Académique Documents
Professionnel Documents
Culture Documents
69–94
Does
Judgment?
Robyn Moroney
SUMMARY: Prior studies have found that industry specialists are more effective when working in specialization. It
has been argued that they should also be more efficient, but this has not been tested to date. The purpose of this
study is to determine whether industry specialists are more efficient at each stage of the decision-making process
when working in specialization. By breaking down and measuring efficiency at each stage of the decision-making
process, the opportunity is afforded to measure and compare the relative efficiencies of two groups of specialists in
very different industries, the manufacturing, and superannuation (pension fund) industries. A controlled experi-
ment was conducted in the offices of each of the Big 4 accounting firms. The two groups of specialists are found to
vary in the decision processes they use, bringing into question the assumption that all specialists behave the same
when working in specialization. It is the differences that are found between the industry groups that have
interesting implications for future research.
INTRODUCTION
I ndustry specialization has been used as a means of distinguishing between the largest
accounting firms (Francis et al. 1999; Hogan and Jeter 1999). It allows firms to improve efficiency, create barriers to
entry, and improve audit quality (Solomon et al. 1999). Prior studies have found that industry specialists are more
effective when working in spe- cialization. They are more knowledgeable about their industry and this allows them to
be more effective when working in specialization. It has been assumed that industry specialists are also more efficient
when working in specialization (for example, Wright and Wright 1997; Solomon et al. 1999; Low 2004). The
purpose of this study is to test whether industry specialists are in fact more efficient when working in specialization.
This is an important issue as it highlights and tests one of the benefits expected to accrue to specialists.
The current analysis breaks down the decision-making process into three stages to determine whether or not
industry specialists are more efficient at each stage of that process.
O’Donnell, Robert Knechel, Arnold Wright, Richard Tubbs, Axel Schulz, George Tanewski, Carlin Dowling and Paul Coram, participants at
research workshops at the University of New South Wales, the University of Melbourne, Monash University, the Australian National University,
time.
Submitted: February 2006 Accepted: May 2007
69
70 Moroney
Breaking down the decision-making process allows a more detailed analysis of the varied processes used by different
groups of industry specialists. Indeed, by breaking down and examining each stage of the decision-making process, a
more complete picture of the relative benefits of specialization is provided.
Two very different groups of industry specialists were chosen for this study. This is consistent with prior
studies in the area of industry specialization that stress the importance of using two groups of specialists whose
knowledge base is very different from each other (for example, Solomon et al. 1999; Owhoso et al. 2002). In the
current study specialists in the manufacturing and superannuation (pension fund) industries are used. Manufacturing
specialists develop knowledge about various topics such as accounting for inventories, re- search and development,
the impact of technology on inventory pricing as well as the various accounting rules that apply to companies in that
industry. Superannuation specialists develop knowledge about the detailed legislation that governs that industry, the
various accounting regulations that apply to superannuation funds, and the methods of auditing the accounts peculiar
to clients in that industry. Both groups of specialists acquire very different sub-speciality knowledge (Bonner and
Lewis 1990) over and above their general domain knowledge, to excel as industry specialist auditors.
While it has been shown in prior studies that two such different groups of specialists should excel on tasks
set within their area of specialization, their relative efficiency gains have not been tested until now. According to the
literature, all groups of industry specialists are expected to benefit from their specialization in much the same way.
Their detailed industry sub-speciality knowledge allows them to have more accurate nonerror frequency knowledge
(Solomon et al. 1999), to be more effective when assessing inherent risk levels for accounts specific to the industry in
which they specialize (Taylor 2000), to be more effective when detecting errors for clients in their industry (Owhoso
et al. 2002), to develop task knowledge differently depending on their area of specialization (Thibodeau 2003), to
assess audit risk more accurately (Low 2004), and to interpret and complete partial cue patterns (Hammersley 2006).
An important purpose of this study is to determine whether the efficiency gains due to specialization also accrue to
specialists at each stage of the decision-making process.
An experiment was conducted in which manufacturing and superannuation industry specialist auditors from
all Big 4 accounting firms completed two industry-based cases. The cases were developed with the aid of industry
specialist Big 4 audit partners (the expert panel), who also selected the industry groups used in this study and
identified participants as being specialists in either of these industries.
The results of this research are of interest to both researchers and practitioners. They highlight the different
aspects of the decision-making process (pre-information search, in- formation search, and decision processing) where
industry specialists can be expected to be more efficient. While both groups are found to perform better when
working in spe- cialization, which is consistent with prior literature, the two groups of specialists are found to vary in
the decision processes they use, with the superannuation industry specialists being more efficient at each stage of the
decision-making process, while the manufacturing industry specialists are not. The manufacturing specialists are
found to be significantly more efficient when reading case materials and when reading information cues but not when
processing their decisions. Inferences are drawn about why these differences between the groups of industry
specialists occur. It is these differences that have interesting implications for future research. Another interesting
finding is that there is a significant relationship between the efficiency variables and performance but not in the
direction expected. All
participants, both when working in and out of specialization, tended to perform better when they were less efficient
(i.e., taking more time, steps, and so on).
The remainder of this paper is set out as follows. The next section contains background literature and
hypotheses in the areas of pre-information search, information search, and decision processing. The research design
follows. The experiment is then described fol- lowed by the results and supplementary analysis. Finally, limitations
are summarized and final conclusions are outlined.
Hypotheses
The hypotheses outlined in this paper are based upon the cognitive perspective of the expertise paradigm.
Figure 1 shows how the various elements of the decision-making proc- ess interact. The experiment tests and
compares efficiency at the pre-information search, information search, and decision-processing stages (the arrows in
Figure 1) of the decision- making process, as well as measuring the performance of auditors when working in and out
of specialization.
Pre-Information Search
Bonner and Pennington (1991) looked at the decision-making process used by auditors, who are found to
retrieve knowledge from memory, do an external information search, comprehend, generate hypotheses, evaluate
these hypotheses, and finally make a choice. Before searching for relevant information, either internally, from
memory, or externally, the first stage in decision-making involves gaining an understanding of the problem at hand.
The pre-information search stage of the decision-making process is the decision-maker’s attempt to understand and
interpret the problem at hand (Christ 1993). Those with a better knowledge of the area in which a problem is set are
better at interpreting the problem and placing the problem information provided into a context (Libby 1981; Choo
1989; Biggs et al. 1988; Christ 1993). Biggs et al. (1988) reported that experts have an internal schema, which allow
them to identify, from their memory, the type of problem confronting them. This knowledge allows an expert to be
more efficient when faced with a problem in a setting familiar to them.
Industry specialization is a type of expertise. Auditors with industry-based knowledge are better at
understanding problems set within the industry they know about. As such, they spend less time gaining an
understanding of a task (Biggs et al. 1988). Experts are superior in their problem representations, utilizing a strategy
of converting a situation into a problem for which a solution can be provided, while novices lack domain-specific
strategies (Choo 1989), which results in poorer efficiencies when problem solving.
Industry specialists are expected to be more efficient during the pre-information search phase of the
decision-making process when working in specialization. Their superior knowl- edge of the industry in which they
specialize allows them to more efficiently attain a comprehensive understanding of the problem at hand when
working in specialization. As the decision-making process is sequential and iterative, efficiency is measured as time
spent reading (and re-reading) case material and the number of times a case is re-read until a complete understanding
of the problem at hand is achieved (Biggs et al. 1988; Choo 1989; Hershey et al. 1990; Salterio 1994). This is tested
by the following hypothesis:
H1: Industry specialist auditors are expected to be more efficient when working in specialization than when
working out of specialization during the pre- information search stage of the decision-making process.
Information Search
When faced with a problem, an auditor will seek information to help understand the relevant issues and
indicate a solution (Bonner and Pennington 1991). The two types of information that can be sought are internal and
external to the individual auditor. Internal information is knowledge held in memory and external information is the
product of client discussions, work papers, databases, and industry data (Koonce 1993).
When compared to novices, experts have in memory more of the knowledge necessary to solve a problem set
within their area of expertise (for example, Davis and Solomon 1989; Koedinger and Anderson 1990; Bonner and
Pennington 1991; Choo and Trotman 1991;
Solomon et al. 1999; Rikers et al. 2002). Industry specialists have in their memory more industry related information
to help them solve problems set in their industry, making them more efficient when searching for and reading
information (Solomon et al. 1999). As such, industry specialists are less likely to access information familiar to them,
such as industry- relevant accounting standards and legislation that they would be familiar with when auditing clients
within their area of specialization.
Industry specialist auditors are expected to be more efficient during the information- search stage of the
decision-making process. Efficiency is measured as time taken searching for and reading information and the number
of information cues selected and read (Be ́dard and Mock 1992; Salterio 1994). This is tested by the following
hypothesis:
H2: Industry specialist auditors are expected to be more efficient when working in specialization than when
working out of specialization during the infor- mation search stage of the decision-making process.
Decision Processing
Experts (including industry specialists) have been found to use a different decision process than novices use.
They tend to be more efficient—using fewer steps, are more goal oriented, and use a forward driven schema (Elio
and Scharf 1990; Bonner and Pennington 1991; Be ́dard and Mock 1992; Be ́dard and Chi 1993; Solomon et al. 1999;
Rikers et al. 2002). Be ́dard and Mock (1992) and Be ́dard and Chi (1993) reported that experts were found to be
more efficient at processing problems set in their area of expertise. Industry specialists are expert auditors, who are
also expected to be more efficient when working in specialization (Solomon et al. 1999).
Industry specialist auditors are expected to use a more efficient decision process when working in
specialization. From the discussion in this section, efficiency is measured as number of steps taken, amount of time
taken throughout each case, and number of recur- sions in solving each case (Bonner and Pennington 1991; Be ́dard
and Mock 1992; Salterio 1994). Recursions refer to going back and re-reading case material and/or information cues.
This is tested by the following hypothesis:
H3: Industry specialist auditors are expected to employ a more efficient deci- sion process when working in
specialization than when working out of specialization.
H4: A greater positive relationship is expected between efficiency at each stage of the decision-making
process and performance for industry specialists when working in specialization than when working out of
specialization.
RESEARCH DESIGN A 2 2 design is used for testing the hypotheses. The first
independent variable is auditor industry specialization (manufacturing or superannuation). The second independent
variable is the client (case) industry setting. Each participant completed two cases, one set in the manufacturing
industry and the other set in the superannuation industry.
The manufacturing case (one page long) deals with research and development expen- diture. The task
comprises some information regarding a client and five questions. The questions include how to audit research and
development expenditure, which items from a list provided in the case material could be capitalized, and how a
government grant should be handled. To correctly answer the problem set in the case, participants require knowledge
of the manufacturing industry and various accounting standards. Excerpts from these stan- dards were provided as
information cues.
The superannuation case (two pages long) deals with the identification of audit pro- cedures necessary to
confirm listed audit assertions and compliance with legislation. An audit report and a list of investments to be audited
are included in the case material. To correctly answer the problem set in the case, participants require knowledge of
the regu- lations impacting the reporting of results in financial statements. The information cues provided to
participants included excerpts from industry specific accounting standards and regulations. These regulations detail
the auditing and disclosure rules particular to super- annuation funds.
Participants
Eighty-six participants from each of the Big 4 accounting firms took part in the ex- periment. The expert panel
selected the industry groups that would be made available, invited participants to take part, and indicated which
industry each specializes in. To be consistent with prior studies (for example, Solomon et al. 1999) and the theory
outlined above regarding expertise, participants with less than two years industry audit experience were excluded
from the current study.1 Details of the participants are provided in Table 1. While the superannuation industry
specialists had significantly more audit experience than the manufacturing industry specialists (7.4 compared to 5
years experience, t 2.378, p .020, not tabulated), there was not a significant difference in their years auditing in their
specialist industry, with the superannuation specialists having an average of 5.2 years industry audit experience
compared to 4.6 years for the manufacturing specialists (t .782, p .437, not tabulated).2 The data in Table 1 also show
that there are differences in the task experience reported by participants.
Task Development
The expertise paradigm (Einhorn 1976; Be ́dard 1989; Libby and Luft 1993; Libby 1995; Solomon et al.
1999) was utilized in testing the decision-making processes and
1
The firms provided 97 participants. The excluded participants came from both the manufacturing (7) and su- perannuation (4) industries, resulting
included. All significant and nonsignificant results remained as reported. 2 To test the impact of experience on the results reported in this
change in the variables identified as significant. Thus, the inclusion of participants with less than four years of industry- based
experience was not a
determining influence on the reported results.
specialize in Minimum
2 yrs Maximum 16 yrs Minimum
2 yrs Maximum 27 yrs Proportion
of participants from each
industry setting with personal experience advising clients on issue used in the manufacturing case (research and
development)
46% 7%
Proportion of participants from each industry setting with personal experience advising clients on issue used in the
The firms provided 8 (4), 13 (11), 19 (13), and 8 (10) manufacturing (superannuation) industry specialists,
a
respectively.
performance of industry specialist auditors. This paradigm stresses the importance of spec- ifying the
knowledge necessary to complete a specified task and that the task must be developed so that the
implications of using (or not using) that knowledge are observable (Libby and Luft 1993).
The cases were developed with the help of the expert panel (the manufacturing and superannuation cases)
and corporate executives (the practice case). The expert panel com- prised industry specialist audit partners
from each of the Big 4. Each firm provided access to one partner from each industry setting. The cases were
initially developed with six partners from three of the Big 4. The cases and solutions were written by some of
these partners and then sent around to the other industry specialist partners and back to the initial partners
until all agreed on the wording and the content. The information cues were selected and agreed upon by all
relevant (industry specialist) partners. Prior to conducting the ex- periment, partners from the fourth firm
confirmed that the cases, information cues, and solutions reflect the types of issues their industry specialist
auditors would confront when working in specialization. All partners confirmed that the case set in their
industry tests industry knowledge and deals with an issue that their staff may encounter when auditing clients
in that industry, while the case set in the other industry would be difficult for their industry specialist staff to
tackle.
According to the expert panel members, the cases are realistic, based upon real prob- lems faced by industry
specialist auditors (Be ́dard and Biggs 1991a, 1991b; Gibbins and Jamal 1993; Trotman 1996). They are
sufficiently ill defined (Abdolmohammadi and Wright 1987; Be ́dard and Biggs 1991a, 1991b; Trotman
1996) in order to test each area of industry specialization (Be ́dard and Chi 1993). Had the cases been too
simple and structured, the auditors would have been expected to perform equally well on both cases (Be ́dard
and Biggs 1991b; Bonner 1990). According to the expertise paradigm, had the cases been too
76 Moroney
complex, the industry specialists may not have been able to perform better on the tasks set in their area of
specialization (Hershey et al. 1990; Spence and Brucks 1997).3
The information cues used in this study contain the type of information that industry specialist auditors
should be aware of (Solomon et al. 1999), such as, excerpts from ac- counting standards and legislation. If the
information cues had included items that industry specialists are not familiar with, they may have selected the items
and spent time reading them. By using cues that the specialists would use when working on problems such as those
contained in the cases, their efficiency during the information search stage of the decision-making process can be
measured.
3
The Internet-based computer program developed for this experiment was pilot tested by five academics, three Big
4 partners, and three junior
members of staff from two Big 4 firms. The purpose of the pilot test was to determine whether the instructions and case materials were understood
and that the Internet program captured the data appropriately, reporting time spent, information cues read, the processes used, and the solutions to
each case were properly saved. As a result of this testing, the questions asked in each case were reproduced at the top
of each solution input screen
to avoid participants having to go back and re-read the case materials when moving on to answer the next part of each question. For clarity, the exit
questions were also re-worded slightly. 4 Participants were not told the reason for this instruction. 5 A test was conducted to measure any possible
order effects. There was no significant difference for time reading the
cases (t 0.489; p .625), re-reading case materials (t 1.141; p .255), time
reading information cues (t 1.253; p .212), number of cues (t 1.209; p .228), number of steps (t 0.725; p .469), time overall
(t 1.678; p .105), and
number of recursions (t 0.732; p .465).
Cue 1
Cue 2
Cue 5
to determine the nature of the information provided. According to the expert panel, those with industry knowledge
were familiar with the information provided and would only have accessed information when working in
specialization to jog their memory. When they fin- ished reading an information cue, they were returned to the
information cue menu where they could select another cue to read, go back and re-read the case, or go on to provide a
solution to the problem. When providing a solution, participants were advised that they could go back and re-read the
case, go to the information menu, or submit their completed solution. The arrows in Figure 2 show the path a
participant could follow when completing a case.Data were captured for each participant as they completed the
questions asked in the cases, and responses to the exit questions used in compiling Table 1.THE RESULTS
The results for the first three hypotheses are presented in Tables 2–4. The tables include the results of a
multivariate analysis of variance (MANOVA) using Wilk’s Lambda for overall efficiency at each stage of the
decision-making process. The results for each de- pendent variable are then analyzed using univariate analysis of
variance (ANOVA). Both within-subject and between-subject simple effects are provided. This allows a
within-subject
6
Two experienced researchers coded the cases to measure performance using the expert panel’s model solutions. Cohen’s
Kappa (1960) was used
to measure inter-rater agreement. The Kappa for the manufacturing (superan- nuation) case was 0.9466 (0.9007). Both are significant at p .001. 7
The click information shows the order and time spent on each aspect of the experiment (reading/re-reading case materials, accessing information
cues, and writing a solution). The click data were transferred to a spreadsheet
and independently checked.
analysis, in line with the hypotheses being tested and a between-subject analysis where each case is held constant.
For each significant interaction between the independent vari- ables, an interaction plot is provided. Descriptive
statistics including the mean, standard deviation, and range for participants in each cell are also provided. Interaction
plots are provided where the results suggest that participants were more efficient when working in specialization.
doing so as indicated above.The results offer some support for the first hypothesis, with the superannuation spe-
cialists behaving as expected for each measure of efficiency at the pre-information search stage of the
decision-making process. The manufacturing specialists are more efficient when measured as time reading cases but
not when measured as number of times case materials are re-read.
Auditor Industry Specialization Manufacturing Manufacturing Superannuation (n 48) (n 48) (n 38) Manufacturing
5.34 (2.29) [1.92–12.02]
[0–14]
[1–10] 5.85
t 0.781 (.219)* 3.13 2.74 t 0.670 (.253)* (3.02) (2.69) (2.65) [1.73–13.52]
[0–12]
(3.28) [0–11.13]
t 1.010 (.158)*
[0–8]
[0–8] 3.36
t 1.338 (.093)* 2.50 1.47 t 1.838 (.035)* (6.11) (2.81) (2.23) [0–29.83]
[0–8]
3.50
0.50
3.00
0.00
Manufacturing Case Superannuation Case
2.50
*, ** one- and two-tailed, respectively.
s euCf or ebmu
N2.00 1.50
84 Moroney
Superannuation specialists also spent less time when working on the superannuation case than when working on the
manufacturing case (t 1.475, p .1, one-tailed test). A between-subject comparison shows that the manufacturing
specialists spent less time than the superannuation specialists (t 1.893, p .04, one-tailed test) when working on the
manufacturing case and the superannuation specialists spent less time than the manufac- turing specialists did, (t
1.338, p .10, one-tailed test) when working on the superan- nuation case.
For number of cues selected and read, Panel A of Table 3 shows a significant interaction between AIS a nd CI
(F(1,84) 11.321, p .002). The plot in Panel D shows that the manufacturing specialists selected the same number of
cues for both cases, while the su- perannuation specialists selected fewer when working in specialization. Simple
effects are carried out on the difference in number of cues. As expected, the superannuation spe- cialists selected
fewer cues when working on the superannuation case than when working on the manufacturing case (t 2.609, p .006,
one-tailed test). A between subject comparison shows that the superannuation specialists selected fewer cues than the
manu- facturing specialists did, (t 1.838, p .04, one-tailed test) when working on the super- annuation case. The
manufacturing specialists selected fewer cues than the superannuation specialists when working on the
manufacturing case; however, the difference is not signif- icant (p .15).
The results offer some support for the second hypothesis. The results are significant for time spent searching
for and reading information cues. While directionally correct, the results for number of cues are only significant for
the superannuation case. Also, only the superannuation specialists select significantly fewer cues when working in
specialization than when working out of specialization.
(16.14) [2–69]
Steps
Time Auditor Industry Specialization
Auditor Industry Specialization Manufacturing Superannuation Manufacturing Superannuation (n 48) (n 38)
(n 48) (n 38)
t 0.047 (.482)*
23.5 t 0.388 29.64 29.52 t 0.045 6.23 6.18 (18.15) (.350)* (10.43) (13.79) (.482)* (4.52) (4.24) [4–63]
[10.92–51.83]
[7.32–65.93]
[0–20]
[1–16]
t 1.203 (.117)* t 1.166 (.123)* t 3.217 (.001)* t 2.297 (.012)* 12.53
t 2.025 24.31 23.03 t 0.444 4.38 3.47 (10.62) (.023)*
[4.13–61.58]
[5.28–64.00]
[0–17]
8
The factor analysis was conducted for the entire data set and then for the auditors working in specialization and
the auditors working out of specialization. Each time the same groups of variables loaded on two factors.
6.00
e roc
S3.00
Manufacturing Specialists
Superannuation Specialists
5.00
Manufacturing Case Superannuation Case
1.00
0.00
Does Industry Expertise Improve the Efficiency of Audit Judgment? 8
9
TABLE 6 Impact of Efficiency Variables on Performance (H4)
Panel A: Regression [t-Statistic (p-Value)*] for All Participants When Working in and out of Specialization
Factor
Auditing: A Journal of Practice & Theory, November 2007 Auditors working in specialization (expected sign ) (n 86)
Auditors working out of specialization (expected sign ) (n 86) Intercept 26.776 (.000) 18.447 (.000) Cases and Time
Panel B: Regression [t-Statistic (p-Value)*] for Each Group of Specialists When Working in and out of
Specialization
Factor
Recursions 5.528
(.000) 4.499
(.000) 2.058
(.047) 5.919
(.001)
* One-tailed.
working in specialization are not significant.9 As the results are consistently positive, with all those who are
less efficient being more effective, the fourth hypothesis is rejected as the results are consistent when
auditors work in specialization and out. Thus, the fourth hypothesis is rejected.
SUPPLEMENTARY ANALYSIS Task Experience
As reported in Table 1, 46 percent of manufacturing specialists and 7 percent of su- perannuation specialists
reported having had experience advising clients on the issue used in the manufacturing case. Also 97 percent
of superannuation specialists and 14 percent of manufacturing specialists reported having had some
experience with the issue contained in the superannuation case. Task specific experience was run as a
covariate for each of the
9
Another way to deal with the high correlation between the efficiency variables is to run the regression for the decision processing
variables only (steps, time, and recursions). When this is done, time overall is found to be positively related to performance for both
those working in specialization and those working out of specialization (p .02, not tabulated). This finding supports the results reported in
efficiency variables and the effectiveness (performance) variable. In each case, task specific experience was found to
be insignificant.
While it is of interest to determine whether task experience impacts the results pre- sented, this variable is
very different to industry specialization. The task experience variable used in the current study does not indicate
frequency or recency of experience. It is used to highlight the fact that, while specialist auditors may encounter a
variety of tasks when auditing clients in the one industry (such as when auditing clients in the manufacturing
industry), when they encounter an unfamiliar task set within the industry that they specialize in, their sub-speciality
knowledge allows them to perform well, even if they are less efficient when working on such a task.
To demonstrate this, a performance comparison is made between the two groups of industry specialists all of
whom reported having no task experience on the manufacturing case (26 manufacturing specialists and 35
superannuation specialists). The manufacturing specialists achieved an average score of 4.83 compared to 2.69 for
the superannuation specialists (t 4.735; p .001, not tabulated). Thus, it is clear that industry specialization provides a
more powerful explanation of superior performance than task experience. It indicates that industry specialist auditors
working on a variety of tasks when working in specialization will outperform nonspecialists even when the
specialists do not have expe- rience with the task at hand.
Information Cues
A number of participants selected no information cues.10 The cues included information such as excerpts
from accounting and auditing standards that are publicly available. Each participant could select up to eight
information cues that were listed in alphabetical order. The cues do not add any new information to the cases and
reading them would only have helped participants not familiar with relevant standards and legislation. A test was
conducted to determine whether those that read more information cues were able to perform better. The result was
found to be significant (F 4.072, p .000, not tabulated). Thus the information cues identified by the expert panel did
help those participants that chose to use them. The information search and decision processing hypotheses were
retested, along with performance, excluding the participants that did not read any information cues. There was no
significant change in the results. Thus, the participants that did not read any information cues did not distort the
results presented here.
At the other extreme, some participants spent over fifteen minutes searching for and reading information
cues (see Table 3). Four manufacturing specialists spent more than fifteen minutes at this stage of the
decision-making process when working on the super- annuation case. Four (two) superannuation specialists spent
more than fifteen minutes at this stage of the decision-making process when working on the manufacturing (superan-
nuation) case. The data for these outliers were excluded from the data set and the analysis for each hypothesis was
retested. The overall results did not change.
The information cues for the manufacturing case (6,794 words in total) were shorter than those for the
superannuation case (11,037 words in total). A comparison of the time taken by all participants when selecting and
reading the information cues shows that they spent an average of 4.36 minutes when reading the manufacturing cues
and an average 4.32 minutes when reading the superannuation cues. The difference is not significant (t .003;
10
Of the 86 participants, 35 read no cues for either case (19 manufacturing specialists and 16 superannuation
specialists).
p .479, one-tailed, not tabulated). Thus the length of the information cues did not distort the results presented here.
SUMMARY AND LIMITATIONS Significant results were found for time spent
reading case materials, time spent search- ing for and reading information cues, and performance. Participants re-read
the manufac- turing case more often (even though the manufacturing case is shorter than the superan- nuation case),
they read more information cues for that case and they took more steps, time, and recursions when completing that
case. Participants as a group also performed better on the manufacturing case.
The case effect may be explained by differences in the number of requirements for each case. While every
attempt was made to keep the cases as similar as possible with regard to degree of difficulty for specialists, number
of information cues, and length of case, there was some variation. The superannuation case was longer than the
manufacturing case. It included attachments to be read and lists of numbers. The question asked was multifaceted. In
both situations, the questions asked were those deemed important by the expert panel.
The manufacturing case was shorter but required a number of steps (5) to answer. A comparison of the time
taken by all participants when reading the case materials shows that they spent an average of 5.82 minutes when
reading the manufacturing case and an average 6.11 minutes when reading the superannuation case. The difference is
not signifi- cant (t .504; p .239, one-tailed, not tabulated). The question(s) for each case were reproduced at the top of
the answer input screen so that participants could glance up and remind themselves of the question or of the next part
of the question, without having to switch back and re-read the case. In spite of this modification made to the program
as a result of pilot testing, participants re-read the case materials more often when completing the manufacturing
case.
Another explanation for the case effect is differences between the groups of specialists. The superannuation
specialists behaved as expected throughout. They were more efficient and more effective when working in
specialization. Only one superannuation specialist had not had experience with the task used in the superannuation
case. On the other hand, only 22 of the 48 manufacturing specialists reported having experience with the task used.
While every care was taken to find tasks that would be familiar to all specialists when working in specialization, as
well as those unique to each industry, that was very difficult for the manufacturing industry.
Interviews were held with industry specialist partners from each of the participating firms after the
experiment was run and the results analyzed. It was reported to them that while the superannuation specialists
behaved as expected, the manufacturing specialists were not always more efficient when working in specialization.
The partners expressed their view that this result is understandable and reflects differences between the two
industries. While superannuation specialists encounter many of the same tasks when working in spe- cialization,
enabling them to build up significant experience, which translates into enhanced efficiencies, the same cannot be said
of manufacturing specialists. These auditors face more variety at their clients, which hampers the development of
improved efficiencies when working in specialization.
CONCLUSION The purpose of the experiment described in this paper was to understand more about how each
stage of the decision-making process of industry specialist auditors varies when
working in and out of specialization. Behavioral decision theory was used to develop test- able hypotheses for the
pre-information search, information search, and decision processing stages of the decision-making process.
Industry-based cases were developed using the ex- pertise paradigm as a guide to the required structure of such tasks.
The results presented in this paper were significant for time taken reading case material, amount of time
taken to search for and read information cues, and for decision quality. The superannuation specialists behaved as
expected throughout, while there were mixed results for the manufacturing specialists. Interviews with industry
specialist partners from each of the participating firms indicate that this result may be explained in part by differences
in the industries being audited. The superannuation industry is highly regulated and auditors encounter many of the
same issues when moving from one client to the next. The manu- facturing industry is more varied. While
industry-specific knowledge is required when au- diting clients in this industry, which was captured by the case used
in this study, there is less scope for manufacturing specialists to build up enhanced efficiencies when working in
specialization. The results reported here have implications for future research in the area of industry specialization.
REFERENCES Abdolmohammadi, M. J., and A. Wright. 1987. An examination of the effects of experience and task
complexity on audit judgments. The Accounting Review 62: 1–13. Ashton, A. H. 1991. Experience and error frequency
knowledge as potential determinants of audit
expertise. The Accounting Review 66: 218–239. Be ́dard, J. 1989. Expertise in auditing: Myth or reality? Accounting,
Organizations and Society 14:
113–131. ———, and T. J. Mock. 1992. Expert and novice problem-solving behavior in audit planning. Au-
diting: A Journal of Practice & Theory 11: 1–20. ———, and M. T. H. Chi. 1993. Expertise in auditing. Auditing: A
Journal of Practice & Theory 12
(Supplement): 21–45. Bedard, J., and S. Biggs. 1991a. The effect of domain-specific experience on evaluation of
manage- ment representations in analytical procedures. Auditing: A Journal of Practice & Theory 10 (Supplement): 77–90.
———, and ———. 1991b. Pattern recognition, hypotheses generation and auditor performance in
an analytical task. The Accounting Review 66: 622–642. Biggs, S., T. Mock, and P. Watkins. 1988. Auditor’s use of
analytical review in audit program design.
XIII: 148–161. Bonner, S. E. 1990. Experience effects in auditing: The role of task-specific
The Accounting Review L
knowledge. The Account-
ing Review 65: 72–92. ———, and B. L. Lewis. 1990. Determinants of auditor expertise. Journal of Accounting
Research
28 (Supplement): 1–19. ———, and N. Pennington. 1991. Cognitive processes and knowledge as determinants of
auditor
expertise. Journal of Accounting Literature 10: 1–50. Choo, F. 1989. Expert-novice differences in judgment/decision
making research. Journal of Account-
ing Literature 8: 106–136. ———, and K. T. Trotman. 1991. The relationship between knowledge structure and
judgments for
experienced and inexperienced auditors. The Accounting Review 6 6: 464–485. Christ, M. Y. 1993. Evidence on the nature of audit
problem representations: An examination of
auditor free recalls. The Accounting Review 6 8: 304–322. Cohen, J. 1960. A coefficient of agreement for nominal
scales. Educational and Psychological Mea-
surement 20: 37–46.
Davis, J. S., and I. Solomon. 1989. Experience, expertise and expert-performance research in public
accounting. Journal of Accounting Literature 8 : 150–164. De Beedle, I. 1997. An exploratory investigation of industry
specialization of large audit firms. The
International Journal of Accounting 3 2: 337–355. Einhorn, H. J. 1976. A synthesis: Accounting and behavioral science.
Journal of Accounting Research
14 (Supplement): 196–206. Elio, R., and P. B. Scharf. 1990. Modeling novice-to-expert shifts in problem-solving
strategy and
knowledge organization. Cognitive Science 14: 579–639. Entwistle, G., and D. Lindsay. 1994. An archival study of the
existence, cause, and discovery of income-affecting financial statement misstatements. Contemporary Accounting Research 1 1:
271–296. Francis, J. R., D. J. Stokes, and D. Anderson. 1999. City markets as a unit of analysis in audit research
and the re-examination of Big 6 market shares. Abacus 3 5: 185–206. Gibbins, M., and K. Jamal. 1993.
Problem-Centered research and knowledge-based theory in the
professional accounting setting. Accounting, Organizations and Society 18: 451–466. Gramling, A. A., and D. N. Stone. 2001.
Audit firm industry expertise: A review and synthesis of the
archival literature. Journal of Accounting Literature 20: 1–29. H
ammersley, J. S. 2006. Pattern identification and
industry-specialist auditors. The Accounting Review
81: 309–336. Hershey, D. A., D. A. Walsh, S. J. Read, and A. S. Chulef. 1990. The effects of expertise on financial
problem solving: Evidence for goal-directed, problem-solving scripts. Organizational Behavior and Human Decision Processes
46: 77–101. Hogan, C. E., and D. C. Jeter. 1999. Industry specialization by auditors. Auditing: A Journal of
Practice & Theory 1 8: 1–17. Koedinger, K. R., and J. R. Anderson. 1990. Abstract planning and perceptual chunks:
Elements of
expertise in geometry. Cognitive Science 1 4: 511–550. Koonce, L. 1993. A cognitive characterization of audit review.
Auditing: A Journal of Practice &
Theory 1 2: 57–76. Libby, R 1981. Accounting and Human Information Processing: Theory and Applications. Upper
Saddle River, NJ: Prentice Hall. ———, and J. Luft. 1993. Determinants of judgment performance in accounting
settings: Ability, knowledge, motivation, and environment. Accounting, Organizations and Society 1 8: 425–450. ———, 1995.
The role of knowledge and memory in audit judgment. In Judgment and Decision Making Research in Accounting and Auditing,
edited by R. H. Ashton, and A. H. Ashton, 176– 206. Cambridge, U.K.: Cambridge University Press. Low, K. Y. 2004. The effects
of industry specialization on audit risk assessments and audit-planning
decisions. The Accounting Review 79: 201–219. Owhoso, V. E., W. F. Messier, Jr., and J. G. Lynch, Jr. 2002. Error
detection by industry-specialized
teams during sequential audit review. Journal of Accounting Research 4 0: 883–900. Rikers, R., H. Schmidt, H.
Boshuizen, G. Linssen, G. Wesseling, and F. Paas. 2002. The robustness of medical expertise: Clinical case processing by medical
experts and subexperts. American Journal of Psychology 1 15: 609–629. Salterio, S. 1994. Researching for accounting precedents:
Learning, efficiency and effectiveness. Con-
temporary Accounting Research 1 1: 515–542. Solomon, I., M. D. Shields, and O. R. Whittington. 1999. What do
industry-specialist auditors know?
Journal of Accounting Research 37: 191–208. Spence, M. T., and M. Brucks. 1997. The moderating effects of problem
characteristics on experts’
and novices’ judgments. Journal of Marketing Research 3 4: 233–247. Taylor, M. H. 2000. The effects of industry
specialization on auditors’ inherent risk assessments and
confidence judgments. Contemporary Accounting Research 1 7: 693–712. Thibodeau, J. C. 2003. The development and
transferability of task knowledge. Auditing: A Journal
of Practice & Theory 22: 47–68.
Trotman, K. T. 1996. Research Methods for Judgment and Decision Making Studies in Auditing.
Melbourne, AU: Coopers and Lybrand. ———, 2001. Design issues in audit JDM experiments. International Journal
of Auditing 5 : 181–
192. Wright, S., and A. M. Wright. 1997. The effect of industry experience on hypothesis generation and
audit planning decisions. Behavioral Research in Accounting 9: 273–294.
Auditing: A Journal of Practice & Theory, November 2007