The Effects of Goal Framing on Auditors' Use of a Decision Aid in Environments of Varied Risk

An auditor performing analytical review must typically diagnose material variances of observed client data from his/her own expectations. The auditor may utilize a decision aid to help in generating potential explanations for a variance; it has, however, the capacity to provide many more explanations than are possible using other means. Under the circumstances of budgetary constraints and limited cognitive load for beginning an information search with these explanations, the auditor may consider the lengthy list and arrive at a more manageable sub-list of the most probable explanations. In doing so, the auditor either eliminates those explanations that are less likely or includes those that are more likely into a reduced list for further consideration. While the goal under either approach is the same-to reduce the list-studies in psychology have shown that those including will reduce the list to a much greater extent than those eliminating. If the auditor begins an information search with this reduced list of explanations, then whether the auditor uses inclusion or elimination may have effectiveness and efficiency implications for the remainder of the analytical review process.

The auditor must also contend with risk in the audit environment, which also may influence the manner in which the auditor reduces the lengthy list of explanations. A risky audit environment is generally related to heightened auditor skepticism and increased audit effort, as predicted by the audit risk model (SAS 47, AICPA 1983). Each of these translates into the desire to pursue a greater number of plausible explanations in a high risk environment than in a low risk environment. Therefore, an auditor would be expected to reduce a decision-aid-provided list of explanations to a lesser degree in a high risk environment than a low risk environment.

The purpose of this dissertation was to investigate the occurrence of a goal framing effect at varied levels of client risk. Using a two-way between subjects design, auditors in this study either eliminated or included explanations from a decision-aid-provided list in a low risk or high risk analytical review setting. As suggested by the goal framing theory, auditors who eliminated concluded with significantly more explanations than those who included. Furthermore, as suggested by the audit risk model, auditors in a high risk environment concluded with significantly more explanations than auditors in a low risk environment.

Because previous auditing literature provides that auditor conservatism, which is heightened in periods of high risk, often mitigates biases and heuristics found in the general decision making or psychology literature, it was also predicted that in the high risk scenario, the influence of high risk in enlarging the set of explanations would overcome the influence of the inclusion goal framing in reducing the set of explanations. No support was found for this interaction.

The results of this study have implications for the implementation of decision aids in practice. This study advises that in various client risk settings, auditors evaluating a lengthy decision-aid-provided list of explanations by inclusion may arrive at a significantly smaller number of explanations than by elimination. Given that the subsequent step of analytical review-information search-is planned according to what the auditor believes are the plausible hypotheses, goal framing may have an impact on the overall efficiency and effectiveness of analytical review, in both high and low risk client scenarios.