Behavioral Research in Accounting
Published annually by the Accounting, Behavior and Organizations Section of the AAA

1997, Volume 9 Supplement

The Effects of Interference and Availability From Hypotheses 
Generated by a Decision Aid Upon Analytical Procedures Judgments

John C. Anderson, Virginia Polytechnic Institute and State University 
Steven E. Kaplan and Philip M. J. Reckers, Arizona State University

This study tests the influence (upon analytical procedures judgment) of having auditors focus on explanations from an error vs. non-error dominated list of explanations. The audit judgment studied was the probability of misstatement due to error or irregularity, based on an unexpected fluctuation in the inventory turnover ratio. A laboratory experiment was conducted using experienced audit seniors and managers, in varied client environments. As predicted by interference and availability theory, having auditors focus on errors from an error dominated list resulted in an increase in auditors' likelihood assessment of error. On the other hand, having auditors focus on non-errors from a non-error dominated list did not significantly decrease auditors' likelihood assessment of error. Results suggest that focusing auditor attention on externally generated errors can make error hypotheses more available and increase the subsequent likelihood assessment of error, and focusing auditor attention on externally generated non-errors does not significantly lower error availability.

  Return to BRIA 1997 Contents

The Relationship Between an Individual's Values and 
Perceptions of Moral Intensity: An Empirical Study

Gail B. Wright, Charles P. Cullinan and Dennis M. Bline 
Bryant College

The accountants' ethical decision-making process occurs in a contextually rich environment which includes the moral imperative of a particular issue (i.e., moral intensity), the values systems of moral agents, and professional influences such as the American Institute of Certified Public Accountants' Code of Professional Conduct. Jones (1991) theorized that issue characteristics, collectively called moral intensity, affect the agent's decision process. Weber (1993) suggests that values may also affect the agent's response to an ethical dilemma. This paper explores the relationship between a moral agent's values system and his or her perceptions of moral intensity. Results indicate that among issues of lower moral intensity, the accountant's values preferences influence perceptions of moral intensity. These results imply that differences in individuals' values preferences are most likely to influence the ethical decision-making process when dealing with issues for which there is a lower moral imperative.

  Return to BRIA 1997 Contents

Some Determinants of Analysts' Forecast Accuracy

Dipankar Ghosh, University of Oklahoma 
Stacey M. Whitecotton, Arizona State University

Financial analysts often make predictions concerning annual earnings, and there is evidence to suggest that some provide more accurate forecasts than others. Despite their importance as both users of information and agents for its dissemination in financial markets, little is currently known about the factors which cause these differences in financial analysts' forecast accuracy. This study provides evidence on this issue by addressing three research questions. First, does ability affect performance in a financial forecasting task? If so, does the effect persist after controlling for experience? Finally, does a decision aid mitigate the effect of ability on earnings forecasting performance?

The results show that two measures of ability, perceptual ability and tolerance for ambiguity, were significantly related to earnings forecasting accuracy. Moreover, these effects were very persistent and not easily overcome by either experience or a decision aid. The decision aid was moderately successful at reducing the effect of tolerance for ambiguity, but not perceptual ability. Suggestions for future research and practical implications are discussed.

  Return to BRIA 1997 Contents

The Effects of Audit Structure and Experience 
on Auditors' Decisions to Isolate Errors

Heather M. Hermanson, Kennesaw State University

This study examines the effect of audit firm structure, control risk and task-related experience on auditors' decisions to project sample errors to the population or to treat sample errors as isolated occurrences. One hundred forty-one practicing auditors from five Big 6 firms evaluated six error scenarios and indicated whether they would project or isolate each error. Results indicated that greater audit firm structure and task-related experience were associated with increased likelihood of projection. Control risk was not related to error projection. These results contribute to a growing body of research attempting to understand the auditor's judgment process in evaluating sample evidence.

  Return to BRIA 1997 Contents

The Effects of Justification, Task Complexity 
and Experience/Training on Problem-Solving Performance

C. Janie Chang, California State University, San Marcos
Joanna L. Y. Ho, University of California, Irvine
Woody M. Liao, University of California

Accounting decision problems require decision makers to not only recognize decision rules but to also apply such decision rules to solve complex problems. This study examines the effects of justification conditions and individuals' general experience and training on problem-solving performance, and whether these effects are contingent on task complexity. Three hundred twelve subjects (145 MBA and 167 undergraduate business students) participated in a laboratory experiment. MBA and undergraduate students were treated as two different subject groups because of their differences in general experience and training. Subjects were randomly assigned to a justification or a no-justification condition and were asked to respond to two sets of three problems. The first set tested their understanding of three basic decision rules. The second set tested their application of the decision rules to solve three complex accounting problems. The results show that general experience and training significantly affected performance. Surprisingly, a justification requirement increased effort but did not significantly improve performance, regardless of task complexity.

  Return to BRIA 1997 Contents

Auditors' Uncertainty Representation and Evidence Aggregation

Theodore J. Mock, University of Southern California and University of Maastricht 
Arnold M. Wright, Boston College
Mary T. Washington, Governors State University
Ganesh Krishnamoorthy, Boston College

Although probabilistic reasoning is an appropriate approach to consider uncertainty, prior research has found little evidence that auditors use explicit probabilistic reasoning processes. In this study, two Bayesian and one Classical judgment models are derived for a task where 12 auditors evaluated the likelihood of a material error. Evidence of use of these models comes from two sources: verbal protocols and comparisons with predictions based on the three models.

The protocol evidence indicates that seven subjects exhibit reasoning consistent with a Bayesian representation and five with a Classical Statistical representation. However, in aggregating the evidence, those employing a Bayesian representation were found to make risk judgments most consistent with a Classical model. By examining two phases of uncertainty evaluation, this study is able to isolate behavior in the representation phase from that in the aggregation phase. The results suggest a need for training or decision aids in the aggregation phase.

  Return to BRIA 1997 Contents

Auditor Judgments: The Effects of the Partner's 
Views on Decision Outcomes and Cognitive Effort

Christine E. L. Tan, Christine A. Jubb and Keith A. Houghton
University of Melbourne

There are political, social and economic incentives embedded within the audit setting which provide stimuli for compliant behavior of subordinate auditors to that preferred by the superior auditor. This study examines this compliant behavior under conditions of accountability. Specifically, this study examines whether subordinate auditors, when they are held accountable for their decisions, align their views with those expressed by the evaluative audience;emtheir superiors. In addition, the degree of cognitive effort associated with this compliance is examined. An experiment involving 70 auditors from one Big 6 firm required the subjects to complete a preliminary risk assessment of inventory obsolescence for five separate inventory items. All auditors were informed that they might have to verbally justify their decision outcomes to the partner and/or the researchers. The risk assessments of the auditors' superior were manipulated. Generally, the risk assessments of the superior significantly influenced the subordinates' risk assessments and the amount of cognitive effort exerted by the subordinate, accountable auditors.

  Return to BRIA 1997 Contents