussed previously were included as the dependent variables. For the “presumptive doubt set,” the MANOVA results show statistical significance at p < 0.01 for CER and the interaction between inversed RIT and CER, and significance at p < 0.05 for inversed RIT. For the “neutrality set,” the MANOVA results show statistical significance at p < 0.001 for CER and significance at p < 0.10 for HPSS. The interaction between HPSS and CER was not significant.
CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
651
A summary of the regressions is provided in Table 3, with a presentation of the “Presumptive doubt regressions” in panel A and of the “Neutrality regressions” in panel B.14 For example, in panel A, concerning the regression of the number of total error explanations, the adjusted R2 (right-most column) is 0.055 and is significant (p < 0.043). Control environment risk (CER) and the inversed Rotter’s Interpersonal Trust Scale (inversed RIT) are not significant (their p-values are 0.304 and 0.433, respectively). However, the interaction between inversed RIT and CER is significant at 0.009, and its coefficient and beta are 0.091 and 0.297, respectively. The sign for the interaction is positive, as expected.15 Five out of the six regression models in panel A (the presumptive doubt set) are significant at p < 0.05 (the likelihood that management explanation is right, the likelihood of fraud, the number of total error explanations, the weight of the error explanations, and the number of budgeted hours).16 In addition, the sixth model is marginally significant at p < 0.10 (the number of alternative explanations). The adjusted R2 of the models vary from 0.048 (the number of alternative explanations) to 0.262 (the likelihood of fraud). In contrast, only one of the regressions in panel B (the neutrality set) is significant at p < 0.001 (the likelihood of fraud). All significant effects in both panels A and B are in the expected direction. Results for the hypothesis: Interaction between CER and skeptical characteristics Overall, the results in panel A of Table 3 are much stronger than the results in panel B of Table 3, suggesting that inversed RIT has a greater impact on auditors’ skeptical judgments and decisions than HPSS. The results appear to be particularly driven particularly by the interaction of inversed RIT and CER in most regressions, as hypothesized.17 The interaction between inversed RIT and CER is significant for regressions concerning the number of alternative explanations, the number of error explanations, the weight of the error explanations, and the number of budgeted hours. In contrast, the interaction between HPSS and CER is not significant in the one significant model in panel B. In all, the results provide strong support for the hypothesis.18 5. Discussion In auditing there are two prevailing perspectives of skepticism: neutrality and presumptive doubt. The results of this study show that inversed RIT (Rotter’s Interpersonal Trust Scale), 14. 15.
16.
17.
18.
Using dichotomized variables (i.e., split by the median score) for inversed RIT and HPSS in the regressions does not qualitatively alter the findings. We also conducted contrast analysis (e.g., Buckless and Ravenscroft 1990; Gold, Knechel, and Wallage 2012) to test whether the form of the ordinal interactions found are consistent with our hypothesis. When we apply a median split to the scores of inversed RIT and HPSS, the results confirm the idea that the dependent variables are particularly influenced by high inversed RIT in the higher-risk setting. For instance, auditors in the higher-risk setting with higher inversed RIT scores exhibited significantly greater skeptical behaviors (p < .05, one-tailed) than auditors in the other three inversed RIT/CER groups, for all six of the dependent variables. In comparison, auditors in the higher-risk setting with higher HPSS scores only exhibited significantly greater skeptical behaviors than those in the other three HPSS/CER groups for three of the six dependent variables. To control for potential confounding effects, demographic variables (as reported in the method section) were included as covariates in the statistical models. The results do not qualitatively change and the demographic variables are therefore omitted from the reported analyses. If inversed RIT, HPSS, and the interaction effects of both measures with CER are all entered into the regressions, the interaction between CER and inversed RIT is (marginally) significant in three of the five (marginally) significant regressions, and the interaction between HPSS and CER is not significant in any of the regressions. This robustness test also confirms our findings. To test for the potential influence of outliers, we removed observations of the dependent variables that are more than three standard deviations from the mean (e.g., Hair, Black, Babin, Anderson, and Tatham 2006) and reran the analyses. The conclusions of our regression analyses are not qualitatively different.
CAR Vol. 31 No. 3 (Fall 2014)
652
Contemporary Accounting Research
TABLE 3 Test of hypothesis: Regression results Panel A: Regressions concerning the presumptive doubt set
CER
Dependent variables (n) The likelihood that management explanation is right (n = 92) The likelihood of fraud (n = 93) The number of alternative explanations (n = 95) The number of total error explanations (n = 95) The weight of the error explanations (n = 93) The number of budgeted hours (n = 88)
Constant (p-value) 38.577 (.000)
20.224 (.000)
3.013 (.000)
1.540 (.000)
28.148 (.000)
140.222 (.000)
Coeff. [beta] {t-value} (p-value)
Inversed RIT
Exp. sign
6.326 [ .145] { 1.447} (.076) 20.370 [.462] {5.145} (.000) .375 [.109] {1.082} (.141) .147 [.052] {.515} (.304) 5.860 [.102] {1.013} (.157) 4.430 [.057] {.552} (.291)
+
+
+
+
+
Inversed RIT * CER
Coeff. [beta] {t-value} (p-value)
Coeff. [beta] Exp. {t-value} Exp. sign (p-value) sign
.900 [ .343] { 2.838} (.003) .420 [.157] {1.447} (.076) .016 [ .077] { .626} (.267) .004 [ .021] { .171} (.433) .262 [.076] {.620} (.269) .079 [.017] {.134} (.447)
.284 [.061] {.503} (.308) .588 [.124] {1.140} (.129) .107 [.289] {2.366} (.010) .091 [.297] {2.437} (.009) 1.465 [.238] {1.953} (.027) 2.626 [.315] {2.522} (.007)
+
+
+
+
+
Adj. R2 (p-value) .094 (.009)
+
.262 (.000)
+
.048 (.059)
+
.055 (.043)
+
.066 (.028)
+
.078 (.020)
Panel B: Regressions concerning the neutrality set CER
Dependent variables (n) The likelihood that management explanation is right (n = 92) The likelihood of fraud (n = 93)
Constant (p-value)
Coeff. [beta] {t-value} (p-value)
38.935 (.000)
6.924 [ .158] { 1.517} (.067)
20.334 (.000)
20.440 [.463] {5.081} (.000)
HPSS
HPSS * CER
Coeff. Coeff. [beta] [beta] Exp. {t-value} Exp. {t-value} Exp. sign (p-value) sign (p-value) sign .127 [ .063] { .459} (.324) +
.500 [.247] {2.044} (.022)
.246 [ .079] { .576} (.283) +
.299 [ .097] { .801} (.213)
Adj. R2 (p-value) .010 (.278)
+
.238 (.000)
(The table is continued on the next page.)
CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
653
TABLE 3 (continued) Panel B: Regressions concerning the neutrality set CER
Dependent variables (n)
Constant (p-value)
The number of alternative explanations (n = 95) The number of total error explanations (n = 95) The weight of the error explanations (n = 93) The number of budgeted hours (n = 88)
3.024 (.000)
1.562 (.000)
28.200 (.000)
140.343 (.000)
Coeff. [beta] {t-value} (p-value) .405 [.118] {1.131} (.131) .156 [.055] {.532} (.298) 6.403 [.112] {1.070} (.146) 4.563 [.059] {.552} (.292)
HPSS
HPSS * CER
Coeff. Coeff. [beta] [beta] Exp. {t-value} Exp. {t-value} Exp. sign (p-value) sign (p-value) sign +
+
+
+
.005 [.029] {.207} (.418) .031 [.233] {1.709} (.046) .275 [.103] {.755} (.226) .698 [.193] {1.377} (.086)
+
+
+
+
.012 [ .049] { .358} (.361) .017 [ .083] { .612} (.271) .185 [.045] {.330} (.371) .257 [.047] {.333} (.370)
Adj. R2 (p-value)
+
.017 (.704)
+
.008 (.296)
+
.000 (.399)
+
.022 (.183)
Notes: This table shows the regression coefficients (Coeff.), betas, the t-value and their significance (p-value) concerning the dependent variables regressed on control environment strength (CER), the inverse of Rotter’s Interpersonal Trust Scale (inversed RIT) in panel A and the Hurtt Professional Skepticism Scale (HPSS) in panel B. In the last column the regressions’ adjusted R2 are presented. Betas are shown in brackets, t-values in braces, and p-values in parentheses. All significance levels reported (p-values) are one-sided. The sample size may vary, due to missing values.
a trait measure of presumptive doubt, is more closely associated with auditors’ skeptical judgments than HPSS (Hurtt Professional Skepticism Scale), a trait measure of neutrality. This relationship particularly holds when control environment risks are higher and, thus, the risk of material misstatement is of particular concern. Since auditing standards direct greater skepticism in a higher-risk setting than in a lower-risk setting, this finding suggests that inversed RIT is more likely to reflect the desired skepticism than HPSS. The study has some limitations that should be borne in mind when interpreting the findings. There is no normative solution to the case, so it is not possible to determine which skepticism measure is more closely related to optimal judgments and decisions. To address this issue, multiple measures of skeptical behavior are employed. Furthermore, HPSS is used to measure neutrality and inversed RIT to measure presumptive doubt. These measures have been widely used and validated and are reflective of the two skepticism constructs examined. However, the scales were not specifically designed to measure auditor neutrality or presumptive doubt. In addition, auditors from one auditing firm participated in the experiment. Although the task is a generic one and we have no reason to expect firm differences, the results may be influenced by firm-specific behavior. There are a number of important implications of the findings for audit practice and future research. For instance, since the findings indicate inversed RIT is a better predictor CAR Vol. 31 No. 3 (Fall 2014)
654
Contemporary Accounting Research
of desired auditor skeptical judgments, this measure warrants further consideration in practice to develop adequate recruitment, staffing, and training guidance. Also a consideration for practice and future research is whether quality control processes, such as audit reviews, mitigate cases where auditors lack sufficient individual skepticism. For example, it is not clear how the review process interacts with skeptical disposition and to what degree the review process can compensate for insufficient testing and care. Exhibit Manipulation of strength of the control environment Lower control environment risk: The management of MAEdic can be described as being conservative in business practices and makes decisions only after considering all risks and possibilities. If necessary, external consultants are asked for advice in making important decisions. Top management and lower management meet on a regular basis, formally as well as informally. The IT department consists of experienced people. The information system is viewed as the instrument to control business activities. Management wants the financial reports to be accurate and reliable and avoids focusing on reporting short-term results. Apart from occasional disputes between management and the external auditor, in general they cooperate harmoniously in order to come to adequate financial reporting. There is a strict policy for following all established internal control procedures. Top management emphasizes several performance measures in evaluating the employees. In addition to short-term measures from the financial information system, there is elaborate attention for long-term developments and qualitative factors. Ethics and integrity are criteria in performance assessment. Directors receive a fixed salary with a bonus of about 20 percent of the fixed salary depending on achieving specified personal or activity targets. Because compensation is only indirectly based on profitability, management has little drive to manipulate short-term results. Higher control environment risk: The management of MAEdic can be described as being aggressive in business practices and emphasizes speed and efficiency when implementing decisions. Management rarely hires external consultants because they are of the opinion that consultants are expensive and often follow a too-conservative approach. Top management and lower management meet during monthly production meetings. Management views the IT department as a necessary evil and considers the accountants and bookkeepers who work there to be bean-counters. Because management has a clear preference for reporting methods that enable earnings management, management has frequent disputes with the external auditor. Although there are a large number of internal control procedures in place, they are sometimes less strictly applied if the progress of the work is suffering from them. Top management mainly focuses on achieving short-term accounting-based performance measures when determining compensation and making promotion decisions. Productivity is the most important criterion in performance assessment. Directors receive a small base salary and a bonus that is based on the profitability of the department in question. Management is convinced that this compensation system encourages healthy competition and personal initiatives.
References Aguinis, H., and R. K. Gottfredson. 2010. Best-practice recommendations for estimating interaction effects using moderated multiple regression. Journal of Organizational Behavior 31 (6): 776–86. Ajzen, I. 2005. Attitudes, personality, and behavior. Berkshire: Open University Press. American Institute of Certified Public Accountants (AICPA). 1977. The independent auditors responsibility for the detection of errors or irregularities. Statement on Auditing Standards (SAS) No. 16. New York: AICPA.
CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
655
American Institute of Certified Public Accountants (AICPA). 1988. The auditor’s responsibility to detect and report errors and irregularities. Statement on Auditing Standards (SAS) No. 53. New York: AICPA. American Institute of Certified Public Accountants (AICPA). 2002. Consideration of fraud in a financial statement audit, statement. Statement on Auditing Standards (SAS) No. 99. New York: AICPA. Asare, S., and W. R. Knechel. 1995. Termination of information evaluation in auditing. Journal of Behavioral Decision Making 8 (1): 21–31. Asare, S. K., G. M. Trompeter, and A. M. Wright. 2000. The effect of accountability and time budgets on auditors’ testing strategies. Contemporary Accounting Research 17 (4): 539–60. Beasley, M. S. 1996. An empirical analysis of the relation between the board of director composition and financial statement fraud. The Accounting Review 71 (4): 443–65. Beasley, M. S., J. V. Carcello, and D. R. Hermanson. 2001. Top 10 audit deficiencies. Journal of Accountancy (April): 63–66. Bedard, J. C., and S. F. Biggs. 1991. The effect of domain-specific experience on evaluation of management representations in analytical procedures. Auditing: A Journal of Practice and Theory 10 (Supp.): 77–95. Behling, O., and K. S. Law. 2000. Translating questionnaires and other research instruments: problems and solutions. Thousand Oaks, CA: Sage Publications. Bell, T. B., M. E. Peecher, and I. Solomon. 2005. The 21st century public company audit: conceptual elements of KPMG’s global audit methodology. New York: KPMG LLP. Benston, G. J., and A. L. Hartgraves. 2002. Enron: What happened and what we can learn from it. Journal of Accounting and Public Policy 21 (2): 105–27. Bernardi, R. A. 1994. Fraud detection: The effect of client integrity and competence and auditor cognitive style. Auditing: A Journal of Practice & Theory 13 (Supp.): 68–84. Bhattacharya, R., T. M. Devinney, and M. M. Pillutla. 1998. A formal model of trust based on outcomes. Academy of Management Review 23 (3): 459–72. Buckless, F. A., and S. P. Ravencroft. 1990. Contrast coding: A refinement of ANOVA in behavioral analysis. The Accounting Review 65 (4): 993–45. Choo, F., and K. Tan. 2000. Instruction, skepticism, and accounting students’ ability to detect frauds in auditing. Journal of the Academy of Business Education 1 (3): 72–87. Cohen, J. R., and D. M. Hanno. 2000. Auditors’ consideration of corporate governance and management control philosophy in preplanning and planning judgments. Auditing: A Journal of Practice & Theory 19 (2): 133–46. Cohen, J., and T. Kida. 1989. The impact of analytical review results, internal control reliability, and experience on auditors’ use of analytical review. Journal of Accounting Research 27 (2): 263–76. Committee of Sponsoring Organizations of the Treadway Commission (COSO). 1992. Internal control–integrated framework. New York: AICPA. Cushing, B. E. 2000. Economic analysis of skepticism in an audit setting. In 14th Symposium on Auditing Research, eds. I. Solomon and M.E. Peecher, 1–3. University of Illinois at UrbanaChampaign: Office of Accounting Research. Dalal, D. K., and M. J. Zickar. 2011. Some common myths about centering predictor variables in moderated multiple regression and polynomial regression. Organizational Research Methods 15 (3): 339–62. Das, T. K., and B. Teng. 2004. The risk-based view of trust: A conceptual framework. Journal of Business and Psychology 19 (1): 85–116. Doucet, M. S., and T. A. Doucet. 1996. Commentary on: Ethics, experience and professional skepticism: A situational analysis. Behavioral Research in Accounting 8 (Supp.): 158–68. Eagly, A. H., and S. Chaiken. 1993. The psychology of attitudes. Belmont, CA: Thomson Wadsworth.
CAR Vol. 31 No. 3 (Fall 2014)
656
Contemporary Accounting Research
Eagly, A. H., and S. Chaiken. 2005. Attitude research in the 21st century: The current state of knowledge. In The Handbook of Attitudes, eds. D. Albarracın, B. T. Johnson, and M. P. Zanna, 743–68. Mahwah, NJ: Lawrence Erlbaum Associates. Financial Reporting Council. 2010. Auditor Scepticism: Raising the Bar. Discussion paper of the Auditing Practices Board. London: Financial Reporting Council. Fullerton, R. R., and C. Durtschi. 2004. The Effect of Professional Skepticism on the Fraud Detection Skills of Internal Auditors. Working paper, Utah State University. Glover, S. M., J. Jiambalvo, and J. Kennedy. 2000. Analytical procedures and audit planning decisions. Auditing: A Journal of Practice and Theory 19 (2): 27–45. Gold, A., W. R. Knechel, and P. Wallage. 2012. The effects of the strictness of consultation requirements on fraud consultation. The Accounting Review 87 (3): 925–49. Hair, J. F. Jr, W. C. Black, B. J. Babin, R. E. Anderson, and R. L. Tatham. 2006. Multivariate Data Analysis, 6th ed. Upper Saddle River, NJ: Pearson Prentice Hall. Haskins, M. E. 1987. Client control environments: An examination of auditors’ perceptions. The Accounting Review 62 (3): 542–63. Hirst, D. E., and L. Koonce. 1996. Audit analytical procedures: A field investigation. Contemporary Accounting Research 13 (2): 457–86. Hoell, R. C. 2004. The effect of interpersonal trust and participativeness on union member commitment. Journal of Business and Psychology 19 (2): 161–77. Hurtt, R. K. 2010. Development of a scale to measure professional skepticism. Auditing: A Journal of Practice & Theory 29 (1): 149–71. Hurtt, R. K., M. Eining, and D. Plumlee. 2012. An experimental examination of professional skepticism. Working paper. Baylor University and University of Utah. IAASB. 2012. Staff questions and answers, professional skepticism in an audit of financial statements. New York: IFAC. IFAC. 2012. Handbook of International Quality Control, Auditing, Review, Other Assurance, and Related Services Pronouncements. New York: IFAC. Johnson-George, C., and W. C. Swap. 1982. Measurement of specific interpersonal trust: Construction and validation of a scale to assess trust in a specific other. Journal of Personality and Social Psychology 43 (6): 1306–17. Kadous, K. 2000. The effects of audit quality and consequence severity on juror evaluations of auditor responsibility for plaintiff losses. The Accounting Review 75 (3): 327–41. Kee, H. W., and R. E. Knox. 1970. Conceptual and methodological considerations in the study of trust and suspicion. Journal of Conflict Resolution 14 (3): 357–66. Knapp, C. A., and M. C. Knapp. 2001. The effects of experience and explicit fraud risk assessment in detecting fraud with analytical procedures. Accounting, Organizations and Society 26 (1): 25–37. Koonce, L., U. Anderson, and G. Marchant. 1995. Justification of decisions in auditing. Journal of Accounting Research 33 (2): 396–84. Kromrey, J. D., and L. Foster-Johnson. 1998. Mean centering in moderated multiple regression: Much ado about nothing. Educational and Psychological Measurement 58 (1): 42–67. Landis, J. R., and G. G. Koch. 1977. The measurement of observer agreement for categorical data. Biometrics 33: 159–74. Libby, R., and J. Luft. 1993. Determinants of judgment performance in accounting settings: Ability, knowledge, motivation, and environment. Accounting, Organizations and Society 18 (5): 425–50. Mautz, R. K., and H. A. Sharaf. 1961. The philosophy of auditing. American Accounting Association Monograph No. 6. Sarasota, FL: American Accounting Association. McMillan, J. J., and R. A. White. 1993. Auditors’ belief revisions and evidence search: The effect of hypothesis frame, confirmation and professional skepticism. The Accounting Review 68 (3): 443–65.
CAR Vol. 31 No. 3 (Fall 2014)
Neutrality versus Presumptive Doubt
657
Nelson, M. 2009. A model and literature review of professional skepticism in auditing. Auditing: A Journal of Practice & Theory 28 (2): 1–34. Nunnally, J. C. 1978. Psychometric theory, 2nd ed. New York: McGraw-Hill. O’Leary, C., E. Iselin, and D. Sharma. 2006. The relative effects of elements of internal control on auditors’ evaluations of internal control. Pacific Accounting Review 18 (2): 69–96. Payne, E. A., and R. J. Ramsay. 2005. Fraud risk assessments and auditors’ professional skepticism. Managerial Auditing Journal 20 (3): 321–30. Peecher, M. E. 1996. The influence of auditors’ justification processes on their decisions: A cognitive model and experimental evidence. Journal of Accounting Research 34 (1): 125–40. Public Company Accounting Oversight Board (PCAOB). 2008. Proposed auditing standards related to the auditor’s assessment of and response to risk and conforming Amendments to PCAOB standards. PCAOB Release No. 2008-006. Stamford, CT: PCAOB. Public Oversight Board (POB). 2000. The panel on audit effectiveness: Report and recommendations. Stamford, CT: POB. Rempel, J. K., J. G. Holmes, and M. P. Zanna. 1985. Trust in close relationships. Journal of Personality and Social Psychology 49 (1): 95–112. Rose, J. M. 2007. Attention to evidence of aggressive financial reporting and intentional misstatement judgments: Effects of experience and trust. Behavioral Research In Accounting 19 (1): 215–29. Rotter, J. B. 1967. A new scale for the measurement of interpersonal trust. Journal of Personality 35 (4): 651–65. Rotter, J. B. 1980. Interpersonal trust, trustworthiness, and gullibility. American Psychologist 35 (1): 1–7. Shaub, M. K. 1996. Trust and suspicion: The effects of situational and dispositional factors on auditors’ trust of clients. Behavioral Research in Accounting 8: 154–74. Shaub, M. K., and J. E. Lawrence. 1996. Ethics, experience, and professional skepticism: A situational analysis. Behavioral Research in Accounting 8 (Supp.): 124–57. Smith, J. E., and T. Kida. 1991. Heuristics and biases: Expertise and task realism in auditing. Psychological Bulletin 109 (3): 472–89. Stack, L. C. 1978. Trust. In Dimensions of personality, eds. H. London, and J. E. Exner Jr, 561–99. New York: John Wiley & Sons. Trompeter, G., and A. Wright. 2010. The world has changed—Have analytical procedure practices? Contemporary Accounting Research 27 (2): 669–700. Webb, W. M., and P. Worchel. 1986. Trust and distrust. In Worchel, S. and W.G. Austin, Psychology of intergroup relations, 213–28. Chicago: Nelson-Hall Publishers. Wrightsman, L. S. 1964. Measurement of philosophies of human nature. Psychological Reports 14 (3): 743–51. Wrightsman, L. S. 1974. Assumptions about human nature: A social-psychological analysis. Monterey, CA: Brooks/Cole. Wrightsman, L.S. 1991. Interpersonal trust and attitudes toward human nature. In Measures of personality and social psychological attitudes, eds. J.P. Robinson, P.R. Shaver and L.S. Wrightsman, 373–412. Vol. 1 in Measures of Social Psychological Attitudes Series. San Diego, CA: Academic Press.
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: S1: Items in Rotter Interpersonal Trust Scale and Hurtt Professional Skepticism Scale.
CAR Vol. 31 No. 3 (Fall 2014)
