Audit Judgments Using Belief versus Probability Assessment

This experimental study of 96 experienced auditors examines whether auditors’ judgments are influenced by the ‘approach’ used by the auditors in uncertainty assessment and by the ‘framing’ of an audit assertion to be verified. Approach was manipulated by using belief functions and probability theory as the manner in which uncertainty was assessed. Differences in framing were manipulated by stating the financial statement assertions to be examined in negative versus positive form. Prior research on framing effects in auditing and in psychology reports somewhat inconsistent results. Such observed inconsistency may have resulted from the nature of framing treatment and/or the approach used to elicit auditors’ uncertainty assessments. Thus we investigate both the effects on auditors’ judgments of ‘assertion framing’ and ‘uncertainty assessment approach’ and also the possibility of significant interactions between these two factors. We find limited evidence of approach effects and pervasive evidence of framing effects. On an auditor by auditor basis, we find a wide range of differences in belief versus probability assessments given the same audit situation. However, only in one case, are the judgments captured by using belief functions statistically different from the judgments captured by using a probability approach. Furthermore, there is some evidence that the auditors exhibit a confirmation bias, that is, have a tendency to place more weight on the aspect of mixed audit evidence that confirms a given assertion than the aspect that disconfirms it. These results are important for both audit theory and practice in that they show that auditors form different likelihood assessments (either belief or probability assessments) based on the same audit evidence depending on the framing of an audit assertion. Such differences may impact both audit effectiveness and efficiency.