Abstract ABSTRACT: This research presents some descriptive evidence of how auditors perform analytical review in a complex and realistic task setting. Specifically, the research investigates how auditors who differ in terms of experience, design and conduct analytical review and revise audit programs in light of their analytical review judgments. To accomplish this purpose, a comprehensive case was administered to four auditors who performed the task while thinking aloud. The results show that both managers and seniors identified the crucial audit problems embedded in the case. However, there was little evidence that analytical review provided a basis for reducing the extent of planned substantive audit procedures even in audit areas where reduction seemed to be warranted. Also, several important behavioral differences were observed in the information acquisition and decision behavior of the managers and seniors. Some of these differences are consistent with differences found in studies of physics experts and novices can be attributed to constructs in human memory and reasoning research. Interestingly, the protocols produced almost no evidence of the use of probabilistic reasoning (e.g., Bayesian inference) by any of the auditors. However, there is evidence of their use of the kinds of reasoning processes being proposed and evaluated in artificial intelligence research.
Biggs et al. (Fri,) studied this question.