This article presents a series of short cases, taken from the human information processing research literature in auditing, which can be used as supplementary teaching material in undergraduate and graduate auditing courses. The purpose of the cases is to illustrate to future auditors some of the subtle judgmental biases that can adversely affect their judgments and decisions. Use of such materials should constitute one step toward a more effective integration of teaching and research in auditing.

This article examines the issue of cue weighting in judgment models used to represent human information processors. Illustrative results are presented which show that predictions produced by linear models are very robust with respect to alternative cue-weighting schemes unless the number of predictors included in the models is small, the mean correlation among the predictors is low, and the dispersion of the weights is large relative to their mean. Implications of these results for judgment modeling research in accounting are discussed. Generally, these implications concern the potential improvement in human decision making that might obtain from the search for alternative predictors and/or measurement methods rather than the search for improved methods of weighting predictors in linear models.

This article presents the consensus concept of objectivity and two problems of the concept which never have been addressed adequately: (1) observed objectivity (agreement among different measurers applying the same measurement system) may be spuriously inflated by the consensus inherent on various other combinations of rules and measurers; (2) suggestions for improving objectivity typically focus entirely upon measurement systems and virtually ignore the impact of measurers on objectivity. A framework which enables the investigation of these problems is presented. The suggested framework, a multirule-multimeasurer matrix, is an adaptation of the multitrait-multimethod matrix often used to assess the validity of psychological tests and enables investigation of these problems.

The article presents a response by the author on comments made by scholar Cornelius J. Casey on his article "The Predictive Ability Criterion and User Prediction Models," published in the October 1974 issue of the periodical "The Accounting Review." Cassey addresses three areas contained in my previous article. My response, which is intended to be constructive and elaborative, considers these areas in the order adopted by Casey. Casey's major contribution lies in his discussion of the severity of criterion problems in accounting contexts. For example. in commenting on my discussion of the imprecision of criterion values, he correctly observes that decision makers act on the basis of criterion values whether they are precise or not. Further, he recognizes that consistent measurement of criterion values is a possible means of minimizing imprecision. These points are well-taken. In fact, research and study subsequent to my original article have led me to question whether criterion problems in accounting contexts are more severe than similar problems in other contexts, for example, the criterion of hospital staff diagnosis, in psychology, or grade point average in education, as I previously had believed. On the other hand, much of Casey's concern apparently arises from misinterpretation of my position.

The article is concerned with improving the utilization of accounting information in prediction situations. In this context improvement is defined in terms of an increase in prediction accuracy. If adequate criterion information is available, increase in prediction accuracy may be "directly" assessed by measuring the correspondence between criterion values and predictions of those values. The article recognizes that there are at least two approaches to improving prediction accuracy. The first approach is to improve the data which are supplied to decision-makers and used as a basis for predictions. The second approach is to improve the way that predictions are made on the basis of whatever data are used. The author presents evidence from the psychology literature that individuals may usefully be replaced by their models in prediction tasks, and he also discusses the reasons for this situation. He presents evidence from the management literature which leads to the same conclusion and discusses implications for accounting of these findings.

This article comments on the article "Behavioral Implications of Taxation," D. Larry Crumbley, published in the October 1973 issue of the journal "The Accounting Review." Crumbley considers some behavioral aspects of income taxation. He makes two major points. First, he suggests that research should be undertaken into the effects of tax laws on behavior. Second, he suggests that accountants should become involved. With respect to the first point, the purpose of this article is to indicate that Crumbley is recommending two distinctly different types of research. He does not always maintain a clear distinction between these two types in his discussion of behavioral tax research. With respect to the second point, the purpose of this article is to examine Crumbley's reasoning for his suggestion that accountants become involved in behavioral tax research. It has been suggested that the primary aim of behavioral science is to identify underlying regularities in human behavior and to determine what antecedent conditions give rise to them and what consequences follow from them. However, research which focuses primarily upon antecedent conditions and human behavior in an effort to predict consequences is very different from research which focuses primarily upon consequences and tries to infer the behavior that led to those consequences.

This article focuses on various models of auditing. One property of accounting numbers which has received a great deal of attention in the past few years is that of predictive ability. Research conducted to date has tended to view predictive ability as an impersonal criterion for the evaluation of accounting numbers. Behavioral reactions to particular sets of accounting numbers by particular decision-makers have typically not been investigated by predictive-ability researchers. However, a committee of the American Accounting Association recently introduced a model which, according to the committee, provides the potential for evaluating internal accounting numbers by examination of the utilization of those numbers by decision-makers. The purpose of this article is to provide a more complete discussion of the proposed model and, more importantly, to indicate potential hazards and other considerations in the implementation of the committee's proposal. The discussion draws heavily upon certain portions of the psychology literature which deal with the manner in which individuals utilize numerous pieces of data in making complex judgments or predictions.

Studies of the error-detection performance of audit procedures have found that three attention-directing procedures--client inquiry, expectations based on prior-year errors, and analytical review--signal almost half of the material errors detected. Further, the cost of performing such procedures appears to be relatively low, suggesting a positive cost/effectiveness relation. The primary purpose of the present study is to report detailed results of applying these attention-directing procedures when errors are identified. The sample consists of 186 engagements involving 368 proposed audit adjustments of a Big Eight audit firm. The results corroborate prior studies in that about half of the errors were signaled by the three attention-directing procedures and the simplest forms of these procedures (e.g., analytical reviews comparing current- and prior-year balances) identify many errors. Moreover, internal controls appear to condition the diagnosticity of audit procedures. When controls are strong, procedures involving internal accounting data are more diagnostic, while with weak controls evidence external to the accounting records signals relatively more errors.

[Studies of the error-detection performance of audit procedures have found that three attention-directing procedures-client inquiry, expectations based on prior-year errors, and analytical review-signal almost half of the material errors detected. Further, the cost of performing such procedures appears to be relatively low, suggesting a positive cost/effectiveness relation. The primary purpose of the present study is to report detailed results of applying these attention-directing procedures when errors are identified. The sample consists of 186 engagements involving 368 proposed audit adjustments of a Big Eight audit firm. The results corroborate prior studies in that about half of the errors were signaled by the three attention-directing procedures and the simplest forms of these procedures (e.g., analytical reviews comparing current- and prior-year balances) identify many errors. Moreover, internal controls appear to condition the diagnosticity of audit procedures. When controls are strong, procedures involving internal accounting data are more diagnostic, while with weak controls evidence external to the accounting records signals relatively more errors.]

[Five experiments that examined sequential belief revision in simplified auditing contexts are reported. The results suggest that auditors' belief revisions depend on both the order in which evidence is received and the manner in which it is presented, and they provide preliminary insights into auditors' "attitudes" toward evidence. The results also suggest that audit judgment may differ from judgment in general in two ways. First, the subjects were "prone" to revise their beliefs when new evidence was received, while the behavioral decision theory literature suggests that persons in general tend to "avoid" new evidence. Second, the subjects revised their beliefs to a greater extent when they received evidence that tended to disconfirm their current beliefs, whereas the literature suggests that persons in general are more strongly influenced by confirming evidence.]