www.emeraldinsight.com/1475-7702.htm
The effects of
accounting
knowledge
15
Ronald D. Picur
University of Illinois at Chicago, Chicago, Illinois, USA
Abstract
Purpose This study aims to examine whether accounting knowledge is associated with a decision
makers tendency to ignore value added information in wealth measurement and distribution
decisions.
Design/methodology/approach A between-subjects laboratory experiment was employed.
Subjects prepared accounting reports that measured and distributed an entitys wealth based upon
given accounting data. Accounting knowledge was measured as: a discrete variable by classifying
subjects into high-, low- and no-accounting knowledge groups, and a continuous variable by
classifying subjects on the number of accounting courses completed.
Findings Findings provide empirical evidence that high levels of accounting knowledge interferes
with a decision makers ability to incorporate value added information (versus accounting profit) in
wealth measurement and distribution decisions.
Research limitations/implications This experiment used subjects from the USA where the
production and disclosure of a value added report is not mandated. The results should be tested in a
country where the statement of value added is routinely produced, disclosed and audited.
Practical implications This study shows the dysfunctional effect of accounting knowledge
which appears to hinder performance in wealth measurement and distribution decisions.
Originality/value This is the first attempt to explain why decision makers may ignore value
added information in wealth measurement tasks and distribution decisions by focusing on the role of
knowledge structures.
Keywords Value added, Knowledge mapping, Accounting, Wealth
Paper type Research paper
1. Introduction
Because shareholders are considered the ultimate owners of the firm, wealth
measurement has consisted of various ways of income measurement that include
schemes of earnings management and/or income smoothing. If the corporation is
instead perceived as an alliance of a concerned team that includes shareholders,
bondholders, employees, and the government, then the measurement of wealth is
reduced to the determination of the value added information that is ultimately
distributed as dividends to the shareholders, interest to bondholders, wages to
employees, taxes to government, and undistributed value added as retained earnings
( Burchell et al., 1985; Meek and Gray, 1988; Morley, 1979).
While accounting practice has focused on shareholders wealth and changes in
wealth with the resulting measurement and distribution of profit, empirical research
does not report any overwhelming use of value added information by decision makers.
The belief is that decision makers tend to ignore or underestimate value added
information (Riahi-Belkaoui, 1991; 1996; 1999). The omission of value added
RAF
6,1
16
This study also adds to prior studies that focused on the dysfunctional effect of
knowledge of a subject matter by examining specific conditions where high-accounting
knowledge may hinder performance (Vera-Munoz, 1998; Libby and Luft, 1993).
The findings of this study suggest that the benefits of high-knowledge or experience
may be hampered by the incongruence between the knowledge structure stored in
memory and the decision problem faced by the decision maker.
2. Theoretical considerations
2.1 Use of value added information in wealth measurement
Value added is the economic measure of wealth. Economists rely on value added for the
wealth of the measurement of a nation. However the accounting profession relies on the
income concept for the measurement of wealth. In effect, generally accepted accounting
principles (GAAP) are primarily used in the case of the income statement for the
measurement of the income of a period as an expression of the wealth generated for
that period. Alternative uses of information used for income measurement can be
easily used to construct value added statements containing the net value added or the
gross value added generated by the activities of the firm in a given period.
Because value added reports are not mandated in the USA, the reconstruction of
value added information from the information used to determine income is not
provided for either external disclosure or internal uses. While optimal wealth
measurement incorporates value added information, externally reported financial
accounting information does not include it. Empirical research on why decision makers
may tend to ignore value added information in wealth measurement tasks when they
are not explicitly provided is non-existent. The current study starts this research by
examining whether differential knowledge of GAAP is the main cause for the
ignorance and/or avoidance of value added information in wealth measurement and
distribution tasks. As discussed below, the knowledge structures created by the mental
storage of GAAP-based information may explain the role of accounting knowledge in
wealth measurement that includes value added information.
2.2 The role of accounting knowledge structures in the use of value added information
The essence of cognitive relativism in accounting is the presence of a cognitive process
that is assumed to guide the judgment/decision process. The cognitive processing
models show that judgments and decisions made about accounting phenomena are the
products of a set of social cognitive operations that include the observation of
information on accounting phenomena and the formulation of schemata or knowledge
structures that are stored in memory and later retrieved to allow the formulation of
judgments and/or decisions when needed ( Libby, 1985; 1992; Libby and Frederick,
1990; Libby and Tan, 1992). Because cognition is highly knowledge- and contextdependent, a specific context will induce the recall of knowledge structure that is
expected to be relevant to the context. Because most formal accounting training focuses
on GAAP, the end result is that the knowledge structures stored and recalled will not
include value added based knowledge for wealth measurement and distribution.
The same phenomenon of a lack of fit between the knowledge structures recalled
from memory and the given problem solving structure has been observed in a chess
context (Chase and Simon, 1973), a tax context (Marchant et al., 1991), and an
opportunity cost context (Vera-Munoz, 1998). In all these contexts, the studies show
evidence of the dysfunctional effects resulting from the lack of fit between the
knowledge structures recalled from memory and the decision situation examined.
The effects of
accounting
knowledge
17
RAF
6,1
18
What may be concluded from the above studies is a similar situation for value
added information in wealth measurement tasks and accounting knowledge. That is,
decision makers with high-accounting knowledge who are faced with a wealth
measurement task may be more inclined to recall GAAP-based knowledge structures
that fail to incorporate value added information. Accordingly the following hypothesis
is proposed:
H1. The number of value added information cues ignored in a wealth measurement
task will be greater for decision makers with high-accounting knowledge than
for decision makers with low- or no-accounting knowledge.
3. Research design
3.1 Subjects
Three groups of subjects were asked to participate in a wealth measurement and
distribution task. The three groups differed in the level of accounting knowledge
possessed. The level of accounting knowledge was measured both as a discrete
variable by classifying subjects into high-, low-, and no-accounting knowledge groups
based on the program of business study in which the participants were enrolled and
the number of accounting courses taken.
A total of 180 students from the graduate and undergraduate accounting programs
of a Midwestern university participated in the experiment. The subjects received a flat
compensation payment of $5 plus a performance-contingent bonus, paid after the
experiment was conducted. The bonus was based on a composite score that includes
the number of times value added information was used by the subject in their wealth
measurement tasks, plus their scores on an analytical ability test. Participants whose
scores were in the top quartile received a bonus of $3, the second quartile received $2,
and the third quartile received $1. The bottom quartile did not receive a bonus.
3.2 Case materials and procedures
An experiment phase and a post-experiment phase characterized the study. In the
experiment phase, the subjects were provided with the instructions, assignment, and
data memorandum and a response memorandum. In the post-experiment phase, the
subjects were provided with a reaction questionnaire, a combined test of analytical
ability and value added knowledge, as well as a background questionnaire. In the
response memorandum, the subjects were provided with revenues and expenses data
for a fictional company for a given year and were instructed to: (a) compute the wealth
generated by the firm for the given year based on the data provided, and (b) to indicate
how that wealth was distributed. The case materials were calibrated in a two-stage
pilot test. In the first stage, three accounting professors evaluated the materials and
suggested changes that were later incorporated in a revised version. In the second
stage, three graduate students completed the tasks required and provided the
comments that were also included in the materials presented to the subjects.
3.3 Independent and dependent variables
The three knowledge groups formed included a high-accounting knowledge group, a
low-accounting knowledge group, and a no-accounting knowledge group. The subjects
in the high-accounting knowledge group were: (a) enrolled in the Master of Science in
Accounting program and (b) had completed more than the minimum six accounting
background courses required before being admitted to the program. The subjects in the
The effects of
accounting
knowledge
19
RAF
6,1
20
Accounting knowledge
Table I.
Descriptive statistics
(Standard deviation in
parenthesis; the
maximum possible
number of value added
information omission
was two, and the
minimum was zero)a
High
Low
No
Overall
0.937,
0.858,
0.560,
0.785,
n = 64b
n = 148
n = 148
n = 360
Notes: aResults of the statistical analysis that control for analytical ability are qualitatively similar
to those that do not control for analytical ability. Thus, for presentation purposes, the means
reported here are the raw (observed) means (i.e. not adjusted for analytical ability); bn = number of
subjects in a group two decisions (one on wealth measurement and one on wealth distribution)
Accounting knowledge
High
Low
No
High
Low
No
Data
30
63
43
28
63
23
(93.7)
(85.1)
(58.1)
(87.5)
(85.1)
(31.08)
Notes: aEntries are number (and percentage) of subjects omitting the value added information.
The percentages are the ratio of the number of subjects omitting the value added information
over the total number of subjects in each category
The results of the planned comparison for purpose of testing H1 after controlling for
analytical ability are shown in Table III. H1 predicts that high-accounting knowledge
decision makers are more likely to ignore value added information in wealth
measurement and distribution calculations than either low-accounting knowledge or
no-accounting knowledge decision makers. The planned comparison for testing H1
shows that, as predicted, the effect of accounting knowledge after controlling for
analytical ability is positive and significant (F 71.61; p 0.00001). This result
indicates that on average, the number of omissions of value added information is
higher for subjects with high-and/or low-accounting knowledge than for subjects with
no-accounting knowledge.
To test the effects of accounting knowledge on the tendencies to ignore value added
information, the following regression analysis was conducted:
Y i 0 1 AKi 2 AAi 3 PEi "i
The effects of
accounting
knowledge
21
Mean
Standard deviation
High-accounting knowledge
Low-accounting knowledge
No-accounting knowledge
1.812
1.698
0.888
0.470
0.544
0.358
Table III.
Variable
Expected sign
Coefficient
t-statistic
Probability (one-tailed)
Intercept
AK
AA
PE
(+)
()
(+)
3.486
0.035
0.990
0.004
8.557
1.994
5.741
2.429
0.0001
0.0401
0.0001
0.0172
Notes: Equation 1; F statistic, 26.355 ( p 0.0001); R2, 0.3452 (adjusted 0.3321), "i, error term for
subject i
Planned comparison
(dependent variable is
the number of value
added information
cues omitted)
Table IV.
Regression results
for the effects of
accounting knowledge
RAF
6,1
22
The effects of
accounting
knowledge
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.