THE EFFECTS OF CONTRASTS IN ACCOUNT-LEVEL FRAUD RISK ASSESSMENTS ON AUDITORS' EVIDENCE EVALUATION

Author

Grace Ngonidzashe Mubako, Southern Illinois University CarbondaleFollow

Date of Award

12-1-2012

Degree Name

Doctor of Philosophy

Department

Accountancy

First Advisor

O'Donnell, Ed

Abstract

Evidence from research in psychology and auditor judgment has shown that perceptions that form early in a sequential judgment process can influence subsequent judgments. Auditing Standard 12 requires auditors to identify fraud risk factors and assess the risk of fraud as part of the process of assessing overall misstatement risk. While it is expected that fraud risk assessments should have a bearing on overall risk assessments, it is possible that perceptions formed from assessments of fraud risk can negatively affect the evaluation of any evidence reviewed thereafter. Because different classes of transactions may be affected by fraud risk factors in different ways, fraud risk assessments may differ across classes of transactions. These differences may make subsequent auditor judgments susceptible to the contrast effects bias, where subjects overreact to the differences such that the fraud risk assessments influence auditor judgment more than they should. This study examines whether auditors who learn that fraud risk is low for one class of transactions immediately after examining a class of transactions that has high fraud risk, can overreact to the contrast such that they reduce their sensitivity to evidence that suggests increased misstatement risk. The study also examines whether these contrast effects can be mitigated by acquiring information about fraud risk assessments later in the sequence of evidence, after auditors have reviewed and assimilated evidence related to other risks. The study finds that, as predicted, auditor judgments are influenced by contrast effects. Auditors who examined classes of accounts for which fraud risk assessments were different were less sensitive to evidence suggesting increased risk in accounts that had been identified as having low fraud risk. However, contrary to predictions, these contrast effects were not mitigated by evidence order.