Description
PriorauditingresearchandPCAOBinspectionreportsnotethatauditorshavedifficultyauditingcom-plexaccountingestimates.Rasso(2015)testswhetherprovidingauditorswithdocumentationinstruc-tions“basedonthepreceptsofconstrual-leveltheory(CLT)”improvesauditoreffectivenessandskep-ticismwhenauditingthesetypesofestimates.Inthisdiscussion,wehighlightsomecommentsaboutRasso’sstudyfromparticipantsattheAccounting,OrganizationsandSocietyConferenceonAccountingEs-timatesandshareourownperspectivesaboutthestudy’sconclusionsandimplicationsforresearchandpractice.
Accounting, Organizations and Society 46 (2015) 56–58
Contents lists available at ScienceDirect
Accounting, Organizations and Society
journal homepage: www.elsevier.com/locate/aos
Discussion of construal instructions and professional skepticism in
evaluating complex estimates
Michele L. Frank
a,b
, Vicky B. Hoffman
a,?
a
Katz Graduate School of Business, University of Pittsburgh, Pittsburgh, PA 15260, United States
b
Miami University, Farmer School of Business, Department of Accountancy, 800 E. High Street, Oxford, OH 45056-3600, United States
a r t i c l e i n f o
Article history:
Received 15 April 2015
Accepted 16 April 2015
Available online 16 May 2015
a b s t r a c t
Prior auditing research and PCAOB inspection reports note that auditors have di?culty auditing com-
plex accounting estimates. Rasso (2015) tests whether providing auditors with documentation instruc-
tions “based on the precepts of construal-level theory (CLT)” improves auditor effectiveness and skep-
ticism when auditing these types of estimates. In this discussion, we highlight some comments about
Rasso’s study from participants at the Accounting, Organizations and Society Conference on Accounting Es-
timates and share our own perspectives about the study’s conclusions and implications for research and
practice.
© 2015 Elsevier Ltd. All rights reserved.
Introduction
Both auditing research (Bratten, Gaynor, McDaniel, Montague, &
Sierra, 2013; Gri?th, Hammersley, & Kadous, 2014) and PCAOB in-
spection reports (PCAOB, 2012) note that auditors have di?culty
auditing complex accounting estimates. Rasso (2015) examines an
intervention designed to improve the way auditors process evi-
dence related to such estimates. The intervention provides audi-
tors with documentation instructions “based on the precepts of
construal-level theory (CLT)” to see if this improves audit effective-
ness and increases auditors’ professional skepticism. Speci?cally,
Rasso applies CLT to predict that auditors provided with instruc-
tions to use high-level construals will be more skeptical than both
auditors provided with instructions to use low-level construals and
those who do not receive documentation instructions. In this dis-
cussion, we highlight some comments on Rasso’s study from par-
ticipants at the Accounting, Organizations and Society Conference on
Accounting Estimates and share our own perspectives about the
study and its implications for research and practice.
CLT was initially developed in the psychology literature to ex-
plain how differences in an event’s temporal distance (i.e., whether
it will occur in the immediate versus distant future) in?uence in-
dividuals’ judgments and choices with respect to the event (Trope
& Liberman, 2011). In its earliest form, CLT predicted that when
thinking about events that are expected to occur in the near fu-
?
Corresponding author.
ture, individuals naturally interpret information about the event at
a low-level, focusing on the details, speci?c steps, and any situ-
ational constraints related to the event’s occurrence. In contrast,
when thinking about events that will occur in the distant future,
individuals tend to interpret information at a high-level, thinking
abstractly about the overall essence of the event. In explaining how
CLT’s use has broadened over time, Trope and Liberman (2011)
note that it is currently used to explain how factors that affect an
individual’s psychological distance from a situation (including tem-
poral distance) in?uence how individuals interpret or construe the
situation.
1
This line of research consistently demonstrates that
people tend to interpret psychologically distant events at a high-
level of abstraction, and interpret psychologically closer events at
a low-level of abstraction.
Rasso’s paper does not measure auditors’ psychological distance
from the situation they are evaluating and does not predict how
differences in psychological distance in?uence the way auditors
construe evidence. Instead, Rasso builds on the CLT literature by
examining whether providing auditors with explicit instructions
prompting them to construe evidence at a high versus low level
causes them to respond differently. Rasso’s study complements a
1
“Psychological distance” is de?ned as an individual’s subjective feeling of
closeness to the situation that he/she is evaluating (Trope & Liberman, 2010). Exam-
ples of factors, other than temporal distance, that in?uence psychological distance
include whether an individual is directly involved in or merely observes a situa-
tion faced by someone else (i.e., is an actor versus an observer) and whether the
situation involves members of the individual’s in-group or out-group.http://dx.doi.org/10.1016/j.aos.2015.04.005
0361-3682/© 2015 Elsevier Ltd. All rights reserved.
M.L. Frank, V.B. Hoffman/ Accounting, Organizations and Society 46 (2015) 56–58 57
growing stream of auditing research (e.g., Backof, Bamber, & Car-
penter, 2014; Gri?th, Hammersley, Kadous, & Young, 2015) that
changes the way auditors interpret information by changing the
level at which they construe information or by manipulating the
mindset they adopt to see how doing so in?uences their judgments
and behaviors. These studies also add to the recent literature ex-
amining how instructional interventions or training can increase
auditors’ professional skepticism (e.g., Plumlee, Rixom, & Rosman,
2015).
Discussion of Rasso’s approach
Overview
Rasso posits that auditors have trouble integrating the various
pieces of evidence they collect when they encounter evidence se-
quentially, and are prone to prematurely terminating their evi-
dence collection process. He predicts that telling auditors to con-
strue or frame evidence at a high level (versus a low level) will in-
crease auditors’ professional skepticism, thereby improving the ef-
fectiveness of the auditors’ evidence collection and evaluation pro-
cess.
In his study, participants are provided with several opportuni-
ties to collect sets of mixed evidence related to a client’s com-
plex estimate. Each set of evidence includes the same proportion
of positive and negative items. Rasso manipulates whether partic-
ipants receive documentation instructions, and if so, what type of
instructions they receive. In the high-level construal condition, par-
ticipants are told to think “broadly about all of the evidence col-
lectively” and to list reasons why management’s estimate could be
presented fairly and why it could be misstated. In contrast, audi-
tors in the low-level construal condition are asked to think “specif-
ically about each evidence item” and to list reasons how manage-
ment’s estimate could be presented fairly and how it could be mis-
stated. Participants in a control condition receive no documenta-
tion instructions.
It is clear that Rasso’s experimental manipulation caused dif-
ferences in auditor participants’ responses. Conference participants’
comments tended to focus on potential reasons for, and implica-
tions of, these differences. We group our comments into four major
areas: (1) What else, besides construal level, might the documenta-
tion instruction manipulation be capturing? (2) Do the changes in
participants’ responses re?ect skepticism and only skepticism? (3)
Do the changes due to high-level construals represent enhanced
audit effectiveness? and (4) Could one apply this approach in prac-
tice to train novices to think more like experts?
What else, besides construal level, might the study’s manipulation
have captured?
Rasso’s central premise is that construing information at a high
level allows “for easier assimilation of information because the
representations become less ambiguous, more coherent, and more
schematic.” In contrast, adopting a low-level construal makes it
more likely that auditors will process information in a piecemeal
fashion without considering how previously received evidence re-
lates to new evidence. Consequently, auditors who interpret infor-
mation at a high level should be better able to see the “big picture”
or overall pattern suggested by the evidence.
There were several comments about exactly how high versus
low-level construals were manipulated, and whether the manipu-
lation might perhaps be too broad or too narrow. The documen-
tation instructions provided to the auditor participants ask them
to do two things. In the high-level construal condition auditors are
asked to both “think broadly about all the evidence collectively”
and about “why” management did something; whereas in the low-
level construal condition they are asked to both “think speci?cally
about each individual evidence item” and about “how” manage-
ment did something. The two parts of each manipulation could
actually represent two distinct constructs. For example, would it
be possible to also think broadly about “how” management’s es-
timate could be misstated? Perhaps, “how” questions capture the
mechanism by which the estimate could be misstated, whereas
“why” questions capture the motivation for misstating the esti-
mate. Listing “why” reasons could be tantamount to asking audi-
tors to consider the “perceived motivation” side of the fraud tri-
angle (SAS No. 99, American Institute of Certi?ed Public Accoun-
tants, 2002), which would naturally result in increased skepticism.
If this is true, auditors instructed to consider why management did
something may be more skeptical of management irrespective of
whether they are better able to assimilate different pieces of audit
evidence. The question is whether instructions to “think broadly”
would have led to increased skepticism if the “why” part of the
instructions had not been present.
It was also noted that in psychology research, priming instruc-
tions are intentionally set in a different context from that of the
primary task in order to avoid demand effects, and to measure
subconscious effects rather than intentional ones. Unlike priming
studies, Rasso’s documentation instructions ask auditors to con-
sciously process information in a particular way, and the instruc-
tions are related to his primary task. However, it is reassuring that
in a subsequent manipulation check, Rasso ?nds that participants
continue to interpret items unrelated to the primary task at the
same level of construal as in the primary task, suggesting that the
construal instructions do have carryover effects that might be sub-
conscious. This ?nding could reduce the concern that the results of
the primary study could be driven by demand effects.
Do the changes in participants’ responses re?ect skepticism and only
skepticism?
To examine participants’ level of skeptical action, Rasso mea-
sures the amount of time participants spent collecting and review-
ing evidence and the number of times they collected additional
sets of evidence. To measure skeptical judgment, he asks partici-
pants to assess the risk that the client’s estimate is materially mis-
stated and the likelihood that they would recommend an adjust-
ment to the client’s estimate.
Several conference participants called into question the study’s
premise that gathering more evidence and taking more time to
perform the task are necessarily evidence of increased skepticism.
It was also suggested that the study’s measures might actually be
capturing participants’ uncertainty about the task, but the speci?c
reasoning underlying this concern was not discussed further at the
conference. We next explain one reason why uncertainty, rather
than differences in construal level, could explain the differences in
auditors’ responses observed in Rasso’s study. It seems plausible
that there is greater uncertainty in the high-level construal condi-
tion because, based on the results of the control condition, Rasso
concludes that auditors naturally tend to process and evaluate in-
formation at a low-level of abstraction. If this is the case, auditors
in the high-level construal condition are being asked to process in-
formation differently from how they normally would on an actual
audit. This could explain why the study ?nds that participants in
the high-level construal condition perceive their task to be more
di?cult than in the low-level construal condition. It would also
explain why they would therefore need to spend more time on
the task than did auditors in the conditions that are more natu-
ral to them. If being forced to process information in an unfamiliar
manner increases participants’ uncertainty regarding the task, it is
possible that the differences observed with respect to “perceived
58 M.L. Frank, V.B. Hoffman/ Accounting, Organizations and Society 46 (2015) 56–58
risk” and “likelihood of adjustment” may re?ect differences in un-
certainty rather than, or in addition to, differences in skepticism.
Do the changes due to high-level construals represent enhanced audit
effectiveness?
Because in this study there was no “correct answer” and no
actual cost to gathering additional sets of evidence, conference
participants questioned what constituted the best, or most effec-
tive, response. For example, some conference participants won-
dered why anyone would choose an incomplete set of evidence
when one could obtain more evidence for free. Others wanted to
better understand the mechanism underlying the high-level con-
strual participants’ request for more information. Speci?cally, if au-
ditors in the high-level construal condition had a greater ability
to see connections and link the pieces of information, as hypothe-
sized, they might be expected to see the big picture more quickly,
and, would thus require fewer sets of evidence, rather than more.
Given the lack of a correct answer, any inferences about the rela-
tive effectiveness of the high-level versus low-level construal con-
ditions should be interpreted with caution.
Could one apply this approach in practice to train novices to think
more like experts?
If “thinking broadly” helps participants think about evidence
more abstractly, integrate it, and recognize patterns in the infor-
mation, as Rasso intends, high-level construal instructions could be
viewed as a technique that promotes thinking more like an expert.
Indeed, the examples of such reasoning that Rasso describes, such
as “chunking” information to remember it better, are classic exam-
ples from the expertise literature (Chase & Simon, 1973). However,
it is possible that these instructions would have different effects on
auditors with different levels of experience. For example, prior re-
search (Hoffman, Joe, & Moser, 2003) ?nds that more experienced
auditors naturally restructure and integrate components of work-
paper evidence, whereas less experienced auditors tend to process
evidence sequentially in the order they encounter it.
Therefore, while the intervention might help staff and senior
auditors to structure their task more like an experienced auditor
would naturally structure it, it is less clear what effect the in-
structions would have on more experienced auditors. The study
includes auditors with four levels of experience (staff, senior, man-
ager, and partner). However, because most participants are senior
auditors and there are only a few participants at each of the other
three levels, it is unclear whether all types of auditors react sim-
ilarly to the instructions. It might be the case that because man-
agers and partners naturally reason more broadly, they would not
respond as much to the high-level construal instructions.
Another issue to consider is whether there is value to hav-
ing some members of the audit team think about the details (i.e.,
low-level construal), while other members think more broadly (i.e.,
high-level construal). For example, Ramsay (1994) ?nds that while
managers and partners are better able than staff and seniors to de-
tect conceptual errors in the audit workpapers, they are less likely
to identify mechanical errors. If we were able to train less senior
auditors to think more broadly, it is possible that we might inad-
vertently sacri?ce what the novices do well – perhaps it is bene-
?cial to have someone think more speci?cally about the detailed
workpaper evidence.
Conclusions
Rasso’s study addresses an important audit issue that is of
great interest to regulators, practitioners, and audit researchers.
The study’s ?ndings suggest that there could be bene?ts to provid-
ing auditors with documentation instructions based on CLT. How-
ever, based on the various issues outlined above, some caution
should be exercised when interpreting the study’s results.
Audit practitioners and PCAOB representatives who attended
the conference were keen to understand how auditing standards
might be changed or how the results of this study could be imple-
mented in practice. Researchers were more cautious about imple-
menting the results of a particular study without further research.
If we could get more consensus from the body of recent research
applying construal level theory and the effect of changing the au-
ditor’s mindset on audits of complex estimates, ultimately the goal
would be to implement the replicable research ?ndings into audit
practice.
Acknowledgement
We thank Deloitte and Touche, LLP for generously funding this
conference. We are also grateful to Hun Tong Tan (editor) and the
conference participants for their helpful comments.
References
American Institute of Certi?ed Public Accountants (2002). Consideration of fraud in
a ?nancial statement audit. Statement of Auditing Standards No. 99. New York,
NY: AICPA.
Backof, A.G., Bamber, E.M., & Carpenter, T.D. (2014). Auditing complex estimates:
Management-provided evidence and auditors’ consideration of inconsistent evi-
dence. Working paper, University of Virginia.
Bratten, B., Gaynor, L. M., McDaniel, L., Montague, N. R., & Sierra, G. E. (2013). The
audit of fair values and other estimates: The effects of underlying environmen-
tal, task and auditor-speci?c factors. Auditing: A Journal of Practice & Theory,
32(Supplement 1), 7–44.
Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive Psychology, 4,
55–81.
Gri?th, E. E., Hammersley, J. S., & Kadous, K. (2014). Audits of complex estimates as
veri?cation of management numbers: How institutional pressures shape prac-
tice. Forthcoming, Contemporary Accounting Research.
Gri?th, E. E., Hammersley, J. S., Kadous, K., & Young, D. (2015). Auditor mindsets
and audits of complex estimates. Journal of Accounting Research, 53(1), 49–77.
Hoffman, V. B., Joe, J. R., & Moser, D. V. (2003). The effect of constrained processing
on auditors’ judgments. Accounting, Organizations, and Society, 28(7–8), 699–714.
PCAOB (2012). Maintaining and applying professional skepticism in audits. PCAOB
Staff Audit Practice Alert No. 10. December 4, 2012. Washington, D.C: PCAOB.
Plumlee, D., Rixom, B., & Rosman, A. (2015). Training auditors to perform analytical
procedures using metacognitive skills. The Accounting Review, 90(1), 351–369.
Ramsay, R. J. (1994). Senior/manager differences in audit workpaper review perfor-
mance. Journal of Accounting Research, 32(1), 127–135.
Rasso, J. (2015). Construal instructions and professional skepticism in evaluating
complex estimates. Accounting, Organizations, and Society, in press.http://dx.doi.
org/10.1016/j.aos.2015.03.003.
Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance.
Psychological Review, 117(2), 440–463.
Trope, Y., & Liberman, N. (2011). Construal level theory. In P. Van Lange,
A. W. Kruglanski, & E. T. Higgins (Eds.), Handbook of theories of social psychology.
London: Sage Publications.
doc_670257201.pdf
PriorauditingresearchandPCAOBinspectionreportsnotethatauditorshavedifficultyauditingcom-plexaccountingestimates.Rasso(2015)testswhetherprovidingauditorswithdocumentationinstruc-tions“basedonthepreceptsofconstrual-leveltheory(CLT)”improvesauditoreffectivenessandskep-ticismwhenauditingthesetypesofestimates.Inthisdiscussion,wehighlightsomecommentsaboutRasso’sstudyfromparticipantsattheAccounting,OrganizationsandSocietyConferenceonAccountingEs-timatesandshareourownperspectivesaboutthestudy’sconclusionsandimplicationsforresearchandpractice.
Accounting, Organizations and Society 46 (2015) 56–58
Contents lists available at ScienceDirect
Accounting, Organizations and Society
journal homepage: www.elsevier.com/locate/aos
Discussion of construal instructions and professional skepticism in
evaluating complex estimates
Michele L. Frank
a,b
, Vicky B. Hoffman
a,?
a
Katz Graduate School of Business, University of Pittsburgh, Pittsburgh, PA 15260, United States
b
Miami University, Farmer School of Business, Department of Accountancy, 800 E. High Street, Oxford, OH 45056-3600, United States
a r t i c l e i n f o
Article history:
Received 15 April 2015
Accepted 16 April 2015
Available online 16 May 2015
a b s t r a c t
Prior auditing research and PCAOB inspection reports note that auditors have di?culty auditing com-
plex accounting estimates. Rasso (2015) tests whether providing auditors with documentation instruc-
tions “based on the precepts of construal-level theory (CLT)” improves auditor effectiveness and skep-
ticism when auditing these types of estimates. In this discussion, we highlight some comments about
Rasso’s study from participants at the Accounting, Organizations and Society Conference on Accounting Es-
timates and share our own perspectives about the study’s conclusions and implications for research and
practice.
© 2015 Elsevier Ltd. All rights reserved.
Introduction
Both auditing research (Bratten, Gaynor, McDaniel, Montague, &
Sierra, 2013; Gri?th, Hammersley, & Kadous, 2014) and PCAOB in-
spection reports (PCAOB, 2012) note that auditors have di?culty
auditing complex accounting estimates. Rasso (2015) examines an
intervention designed to improve the way auditors process evi-
dence related to such estimates. The intervention provides audi-
tors with documentation instructions “based on the precepts of
construal-level theory (CLT)” to see if this improves audit effective-
ness and increases auditors’ professional skepticism. Speci?cally,
Rasso applies CLT to predict that auditors provided with instruc-
tions to use high-level construals will be more skeptical than both
auditors provided with instructions to use low-level construals and
those who do not receive documentation instructions. In this dis-
cussion, we highlight some comments on Rasso’s study from par-
ticipants at the Accounting, Organizations and Society Conference on
Accounting Estimates and share our own perspectives about the
study and its implications for research and practice.
CLT was initially developed in the psychology literature to ex-
plain how differences in an event’s temporal distance (i.e., whether
it will occur in the immediate versus distant future) in?uence in-
dividuals’ judgments and choices with respect to the event (Trope
& Liberman, 2011). In its earliest form, CLT predicted that when
thinking about events that are expected to occur in the near fu-
?
Corresponding author.
ture, individuals naturally interpret information about the event at
a low-level, focusing on the details, speci?c steps, and any situ-
ational constraints related to the event’s occurrence. In contrast,
when thinking about events that will occur in the distant future,
individuals tend to interpret information at a high-level, thinking
abstractly about the overall essence of the event. In explaining how
CLT’s use has broadened over time, Trope and Liberman (2011)
note that it is currently used to explain how factors that affect an
individual’s psychological distance from a situation (including tem-
poral distance) in?uence how individuals interpret or construe the
situation.
1
This line of research consistently demonstrates that
people tend to interpret psychologically distant events at a high-
level of abstraction, and interpret psychologically closer events at
a low-level of abstraction.
Rasso’s paper does not measure auditors’ psychological distance
from the situation they are evaluating and does not predict how
differences in psychological distance in?uence the way auditors
construe evidence. Instead, Rasso builds on the CLT literature by
examining whether providing auditors with explicit instructions
prompting them to construe evidence at a high versus low level
causes them to respond differently. Rasso’s study complements a
1
“Psychological distance” is de?ned as an individual’s subjective feeling of
closeness to the situation that he/she is evaluating (Trope & Liberman, 2010). Exam-
ples of factors, other than temporal distance, that in?uence psychological distance
include whether an individual is directly involved in or merely observes a situa-
tion faced by someone else (i.e., is an actor versus an observer) and whether the
situation involves members of the individual’s in-group or out-group.http://dx.doi.org/10.1016/j.aos.2015.04.005
0361-3682/© 2015 Elsevier Ltd. All rights reserved.
M.L. Frank, V.B. Hoffman/ Accounting, Organizations and Society 46 (2015) 56–58 57
growing stream of auditing research (e.g., Backof, Bamber, & Car-
penter, 2014; Gri?th, Hammersley, Kadous, & Young, 2015) that
changes the way auditors interpret information by changing the
level at which they construe information or by manipulating the
mindset they adopt to see how doing so in?uences their judgments
and behaviors. These studies also add to the recent literature ex-
amining how instructional interventions or training can increase
auditors’ professional skepticism (e.g., Plumlee, Rixom, & Rosman,
2015).
Discussion of Rasso’s approach
Overview
Rasso posits that auditors have trouble integrating the various
pieces of evidence they collect when they encounter evidence se-
quentially, and are prone to prematurely terminating their evi-
dence collection process. He predicts that telling auditors to con-
strue or frame evidence at a high level (versus a low level) will in-
crease auditors’ professional skepticism, thereby improving the ef-
fectiveness of the auditors’ evidence collection and evaluation pro-
cess.
In his study, participants are provided with several opportuni-
ties to collect sets of mixed evidence related to a client’s com-
plex estimate. Each set of evidence includes the same proportion
of positive and negative items. Rasso manipulates whether partic-
ipants receive documentation instructions, and if so, what type of
instructions they receive. In the high-level construal condition, par-
ticipants are told to think “broadly about all of the evidence col-
lectively” and to list reasons why management’s estimate could be
presented fairly and why it could be misstated. In contrast, audi-
tors in the low-level construal condition are asked to think “specif-
ically about each evidence item” and to list reasons how manage-
ment’s estimate could be presented fairly and how it could be mis-
stated. Participants in a control condition receive no documenta-
tion instructions.
It is clear that Rasso’s experimental manipulation caused dif-
ferences in auditor participants’ responses. Conference participants’
comments tended to focus on potential reasons for, and implica-
tions of, these differences. We group our comments into four major
areas: (1) What else, besides construal level, might the documenta-
tion instruction manipulation be capturing? (2) Do the changes in
participants’ responses re?ect skepticism and only skepticism? (3)
Do the changes due to high-level construals represent enhanced
audit effectiveness? and (4) Could one apply this approach in prac-
tice to train novices to think more like experts?
What else, besides construal level, might the study’s manipulation
have captured?
Rasso’s central premise is that construing information at a high
level allows “for easier assimilation of information because the
representations become less ambiguous, more coherent, and more
schematic.” In contrast, adopting a low-level construal makes it
more likely that auditors will process information in a piecemeal
fashion without considering how previously received evidence re-
lates to new evidence. Consequently, auditors who interpret infor-
mation at a high level should be better able to see the “big picture”
or overall pattern suggested by the evidence.
There were several comments about exactly how high versus
low-level construals were manipulated, and whether the manipu-
lation might perhaps be too broad or too narrow. The documen-
tation instructions provided to the auditor participants ask them
to do two things. In the high-level construal condition auditors are
asked to both “think broadly about all the evidence collectively”
and about “why” management did something; whereas in the low-
level construal condition they are asked to both “think speci?cally
about each individual evidence item” and about “how” manage-
ment did something. The two parts of each manipulation could
actually represent two distinct constructs. For example, would it
be possible to also think broadly about “how” management’s es-
timate could be misstated? Perhaps, “how” questions capture the
mechanism by which the estimate could be misstated, whereas
“why” questions capture the motivation for misstating the esti-
mate. Listing “why” reasons could be tantamount to asking audi-
tors to consider the “perceived motivation” side of the fraud tri-
angle (SAS No. 99, American Institute of Certi?ed Public Accoun-
tants, 2002), which would naturally result in increased skepticism.
If this is true, auditors instructed to consider why management did
something may be more skeptical of management irrespective of
whether they are better able to assimilate different pieces of audit
evidence. The question is whether instructions to “think broadly”
would have led to increased skepticism if the “why” part of the
instructions had not been present.
It was also noted that in psychology research, priming instruc-
tions are intentionally set in a different context from that of the
primary task in order to avoid demand effects, and to measure
subconscious effects rather than intentional ones. Unlike priming
studies, Rasso’s documentation instructions ask auditors to con-
sciously process information in a particular way, and the instruc-
tions are related to his primary task. However, it is reassuring that
in a subsequent manipulation check, Rasso ?nds that participants
continue to interpret items unrelated to the primary task at the
same level of construal as in the primary task, suggesting that the
construal instructions do have carryover effects that might be sub-
conscious. This ?nding could reduce the concern that the results of
the primary study could be driven by demand effects.
Do the changes in participants’ responses re?ect skepticism and only
skepticism?
To examine participants’ level of skeptical action, Rasso mea-
sures the amount of time participants spent collecting and review-
ing evidence and the number of times they collected additional
sets of evidence. To measure skeptical judgment, he asks partici-
pants to assess the risk that the client’s estimate is materially mis-
stated and the likelihood that they would recommend an adjust-
ment to the client’s estimate.
Several conference participants called into question the study’s
premise that gathering more evidence and taking more time to
perform the task are necessarily evidence of increased skepticism.
It was also suggested that the study’s measures might actually be
capturing participants’ uncertainty about the task, but the speci?c
reasoning underlying this concern was not discussed further at the
conference. We next explain one reason why uncertainty, rather
than differences in construal level, could explain the differences in
auditors’ responses observed in Rasso’s study. It seems plausible
that there is greater uncertainty in the high-level construal condi-
tion because, based on the results of the control condition, Rasso
concludes that auditors naturally tend to process and evaluate in-
formation at a low-level of abstraction. If this is the case, auditors
in the high-level construal condition are being asked to process in-
formation differently from how they normally would on an actual
audit. This could explain why the study ?nds that participants in
the high-level construal condition perceive their task to be more
di?cult than in the low-level construal condition. It would also
explain why they would therefore need to spend more time on
the task than did auditors in the conditions that are more natu-
ral to them. If being forced to process information in an unfamiliar
manner increases participants’ uncertainty regarding the task, it is
possible that the differences observed with respect to “perceived
58 M.L. Frank, V.B. Hoffman/ Accounting, Organizations and Society 46 (2015) 56–58
risk” and “likelihood of adjustment” may re?ect differences in un-
certainty rather than, or in addition to, differences in skepticism.
Do the changes due to high-level construals represent enhanced audit
effectiveness?
Because in this study there was no “correct answer” and no
actual cost to gathering additional sets of evidence, conference
participants questioned what constituted the best, or most effec-
tive, response. For example, some conference participants won-
dered why anyone would choose an incomplete set of evidence
when one could obtain more evidence for free. Others wanted to
better understand the mechanism underlying the high-level con-
strual participants’ request for more information. Speci?cally, if au-
ditors in the high-level construal condition had a greater ability
to see connections and link the pieces of information, as hypothe-
sized, they might be expected to see the big picture more quickly,
and, would thus require fewer sets of evidence, rather than more.
Given the lack of a correct answer, any inferences about the rela-
tive effectiveness of the high-level versus low-level construal con-
ditions should be interpreted with caution.
Could one apply this approach in practice to train novices to think
more like experts?
If “thinking broadly” helps participants think about evidence
more abstractly, integrate it, and recognize patterns in the infor-
mation, as Rasso intends, high-level construal instructions could be
viewed as a technique that promotes thinking more like an expert.
Indeed, the examples of such reasoning that Rasso describes, such
as “chunking” information to remember it better, are classic exam-
ples from the expertise literature (Chase & Simon, 1973). However,
it is possible that these instructions would have different effects on
auditors with different levels of experience. For example, prior re-
search (Hoffman, Joe, & Moser, 2003) ?nds that more experienced
auditors naturally restructure and integrate components of work-
paper evidence, whereas less experienced auditors tend to process
evidence sequentially in the order they encounter it.
Therefore, while the intervention might help staff and senior
auditors to structure their task more like an experienced auditor
would naturally structure it, it is less clear what effect the in-
structions would have on more experienced auditors. The study
includes auditors with four levels of experience (staff, senior, man-
ager, and partner). However, because most participants are senior
auditors and there are only a few participants at each of the other
three levels, it is unclear whether all types of auditors react sim-
ilarly to the instructions. It might be the case that because man-
agers and partners naturally reason more broadly, they would not
respond as much to the high-level construal instructions.
Another issue to consider is whether there is value to hav-
ing some members of the audit team think about the details (i.e.,
low-level construal), while other members think more broadly (i.e.,
high-level construal). For example, Ramsay (1994) ?nds that while
managers and partners are better able than staff and seniors to de-
tect conceptual errors in the audit workpapers, they are less likely
to identify mechanical errors. If we were able to train less senior
auditors to think more broadly, it is possible that we might inad-
vertently sacri?ce what the novices do well – perhaps it is bene-
?cial to have someone think more speci?cally about the detailed
workpaper evidence.
Conclusions
Rasso’s study addresses an important audit issue that is of
great interest to regulators, practitioners, and audit researchers.
The study’s ?ndings suggest that there could be bene?ts to provid-
ing auditors with documentation instructions based on CLT. How-
ever, based on the various issues outlined above, some caution
should be exercised when interpreting the study’s results.
Audit practitioners and PCAOB representatives who attended
the conference were keen to understand how auditing standards
might be changed or how the results of this study could be imple-
mented in practice. Researchers were more cautious about imple-
menting the results of a particular study without further research.
If we could get more consensus from the body of recent research
applying construal level theory and the effect of changing the au-
ditor’s mindset on audits of complex estimates, ultimately the goal
would be to implement the replicable research ?ndings into audit
practice.
Acknowledgement
We thank Deloitte and Touche, LLP for generously funding this
conference. We are also grateful to Hun Tong Tan (editor) and the
conference participants for their helpful comments.
References
American Institute of Certi?ed Public Accountants (2002). Consideration of fraud in
a ?nancial statement audit. Statement of Auditing Standards No. 99. New York,
NY: AICPA.
Backof, A.G., Bamber, E.M., & Carpenter, T.D. (2014). Auditing complex estimates:
Management-provided evidence and auditors’ consideration of inconsistent evi-
dence. Working paper, University of Virginia.
Bratten, B., Gaynor, L. M., McDaniel, L., Montague, N. R., & Sierra, G. E. (2013). The
audit of fair values and other estimates: The effects of underlying environmen-
tal, task and auditor-speci?c factors. Auditing: A Journal of Practice & Theory,
32(Supplement 1), 7–44.
Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive Psychology, 4,
55–81.
Gri?th, E. E., Hammersley, J. S., & Kadous, K. (2014). Audits of complex estimates as
veri?cation of management numbers: How institutional pressures shape prac-
tice. Forthcoming, Contemporary Accounting Research.
Gri?th, E. E., Hammersley, J. S., Kadous, K., & Young, D. (2015). Auditor mindsets
and audits of complex estimates. Journal of Accounting Research, 53(1), 49–77.
Hoffman, V. B., Joe, J. R., & Moser, D. V. (2003). The effect of constrained processing
on auditors’ judgments. Accounting, Organizations, and Society, 28(7–8), 699–714.
PCAOB (2012). Maintaining and applying professional skepticism in audits. PCAOB
Staff Audit Practice Alert No. 10. December 4, 2012. Washington, D.C: PCAOB.
Plumlee, D., Rixom, B., & Rosman, A. (2015). Training auditors to perform analytical
procedures using metacognitive skills. The Accounting Review, 90(1), 351–369.
Ramsay, R. J. (1994). Senior/manager differences in audit workpaper review perfor-
mance. Journal of Accounting Research, 32(1), 127–135.
Rasso, J. (2015). Construal instructions and professional skepticism in evaluating
complex estimates. Accounting, Organizations, and Society, in press.http://dx.doi.
org/10.1016/j.aos.2015.03.003.
Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance.
Psychological Review, 117(2), 440–463.
Trope, Y., & Liberman, N. (2011). Construal level theory. In P. Van Lange,
A. W. Kruglanski, & E. T. Higgins (Eds.), Handbook of theories of social psychology.
London: Sage Publications.
doc_670257201.pdf