DEVELOPMENT OF AN INSTRUMENT TO MEASURE PROFESSIONAL
SKEPTICISM
R. Kathy Hurtt
University of Wisconsin – Madison
Grainger Hall
975 University Avenue
Madison, WI 53706-1323
(608) 265-6083
khurtt@bus.wisc.edu
November 2001
Development of an instrument to measure professional skepticism
ABSTRACT
This paper presents the development of an instrument designed to measure professional
skepticism. Professional skepticism, while acknowledged in the professional literature as being a
fundamental requirement of an audit, has received only limited academic research, and this
research has often called for development of a professional skepticism scale. Theoretically
grounded in philosophy with additional insights from professional accountants, a 30-item scale is
presented that has been tested on over 200 student and 200 professional subjects. The scale
demonstrates inter-item reliability and test-retest stability.
KEYWORDS: professional skepticism, skepticism, auditor skepticism, scale development
1
Development of an Instrument to Measure Professional Skepticism
INTRODUCTION
This paper presents the development of an instrument designed to measure professional
skepticism. Professional skepticism is a fundamental precept for the auditing profession, and the
first Statement of Auditing Standards (SAS) stressed this concept. Professional skepticism was
re-emphasized in the 1997 revisions to SAS No.1, which states:
Due professional care requires the auditor to exercise professional skepticism.
Professional skepticism is an attitude that includes a questioning mind and a
critical assessment of audit evidence. The auditor uses the knowledge, skill, and
ability called for by the profession of public accounting to diligently perform, in
good faith and with integrity, the gathering and objective evaluation of evidence.
Gathering and objectively evaluating audit evidence requires the auditor to
consider the competency and sufficiency of the evidence. Since evidence is
gathered and evaluated throughout the audit, professional skepticism should be
exercised throughout the audit process. The auditor neither assumes that
management is dishonest nor assumes unquestioned honesty. In exercising
professional skepticism, the auditor should not be satisfied with less than
persuasive evidence because of a belief that management is honest. (SAS No. 1,
AU sec. 230, paragraphs 7 through 9 (AICPA) 1997)) (highlighting added).
This emphasis in the standards suggests that research into professional skepticism is appropriate,
but it has been difficult to investigate this characteristic and its impact on the audit process.
When investigating individual differences, accounting researchers often measure the
construct of interest using scales developed in other disciplines. (A few recent examples of this
type of research include: (Bierstaker and Wright 2001; Hyatt and Prawitt 2001; Kaplan and
Whitecotton 2001; Wright and Davidson 2000).) However, professional skepticism is not a
construct that is significantly used in other disciplines. The accounting profession is unique in its
2
mandate of professional skepticism1, and consequently, no other discipline has developed an
instrument to measure this2. Even in accounting research, researchers have defined and measured
skepticism distinctively in the few accounting studies that attempt to measure this construct. A
brief summary of the extant accounting literature, the definition of skepticism used in each study
and the method used to measure the skepticism construct follows.
McMillan and White (1993) indicate, “...professional skepticism implies that auditors
focus more on error-related evidence.” They gauged professional skepticism by categorizing
subjects’ initial explanations for a hypothetical company’s financial ratio fluctuations as either an
environmental (i.e., naturally occurring) or error-based explanation. Those subjects who used an
error-based explanation were judged to be more skeptical. Shaub (1996) characterized
skepticism as the opposite of trust and operationalized the construct using the trust and
independence portions of Wrightsman’s Trustworthiness Scale (1974). Shaub and Lawrence
(1996) defined skepticism as the behavior of confronting a client or engaging in expanded audit
work and operationalized this construct through their subjects’ responses to nine scenarios. They
used a similar definition and method of measuring skepticism in their 1999 work. Choo and Tan
(2000) measured professional skepticism using ten items from the Rempel, Holmes and Zanna
(1985) Trust Scale. Choo and Tan modified the scale so that the questions referred to a “client”
rather than the life-partner referenced in the original scale. Subjects with low scores on the Trust
Scale were assumed to have high skepticism. Choo and Tan conclude their study with a
1
Other professions such as attorneys, police officers, and bank examiners can benefit from skepticism; however, no
other discipline has a requirement for professional skepticism or the emphasis on professional skepticism that is
present in accounting.
2
There have been a few instances in marketing where researchers investigated skepticism, but these scales are so
specifically directed towards skepticism about advertising that they are of no assistance in measuring professional
skepticism.
3
discussion of the problems of using the Trust Scale and call for the development of a
psychometrically reliable scale designed specifically to measure skepticism.
Using unique measures of skepticism within each study makes it difficult to develop an
understanding of the behavioral implications of professional skepticism. In some instances,
because of the embedded nature of the scales in the experiment, the scale used cannot easily be
replicated in a subsequent study3. In other instances, the amount of time required for subjects to
complete the skepticism measure is so great, that it would be difficult to use the measure in most
experimental contexts4. In other instances, the authors described scales they used in their study
as easily available proxies that were used only because an instrument designed specifically to
measure professional skepticism had not been developed (Shaub, 1996, Choo and Tan, 2000).
This paper presents the development of such an instrument: one designed specifically to measure
professional skepticism. In the remainder of the paper I will present a brief overview of the steps
required for instrument development, describe how those steps were completed for the
professional skepticism instrument developed here, and discuss future research directions.
INSTRUMENT DEVELOPMENT
Scale development must flow from theory (DeVellis, 1991). Hurtt, Eining and Plumlee
(2001) present a theoretical model of professional skepticism that is grounded in psychological
literature and philosophical literature on methodological skepticism. Their model (presented in
Figure 1) suggests professional skepticism is a multi-dimensional construct comprised of six
individual characteristics: curiosity; self-confidence; interpersonal understanding, questioning;
self-determining and deliberating.
3
In private conversation, Jeffrey McMillan indicated that the embedded nature of their instrument would make
replicating their skepticism measure for comparison to other measures almost impossible.
4
In private correspondence with Shaub and Lawrence, they estimated that it took subjects approximately 25 minutes
to complete the scenario section of their instrument.
4
Insert Figure 1 about here
They describe their curiosity characteristic as searching for knowledge and investigating.
Hurtt et al. (2001) differentiate curiosity from questioning by characterizing questioning as a
requirement for reasons, evidence, justification or proof, rather than the personal investigation
that characterizes curiosity. Self-confidence, while self-explanatory, is included in their model
because of the necessity of a skeptical individual to have the courage to act on their curiosity and
questioning. Interpersonal understanding is defined as a curiosity about people that is more
specific than the general trait of curiosity. Self-determining is characterized as a reluctance to
accept others’ statements or claims without making a self-determination, and deliberating is the
characteristic of reflective decision-making. Using these elements, I followed standard scale
development techniques to create a scale designed specifically to measure professional
skepticism.
Developing an initial item pool
Clark and Watson (1998) suggest that after a theoretical basis for the scale is identified,
existing literature should be reviewed to determine how others have assessed the same or closely
related constructs. I reviewed the psychological scales suggested by Hurtt, et al. for items
designed to measure constructs closely related to one or more of the six elements of skepticism,
and I also performed an independent search for any existing scales designed to measure
professional skepticism or a similar construct. Although I did not find an existing scale, from the
scales examined I identified over 170 individual items that could be modified to capture one of
the six skepticism elements. Table 1 indicates the scales from which items were extracted for the
initial item pool5. Some elements of professional skepticism did not correspond well to items
5
In many instance, I theorized that individual items from an existing scale would likely map to more than one
construct in the professional skepticism scale. This splitting of items from existing scales is due to sub-constructs
5
from existing scales and to resolve such instances I wrote an additional 50 items. This created an
initial item pool of 220 questions. Two Ph.D. students with experience teaching scale
development and I reviewed each of the items for clarity and to eliminate redundant questions.
This process reduced the number of items to 85.
Insert Table 1
To allow for consistent presentation, both those items that had been used in previous
scales and all the new items were written for use with a 6-point6 Likert type scale anchored by
“strongly agree” and “strongly disagree.” Changing item wording, scaling methods or anchors
invalidates established reliabilities for measures (Converse and Presser 1990, Schwarz 1996);
therefore, all the items used in this new instrument needed to be re-validated, regardless of the
reliability or validity of the original scale.
One of the major threats to construct validity is construct under-representation. This
occurs when assessment of a construct is too narrow and fails to include important dimensions of
the construct (Messick 1998). Because the Hurtt et al. (2001) model was developed from
psychology research and from philosophical skepticism, I wanted to see if this theoretical view
of professional skepticism differed from that of practicing accountants. I solicited written
comments on skepticism from thirty-five professionals (corporate accountants and auditors) and
asked them to answer open-ended questions about what they believed comprised skepticism and
within the original scale. For example, items in the Melbourne Curiosity Inventory (MCI), (Bunge 1991), that dealt
with general curiosity were initially included as curiosity questions in the professional skepticism scale, while those
items that dealt with curiosity about people were included as interpersonal understanding questions. This results in
the same scale being listed under different constructs in Table 1.
6
There is considerable discussion in the scale development literature about whether an odd or even number of
response points is preferable (see for example, Moser and Kalton, 1972, 344[Moser, 1972 #215] and Rugg and
Cantril, 1944, 33). With an odd number of response alternatives, the subjects are offered a neutral middle. I used an
even number to avoid the issue of subjects simply selecting the mean value. This is consistent with the
recommendation of Converse and Presser (1990, 37)[Converse, 1990 #209] who indicate, “Do not explicitly
provide the middle category and thereby avoid losing information about the direction in which some people lean….”
6
what they believed would be the opposite of skepticism. I compared these comments to both the
theorized constructs in the model and to the individual scale items. This review did not identify
any dimensions of professional skepticism that were not included in the model, and the
definitions of skepticism were consistent with the items used in the scale.
DeVellis (1991) also recommends that experts review the scale. This process is intended
to improve the instrument’s content validity. I met with three accounting faculty members with
backgrounds and education in auditing who reviewed the scale items for relevance and to ensure
that they agreed that each item was properly classified regarding the construct the item was
designed to measure. I eliminated items identified by the accounting faculty members as
irrelevant or inconsistent.
Pre-testing the instrument
From this process I reduced the scale to 49 items and administered the instrument to 89
graduate and undergraduate business students in what Converse and Presser (1990) describe as a
“developmental pretest.” The intent of this pretest is to smooth the wording of the questions and
to see which questions measure the constructs under consideration. They recommend that this
first pretest contain less than twice as many questions as the final instrument will contain, that
the respondents resemble the target population and that a reasonable number of subjects for this
stage pretest is 25 – 75. Because the instrument is designed to be used both with business
students, and also with professional accountants, the use of a business student subject pool for
this pretest is justified.
I examined the results of the pretest using factor analysis to assist me in identifying items
that did not load on expected constructs. My initial examination of the correlation matrix
They further suggest measuring intensity by anchoring with words such as “strongly agree” and “strongly disagree”
as was done in this scale.
7
indicated that many items were significantly correlated; therefore, I used oblique rotation (direct
oblimin), which relaxes the assumption of orthogonal factors. After eliminating items that did
not load on any of the hypothesized factors, 22 items remained. These 22-items loaded on the
six theoretical factors with item factor-loadings generally of .70 or higher (Table 2). Factor
loadings above .50 are considered acceptable, (DeVellis, 1991).
Insert Table 2 about here
Cronbach’s alpha, which indicates whether the questions in the scale are reliably
measuring the same construct, for the entire scale was .827. Cronbach’s alpha was also
calculated for each of the sub-scales. For five of the six dimensions of skepticism, the alpha for
the sub-scale exceeded .80. Alphas greater than .80 are considered adequate results, although the
small sample size and fairly homogeneous composition of the subject pool could lead to
overestimating the reliability of the items. Although the sub constructs are not to be used
individually, and the alpha for the entire scale was adequate, the alpha for the “questioning”
construct was very low at .57. To address this problem, I generated a number of new items for
this construct that could be used in a subsequent pilot test. I also closely examined each item by
construct to identify any items that had been confusing to the subjects, and I wrote or revised
additional items towards the goal of obtaining an equal number of items (5) for each of the six
factors. Because there is no theoretical basis for assuming that any single aspect of skepticism is
more important than another, developing a scale with equal weighting for all factors seemed
DeVellis (1991) indicates the following acceptable ranges for Cronbach’s alpha: “between .80 and .90 is very
good; between .70 and .80, respectable; between .65 and .70 minimally acceptable.”
7
8
most appropriate8. This process resulted in a 40-item measure for the pilot test. The pilot test had
more items than I wanted for the final test to allow for the possibility of invalid questions.
Pilot and reliability testing
The revised 40-item scale was administered to 250 undergraduate business students and
useable responses were received from 247 students. Seventeen responses had at least one item
skipped. The mean score for the item was substituted in these instances so that the remainder of
the response could be analyzed9. Analysis of the data revealed no significant differences in
skepticism scores related to any of the demographic information such as gender, GPA, class
standing, class where the instrument was administered, native English speaker, or age.
The correlation matrix of the responses indicated significant or marginally significant
correlations for many of the items; therefore, I again used oblique rotation for the factor analysis
of the responses. The rotated-factor analysis indicated that the majority of items loaded on the
six theoretical factors. I eliminated items that loaded on constructs other than those anticipated
(this occurred primarily with new or reversed items), which resulted in a 25-item scale. Table 3
lists the factor loadings for these items.
Insert Table 3 about here
Cronbach’s alpha for the entire scale was .85, but, as expected given the small sample
used for the previous test, both factor loadings and Cronbach’s alpha decreased for most of the
sub-constructs. These results indicate that overall the professional skepticism scale has adequate
inter-item reliability, although the sub-construct “questioning” remained below the normally
8
30-items is a parsimonious number for a scale containing 6 sub-constructs; however, my goal was to develop an
instrument that could be quickly administered before or after behavioral experiments, thus a fairly brief scale was
required.
9
accepted range. Although not ideal, the low alpha for this sub-construct indicates that these
questions are not all measuring the same characteristic. Because the scale as a whole
demonstrates adequate reliability, the overall score should be a reliable measure. However, I
again wrote additional questions to attempt to raise the alpha score for the questioning construct
and to ensure that there were an equal number of questions for each construct.
Intra-subject test-retest reliability measures the stability of an individual’s responses over
time. The test-retest method is recommended when developing a measure that cannot be
validated using known group10 validation techniques, or when developing a measure with subconstructs that preclude using split-half reliability (DeVellis, 1991, Fink, 1995, Litwin, 1995).
After an approximately 8-week delay, the 250 subjects were asked to complete the professional
skepticism instrument for a second time11. The subjects were also asked to complete
Wrightsman’s (1974) Trustworthiness and Independence Scales that were used by Shaub
(1996)12. Having subject’s scores on the professional skepticism instrument as well as the two
Wrightsman’s scales allowed me to begin the process of determining whether the skepticism
scale measures a different construct than the Wrightsman scales, (i.e., discriminant validity).
Responses were received from 147 subjects, but because of incomplete information only
92 matched sets of tests were identified. The correlation between the total skepticism score (sum
of the 25 items) on the first and second test was significant, r (92) = .89, p<. 001. Litwin (1995)
indicates that in reliability testing, “r values are considered good if they equal or exceed .70,”
9
Although the factor loadings changed, there were no significant differences if the median was substituted or if
these questions were dropped.
10
Known group validation involves administering the instrument to a group of subjects known to have the
characteristic that the scale is designed to measure. Because there is not a “known group” that is skeptical, this
preferred validation technique cannot be used.
11
Although questions were added to bring the total number of scale questions again to 40, these additional questions
were not included in the test-retest analysis. Only the twenty-five questions displayed in Table 3 were included in
the analysis.
12
Shaub found that these scales were not predictive of the skeptical behaviors that he expected from his subjects.
10
while Fink (1995) suggests “a conservative rule of thumb: over .75 is a very good to excellent
relationship.” Cronbach’s alpha for the combined test-retest was .95. The 15 “new” items were
examined to identify questions that loaded on the appropriate factor. From this a 30-item test
was constructed with an equal number of items for each construct. (See appendix A for the 30item Professional Skepticism Instrument.) On this 30-item instrument the subjects had a mean
score of 132.7 with a standard deviation of 15.9, and scores ranged from 77 to 175.
As shown in Table 4, subjects’ score on the professional skepticism scale was not
significantly correlated with either Wrightsman’s Trustworthiness or Wrightsman’s
Independence Scales. This indicates that the professional skepticism scale is measuring a
different construct than either trustworthiness or independence, and might help to explain why
Shaub (1996) found neither test was predictive of his subjects’ decision to exercise skeptical
behavior. This presents preliminary evidence that a scale designed specifically to measure
professional skepticism can assist in further study of this important construct.
Insert Table 4 about here
Testing with professional accountants
I received permission from a major international accounting firm to administer the
professional skepticism instrument to members of their firm. Managing partners at thirteen U.S.
offices of this firm agreed to provide email addresses for their staff members13 and also agreed to
send a memo to those individuals encouraging them to respond to my email. I emailed 350
auditors and asked them to participate in a two-part study whose purpose was represented as
being research on the work paper review process. The email included the URL for a web site that
13
The partners were asked to supply email addresses for individuals with four or more years of audit experience.
This was requested as part of an experimental study that was also conducted at this time; however, it also served to
ensure that only those individuals who have experience with auditing and the requirement that auditors exhibit
professional skepticism were included in this instrument validation.
11
contained the professional skepticism instrument, and questions on demographic information. To
increase the response rate, I sent two follow-up emails during the next two months. Two hundred
auditors completed the questions at the designated URL, a 57% response rate. Tests of
differences between early and late responders did not reveal any differences between the groups.
Demographics on the auditors are presented in Table 5. Their mean score on the professional
skepticism scale was 138.6 with a range from 111 to 173 (of 180 points possible), and a standard
deviation of 12.6. There was no significant difference between the subject’s score on the scale
and any of their demographics (e.g., age, amount of experience, or gender.) The mean experience
level of the auditors was 58 months, so all of the subjects should be familiar with the requirement
for professional skepticism. However, the purpose of the scale was not disclosed to the subjects
to lessen the likelihood of subjects responding in a manner that they believed was “correct.”
Insert Table 5 about here
The internal consistency coefficient for this administration of the 30-item scale using
Cronbach’s alpha is .86. As with previous administrations of the instrument, and as expected for
a scale designed to measure a single construct, the correlation matrix indicated that many of the
items were significantly correlated; therefore, oblique rotation was used in the factor analysis of
the responses. As shown in Table 6, the questions loaded on the six theorized factors, but two
questions did not load on the construct they were theorized to measure, and one of these items
loaded at only .49, below the .50 minimum suggested by DeVellis (1991)14. Because this scale is
designed to provide a single score for professional skepticism, and the internal consistency of the
14
It is unknown if this is a problem unique to this sample population, or a problem with the instrument. Only
further testing with a broad variety of subjects will be able to answer that question. However, I did rerun the
analysis using the combined first and second scores of the subjects and all items loaded on the appropriate
constructs.
12
questions is appropriate, this shift in where the two questions loaded might be simply a function
of the sample population.
After a brief wait, subjects were again emailed and asked to complete an online work
paper review, answer a few questions and then complete the skepticism scale. Scores on the
skepticism scale were collected from 88 subjects. Their mean score on the second test was 135.4,
with a range of 105 to 177 and a standard deviation of 14.7. This contrasts with these 88 subjects
scores on the first administration of the professional skepticism scale of a mean of 138.5, a range
of 113 to 170 and a standard deviation of 12.4. The mean elapsed time between completion of
the first and second scale was 22 days, with a range from 2 days to 56 days, and a mode of 15
days. The correlation between the first and second test score was .85, which is significant at the
.01 level. Cronbach’s alpha, measuring internal consistency between the two total scores was
.91.
SUMMARY DISCUSSION
This paper presents the development of an instrument designed to measure professional
skepticism. Previous researchers have used proxy measures or experimental type measures that
are lengthy to administer and time-consuming to score. The scale presented here is an initial
effort at developing a psychometrically valid instrument that can be easily administered and
quickly scored by researchers interested in the area of professional skepticism. A limitation of
this development process is that the instrument was validated using business students at a single
school and professionals from a single firm. Research using other psychological-based
instruments has demonstrated that there can be significant differences between the average
scores between firms, and that possibility should be explored in future research.
13
The theoretical foundation for the professional skepticism instrument was an examination
of philosophical writings about skepticism for elements that are descriptive of professional
skepticism. In an effort to identify any elements of professional skepticism that were not evident
in philosophical skepticism, accounting research was reviewed and input was solicited from
accounting professionals. These efforts did not identify any characteristics that were not present
in the philosophical model.
The analysis of the instrument indicates that the scale has adequate inter-item
consistency, and adequate test-retest reliability with both student and professional subjects. This
is a necessary first step in instrument development; however, significantly more research needs
to be completed. The validity of the instrument, which is whether it actually measures the
professional skepticism that it purports to measure, must be determined. The model in Figure 1
predicts that the characteristics of skepticism should lead to four behaviors. Research needs to be
undertaken to determine if this instrument differentiates between subjects and whether different
scores lead to these different behaviors. Additional issues such as the longitudinal stability of the
measure over a professional career and whether a subject’s score changes through education and
coursework also should be examined.
Research on professional skepticism has been difficult because of the lack of a standard,
easily administered method of determining differences in this characteristic between subjects.
With the availability of an instrument, a host of interesting research questions is possible such as
“Does a higher level of professional skepticism lead to increased fraud detection?” “Can an
auditor be too skeptical?” “Are more skeptical auditors slower at performing audit work: does
professional skepticism lead to over-auditing and inefficient audits?” “Are there differences in
the review procedures of more and less skeptical auditors?” “Can professional skepticism be
14
taught or learned?” “Is professional skepticism related to expertise, or schemas, or some other
type of cognitive process?” The development of a psychometrically valid instrument to measure
professional skepticism is perhaps overdue, and will hopefully spur additional research into an
area that is both professionally mandated and is often cited as lacking in SEC actions against
auditors.
15
Figure 1
Model of Professional Skepticism
(Hurtt, Eining and Plumlee, 2001)
Behaviors of Skeptics
Characteristics of Skeptics
Expanded
Information
Search
Curiosity
Self-confidence
Interpersonal
understanding
Questioning
Individual’s
Level of
Professional
Skepticism
Increased
Contradiction
Detection
Increased
Alternative
Generation
Self Determining
Increased Scrutiny
of Interpersonal
Information
Deliberating
The model of professional skepticism addresses both characteristics and behaviors. The
professional skepticism instrument only measures the individual characteristics. These
characteristics are in the dotted portion of the model.
16
Table 1
Characteristics of Skeptics
(Adapted from Hurtt, Eining and Plumlee, 2001)
Characteristics of
Skeptics
Related Philosophical
and Psychological
Literature
Psychological Scales Examined
for initial item pool
Curiosity
Bunge (1991)
Johnson (1978)
Naylor (1981)
Spielberger (1995)
Melbourne Curiosity Inventory (MCI)
(Naylor 1981)
State-Trait Personality Inventory (STPI)
(Spielberger 1995)
Self-confidence
Forsman and Johnson (1996)
Fleming and Courtney (1984)
Helmreich and Stapp (1974)
Linn et al. (1982)
Interpersonal
understanding
Questioning
Self-determining
Deliberating
Burnyeat (1983)
Hallie (1985)
Hookway (1990)
Johnson (1978)
Kurtz (1992)
McGinn (1989)
Popkin (1979)
Dimensions of Self Esteem
(Forsman and Johnson 1996)
Dimensions of Self Esteem II
(Fleming and Courtney 1984)
Texas Social Behavior Inventory
(Helmreich and Stapp 1974)
Melbourne Curiosity Inventory (MCI)
(Naylor 1981)
State-Trait Personality Inventory (STPI)
(Spielberger 1995)
Fogelin (1994)
Kurtz (1992)
Rempel et al. (1985)
Rotter (1967)
Rotter (1980)
Spielberger (1995)
Stough (1969)
Wrightsman (1974)
State-Trait Personality Inventory (STPI)
Spielberger, 1995
10 Item Trust Scale (Rempel et al. 1985)
Trustworthiness and Independence Scale
(Wrightsman 1974)
Interpersonal Trust Scale (Rotter 1967)
Interpersonal Trust, Trustworthiness and
Gullibility Scale (Rotter 1980)
Carver (1997)
Lefcourt (1982)
Lefcourt (1991)
Stickland (1989)
Rotter (1966)
Locus of Control (Rotter 1966)
Bunge (1991)
Hallie (1985)
Kruglanski (1990)
Kruglanski et al. (1993)
Kurtz (1992)
MacDonald (1970)
McGinn (1989)
Mikulincer (1997)
Webster & Kruglanski (1994)
Tolerance of Ambiguity
(MacDonald 1970)
Need for Cognitive Closure
(Kruglanski 1990)
17
Table 2
Rotated Component Matrix for Pretest - 23-items
1
2
Factor
3
4
Scale  = .82; n=89
5
6
.88
.87
-.22
.06
.15
.23
.15
.24
-.12
-.29
.25
.21
.85
.84
-.00
-.09
.17
.26
.26
.16
-.29
-.35
.29
.20
.67
.05
.14
-.01
-.34
.34
-.00
-.03
.03
-.06
-.19
.89
.85
.80
.79
.76
.16
.09
.06
.12
-.05
-.23
-.09
-.09
-.07
-.10
-.16
-.02
.13
-.15
-.02
.09
.03
-.11
-.21
.03
.12
.00
.89
.22
-.31
.21
.12
.21
.09
.08
.82
.78
.06
.21
-.44
-.18
.18
.22
.28
.13
.74
.25
-.18
.33
.19
.17
.17
.30
.33
.41
-.06
-.08
-.23
.04
.08
-.06
.25
.22
.20
.35
.32
.36
.94
.94
.91
.20
.22
.33
-.24
-.25
-.11
-.92
-.90
-.79
Item
Interpersonal Understanding
( = .88)
I like to understand the reason for other people’s behavior.
I am interested in what causes people to behave the way that
they do.
I like thinking about how individual people influence events.
The actions people take and the reasons for those actions are
fascinating.
I often find myself imagining scenarios to explain why people
behave in a certain way.
Deliberating
( = .87)
I seldom notice inconsistencies in explanations.
Most often, I agree with what the others in my group think.
I tend to immediately accept what other people tell me.
I usually accept things I see, read, or hear at face value.
Bad luck or good luck can really follow you around.
Self-determining
( = .82)
I like to ensure that I’ve considered most available
information before making a decision.
I take my time when making decisions.
I don’t like to decide until I’ve looked at all of the readily
available information.
I carefully consider most information before I accept it as
true.
.11
.07
.13
Self-confidence
( = .93)
I have confidence in myself.
I feel sure of myself.
I feel good about me.
.17
.14
.16
Curiosity
( = .87)
I think that learning is exciting.
The prospect of learning excites me.
Discovering new information is fun.
.17
-.05
.23
.08
-.13
.79
.24
.32
.04
-.03
.43
.12
.14
.10
-.31
.01
.71
.69
Questioning
( = .57)
My friends tell me that I often question things that I see or
hear.
I enjoy trying to determine if what I read or hear is true.
I often doubt other people’s explanations.
18
Table 3
Rotated Component Matrix for the Pilot - 25-items
1
2
Factor
3
4
Scale  .85; n = 247
5
6
Item
(“R” = Reversed item)
.90
.90
.81
.76
.16
.17
.15
.41
.10
.13
.11
.02
-.22
-.30
-.44
-.30
.10
.07
.12
.27
.35
.44
.48
.38
.24
.01
.31
.79
.74
.74
.08
.00
.23
-.22
-.03
-.13
.21
.02
.53
.08
-.12
.11
.42
.73
.06
-.11
.33
.22
.06
.00
.06
.25
.11
.13
-.13
.00
.24
.17
.77
.73
.71
.66
.66
-.16
-.07
-.16
-.18
-.10
.08
.01
.20
.09
.32
.35
.16
.00
-.86
.08
.31
.26
.19
.22
.10
.06
.08
.22
.29
-.86
-.77
-.77
.21
.03
-.06
.02
.06
.15
.42
.46
.37
.42
.30
.09
.24
.21
.27
.04
.11
.02
-.04
.11
.11
.20
.18
.03
.14
.16
.26
-.07
-.07
.19
-.28
-.18
-.16
-.19
-.21
.69
.68
.63
.62
.12
.06
.21
-.08
Curiosity
( = .87)
I think that learning is exciting.
The prospect of learning excites me.
Discovering new information is fun.
I like learning more about many situations.
Deliberating
( = .75)
I take my time when making decisions.
I dislike having to make decisions quickly.
I don’t like to decide until I’ve looked at all of the readily
available information.
I like to ensure that I’ve considered most available information
before making a decision.
.26
.08
.12
.14
.09
Self-determining
( = .76)
It is easy for other people to convince me. (R)
I let what others say influence my decision making. (R)
Most often I agree with what the others in my group think. (R)
I usually I accept things I see, read, or hear at face value. (R)
I tend to immediately accept what other people tell me. (R)
.20
96
.22
.21
.10
Interpersonal Understanding
( = .84)
I am interested in what causes people to behave the way that they
do.
I like to understand the reason for other people’s behavior.
I seldom consider why people behave in a certain way. (R)
Other people’s behavior doesn’t interest me. (R)
.14
-.17
.04
.31
.88
.86
.86
.79
Questioning
( = .60)
My friends tell me that I often question things that I see or hear.
I usually notice inconsistencies in explanations.
I often reject statements unless I have proof that they are true.
I enjoy trying to determine if what I read or hear is true.
Self-confidence
( = .87)
I have confidence in myself.
I feel good about myself.
I am self-assured.
I don’t feel sure of myself. (R)
19
Table 4
Correlation Matrix of Skepticism Score with Wrightsman’s
Trustworthiness and Independence Scales
Skepticism
Trustworthiness
Independence
*
p < .001, n = 140
Skepticism
Trustworthiness
.115
.057
.507*
20
Table 5
Demographics on Professional Accounting Respondents
Mean
Standard Deviation
Months of professional experience
Age
Education:
Bachelor’s degree *
Advanced degree
Certifications held:
CPA
Other
Male
Female
Declined to respond
Total
n = 112
n = 87
n=1
n = 200
60
26
28.7
3.3
57
21
27.8
3.0
75
58.4
24
28.3
3.2
90
22
72
14
99
6
79
2
30
1
1
162
37
179
8
* One subject did not complete the information on education.
21
Table 6
Rotated Component Matrix for the Professional Subjects
1
2
Factor
3
4
Scale  = .86; n = 200
5
6
Item
(“R” = Reversed item)
Curiosity ( = .88.)
I think that learning is exciting.
I relish learning.
Discovering new information is fun
I like searching for knowledge.
The prospect of learning excites me.
I enjoy trying to determine if what I read or hear is true.
.91
.83
.83
.82
.77
.55
.09
.07
.07
.16
.03
.22
.15
.18
.07
.20
.17
.18
.21
.25
.20
.25
.08
.34
.26
.21
.28
.32
.28
.27
-.14
-.22
-.11
-.10
-.12
-.45
.07
.03
.82
.80
-.06
.14
.18
.22
-.01
.02
-.14
-.16
.03
.17
.77
.76
-.03
.23
.08
.12
-.22
.09
-.25
-.13
.11
.69
.01
.06
-.14
.02
Deliberating ( = .83.)
I take my time when making decisions.
I don’t like to decide until I’ve looked at all of the readily available
information.
I dislike having to make decisions quickly.
I like to ensure that I’ve considered most available information before
making a decision.
I wait to decide on issues until I can get more information.
.07
.12
.13
.12
.08
.25
.07
.19
-.03
.08
-.03
.11
.76
.76
.74
.64
.56
.49
.07
.11
.07
.10
-.03
.05
.25
.06
.22
.29
.12
.17
-.37
-.25
-.37
-.57
.05
-.11
Self-determining ( = 76.)
I tend to immediately accept what other people tell me. (R)
I usually I accept things I see, read, or hear at face value. (R)
I often accept other people’s explanations without further thought. (R)
It is easy for other people to convince me. (R)
Most often I agree with what the others in my group think. (R)
I usually notice inconsistencies in explanations.
.28
.22
.33
.19
.19
.12
.10
-.00
-.04
.90
.88
.88
.13
.06
.02
-.21
-.18
-.10
.14
.05
.06
.12
.17
-.02
.79
.75
.08
.11
-.03
.14
Interpersonal Understanding ( = 90.)
I like to understand the reason for other people’s behavior.
I am interested in what causes people to behave the way that they do.
The actions people take and the reasons for those actions are
fascinating.
I seldom consider why people behave in a certain way. (R)
Other people’s behavior doesn’t interest me. (R)
.29
.25
.34
.22
.24
-.10
-.03
-.01
-.01
-.08
.17
.25
.23
.11
.18
.02
.11
.14
.11
.04
.90
.87
.87
.84
.81
-.06
-.07
-.12
-.15
.06
Self-confidence ( = 91.)
I have confidence in myself.
I don’t feel sure of myself. (R)
I am self-assured.
I am confident of my abilities.
I feel good about myself.
.23
.32
.00
.06
.07
.31
.21
.46
.21
.11
.12
-.02
.07
.17
.01
-.85
-.78
-.60
Questioning ( = .67.)
My friends tell me that I often question things that I see or hear.
I frequently question things that I see or hear.
I often reject statements unless I have proof that they are true
22
American Institute of Certified Public Accountants (AICPA), 1997. Due care in the performance
of work, in Statement on Auditing Standards No. 1 (New York, NY).
Bierstaker J. L., S. Wright. 2001. A research note concerning practical problem-solving ability as
a predictor of performance in auditing tasks. Behavioral Research in Accounting 13 : 4959.
Bunge M. 1991. A skeptic's beliefs and disbeliefs. New Ideas in Psychology 9 (2): 131-149.
Burnyeat M. 1983. The Skeptical Tradition.(Berkeley: University of California Press).
Carver C. S. 1997. The Internal-External Scale Confounds Internal Locus of Control with
Expectancies of Positive Outcomes. Personality and Social Psychology Bulletin 23 (6):
580-585.
Choo F., K. Tan. 2000. Instruction, skepticism, and accounting students' ability to detect frauds
in auditing. The Journal of Business Education 1 (Fall): 72 - 87.
Clark L. A., D. Watson. 1998. “Constructing Validity: Basic issues in objective scale
development”, in Methodological issues and strategies in clinical research. Ed A. E.
Kazdin (Washington, DC: Washington, DC) 215-239.
Converse J. M., S. Presser. 1990. Survey Questions Handcrafting the standardized
questionnaire.(Newbury Park, CA: SAGE Publications, Inc.).
DeVellis R. F. 1991. Scale Development: Theory and Applications.(Thousand Oaks, California:
SAGE Publications).
Fink A. 1995. How to Design Surveys.(Thousand Oaks, California: SAGE Publications, Inc.).
Fleming J. S., B. E. Courtney. 1984. The dimensionality of self-esteem II: Heirarchical facet
model for revised measurement scales. Journal of Personality and Social Psychology 46 :
404-421.
Fogelin R. J. 1994. Pyrrhonian reflections on knowledge and justification.(New York: Oxford
University Press).
Forsman L., M. Johnson. 1996. Dimensionality and validity of two scales measuring different
aspects of self-esteem. Scandinavian Journal of Psychology 37 : 1-15.
Hallie P. P., 1985. Sextus Empiricus: Selections from the major writings on skepticism, man &
God, (Indianapolis) pp 238.
Helmreich R., J. Stapp. 1974. Short forms of the Texas Social Behavior Inventory (TSBI), an
objective measure of self-esteem. Bulletin of the Psychonomic Society 4 : 473-475.
23
Hookway C. 1990. Scepticism.(New York: Routledge).
Hurtt R. K., M. Eining, R. D. Plumlee. 2001. What is Professional Skepticism? Working Paper,
University of Wisconsin: .
Hyatt T. A., D. F. Prawitt. 2001. Does Congruence between audit structure and auditors' locus of
control affect job performance? The Accounting Review 76, No. 2 (April): 263-274.
Johnson O. A. 1978. Skepticism and cognitivism.(Berkeley: University of California Press, Ltd.).
Kaplan S. E., S. M. Whitecotton. 2001. An examination of Auditors' Reporting Intentions when
another auditor is offered client employment. Auditing: A journal of practice and theory
20 (1): 45-63.
Kruglanski A. W. 1990. “Motivations for judging and knowing: Implications for causal
attribution”, in The handbook of motivation and cognition: Foundation of social
behavior. Eds E. T. Higgins and R. M. Sorrention (New York: New York) 333-368.
Kruglanski A. W., D. M. Webster, A. Klem. 1993. Motivated resistance and openness to
persuasion in the presence or absence of prior information. Journal of Personality and
Social Psychology 65 (5): 864-876.
Kurtz P. 1992. The New Skepticism: inquiry and reliable knowledge.(Buffalo, New York:
Prometheus Books).
Lefcourt H. M. 1982. Locus of Control: Current Trends in Theory and Research.(Hillsdale, NJ:
Lawrence Erlbaum Associates).
Lefcourt H. M. 1991. “Locus of Control”, in Measures of personality and social psychological
attitudes. Eds J. P. Robinson, P. R. Shaver and L. S. Wrightsman (San Diego: San Diego)
413-499.
Linn M. C., T. de Benedictis, K. Delucchi. 1982. Adolescent reasoning about advertisements:
Preliminary investigations. Child Development 53 : 1599 - 1613.
Litwin M. S. 1995. How to measure survey reliability and validity.(Thousand Oaks, California:
SAGE Publications, Inc).
MacDonald A. P. J. 1970. Revised Scale for Ambiguity tolerance: reliability and validity.
Psychological Reports 26 : 791-798.
McGinn M. 1989. Sense and Certainty: A dissolution of scepticism.(New York: Basil Blackwell
Inc.).
24
McMillan J. J., R. A. White. 1993. Auditors' Belief Revisions and Evidence Search: The Effect
of Hypothesis Frame, Confirmation and Professional Skepticism. The Accounting Review
July : 443-465.
Messick S. 1998. “Validity of psychological assessment: validation of inferences from persons
responses and performances as scientific inquiry into score meaning”, in Methodological
issues and strategies in clinical research. Ed A. E. Kazdin (Washington, DC:
Washington, DC) 241-260.
Mikulincer M. 1997. Adult attachment style and information processing: individual differences
in curiosity and cognitive closure. Journal of Personality and Social Psychology 72 (5):
1217 - 1230.
Naylor F. D. 1981. A state-trait curiosity inventory. Austrailian Psychologist 16 (2): 172 - 183.
Popkin R. H. 1979. The history of scepticism from Erasmus to Spinoza.(Berkeley: University of
California Press).
Rempel J. K., J. G. Holmes, M. P. Zanna. 1985. Trust in close relationships. Journal of
Personality and Social Psychology 49 : 95-112.
Rotter J. B. 1966. Generalized Expectancies for Internal versus external control of
reinforcement. Psychological Monographs: General and Applied (Whole No. 601) 80,
no. 1.
Rotter J. B. 1967. A new scale for the measurement of interpersonal trust. Journal of Personality
35 : 651-665.
Rotter J. B. 1980. Interpersonal trust, trustworthiness, and gullibility. American Psychologist 35
(1): 1 - 7.
Schwarz N. 1996. Cognition and Communication: Judgmental Biases, Research Methods, and
the logic of conversation.(Mahwah, New Jersey: Lawrence Erlbaum Associates).
Shaub M. K. 1996. Trust and suspicion: The effects of situational and dispositional factors on
auditors' trust of clients. Behavioral Research in Accounting 8 : 154-174.
Shaub M. K., J. E. Lawrence. 1996. Ethics, experience and professional skepticism: A
situational analysis. Behavioral Research in Accounting 8 (Supplement): 124-157.
Shaub M. K., J. E. Lawrence. 1999. Differences in auditors' professional skepticism across
career levels in the firm. Advances in Accounting Behavioral Research 2 : 61 - 83.
Spielberger C. D., 1995. State-Trait Personality Inventory (STPI), (Redwood City, California) .
25
Stickland B. R. 1989. Internal-external control expectancies: From contingency to creativity.
American Psychologist 44 : 1-12.
Stough C. L. 1969. Greek Skepticism.(University of California Press, Ltd.).
Webster D. M., A. W. Kruglanski. 1994. Individual differences in need for cognitive closure.
Journal of Personality and Social Psychology 67 (6) : 1049-1062.
Wright M. E., R. Davidson. 2000. The effect of auditor attestation and tolerance for ambiguity
on commercial lending decisions. Auditing: A journal of practice and theory 19 (2).
Wrightsman L. S. 1974. Assumptions about human nature: A social-psychological
approach.(Monterey, California: Brooks/Cole Publishing Company).
26