Auditor Mental Representations and Hypothesis Testing of the Control Environment

Abstract

In this thesis, I examine how auditors construct their mental representations and test their hypotheses about the strength of a client’s control environment. With regard to the former, I hypothesize that management’s frame of the control system and auditor’s retrieval of control environment information from memory may influence the auditor’s control environment mental representation and impact subsequent audit judgments. Consistent with my theoretical predictions, I find that retrieval of control environment information from memory biases an auditor’s mental representation, and that this biased mental representation impacts subsequent fraud assessment. In addition, there is limited evidence to support the conjecture that auditors may be susceptible to management’s framing of the internal control system resulting in relatively positive control environment evaluations which was found to transfer to some subsequent audit judgments. With regard to the latter, prior audit literature has examined how auditors evaluate person specific characteristics, such as competence, of other auditors, however there has been no research that has examined how auditors test such characteristics of client management. I disentangle whether auditors utilize a diagnostic and/or a conservative hypothesis testing strategy when testing client management’s ethicality and competence as these are fundamental components of the client’s control environment. A diagnostic testing strategy is evidenced by the auditor searching for the most informative information, whereas a conservative testing strategy is evidenced by the auditor searching for risks. I examine how a checklist decision aid contained in the current institutional context may inhibit auditors’ utilization of a diagnostic testing strategy, and examine how a schematic decision aid is able to enhance diagnostic testing. The results indicate that auditors utilize both diagnostic and conservative testing strategies when testing client management ethicality; however, the auditor’s testing strategy is only diagnostic when testing client management competence. In regard to decision aids, I found that when testing client management ethicality and competence, a schematic decision aid was able to increase the auditor’s extent of diagnostic testing. The checklist decision aid decreased the auditor’s extent of diagnostic testing only when testing client management ethicality, and was not different from unaided judgments when testing client management competence.