Frame, Confirmation Bias, and Professional Skepticism
Jeffrey J. McMillan; Richard A. White
The Accounting Review, Vol. 68, No. 3 (ul., 1993), 443-465.
Stable URL:
tp: flinks,jstor-org/sici?sici~0001-4826%28 19930729583 AIZICMIZIAABRAES Z3E2.0.CO®IB2-E,
‘The Accounting Review is currently published by American Accounting Association.
‘Your use of the ISTOR archive indicates your acceptance of ISTOR’s Terms and Conditions of Use, available at
htp:sseww jstor org/aboutiterms.html. ISTOR’s Terms and Conditions of Use provides, in part, that unless you
hhave obtained prior permission, you may aot download an entie issue of a journal or multiple copies of articles, and
‘you may use content in the ISTOR archive only for your personal, non-commercial use
Please contact the publisher eegarding aay futher use ofthis work. Publisher contact information ray he abained at
ftps jtor.orgourmaleaaaso. hl.
Each copy of any part ofa JSTOR transenission must contain the same copyright tice that appears on the screen or
printed page of such transtnission,
ISTOR isan independent not-for-profit organization dedicated to creating and preserving a digital archive ot
scholarly journals. For more information regarding ISTOR, please contact suppom@jstor org.
hup:thwww.jstor.orgy
Sat May 29 06:58:47 2004Mien
Auditors’ Belief Revisions and
Evidence Search: The Effect of
Hypothesis Frame, Confirmation
Bias, and Professional Skepticism
Jeffrey J. McMillan
Clemson University
Richard A. White
University of South Carolina
SYNOPSIS AND INTRODUCTION: The primary purpose of this study
was to investigate how auditors’ belief revisions and evidence search are in-
fluenced by the frame of the hypothesis being tested and by confirmation
bias and professional skepticism (conservative bias). This study is dis
guished from prior research in three main ways. First, the effects of confi
‘mation bias and professional skepticism on the complementary audit func-
tions of evidence evaluation and evidence search are examined jointly in the
same experiment. Second, unlike many studies which have examined belief
revisions of auditors solely under an error frame (Ashton and Ashton 1988,
1990; Tubbs et al. 1990), this study examines the judgments and behavior
of auditors operating with both environmental (nonertor) and error-framed
hypotheses. Third, the full effects of confirmation bias and professional
skepticism were enhanced by having the subjects establish their own hy-
pothesis frame and likelihood assessments, rather than respond to preset
conditions.
Confirmation bias implies that auditors may seek to confirm their
hypotheses and so may favor information thet confirms rather than refutes,
their initial assessments. This approach could lead to premature closure on
a belief or hypothesis. Professional skepticism implies that auditors focus,
We agpreclate the criticisms and suggestions ofthe ananymaus reviewers. associste editor, and editar. Wa
Bruce Hadlock, Paul Hartson, Arnold Schoetder, Brad Tuttle, and for the guidance of disetation
committer members Thomas Cafferty, Gene Chewing, James &, Eawards, and Alen Winters, Special grate
{de s expressed ta james F. Fdwards, wha designed the computer program nacessery 0 carry aut the experi
‘mental task. An earlier version ofthis article was presented at the 1991 Antal Meeting of the American
‘Accounting Association.
Submitted August 1991
Accepted December 1992.
443aaa ‘The Accounting Review, July 1983
more on error-related evidence. An approach that is too conservative may
lead to the performance of unnecessary audit procedures and thereby
reduce audit efficiency. Environmental conditions refer to economic
changes, changes in the industry and geographic area in which the com-
pany operates, or changes in company policies ragarding investment,
marketing, and financing strategies. Error conditions refer to intentional or
unintentional misstatements in the financial statements.
A field experiment was conducted in which auditors reviewed prelimin-
ary audit information and then indicated whether they favored an en-
vironmental or an error-framed hypothesis as the most likely explanation of
an observed fluctuation in financial statement ratios. A likelihood assess-
‘ment was assigned to the favored hypothesis frame, and the auditors were
then asked to seek information to test their initial hypothesis from a list of
audit questions. After evaluating the audit cues, the auditors updated their
hypothesis beliefs and continued their evidence search.
The results of this study indicate that auditors reacted differently to
audit evidence, depending upon the frame of the hypothesis they favored
and their belief extremity. Specifically, auditors who favored the error
frame reacted more strongly to both confirming and disconfirming evidence
than did those who favored the environmental frame. Furthermore, in con=
formance with the findings of Ashton and Ashton (1988, 1990) and contrary
to Bamber et al. (1991), the auditors were more responsive to discanfirming
evidence than to confirming evidence when belief revision was measured
with an absolute scale. Howaver, when the relative change in belief revision
was measured with a proportional scale, the magnitude of response with
confirming evidence was not significantly different from that with discon-
firming evidence.
The auditors’ continued avidence search was conditioned by @ conser-
vative bias irréspective of the hypothesis frame favored or belief extremity;
that is, their search strategy emphasized the uncovering of potential
material errors (Smith and Kida 1991). Because the conservative bias was
stronger for auditors who favored the error frame, confirmation bias may
partially account for this effect by enhancing the emphasis on error. Con-
versely, confirmation bias may have somewhat weakened the effect of con-
servative bias for those auditors who favored the environmental frame.
Key Wards: Belief revision, Evidence search, Hypothesis frame, Confirma-
tion bias, Professional skepticism, Conservative bias.
Data Availability: Data used in this study are available from the first
author upon request.
N experiment to examine the effects of hypothesis frame, confiematian bias, and
A professional skepticism consisted of twa phases. In the first phase, reported in.
McMillan and White (1992), we established the manipulation of hypothesis
frame and examined whether an auditor’s initial bypotbesis affected the initial evi
dence search and reflected a confirmaiory ot conservative search strategy. In the
second phase, we continued to investigate the effects these factors had on the auditor's
decision making following the evaluation of audit evidence. Specifically, we examined