The effect of auditor interaction on decision making in the going-concern task

Purpose - To investigate the effects of auditor interaction on problem representation, information acquisition, and performance in the going-concern task. Design/methodology/approach - Participants were asked to evaluate the going-concern status of a company. The study used a pretest-posttest control group (CG) experimental design. Half of participating accountants were randomly assigned to the CG and the other half to the experimental group (EG). In stage one, participants in both CG and EG performed the entire task individually. In stage two, participants in EG performed the experimental task as an interacting dyad while participants in CG performed the experimental task again individually. Findings - The results showed that interacting auditors' "shared problem representation" focus more on relationship between information and less focus on the simple facts or abstraction. Interacting auditors acquired fewer total number of cues, spent more time, visited financial cues fewer times, and acquired fewer liquidity and management cues than did individual auditors. The results also showed that the effects of auditor interaction were maximized when the member of a dyad were heterogeneous. Research limitations/implications - The paper used a going-concern task and non-hierarchical dyads. Future studies might investigate the effects of: different tasks other than going-concern evaluation, hierarchical dyads (e.g. senior-manager) after statistically controlling for the power variable, and different types of dyads (e.g. traditioned vs staticized) on group decision making. Practical implications - The results can be used for training purposes for auditors to increase their performance. Originality/value - The study is one of the first to work in the area of multi-person audit judgment (especially in interacting auditor judgment).