Female Auditor1

8/4/2019 Female Auditor1

1/46Electronic copy available at: http://ssrn.com/abstract=1409964

Are Female Auditors Still Women?

Analyzing the sex differences affecting audit quality

Kris Hardiesa, 1

, Diane Breescha, Jol Branson

a

a Faculty of Economic, Social and Political Sciences, and Solvay Business School, Free

University of Brussels, Pleinlaan 2, 1050 Brussels, Belgium

Submitted Manuscript

Abstract

Previous research has hinted a potential impact of auditor gender on audit quality. It appears

that, for example, men are less risk-averse than women. Female auditors may, therefore,

express more severe audit opinions than male auditors. This paper addresses a potential major

bias underlying the gender auditing research as it is not obvious that stereotypical believes

about men and women are true or that findings from literature about the general population can

be interpolated to the specific context of auditors.

Keywords: audit quality, gender, sex, gender differences, sex differences, psychological

research

Acknowledgments

We would like to thank Donald Wygal and Alison Woodward for comments on earlier versions

of this paper. This paper has also benefited from the comments of participants in research

seminars at the AAA Mid-Atlantic Regional Meeting, Long Branch (NJ), April 2009, and the

EAA Congress, Tampere, May 2009. Remaining errors are the responsibility of the authors.

1 Corresponding author. Tel.: +32 (0) 2 629 14 32; fax: +32 (0) 2 629 20 60.

E-mail address: [email protected] (K. Hardies).


2/46Electronic copy available at: http://ssrn.com/abstract=1409964

1

Are Female Auditors Still Women?

Analyzing the sex differences affecting audit quality

1. Introduction

As auditing is inherently a judgment and decision-making process, ultimately audit quality (i.e.

the probability that, within reasonable limits, the auditor discovers and reports a material

misstatement in the financial statements) is dependent upon the auditors judgment and

decision-making qualities (Knechel, 2000). The quality of an auditors judgment and decision-

making depends on its turn on certain auditor characteristics. The probability of discovering a

material misstatement depends, among others, on the auditors technical expertise, problem-

solving ability, risk profile, and experience. The probability of reporting a material

misstatement depends, among others, on its discovery, the auditors risk profile, and the

auditors independence from the client.

Based on psychological literature, very recently some researchers (e.g. Gold et al., 2009)

have posited that there are sex differences in personal auditor characteristics (e.g. risk-

aversion), leading to sex-differentiated audit judgments and decisions. Among other things, this

may cause the height of the audit fee to be associated with the sex of the audit engagement

partner (Ittonen and Peni, 2009).

Beliefs and speculations about whether there are fundamental differences between men and

women are pervasive, and are further augmented by the pile of scientific literature on sex

differences and by popular culture and media endorsing the existence of such differences. Our

cultural knowledge that is embedded in such stereotypes1 (as in men are from Mars and

women are from Venus) is our belief about how most people view the typical man or woman

(Ridgeway, 2009). Stereotypical beliefs relate personal characteristics to sex (so that men are

rather masculine and women rather feminine) suggesting that sex differences exist in the areas

of mathematical problem-solving, risk taking, and independence.

Math is believed to be a boy thing (Rowley et al., 2007) and, although meta-analyses

show that sex differences in mathematical performances are somewhere between non-existent

and almost non-existent (Hyde and Linn, 2006), most studies (e.g. Penner and Paret, 2008) find

men, on average, to be somewhat better mathematical problem-solvers than women. Also risk

taking is in our contemporary society associated with masculinity (Wilson and Daly, 1985).

Most studies report more risk-aversion amongst females (e.g. Jianakoplos and Bernasek, 1998),

even when these females are trained financial professionals (Olsen and Cox, 2001). There is


3/46

2

also evidence that suggests that males and females differ in determinants of independence, i.e.

in ethical behaviour, empathy, and proneness to cognitive biases. Females appear to be more

ethical (OFallon and Butterfield, 2005), more empathetic (Baron-Cohen, 2004), and less prone

to cognitive biases (especially overconfidence [Beyer, 2002]).

This is an interesting field of research as the conclusions of the findings can have important

implications on the auditor choice by companies, the assignment of personnel to audit tasks,

and quality control issues. However, before expanding this line of research we need to control

for a potential major bias as it is not obvious that sex differences that may be present in the

general population are also present between female and male auditors. We need to be sure that

conclusions that stem from psychological research about the average man and woman (about

aspects that potentially influence audit quality) are also applicable for auditors. In general,

psychological research can be very useful, as well for experimental accounting research (Libby

et al., 2002) as for archival accounting research (Koonce and Mercer, 2005). For the

generalization of sex differences across populations, tasks, and settings, carefulness is however

warranted. It is not self-evident that findings from the general population can be easily

interpolated to the specific context of auditors since it is clear from the large amount of

literature on sex differences that their existence and size are context-dependent. Hence, it is far

from sure that psychological research on sex differences is predictive of sex differences within

an auditor population. The purpose of this is paper is to seek an answer to the question, Can

psychological research on sex differences be a valid basis for auditing research?

The remainder of this paper is organized as follows. In the section following this

introduction, we review scientific literature from various domains to clarify the potential

impact of auditors sex on audit quality and we define our research question. Next, we describe

the research design and sample selection. Finally, we discuss our results, limitations of our

research, and possibilities for further research.

2. Literature, Theory, and Hypothesis Development

In this section, we first discuss the topic of audit quality and auditor characteristics that could

have an impact on audit quality. We then outline why sex differences in such characteristics

would be of importance for audit quality research. We further review the existing literature on

sex differences which are most crucial for auditing (viz. problem-solving, risk profile, and

independence) and analyze what impact they could have on audit quality. Finally, we formulate

the motivation for our research and our hypotheses.


4/46

3

2.1 Audit quality and auditor characteristicsIn the past, it was common to define audit quality to be the probability that the auditor

discovers and reports a material misstatement in the financial statements (DeAngelo, 1981b).

More recently however (e.g. Knechel, 2007), audit quality has been defined in terms of the

level of assurance that the auditor obtains. As the basis for the auditors opinion, professional

standards require the auditor to obtain reasonable assurance about whether the financial

statements as a whole are free from material misstatement, whether due to fraud or error.

Reasonable assurance is a high level of assurance. It is obtained when the auditor has obtained

sufficient appropriate audit evidence to reduce audit risk (that is, the risk that the auditor

expresses an inappropriate opinion when the financial statements are materially misstated) to

an acceptably low level (IFAC, 2009, ISA 200, 5).

As summarized in Fig. 1 audit quality is dependent on several personal auditor

characteristics such as the auditors technical expertise, problem-solving abilities, risk profile,

experience, and independence. Audit quality is also contingent upon several firm related

factors (e.g. the culture within an audit firm, the firm size, and the supply of non-audit

services), several audit team characteristics (e.g. the composition of the audit team and the

degree of familiarity among team members), and some factors outside the control of the

auditors (e.g. audit committees active in the reporting entity and the audit regulatory

environment) (FRC, 2008; Francis, 2004; Watkins et al., 2004). Moreover, task related

circumstances such as time pressure (DeZoort and Lord, 1997) and the complexity of

information to be processed (Bonner, 1994) might also affect audit quality. Finally, it should be

understood that all these determinants might interfere with each other.

[Insert Fig. 1 around here]

As this paper aims to identify sex differences that might affect audit quality, we will fromhereupon focus on personal auditor characteristics.2

As auditing is inherently a judgment and decision-making process, audit quality is

ultimately contingent upon the auditors judgment and decision-making qualities (Knechel,

2000). Applicable auditing standards (e.g. IFAC, 2009) and researchers (e.g. Ashton and

Ashton, 1995) have acknowledged that the work undertaken by the auditor to form an audit

opinion is permeated by judgment, in particular regarding risks of material misstatement, the

gathering of audit evidence, and the drawing of conclusions based on this evidence. The qualityof an auditors judgment and decision-making depends on personal auditor characteristics such


5/46

4

as problem-solving ability (e.g. Bierstaker and Wright, 2001; Libby and Tan, 1992), technical

expertise (e.g. Bdard and Chi, 1993; Tan and Kao, 1999), risk profile (e.g. Farmer, 1993; van

Nieuw Amerongen, 2007), experience (e.g. Early, 2002; Shelton, 1999), and independence

(e.g. DeAngelo, 1981b; Moore et al., 2006).

2.2 Sex differences in auditor characteristics

In their review article on the auditors reporting model Church et al. (2008: 75) suggested that

researchers investigate whether there is a correlation or systematic relationship between

individual characteristics (e.g., age, gender, personality, and appearance) and the recommended

report. Since the recommended report is an output variable of the audit (and thus a proxy for

audit quality) such a relationship could exist because of mediating relationships between an

individual characteristic (e.g. sex) and the auditors judgment and decision-making and

personal auditor characteristics (Fig. 2). As the audit process is affected by the specific auditor

performing a task, we believe that it would indeed be interesting for auditing researchers to

know if there are such systematic associations. Systematic associations between, for example,

sex and one or more personal auditor characteristics enable an observer in possession of

information about an auditors sex to make, ceteris paribus, a statistically more accurate

prediction as to that auditors personal characteristics (and thus audit quality) than an observer

ignorant of that auditors sex. If there are indeed significant sex differences in personal auditor

characteristics, it is reasonable to expect that an auditors sex is systematically associated with

audit quality.

[Insert Fig. 2 around here]

We shortly review some of the scientific literature that supports the idea that sex

differences exist in areas which potentially affect audit quality (viz. problem-solving, riskprofile, and independence)3.

2.3 Problem-solving

In an audit context the focus should be on mathematical problem-solving since the

understanding of financial statement matters and audit reports is most likely to be influenced

by logico-mathematical abilities (Anandarajan et al., 2008). If in the population of auditors

there would be, for example, a sex difference in mathematical problem-solving favouring


6/46

5

males it might be that male auditors discover more potential misstatements than female

auditors.

As noted by Rowley et al. (2007: 151) there is a fair amount of evidence that within our

culture, math is viewed as a male domain. If one, for examples, asks children to draw a

mathematician they will, almost without exception, draw a male figure (Picker and Berry,

2000, 2001). Meta-analyses have nevertheless shown sex differences in mathematical

performances to be somewhere between non-existent and almost non-existent (Hyde and Linn,

2006). Nonetheless it appears from a wide variety of studies that women vis--vis men have a

relatively negative perception of their own mathematical abilities (e.g. OLaughlin and

Brubaker, 1998; Miller and Bichsel, 2004). Such attitudes cause stress and anxiety. A special

case of this is the so-called stereotype threat. The existence of a stereotype leads to

heightened performance anxiety experienced by individuals who must perform a task for which

their group is thought to not be qualified (e.g. women and mathematics). The meta-analysis of

Signorella and Jamison (1986) concluded that in women an association exists between self-

image and cognitive performance. Self-identification as being a woman undermines the

mathematical desires, expectations, and abilities because math is thought of as being male

(Nosek et al., 2002). As it was eloquently put by Barbie (Mattel Inc., 1992) Math class is

tough! Hence, it is no surprise that it is well documented that women underestimate their

mathematical performances (Spencer et al., 1999). It also explains why Asian women score

better on tests that try to capture mathematical abilities when they identify themselves in the

first place as Asian vis--vis when they identify themselves in the first place as women

(Shih et al., 1999). Stereotypes influence the self-image and the behaviour of stereotyped

individuals. Thus, the mathematical expectations, preferences, and performances of women are

influenced by their implicit stereotyping as women (I am female thus math is not for me).

In sum, the existence of the stereotype image that women and mathematics do not accord

very well undermines (unconsciously) womens desire to pursue outstanding mathematicalperformances (Kiefer and Sekaquaptewa, 2007). Accordingly, it appears that sex differences in

math achievement are uniquely related to a nations average implicit stereotyping (Noseket al.,

2009) beyond the mere influence of generalized national gender inequality (i.e. the gender gap

in math is smaller in countries with a more gender-equal culture [Guiso et al., 2008]).

Moreover, since beliefs about math and women are widespread believed in society women see

their ideas and stereotypes about women and math being confirmed, reinforced, and stimulated

by their environments. Teachers, for example, ascribe mathematical excellence of femalesrather to effort while they ascribe the mathematical excellence of males rather to talent


7/46

6

(Hamilton, 2008). Such different approach is, of course, not without consequences since

effort is, by definition, restricted to certain limits, while talent is not. So, it is no surprise

that the mathematical performances of women are negatively influenced when they are told

that sex differences in mathematical performances are caused by genetic differences while

there is no effect when women are told that such differences are due to experiential causes

(Dar-Nimrod and Heine, 2006). Males are therefore more than females stimulated to further

develop their mathematical abilities, resulting in a vicious stimulus-response-cycle. Thus,

while this remains a very complex area with many unresolved issues (i.e. unknown causes and

relationships that are not well understood, cf. Ceci et al. [2009] for a recent review) it is no

surprise that most research (e.g. Penner and Paret, 2008) finds men, on average, to be

somewhat better mathematical problem-solvers than women.

2.4 Risk profile

Just as mathematical ability, risk taking is perceived as a masculine attribute, at least in

Western cultures (Wilson and Daly, 1985). Notwithstanding the fact that there are some claims

(e.g. Schubert et al., 1999) that differences in risk attitudes between males and females are

more prejudice than fact, and that there are some claims (e.g. Roszkowski and Grable, 2005)

that sex differences in risk attitude are real but smaller than most of the time is assumed, there

seems to be evidence that women, in general, are more risk-averse than men (Barsky et al.,

1997; Byrnes et al., 1999; Damodaran, 2008; Jianakoplos and Bernasek, 1998).

Plenty of experiments on risk taking have shown that women are more risk-averse than

men in a wide variety of domains. Women are less likely to donate blood (Nonis et al., 1996),

take fewer risks in gambling (Eckel and Grossman, 2002), and are more risk-averse when

investing (Barber and Odean, 1995; Graham, 2002), even when they are financial professionals

(Olsen and Cox, 2001). However, the average effect sizes for sex differences in risk taking are

not large and vary according to context (Byrnes et al., 1999). In fact, males are no more likelythan females to have a domain-general tendency towards risk-taking. Sex differences in risk-

taking may have more to do with the fact that males are more often confronted or seek out

risky situations than females (Boyer and Byrnes, 2009). Indicative for the importance of

context, Schubert et al. (1999) showed that under controlled economic conditions females do

not make less risky financials choices than males.

Weber et al. (2002) suggested that risk-taking is indeed domain-specific. With exception

however of the social domain, their study found females to be less likely to engage in riskybehaviour than males in all domains. A pilot study conducted by Gold et al. (2009) to test


8/46

7

various assumptions on which their research hypotheses and experimental manipulations

relied, suggested that also female auditors are more risk averse than male auditors.

In risk-assessment the goal of the auditor should be to identify those domains that are more

likely to contain material misstatements. However, the fact that women tend to be more risk-

averse possibly leads to a different selection of samples. The existence of the misstatements in

the financial statements of the reporting entity is unaffected by the auditors sex, but if men and

women select other or larger/smaller samples (because of differences in risk propensity), then

the probability that material misstatements will be detected might be affected by the auditors

sex. As female auditors tend to be more risk-averse, female auditors might be expected to set a

lower materiality threshold and select larger samples than male auditors. This could result in a

higher number of material misstatements detected and reported by female auditors than by

male auditors. The question then arises however to which extent female auditors might violate

the cost-benefit criterion more than male auditors, as they may desire more evidence than what

is sufficient appropriate audit evidence needed to obtain reasonable assurance.

Furthermore, even if the data considered by female and male auditors is identical, female

auditors (being more risk-averse) might assess the same type of information or misstatement as

more serious, thereby expressing more severe audit opinions, including going-concern

disclosures4, than male auditors.

2.5 Independence

Reporting material misstatements depends, among other things, on auditor independence.

Concerning independence two issues are at stake: malevolent behaviour and unconscious

behaviour. Auditor independence may not be possible at all (Moore et al., 2006), but that is an

irrelevant consideration as in the context of sex differences the question is: Are female

auditors less/more likely to be independent from their client?

If female auditors are less likely to display malevolent behaviour, there has to be evidencethat they are less often tempted to pursue some interest of their own and/or that they are bribed

less easily. Some evidence indeed points in that direction: women are less likely to be asked for

a bribe than men (Mocan, 2008) and women hold more negative attitudes toward cheating than

men (Whitley, 2001). In their review on ethical decision-making OFallon and Butterfield

(2005: 379) concluded that There are often no differences found between males and females,

but when differences are found, females are more ethical than males. Specifically with regards

to the field of accounting, Venezia (2008) found that female accounting students possessed


9/46

8

higher levels of ethical reasoning than male accounting students and Eynon et al. (1997) found

far greater moral reasoning skills displayed by female accountants than by male accountants.

Rather than malevolent practice, the real problem however is an unconscious lack of

independence (Moore et al., 2006). In the context of unconscious behaviour, the primary focus

has to be on a priori-assumptions and (cognitive) biases that interfere with an objective

independent judgment.5 It is widely known and documented that prior beliefs bias the

evaluation of arguments and data. In evaluating new information, we draw upon our

background knowledge and use schemata in order to fill in details that are not there. In doing

so our thought patterns can become faulty as we suffer from cognitive distortions (i.e. our

thinking is not rational) (Gazzaniga and Heatherton, 2003).6 Cognitive biases may, for

example, hinder the issuing of a going-concern opinion, even when the auditor is aware of a

clients uncertain financial position (Kleinman et al., in press).

Some evidence points out that men and women tend to differ in this unconscious

behaviour. Psychological research indicates that men are more likely to suffer from cognitive

distortions and to exhibit dichotomous thinking than women (Chung and Monroe, 1998; Soutar

and Sweeney, 2003). The study of Chung and Monroe (1998) found evidence indicating that

male students enrolled in a third year undergraduate auditing class displayed confirmation bias,

while female students did not. Therefore, female auditors might discover more material

misstatements than male auditors. Especially with regards to overconfidence this sex difference

is well-documented (e.g. Bengtsson et al., 2005; Beyer, 2002). The overconfidence effect

describes, among others, the tendency of people to believe that their judgment is more accurate

than it really is. As a result, overconfidence can create a mismatch between ones confidence in

ones own judgments and the real accuracy of these judgments. Because women are less

overconfident than men, female auditors could be expected to be more reluctant than male

auditors when deciding to report a material misstatement.

A second issue that may unconsciously undermine auditor independence is empathy. It isbelieved that long-term relationships result in increasing empathy between parties; therefore

audit quality would get undermined when an auditor gets too familiar with a company

(Richard, 2006). To ensure auditor independence mandatory rotation of the audit firm, or of

key personnel, is therefore frequently advocated. Scientific evidence from the Empathizing-

Systemizing theory (Baron-Cohen, 2004; Chapman et al., 2006; Nettle, 2007) and from the

psychological Big Five taxonomy of personality traits (Costa et al., 2001; Feingold, 1994;

Schmitt et al., 2008) supports that women are more empathetic than men. Female auditorsmight therefore identify themselves more with their clients than male auditors and therefore


10/46

9

report less material misstatements and/or append fewer going-concern disclosures to a clients

audit report than male auditors.

2.6 Hypothesis formulation

The recent finding of Ittonen and Peni (2009) (i.e. in the Nordic countries firms with female

audit partners seem to have higher audit fees) gives support to the idea that an auditors sex is

systematically associated with audit quality. Some preliminary research on sex differences in

the domain of auditor judgment and decision-making has, however, generated mixed results.

Gold et al. (2009) found the final proposed adjusting journal entry recommendation of male

auditors to be more accurate than that of female auditors. Chung and Monroe (2001) found

females audit judgments to be more accurate in more complex tasks than those of males, while

in the less complex tasks they found males to be more accurate than females. ODonnell and

Johnson (2001) found that female auditors were also more efficient than male auditors in

completing high complexity tasks, while male auditors were more efficient than female

auditors in completing tasks of low complexity. Breesch and Branson (2009) however, reported

no association between sex and the number of reported material misstatements or with the

severity of the expressed audit opinion.

The above-mentioned studies took it for granted that the sex of an auditor is somehow

systematically associated with some personal auditor characteristic (e.g. risk-aversion). While

stereotypical beliefs about sex differences in personal characteristics our affluent in our

contemporary society, it can not be safely assumed that the behaviour of auditors is biased in

the same gendered directions as the behaviour of the average man and the average woman.

There are two distinct contextual effects that may prevent this. First, because of two reasons

there might be correlational effects between certain personal characteristics and the sex of an

auditor. Firstly, there is a statistical reason, i.e. auditors are not a random sample. This makes it

impossible to interpolate findings about sex differences in personal characteristics from thewhole population to the specific context of female and male auditors. After all, considerable

self-selection might be present, i.e. people (males and females) with more or less the same

profile chose to become auditors. Secondly, certain sex differences may be absent in specific

populations as a result of (parental, educational, and occupational) socialization and the

resulting stimulus-response-cycle, i.e. people (males and females) are learned certain roles and

acquire certain abilities and since environments are not experienced randomly they reinforce

these roles and strengthen these abilities. In this respect, Smith and Rogers (2000: 76) wrotethat occupational socialization theory illustrates that differences tend to disappear as


11/46

10

employees are socialized within the work environment, the same may be said for those

working in the accounting profession. Second, there might be interactional effects between

auditors gender and the organizational logic of audit firms, i.e. the way that gender stereotypes

bias behaviour and performances in gendered directions depends on how strongly a certain

setting is culturally typed as male or female (Ridgeway, 2009). Thus, even if auditors were a

random sample of the population we would not be able to conclude that sex differences in

certain personal characteristics are present between males and females in an audit context. One

would have to examine how the organizational logic interacts with the background effects of

the gender frame (Ridgeway, 2009: 156) since the prevalence of such sex differences is

context-dependent. This is true as well for sex differences in mathematical problem-solving,

risk taking, and different components of independence (ethical behaviour, cognitive biases, and

empathy).

Altogether this makes it impossible to determine a priori if an auditors sex is relevant

information when one is interested in audit quality, i.e. does knowing that an auditor is male or

female, ceteris paribus, enable an observer to make a statistically more accurate prediction as

to the audit quality supplied by that person vis--vis an observer who is ignorant of that fact.

On the one hand it appears that the answer is yes since it is believed that sex differences exist

in certain personal characteristics (viz. problem-solving, risk profile, and independence) which

potentially affect audit quality. On the other hand one will be inclined to say no since

(parental, educational, and occupational) socialization can override the influence of sex (i.e.

women and men in the same profession are more alike than men and women randomly

sampled from the whole population). Certainly in the world of auditing where there is a

longstanding tradition of the large firm quality homogeneity assumption (i.e. the assumption

that audit quality within large audit firms is reasonably homogeneous) one will be willing to

accord to the second answer. However, based on the maleness of the auditing world (cf.

Hardies et al., 2009) it would be just as good to accord to the first answer because this maylead one to believe that the background gender frame is powerfully relevant in the context of

auditing. It is however impossible to favour one position more than the other solely on

theoretical grounds. There is no reason to believe that in the context of the audit profession

women will be women in their profession (Beckmann and Menkhoff, 2008) is truer than

female auditors help their colleagues to forget that they are women (Anderson-Gough et al.,

2005).

Thus, knowing an auditors sex may enable an observer to make more accurate predictionsvis--vis an observer who is ignorant of this information in some cases, but it may just as well


12/46

11

be no more relevant than the hair colour of the auditor. So before any further research in the

area of gender and auditing is undertaken we have to address the question: Are female

auditors still women? The magnitude of sex differences might be the same, smaller, larger or

even reverse for the population of auditors as it is for the general populationgender

researchers have emphasized the importance of context in creating, erasing, or even reversing

(psychological) sex differences (e.g. Bussey and Bandura, 1999; Deaux and Major, 1987;

Eagly and Wood, 1999). Based upon the stereotypical beliefs that are embedded in our

contemporary society and the enormous amount of scientific literature that has reported sex

differences in mathematical performances, risk taking, cognitive distorted thinking, and

empathy, we expect to find such sex differences between males and females randomly sampled

from the population as a whole. If auditors differ from the general population and sex

differences are, due to self-selection and/or socialization, not present in the population of

auditors this may (to some degree) also be the case in the population of business students since

auditors are only a very specific subgroup of (former) business students. Indeed it appears that

students in fields of economics are not even representative of the population of students, both

in terms of their attitudes (e.g. their appraisal of the virtues of the market system [Cipriani et

al., 2009]) as in terms of their personal characteristics (e.g. their selfishness [Frey and Meier,

2003, 2005], their tolerance to economic risk taking [Sjberg and Engelberg, 2009]). Since this

is probably due to self-selection rather than due to indoctrination (Cipriani et al., 2009; Frey,

2003, 2005), it can be assumed that people (males and females) who attend business schools

are rather similar. Thus we expect sex differences in a population of business students to be of

a magnitude somewhere in-between those of the general population and the population of

auditors.

3. Research design and sample description

The utilized data consists of a unique set of responses gathered in Belgium and the Netherlandsfrom 677 non-business students, 173 business students, and 1051 auditors. While university

students can not be expected to be representative on these traits for the general population, we

used in our research a sample of (non-business) students as a proxy for the population as a

whole because psychological research has heavily relied on the study of college students

(Peterson, 2001; Sherman et al., 1999). Hence, using such a sample yields the greatest

possibility to capture similar results, and makes it possible to argue strongly in favour of

contextual reasons for the presence or absence of sex differences (if in one or more, but not all,of our subpopulations they would be absent; since in that case they can not solely be ascribed


13/46

12

to the specifics of our research instrument). For the same reason, our research instrument was

composed out of generic (i.e. non-audit related) questions. The usage of such generic

questions, that have been widely used in psychological research on sex differences, made it

possible to compare our results across our different subpopulations as well as with previous

(psychological) research (that has served as a starting point for the research on auditor gender

undertaken so far). Hence, our research instrument is well-suited to examine if psychological

research can serve as a valid foundation for hypotheses about (and explanations for) sex

differences among auditors.7

Both student groups were surveyed by means of a paper or online version of our

questionnaire (included in Appendix) in November 2008. We compared results from both

versions to make sure that our research instrument did not yield different results depending on

the fact if one answered it online or on paper. Since no such differences were found, we used

the online version for our examination of the auditor population. For subpopulations with high

levels of Internet access, Internet-based research has the potential to yield higher-quality data

with lower non-response rates and at a lower cost than traditional methods (Schmidt, 1997). In

Belgium the external auditor is chosen from the members of the professional body of Belgian

auditors IBR (Instituut van de Bedrijfsrevisoren). We distributed our survey among the Dutch

speaking Belgian auditors (which represents almost 75% of the total population of Belgian

auditors). In the Netherlands 98% of all external audits is performed by a member of the

professional body of Dutch auditors NIVRA (Nederlands lnstituut van Registeraccountants).

We used the member lists of the IBR and the NIVRA and invited 7400 auditors in this manner

to participate in our investigation, 1241 of them responded (17% response rate). 8 It is not very

likely that our results are affected by non-response bias (which occurs when the individuals

responding to a survey differ from nonrespondents on variables relevant to the survey topic

[Rogelberg and Luong, 1998: 6061]) since we are dealing with a very homogeneous group.

Hence, traditional variables that cause non-response bias are identical (e.g. occupation) or atleast very similar (e.g. amount of education) for the whole population under examination and

can not be relevant. Approximately 18% of all the auditors in the Low Countries are female;

our sample is representative in this respect (p = .514).

[Insert Table 1 around here]

Table 1 presents descriptive statistics of our entire sample. We found female auditors, onaverage, to be younger and less experienced than their male counterparts. The proportion of


14/46

13

female and male auditors that were interns at a B4 audit firm did not significantly differ.

Finally, we found that female auditors hold lower positions in audit firms than male auditors.

Altogether we generated 1901 valid questionnaires. The questionnaire was designed to

examine three dimensions: problem-solving ability (including mathematical abilities), risk

profile (risk-aversion), and independence (through empathy and cognitive biases). To measure

problem-solving ability we constructed a scale ( = .609) counting the number of correct

answers on the ten questions regarding problem-solving. Eleven Likert-items were cumulated

into a Likert-scale ( = .720) to measure risk-aversion. To evaluate empathy we used one item

of Baron-Cohen and Wheelwrights (2004) Empathy Quotient. Independence was further

examined with seven questions testing for cognitive biases in thinking. In the population of

non-business students sex differences were expected to be found in all these areas. These sex

differences would or would not be replicated throughout the other two populations. The results

of the survey are discussed below.

4. Results

It is now well-established and broadly acknowledged that statistical significance testing is often

misused and misinterpreted (Nickerson, 2000; Thompson, 2007). Following the advices of

Cohen (1990) and Feingold (1995), therefore, when we present univariate comparisons we

supplement the reporting of statistical significance with confidence intervals (CI.975), and with

standardized effect sizes (Cohens dor odds ratios, depending on the appropriateness) and their

confidence intervals (CI.975). Since estimates of effect-size provide a means for interpreting

statistical results regardless of statistical significance (Robey, 2004), we report effect sizes

even when a statistical test failed to achieve statistical significance. Furthermore, to enhance

comparability, we also computed standardized effect sizes (Cohens d) when regression

coefficients (i.e. unstandardized measures) are given. We derived 95% confidence intervals for

(based on Fishers z-transformation) from partial point-biserial correlations, which we

converted into Cohens ds (using Friedmans [1968: 246] formula).

4.1 Problem-solving

In the population of non-business students we found men to be better problem-solvers than

women (Table 2).



15/46

14

We also found the sex difference in problem-solving to narrow when considering more

specified population categories: the difference was smaller for the business students (Cohens d

= 0.23 0.30) (p = .133) than for the non-business students (Cohens d= 0.45 0.17) (p =

.000), and negligible for the auditors (Cohens d= 0.02 0.22) (p = .811) (Table 3).


To control for mediating influences we incorporated the demographic and career

characteristics (Table 1) in a multivariate framework (Table 4). Controlling for these

demographic and career characteristics confirmed the results from the univariate analysis:

while in the populations of non-business students (Cohens d= 0.34 0.17) (p = .001) and

business students (Cohens d= 0.20 0.30) (p = .001) (p = .065) females scored lower on the

problem-solving test than their male counterparts, this was not true for the female auditors

(Cohens d= 0.13 0.16) (p = .273). As could be expected (cf. Verhaeghen and Salthouse,

1997) problem-solving ability was negatively influenced by age, although the effect was

small.9


4.2 Risk profile

As expected, women were, on average, found to be more risk-averse than men. Female non-

business students (Cohens d= 0.31 0.16) (p = .000) as well as female business students

(Cohens d = 0.47 0.31) ( p = .008) were found to be more risk-averse than their male

counterparts. We found, however, no difference in risk-aversion between female and male

auditors (Cohens d= 0.08 0.16) (p = .354) (Table 5).


In order to control for possible mediation of demographic and career characteristics, we

analyzed sex differences in risk-aversion in a multivariate framework (Table 6) similar to the

above. The multivariate framework confirmed the results of our univariate analysis only partly:

in the populations of non-business students (Cohens d= 0.31 0.16) (p = .000) and business

students (Cohens d= 0.47 0.31) (p = .005), females were indeed found to be significantlymore risk-averse than males. However, as could be expected, older individuals were found to


16/46

15

be more risk-averse than younger ones (cf. Steinberg, 2007). It was also found that Dutch

auditors indicated a greater willingness to take economic risks than their Belgium colleagues.

When controlling for these mediating characteristics, female auditors were found to be

significantly more risk-averse than male auditors (Cohens d= 0.14 0.12) (p = .025).


4.3 Independence

In none of our three subpopulations statistical testing revealed women to be significantly more

empathetic than men (Table 7) and the effect sizes were small as well in the population of non-

business students (Cohens d = -0.12 0.18), as in the population of business students

(Cohens d= -0.21 0.32), as in the population of auditors (Cohens d= -0.18 0.19).


As expected, our experiment revealed significant differences between men and women for

some of the cognitive biases (Table 8). However, in the population of non-business students,

we found females to be somewhat more cognitive biased than their male counterparts: they

tended to be somewhat more influenced by the conjunctive fallacy (OR = 1.91) (p = .009),showed greater insensitivity to sample size (OR = 1.42) (p = .098), and suffered more from the

information bias (OR = 1.51) ( p = .073) than there male counterparts. In the population of

business students such differences were only found with regards to the conjunctive fallacy (OR

= 2.12) ( p = .077). On the other hand, as expected, in the non-business students population

females were found to be somewhat better calibrated (i.e. less overconfident)10 than males (OR

= 0.60) (p = .025). This result was replicated throughout the population of business students

(OR = 0.38) (p = .012), but not in the auditor subpopulation (OR = 1.23) (p = .338). In the

auditor population females were also found to be somewhat more influenced by the

conjunctive fallacy (OR = 1.99) (p = .063), and showed greater insensitivity to sample size

(OR = 1.61) (p = .055) than there male counterparts.


5. Conclusion


17/46

16

Throughout this paper we analyzed a number of personal auditor characteristics that are

potentially associated with the auditor being male or female. For a non-biased gender auditing

research, we need however to be sure that conclusions for the average men and women also

apply to auditors. It is not straightforward to determine a priori if auditors are a random sample

of the general population with respect to relevant characteristics (see supra) for audit quality.

Therefore, we conducted a survey whereby different groups of males and females were asked

to fill in a questionnaire. Three groups of people were distinguished: non-business students,

business students, and auditors. The questionnaire examined three dimensions: problem-

solving ability (including mathematical abilities), risk profile (risk-aversion), and independence

(through empathy and cognitive biases). The results of our experiment were somewhat mixed

(Table 9).


On the basis of our experiment we can not unambiguously answer the question Are female

auditors still women? We found, in accordance with Gold et al. (2009), female auditors to be

somewhat more risk-averse than male auditors, but we did not find sex differences in other

personal auditor characteristics. To a great degree our results accord with those of Beckmann

and Menkhoff (2008) who found female fund managers to be more risk averse, but not less

overconfident than their male colleagues. Our conclusion, however, differs from theirs. Since

female auditors differ less from their male colleagues than the average women from the

average men and since female auditors are more alike than different from their male

colleagues, we conclude that women will not be women in the audit profession: being an

auditor dominates doing gender. Limitations of our study have however to be taken into

account. It is without doubt that our research instrument is a source for criticism, and it can not

be ruled out that we would have found some significant sex differences if we had used anotherresearch instrument; for example, no sex differences in empathy were found, but we measured

empathy by a measure that is self-report while it is not yet clear how self-report of empathizing

relate to actual performance (Wakabayashi et al., 2006). More important, while our research

points to the importance of context when examining sex differences, and indicates that sex

differences may not be present between male and female auditors (even when they exist in the

general population), it can not rule out that there are sex differentiated audit quality differences

(due to differences in problem-solving ability, risk-aversion, and/or independence betweenmale and female auditors). After all, the prevalence of sex differences depends on the way that


18/46

17

behaviour and performances are biased in gendered directions, which depends on how strongly

a certain setting is culturally typed as male or female (Ridgeway, 2009). Therefore, the priming

of one aspect of someones identity (e.g. women stereotype) may undermine performance

while another aspect of someones identity (e.g. accountant stereotype) may enhance it. We

can only speculate about which aspect of their identity was primed when auditors took our

survey. Thus, it may be that stereotypical gender-roles are made salient in an audit context,

while they were not activated in our experimental setting.

Furthermore, one should be very careful to jump to conclusions based on this experiment

alone. It is wise to remember Schmidts (1996: 127) admonition that meta-analysis has

revealed how little information there typically is in any single study. It has shown that, contrary

to widespread belief, a single primary study can rarely resolve an issue or answer a question.

Any individual study must be considered a data point to be contributed to a future meta-

analysis. More research into this topic is certainly necessary. Further research could also focus

on more traditional measures of audit quality. Chin and Chi (2008) found for example an

association between the sex of the auditor and the likelihood that a going-concern opinion was

issued. Finally, even if auditor gender does not differentiate actual audit quality it might have

an impact on how audit quality is perceived. After all, it is a well-known fact that performance

judgments (cf. Bauer and Baltes 2002) and perceptions on ethical behaviour (cf. McCabe et al.,

2006) are gender-biased. Female and male auditors might therefore not be perceived as

supplying audits of the same quality (cf. Hardies et al., 2009).

Notes

1Stereotypes distinguish a particular group from other groups; they need not be negative or inaccurate (Judd

and Park, 1993).

2

We acknowledge that it would be interesting to study auditor gender in relationship with task related

circumstances and audit team characteristics, ifthere are differences between male and female auditors. Such

relationships are, however, not discussed in this paper since the central claim of this paper is that we do not yet

know if there are differences between male and female auditors.

3As technical expertise and experience are individual characteristics that are rather acquired during lifespan

than sex differentiated characteristics we omit them from our analysis here and incorporate them back in our

empirical analysis (see infra) as independent variables.


19/46

18

4If a client is nearing bankruptcy, it is less costly for auditors to be conservative since the type II error cost (i.e.

misclassification of a failing company as a non-failing company) is typically larger than the type I error cost

(i.e. misclassification of a non-failing company as a failing company) (Matsumara et al., 1997).

5Auditor independence is the conditional probability that an auditor reports a discovered breach in a clients

accounting system (DeAngelo, 1981a). Therefore, we regard any cognitive bias that may undermine the

probability that an auditor reports a discovered material misstatement as a potential threat to auditor

independenceauditor independence is broader than the ability to withstand clients pressure when forming

judgments and making decisions.

6The widely cited review of Smith and Kida (1991: 485) concluded that although the evidence indicates that

the heuristics and biases common to many experiments using student subjects and generic tasks are also

present in the judgments of professional auditors performing familiar, job-related tasks, the nature of these

heuristics or the extent of their presence is often notably different. From the more recent review of Koch and

Wstemann however (2009) it appears that auditors make as much use of heuristics as laymen, and

consequently, are also prone to cognitive biases, even when conducting audit tasks.

7Clearly, this would not be the case if we made use of audit related questions or an actual audit task

comparisons with non-audit subjects and non-auditing literature would be meaningless.

8600 Belgian and 6800 Dutch auditors were invited. This big numerical difference is partly due to a difference

in country size (approximately 6 million people live in the Dutch speaking part of Belgium compared to more

than 16 million in the Netherlands), and partly due to the fact that the member list of the IBR only contains

practicing external auditors while the member list of the NIVRA also contains individuals who are legally

qualified to perform statutory audits but do not work as external auditor.

9These results, as well as those reported later, are robust to the exclusion of different variables from the

regression.

10Calibration is a criterion for evaluation of probability judgments. Overconfident individuals have been found

to miscalibrate by giving too narrow confidence intervals (e.g. Beckmann and Menkhoff, 2008).

References

Anandarajan, A., Kleinman, G. and Palmon, D. (2008) Novice and expert judgment in the presence of

going concern uncertainty: The influence of heuristic biases and other relevant factors,

Managerial Auditing Journal, 23(4/5), pp. 345366.


20/46

19

Anderson-Gough, F., Grey, C. and Robson, K. (2005) Helping them to forget..: the organizational

embedding of gender relations in public audit firms, Accounting, Organizations and Society,

30(5), pp. 469490.

Ashton, R. H. and Ashton, A. H. (1995) Perspectives on Judgment and Decision-Making Research in

Accounting and Auditing, in: Judgment and Decision-Making Research in Accounting and

Auditing, pp. 325 (New York: Cambridge University Press).

Barber, B. M. and Odean, T. (2001) Boys Will be Boys: Gender, Overconfidence, and Common Stock

Investment, Quarterly Journal of Economics, 116(1), pp. 261292.

Baron-Cohen, S. (2004) The Essential Difference. Male and Female Brains and the Truth About

Autism (New York: Basic Books).

Baron-Cohen, S. and Wheelwright, S. (2004) The Empathy Quotient: An Investigation of Adults with

Asperger Syndrome or High Functioning Autism, and Normal Sex Differences, Journal of

Autism and Developmental Disorders, 34(2), pp. 163175.

Barsky, R. B., Juster, F. T., Kimball, M. S. and Shapiro, M. D. (1997) Preference Parameters and

Behavioral Heterogeneity: An Experimental Approach in the Health and Retirement Survey,

The Quarterly Journal of Economics, 112(2), pp. 537580.

Bauer, C. C. and Baltes, B. B. (2002) Reducing the Effects of Gender Stereotypes on Performance

Evaluations, Sex Roles, 47(9/10), pp. 465476.

Beckmann, D. and Menkhoff, L. (2008) Will Women Be Women? Analyzing the Gender Difference

among Financial Experts, Kyklos, 61(3), pp. 364384.

Bdard, J. and Chi, M. T. H. (1993) Expertise in Auditing, Auditing - a Journal of Practice and

Theory, 12(supplement), pp. 2145.

Bengtsson, C., Persson, M. and Willenhag, P. (2005) Gender and Overconfidence, Economics letters,

86(2), pp. 199203.

Beyer, S. (2002) The Effects of Gender, Dysphoria, and Performance Feedback on the Accuracy of

Self-Evaluations, Sex Roles, 47(9), pp. 453464.

Bierstaker, J. L. and Wright, S. (2001) A Research Note Concerning Practical Problem-Solving

Ability as a Predictor of Performance in Auditing Tasks, Behavioral Research in Accounting,

13, pp. 4962.

Bonner, S. E. (1994) A Model of the Effects of Audit Task Complexity, Accounting, Organizations

and Society, 19(3), pp. 213234.

Boyer, T. W. and Byrnes, J. P. (2009) Adolescent risk-taking: Integrating personal, cognitive, and

social aspects of judgment,Journal of Applied Developmental Psychology, 30(1), pp. 2333.


21/46

20

Breesch, D. and Branson, J. (2009) The Effects of Auditor Gender on Audit Quality, The IUP Journal

of Accounting Research & Audit Practices, 8(3/4), pp. 78107.

Byrnes, J. P., Miller, D. C. and Schafer, W. D. (1999) Gender Differences in Risk Taking: A Meta-

Analysis, Psychological Bulletin, 125(3), pp. 367383.

Bussey, K. and Bandura, A. (1999) Social Cognitive Theory of Gender Development and

Differentiation, Psychological Review, 106(4), pp. 676712.

Ceci, S. J., Williams, W. M. and Barnett, S. M. (2009) Womens Underrepresentation in Science:

Sociocultural and Biological Considerations, Psychological Bulletin, 135(2), pp. 218261.

Chapman, E., Baron-Cohen, S., Auyeung, B., Knickmeyer, R., Taylor, K. and Hackett, G. (2006) Fetal

Testosterone and Empathy: Evidence from the Empathy Quotient (EQ) and the Reading the

Mind in the Eyes Test, Social Neuroscience, 1(2), pp. 135148.

Chin, C. and Chi, H. (2008) Sex Matters: Gender Differences in Audit Quality. Working paper

presented at the 2008 Annual Meeting of the American Accounting Association, August 36,

Anaheim.

Chung, J. and Monroe, G. (1998) Gender Differences in Information Processing: An Empirical Test of

the Hypothesis-Confirming Strategy in an Audit Context, Accounting & Finance, 38(2), pp.

265279.

Chung, J. and Monroe, G. (2001) A Research Note on the Effects of Gender and Task Complexity on

an Audit Judgment,Behavioral Research in Accounting, 13, pp. 111126.

Church, B. K., Davis, S. M. and McCracken, S. A. (2008) The Auditors Reporting Model: A

Literature Overview and Research Synthesis,Accounting Horizons, 22(1), pp. 6990.

Cipriani, G. P., Lubian, D. and Zago, A. (2008) Natural born economists? Journal of Economic

Psychology, 30(3), pp. 455468.

Cohen, J. (1990) Things I Have Learned (So Far),American Psychologist, 45(12), pp. 13041312.

Costa, P. T. Jr., Terracciano, A. and McCrae, R. R. (2001) Gender Differences in Personality Traits

Across Cultures: Robust and Surprising Findings, Journal of Personality and Social

Psychology, 81(2), pp. 322331.

Damodaran, A. (2008) Strategic Risk Taking. A Framework for Risk Management (New Jersey:

Pearson Education).

Dar-Nimrod, I. and Heine, S. J. (2006) Exposure to Scientific Theories Affects Womens Math

Performance, Science, 314(5798), p. 435.

DeAngelo, L. E. (1981a) Auditor Independence, low balling, and disclosure regulation, Journal of

Accounting and Economics, 3(2), pp. 113127.


22/46

21

DeAngelo, L. E. (1981b) Auditor Size and Audit Quality,Journal of Accounting and Economics, 3(3),

pp. 183199.

Deaux, K. and Major, B. (1987) Putting gender into context: An interactive model of gender-related

behavior, Psychological Review, 94(3), pp. 369389.

DeZoort, T. F. and Lord, A. T. (1997) A Review and Synthesis of Pressure Effects Research in

Accounting,Journal of Accounting Literature, 16, pp. 2885.

Eagly, A. H. and Wood, W. (1999) The Origins of Sex Differences in Human Behavior. Evolved

Dispositions versus Social Roles,American Psychologist, 54(6), pp. 408423.

Early, C. E. (2002) The Differential Use of Information by Experienced and Novice Auditors in the

Performance of Ill-Structured Audit Tasks, Contemporary Accounting Research, 19(4), pp.

595614.

Eckel, C. C. and Grossman, P. J. (2002) Sex Differences and Statistical Stereotyping in Attitudes

Toward Financial Risk,Evolution and Human Behavior, 23(4), pp. 281295.

Eynon, G., Hill, N. T. and Stevens, K. T. (1997) Factors that Influence the Moral Reasoning Abilities

of Accountants: Implications for Universities and the Profession, Journal of Business Ethics,

16(12/13), pp. 12971309.

Farmer, T. A. (1993) Testing the Effect of Risk Attitude on Auditor Judgment Using Multiattribute

Utility Theory,Journal of Accounting, Auditing & Finance, 8(1), pp. 91110.

Feingold, A. (1994) Gender Differences in Personality: A Meta-Analysis, Psychological Bulletin,

116(3), 429456.

Feingold, A. (1995) The Additive Effects of Differences in Central Tendency and Variability Are

Important in Comparisons Between Groups,American Psychologist, 50(1), pp. 513.

Francis, J. R. (2004) What Do We Know About Audit Quality?British Accounting Review, 36(4), pp.

345368.

FRC (2008) The Audit Quality Framework(London: Financial Reporting Council).

Frey, B. S. and Meier, S. (2003) Are Political Economists Selfish and Indoctrinated? Evidence from aNatural Experiment,Economic Inquiry, 41(3), pp. 448462.

Frey, B. S. and Meier, S. (2005) Selfish and Indoctrinated Economists?European Journal of Law and

Economics, 19(2), pp. 165171.

Friedman, H. (1968) Magnitude of experimental effect and a table for its rapid estimation,

Psychological Bulletin, 70(4), pp. 245251.

Gazzaniga, M. S. and Heatherton, T. F. (2003) Psychological Science: Mind, Brain, and Behaviour

(New York: W.W. Norton & Company).


23/46

22

Gold, A., Hunton, J. E. and Gomaa, M. (2009) The Impact of Client and Auditor Gender on Auditors

Judgments,Accounting Horizons, 23(1), pp. 118.

Graham, J. F. (2002) Gender Differences in Investment Strategies: An Information Processing

Perspective,International Journal of Bank Marketing, 20(1), pp. 1726.

Guiso, L., Monte, F., Sapienza, P. and Zingales, L. (2008) Culture, Gender, and Math, Science,

320(5880), pp. 11641165.

Hamilton, C. (2008) Cognition and sex differences (Basingstoke: Palgrave Macmilla).

Hardies, K., Breesch, D. and Branson, J. (2009) Male and Female Auditors: Who in this Land is the

Fairest of All?Accountancy & Bedrijfskunde, 29(7), pp. 2230.

Hyde, J. S. and Linn, M. C. (2006) Gender Similarities in Mathematics and Science, Science,

314(5799), pp. 599600.

International Federation of Accountants (IFAC) (2009) Handbook of International Standards on

Auditing and Quality Control (New York: IFAC).

Ittonen, K. and Peni, E. (2009)Auditors Gender and Audit Fees. Working paper presented at the 32nd

Annual Congress of the European Accounting Association, May 1215, Tampere.

Jianakoplos, N. A. and Bernasek, A. (1998) Are Women More Risk Averse?Economic Inquiry, 36(4),

pp. 620630.

Judd, C. M. and Park, B. (1993) Definition and Assessment of Accuracy in Social Stereotypes,

Psychological Review, 100(1), pp. 109128.

Kiefer, A. K. and Sekaquaptewa, D. (2007) Implicit stereotypes and womens math performance: How

implicit gender-math stereotypes influence womens susceptibility to stereotype threat,Journal

of Experimental Social Psychology, 43(5), pp. 825832.

Kleinman, G., Anandarajan, A., Medinets, A. and Palmon, D. (in press) A Theoretical Model of

Cognitive Factors that Affect Auditors Performance and Perceived Independence.

International Journal of Behavioural Accounting and Finance.

Knechel, R. W. (2000) Behavioral Research in Auditing and Its Impact on Audit Education, Issues in

Accounting Education , 15(4), pp. 695712.

Knechel, R. W. (2007) The Business Risk Audit: Origins, Obstacles and Opportunities, Accounting,

Organizations and Society, 32(4), pp. 383408.

Koch, C. and Wstemann, J. (2009) A review of bias research in auditing: Opportunities for

combining psychological and economic research, Working paper presented at the Midyear

Conference of the American Accounting Association Auditing Section, January 1517, St.

Petersburg (FL).


24/46

23

Koonce, L. and Mercer, M. (2005) Using Psychology Theories in Archival Financial Accounting

Research,Journal of Accounting Literature, 24, pp. 175214.

Libby, R. and Tan, H. T. (1992) Modeling the Determinants of Audit Expertise, Accounting,

Organizations and Society, 19(8), pp. 701716.

Libby, R., Bloomfield, R. and Nelson, M. W. (2002) Experimental research in financial accounting,

Accounting, Organizations and Society, 27(8), pp. 775810.

Matsumura, E. M., Subramanyam, K. R. and Tucker, R. R. (1997) Strategic Auditor Behavior and

Going-Concern Decisions,Journal of Business Finance & Accounting, 24(6), pp. 727758.

McCabe, A. C., Ingram, R. and Dato-on, M. C. (2006) The Business of Ethics and Gender,Journal of

Business Ethics, 64(2), pp. 101116.

Miller, H. and Bichsel, J. (2004) Anxiety, working memory, gender, and math performance,

Personality and Individual Differences, 37(3), pp. 591606.

Mocan, N. (2008) What Determines Corruption? International Evidence from Microdata, Economic

Inquiry, 46(4), pp. 493510.

Moore, D. A., Tetlock, P. E., Tanlu, L. and Bazerman, M. H. (2006) Conflicts of Interest and the Case

of Auditor Independence: Moral Seduction and Strategic Issue Cycling, Academy of

Management Review , 31(1), pp. 1029.

Nettle, D. (2007) Empathizing and Systemizing: What are they, and what do they Contribute to our

Understanding of Psychological Sex Differences? British Journal of Psychology, 98(2), pp.

237255.

Nickerson, R. S. (2000) Null Hypothesis Significance Testing: A Review of an Old and Continuing

Controversy, Psychological Methods, 5(2), pp. 241301.

Nonis, S. A., Ford, C. W., Logan, L. and Hudson, G. (1996) College Students Blood Donation

Behavior: Relationship to Demographic, Perceived Risk, and Incentives, Health Marketing

Quarterly, 13(4), pp. 3346.

Nosek, B. A., Banaji, M. R. and Greenwald, A. G. (2002) Math = Male, Me = Female, Therefore Math

Me,Journal of Personality and Social Psychology, 83(1), pp. 4459.

Nosek, B. A., Smyth, F. L., Sriram, N., Lindner, N. M., Devos, T., Ayala, A., Bar-Anan, Y, Bergh, R.,

Cai, H., Gonsalkorale, K., Kesebir, S., Maliszewski, N., Neto, F., Olli, E., Park, J., Schnabel,

K., Shiomura, K., Tulbure, B. T., Wiers, R. W., Somogyi, M., Akrami, N., Ekehammar, B.,

Vianello, M., Banaji, M. R. and Greenwald, A. G. (2009) National differences in gender-

science stereotypes predict national sex differences in science and math achievement,

Proceedings of the National Academy of Sciences of the USA, 106(26), pp. 1059310597.


25/46

24

ODonnell, E. and Johnson, E. N. (2001) The Effects of Auditor Gender and Task Complexity on

Information Processing Efficiency,International Journal of Auditing, 5(2), pp. 91105.

OFallon, M. J. and Butterfield, K. D. (2005) A Review of the Empirical Ethical Decision-Making

Literature: 19962003,Journal of Business Ethics, 59(4), pp. 375414.

OLaughlin, E. M. and Brubaker, B. S. (1998) Use of landmarks in cognitive mapping: Gender

differences in self report versus performance, Personality and Individual Differences, 24(5),

pp. 595601.

Olsen, R. A. and Cox, C. M. (2001) The Influence of Gender on the Perception and Response to

Investment Risk: The Case of Professional Investors, The Journal of Psychology and Financial

Markets, 2(1), pp. 2936.

Penner, A. M. and Paret, M. (2008) Gender Differences in Mathematics Achievement: Exploring the

Early Grades and the Extremes, Social Science Research, 37(1), pp. 239253.

Peterson, R. A. (2001) On the Use of College Students in Social Science Research: Insights from a

Second-Order Meta-Analysis,Journal of Consumer Research, 28(3), pp. 450461.

Picker, S. H. and Berry, J. S. (2000) Investigating pupils images of mathematicians, Educational

Studies in Mathematics, 43(1), pp. 6594.

Picker, S. H. and Berry, J. S. (2001) Your students images of mathematicians and mathematics,

Mathematics Teaching in the Middle School, 7(4), pp. 202208.

Richard, C. (2006) Why an Auditor cant be Competent and Independent: A French Case Study,

European Accounting Review , 15(2), pp. 153179.

Ridgeway, C. L. (2009) Framed Before We Know It: How Gender Shapes Social Relations, Gender &

Society, 23(2), pp. 145160.

Robey, R. R. (2004) Reporting point and interval estimates of effect-size for planned contrasts: fixed

within effect analyses of variance,Journal of Fluency Disorders, 29(4), pp. 307341.

Rogelberg, S. G. and Luong, A. (1998) Nonresponse to mailed surveys: A review and guide, Current

Directions in Psychological Science, 7(2), pp. 6065.Roszkowski, M. J. and Grable, J. (2005) Gender Stereotypes in Advisors Clinical Judgments of

Financial Risk Tolerance: Objects in the Mirror Are Closer than They Appear, The Journal of

Behavioral Finance, 6(4), pp. 181191.

Rowley, S. J., Kurtz-Costes, B., Mistry, R. and Feagans, L. (2007) Social Status as a Predictor of Race

and Gender Stereotypes in Late Childhood and Early Adolescence, Social Development, 16(1),

pp. 50168.

Schmidt, F. L. (1996) Statistical Significance Testing and Cumulative Knowledge in Psychology:Implications for Training of Researchers, Psychological Methods, 1(2), pp. 115129.


26/46

25

Schmidt, W. C. (1997) World-Wide Web Survey Research: Benefits, Potential Problems, and

Solutions,Behavior Research Methods Instruments and Computers, 29(2), pp. 274279.

Schmitt, D. P., Realo, A., Voracek, M. and Allik, J. (2008) Why Cant a Man Be More Like a

Woman? Sex Differences in Big Five Personality Traits Across 55 Cultures, Journal of

Personality and Social Psychology, 94(1), pp. 168182.

Schubert, R., Brown, M., Gysler, M. and Brachinger, H. W. (1999) Gender and Economic

Transactions - Financial Decision-Making: Are Women Really More Risk Averse? American

Economic Review, 89(2), pp. 381385.

Shelton, S. W. (1999) The Effect of Experience on the Use of Irrelevant Evidence in Auditor

Judgment, The Accounting Review, 74(2), pp. 217-224.

Sherman, R. C., Buddie, A. M., Dragan, K. L., End, C. M. and Finney, L. J. (1999) Twenty years of

PSPB: Trends in Content, Design, and Analysis, Personality and Social Psychology Bulletin,

25(2), pp. 177187.

Shih, M., Pittinsky, T. L. and Ambady, N. (1999) Stereotype Susceptibility: Identity Salience and

Shifts in Quantitative Performance, Psychological Science, 10(1), pp. 8083.

Signorella, M. L. and Jamison, W. (1986) Masculinity, Femininity, Androgyny, and Cognitive

Performance: A Meta-Analysis, Psychological Bulletin, 100(2), pp. 207228.

Sjberg, L. and Engelberg, E. (2009) Attitudes to Economic Risk Taking, Sensation Seeking and

Values of Business Students Specializing in Finance, Journal of Behavioral Finance, 10(1),

pp. 111.

Smith, J. F. and Kida, T. (1991) Heuristics and Biases: Expertise and Task Realism in Auditing,

Psychological Bulletin, 109(3), pp. 472489.

Smith, A. and Rogers, V. (2000) Ethics-Related Responses to Specific Situation Vignettes: Evidence

of Gender-Based Differences and Occupational Socialization, Journal of Business Ethics,

28(1), pp. 7386.

Soutar, G. N. and Sweeney, J. C. (2003) Are There Cognitive Dissonance Segments? Australian

Journal of Management, 28(3), pp. 227249.

Spencer, S., Steele, C. and Quinn, D. (1999) Stereotype Threat and Womens Math Performance,

Journal of Experimental Social Psychology, 35(1), pp. 428.

Steinberg, L. (2007) Risk Taking in Adolescence: New Perspectives From Brain and Behavioral

Science, Current Directions in Psychological Science, 16(2), pp. 5559.

Tan, H. T. and Kao, A. (1999) Accountability Effects on Auditors Performance: The Influence of

Knowledge, Problem-Solving Ability, and Task Complexity,Journal of Accounting Research,37(1), pp. 209223.


27/46

26

Thompson, B. (2007) Effect sizes, confidence intervals, and confidence intervals for effect sizes,

Psychology in the Schools, 44(5), pp. 423432.

van Nieuw Amerongen, C. M. (2007) Auditors Performance in Risk and Control Judgments: An

Empirical Study. Ph. D. dissertation, Vrije Universiteit Amsterdam.

Venezia, C. C. (2008) Are Female Accountants More Ethical Than Male Accountants: A Comparative

Study Between The U.S. And Taiwan,International Business & Economics Research Journal,

7(4), pp. 110.

Verhaeghen, P. and Salthouse, T. A. (1997) Meta-Analyses of Age-Cognition Relations in Adulthood:

Estimates of Linear and Nonlinear Age Effects and Structural Models, Psychological Bulletin,

122(3), pp. 231248.

Wakabayashi, A., Baron-Cohen, S., Wheelwright, S., Goldenfeld, N., Delaney, J., Fine, D., Smith, R.

and Weil, L. (2006) Development of short forms of the Empathy Quotient (EQ-Short) and the

Systemizing Quotient (SQ-Short), Personality and Individual Differences, 41(5), pp. 929940.

Watkins, A. L., Hillison, W. and Morecroft, S. E. (2004) Audit Quality: A Synthesis of Theory and

Empirical Evidence,Journal of Accounting Literature, 23, pp. 153193.

Weber, E. U., Blais, A. R. and Betz, N. E. (2002) A Domain-specific Risk-attitude Scale: Measuring

Risk Perceptions and Risk Behaviors,Journal of Behavioral Decision Making, 15(4), pp. 263

290.

Whitley, B. E. Jr. (2001) Gender Differences in Affective Responses to Having Cheated: The

Mediating Role of Attitudes,Ethics & Behaviour, 11(3), pp. 249259.

Wilson, M. and Daly, M. (1985) Competitiveness, risk-taking, and violence: The young male

syndrome,Ethnology and Sociobiology, 6(1), pp. 5973.


28/46

27

Fig. 1. An overview of the determinants of audit quality.

Culture within the firmFirm sizeNon-audit services

CompositionFamiliarity

Technical expertiseProblem-solving abilityRisk profileExperienceIndependence

Firm RelatedFactors

Factors Outside

Control of AuditorsAudit CommitteesAudit regulation

Task Related

CircumstancesComplexityTime pressure

Personal Auditor

Characteristics

Audit Team

Characteristics

Audit quality


29/46

28

Fig. 2. Schematic overview of the relationship between individual characteristics, personal

auditor characteristics, auditor's judgment and decision-making, and audit quality. Solid lines

indicate a direct relationship, dashed lines indicate moderation.

Table 1. Descriptive statistics of non-business students, business students, and auditors characteristics a

Audit quality

Personal auditor characteristics(e.g. risk profile)

Individual characteristics(e.g. sex)

Auditors judgment and decision-making(e.g. risk assessments)

Socio-cultural and

biological factors

Task relatedfactors


30/46

29

Characteristics split by subpopulation

and sex Non-business students Business students Auditors

Men Women Men Women Men Women

Sex 241(35.6%) 436(64.4%) 97(56.1%) 76(43.9%) 862 (82%) 189 (18%)

56.167 p= .000 2.549 p= .110 4.31 p= .000

Age22.85 21.26 20.70 20.47 47.66 38.57

3.053 p= .002 1.015 p= .311 15.699 p= .000

Experienceb 17.01 9.95

13.006 p= .000

B4 542 (62.9%) 121(64.4%)Internshipc

NB4 320 (37.1%) 67(35.6%)

0.146 p= .702

B4 related 151(17.5%) 28(15.1%)

NB4 related 219(25.4%) 52(28.1%)

0.927 p= .336

Current statusd

Self-employed 79(9.2%) 10(5.4%)

Othere 413(47.9%) 95(51.4%)

Employee 82(22.2%) 26(32.5%)

Co-worker 135(36.5%) 32(40.0%)

Office positionf

Partner 153(41.4%) 22(27.5%)

6.393 p= .041

aThe table gives the mean value for age and experience for males and females in comparison as well as t-value of

the respective Independent Samples Test and respective p-value regarding sex specific differences. For the other

variables, the distribution per value is given for males and females as well as the Chi-value of the respective Chi

Test regarding sex specific differences.b

Experience is derived from the year one took the oath to be appointed as an external auditor (in years).

cIn Belgium and the Netherlands, auditor trainees must complete a minimum of three years of practical training in

an audit office.d

The external auditor can be self-employed or employed by an audit office.

e Members of the NIVRA do not exclusively work as external auditors, a majority works in other professions (e.g.

internal auditor).

fAuditors can have one of the following three statuses in an audit firm: partner, co-worker or employee.


31/46

30

Table 2. Men and women as problem-solvers

Population category Sex N Meana Std. Dev.

Men 213 7.35 [7.06 ; 7.64] 2.13Non-business students

Women 372 6.33 [6.09 ; 6.57] 2.35

Men 96 7.61 [7.26 ; 7.96] 1.77Business students

Women 76 7.18 [6.75 ; 7.61] 1.93

Men 482 7.38 [7.23 ; 7.53] 1.71Auditors

Women 100 7.34 [6.95 ; 7.73] 1.97

aMean problem-solving score is derived from the number of correct answers on the ten questions regarding

problem-solving.


32/46

31

Table 3. Men and women as problem-solvers (Independent Samples Test)

t-test for Equality of Means

Levenes

Test for

Equality of

Variances 0H : no sex difference

95%

Confidence

Interval of the

Difference

Population category F Sig. t df

Sig.

(2-

tailed)

Mean

Difference

Std. Error

Difference Lower Upper

Equal variances

assumed2.348 .126 5.175 583 .000 1.01 .20 .63 1.40

Non-business

students

Equal variances

not assumed

5.314 477.603 .000 1.01 .19 .64 1.39

Equal variances

assumed.314 .576 1.510 170 .133 .43 .28 -.13 .98

Business students

Equal variances

not assumed1.495 153.902 .137 .43 .29 -.14 .99

Equal variances

assumed3.716 .054 .239 580 .811 .046 .19 -.33 .42

Auditors

Equal variances

not assumed.218 131.685 .828 .046 .21 -.37 .46

Table 4. Problem-solving in a multivariate frameworka


33/46

32

Non-business students Business students Auditors

Female -.313660 (p= .001) -.296691 (p= .065) -.031750 (p= .273)

More math educationb .181193 (p= .000) .107205 (p= .044)

More mathematical studyc .751234 (p= .000)

Time pressured -.744891 (p= .000)

Older -.00265089 (p= .010)

NL dummye

-.047655 (p= .125)

aThe table gives the coefficients of the Poisson regression models with p-values in parentheses.

bHours of mathematical education per week at the end of secondary education.

cA dummy variable distinguished between natural sciences and social sciences.

dFor 39 business students the time available to finish the questionnaire was limited.

eA dummy variable was used to code for nationality (Belgian = 0, Dutch = 1).

Table 5. Sex differences in risk-aversion


34/46

33


0H : no sex difference

Asymp. Sig.

Z (2-tailed)

Men 224 19.19 [18.58 ; 19.80] 4.63Non-business students

Women 383 17.71 [17.23 ; 18.19] 4.76-4.003 .000

Men 92 21.18 [20.18 ; 22.18] 4.90Business students

Women 76 19.11 [18.28 ; 19.94] 3.71-2.651 .008

Men 818 18.93 [18.56 ; 19.30] 5.44Auditors

Women 175 18.47 [17.68 ; 19.26] 5.34-.928 .354

aMean is derived from the Likert-scale measuring risk-attitude. A higher score indicates a higher willingness to take

risks (i.e. a less risk-averse attitude).


35/46

34

Table 6. Risk-aversion in a multivariate frameworka


Female -.336868 (p= .000) -.453570 (p= .005) -.200110 (p= .025)

Older -.00794270 (p= .339) .042745 (p= .431) -.016367 (p= .000)

NL dummyb .618536 (p= .000)

aThe table gives the coefficients of the linear ordered probit regressions with p-values in parentheses.

bA dummy variable was used to code for nationality (Belgian = 0, Dutch = 1).


36/46

35

Table 7. Sex differences in empathy



Asymp. Sig.

Z (2-tailed)

Men 181 1.25 [1.15 ; 1.35] .78Non-business students

Women 317 1.33 [1.26 ; 1.40] .62-1.325 .185

Men 84 1.17 [1.01 ; 1.33] .75Business students

Women 67 1.33 [1.15 ; 1.51] .74-1.588 .112

Men 620 1.30 [1.25 ; 1.35] .65Auditors

Women 128 1.41 [1.30 ; 1.52] .63-1.751 .080

aMean is derived from the question on empathy. A higher score indicates a higher level of empathy.


37/46

36

Table 8. Sex differences in cognitive biases (Test Statistics)


Population

category

Conjunctive

fallacy a

Confirmation

bias 1 b

Confirmation

bias 2 b

Insensitivity

to sample

size c

Information

bias d

Disjunctive

fallacy e

Overconfidence

(miscalibration) f

Z -2.626 -1.136 -.720 -1.655 -1.792 -1.027 -2.244

Non-

business

students

Asymp.

Sig. (2-

tailed)

.009 .256 .471 .098 .073 .305 .025

Odds

ratio

1.91

[1.17; 3.10]

1.61

[0.70; 3.70]

1.31

[0.63; 2.76]

1.42

[0.94; 2.16]

1.51

[0.96; 2.37]

1.33

[0.77; 2.27]

0.60

[0.38; 0.94]

Z -1.768 -.247 -2.215 -.514 -1.181 -.100 -2.501

Business

students

Asymp.

Sig. (2-

tailed)

.077 .805 .027 .607 .237 .920 .012

Odds

ratio

2.12

[0.92; 4.90]

1.15

[0.38; 3.52]

0.12

[0.14; 1.05]

1.22

[0.57; 2.63]

0.62

[0.29; 1.37]

0.09

[0.20; 4.38]

0.38

[0.18; 0.82]

Z -1.859 -.072 -.595 -1.919 -.1064 -1.378 -.959

Auditors Asymp.

Sig. (2-

tailed)

.063 .943 .552 .055 .287 .168 .338

Oddsratio

1.99[0.95; 4.12]

1.03[0.45; 2.38]

1.57[0.35; 7.03]

1.61[0.99; 2.61]

1.30[0.80; 2.11]

2.07[0.72; 5.96]

1.23[0.80; 1.90]

aConjunctive fallacy: false belief that specific conditions are more probable than a single general one.

bConfirmation bias: the tendency to look for believes-supporting evidence and to attach relatively more value to

evidence that confirms believes than to evidence that contradicts believes.

cInsensitivity to sample size: ignoring sample size when assessing the likelihood of certain results.

dInformation bias: believing that the more information that can be acquired to make a decision, the better, even if

that extra information is irrelevant for the decision.

eDisjunctive fallacy: false belief that the probability of a general event is less than the sum of the probabilities of

its separate components.

fOverconfidence: the tendency to believe that judgments are more accurate than they really are.

Table 9. Summary


38/46

37


Problem-solving

F < M F < M F = M

Risk-aversion

F > M F > M F > M

Empathy F = M F = M F = M

Independence

Cognitive biases Overconfidence: F < M

Conjunctive fallacy +

Insensitivity sample size +

Information bias: F > M

Overconfidence: F < M

Other biases: F = M

Conjunctive fallacy +

Insensitivity sample size:

F > M

Other biases: F = M


39/46

38

Appendix

This questionnaire examines three dimensions: risk profile (risk-aversion and empathy),

problem-solving ability (including mathematical abilities), and independence (through a priori-

assumptions and cognitive biases). The purpose of this questionnaire is to examine if sex

differences present at the level of the general population are replicated throughout different

subpopulations of men and women.

This questionnaire was produced based on the following sources:

Baron-Cohen, S. and Wheelwright, S. (2004) The Empathy Quotient: An Investigation of

Adults with Asperger Syndrome or High Functioning Autism, and Normal Sex

Differences,Journal of Autism and Developmental Disorders, 34(2), pp. 163175.

Goldstein, E. B. (2008) Cognitive Psychology: Connecting Mind, Research, and Everyday

Experience (Belmont: Wadsworth Publishing).

Tversky, A. and Kahneman, D. E. (1973) Availability: A Heuristic for Judging Frequency and

Probability, Cognitive Psychology, 5(2), pp. 207232.

Tversky, A. and Kahneman, D. E. (1974) Judgment Under Uncertainty: Heuristics and Biases,

Science, 185(4157), pp. 11241131.

Weber, E. U., Blais, A. R. and Betz, N. E. (2002) A Domain-specific Risk-attitude Scale:

Measuring Risk Perceptions and Risk Behaviors, Journal of Behavioral Decision

Making, 15(4), pp. 263290.

Wickelgren, W. A. (1995) How to Solve Mathematical Problems (New York: Dover

Publications).

Risk profile (11 questions on risk-aversion)

For each of the following statements, please indicate the likelihood of engaging in each

activity. Provide a rating from 1 to 5, using the following scale:


40/46

39

1. ____ Betting a days income at a sports game.

2. ____ Co-signing a new car loan for a friend.

3. ____ Investing 10% of your annual income in a blue chip stock.

4. ____ Investing 10% of your annual income in a very speculative stock.

5. ____ Investing 10% of your annual income in government bonds (treasury bills).

6. ____ Investing in a business that has a good chance of failing.

7. ____ Lending a friend an amount of money equivalent to one months income.

8. ____ Spending money impulsively without thinking about the consequences.9. ____ Taking a days income to play the slot-machines at a casino.

10.____ Taking a job where you get paid exclusively on a commission basis.

11.____ Betting one months income at a double or nothing coin toss game.

Problem-solving capacities (10 questions)

1. What is the value of the expression (x y) when x = 5 and y = 1

A) 4

B) 6

C) 16

D) 24

E) 36

2. Which of the following is equivalent to (4x)

A) 864x

B) 664x

C) 612x

D) 512x

E)

6

4x

1 2 3 4 5

Extremely unlikely Not sure Extremely likely


41/46

40

3. Which of the following is an equivalent simplified expression for 2( x4 + 7) 3( x2 - 4)

A) 2+x

B) 22 +x

C) 262 +x

D) 103 +x

E) 113 +x

4. Ms. Cox plans to drive 900 miles to her vacation destination, driving an average of 50

miles per hour. How many miles per hour faster must she average, while driving, to

reduce her total driving time by 3 hours.

A) 5

B) 8

C) 10

D) 15

E) 18

5. Evi cuts a board in the shape of a regular hexagon and pounds in a nail at an equal

distance from each vertex, as shown in the figure below. How many rubber bands will

she need in order to stretch a different rubber band across every possible pair of nails?

A) 15

B) 14

C) 12

D) 9

E) 6

6. There are 280 runners registered for a race, and the runners are divided into 4 age

categories, as shown in the table below.

Age category: Under 16 16-25 26-35

Female Auditor1

Documents

Transcript of Female Auditor1