Evaluation of alternative idea generation techniques in audit fraud risk assessments.

Recent international regulatory reforms have put more responsibility on auditors for detecting financial statement fraud. The recent changes to international auditing standards make it compulsory for members of the audit team to discuss the susceptibility of an entity to fraudulent misstatements at the audit planning stage. These standards do not stipulate the format of these group discussions, nor provide anyexplicit guidelines on how to carry out effective discussions. This dissertation reports on two studies that examine different discussion techniques which could lead to a betteridentification of audit fraud risks at the audit planning stage, and may in turn help in frauddetection.Study One tests whether two enhanced forms of interacting group processes (brainstorming guidelines and premortem instructions) have advantages over an interacting group (without brainstorming guidelines). The brainstorming group is provided with Osborns (1957) original brainstorming guidelines. Premortem is a variation of the mental simulation idea developed by Klein (1999) invoking a scenario where participants actively search for a flaw in their plans. The auditors in both the brainstorming group treatment and the premortem group treatment generated a larger number of potential frauds than the interacting group treatment. These results wererobust across a range of dependent variables used to measure quantity and quality of fraud items. The results also show that more premortem groups listed rare frauds compared to the other two group structures. There was a negative relationship between the number of items listed and the assessed likelihood of fraud. Participants who estimated a higher likelihood of fraud provided a more complete mental simulation of how a fraud could be perpetrated.Study Two focuses on the individual brainstorming phase which may occur before the group brainstorming session. It examines whether giving additional brainstorming guidelines and task decomposing (in addition to the original brainstorming guidelines), improve performance compared to having only the original brainstorming guidelines.Both the additional brainstorming guidelines and the task decomposition treatments generated a larger number of potential frauds than the brainstorming guidelines treatment. Task decomposition helps participants focus equally on all task subcategories leading to a more diverse set of ideas