Title

A multi-method study of auditors' knowledge structures of financial statement errors

Date of Completion

January 2002

Keywords

Business Administration, Accounting

Degree

Ph.D.

Abstract

This dissertation examined auditors' knowledge structures of financial statement errors. Two different experiments involving data sorting exercises and reaction time tests were conducted. In the sorting experiment, auditors were asked to sort financial statement errors as they thought the errors should go together. In the reaction time experiment, auditors were asked to identify financial statement errors after being first prompted with either an audit objective or a transaction cycle cue. All subjects participated in both experiments in order to strengthen the case for determining whether the knowledge structure exhibited varies depending on the methodology used. ^ There were three primary findings. First, the sorting methodology results provided support for the theory that the dominant knowledge structure utilized by experienced auditors is an audit objective structure. This result was consistent with other studies using a sorting methodology. Conversely, the second finding was that experienced auditors exhibited a transaction cycle structure when tested under the reaction time method. The third finding was that both experienced and novice auditors were more accurate in identifying financial statement errors after being prompted with transaction cycle cues than after being prompted with audit objective cues. ^ There are two primary contributions of this dissertation. First, it provides support for the theory that the more dominant knowledge structure of experienced auditors is the transaction cycle structure. This is an important contribution because a series of audit training recommendations and assumptions regarding accounting research have been made based on the assumption that experienced auditors have audit objective dominant knowledge structures. Consequently, these findings may lead to a reconsideration of these recommendations and assumptions and also prompt future researchers to reexamine this issue. ^ This dissertation also demonstrates that the reaction time method is a methodology well-suited for studying issues in accounting and auditing domains. The fact that the reaction time method found different results than the sorting experiments related to an issue that was considered “resolved” may lead to further use of this and other new methodologies. The use of new and innovative research techniques could also possibly improve the quality of future accounting research. ^