Date of Award

2015

Document Type

Dissertation

Degree Name

PhD in Accountancy

Department

Department of Accountancy

First Advisor

Mohammad Abdolmohammadi

Second Advisor

Jay Thibodeau

Third Advisor

Alex Pentland

Abstract

This dissertation investigates the impact of collective intelligence and honest signaling on the fraud risk brainstorming effectiveness of audit groups. Emerging research in psychology shows that collective intelligence, which captures the general ability of a group to perform a wide variety of tasks, is a strong predictor of group performance. Honest signals are nonverbal communication cues that are either so costly to make or difficult to suppress that they are reliable in signaling intent.

The first study of this dissertation provides a summary of extant literature examining collective intelligence and honest signaling, with an emphasis on those studies that have implications for the task performance of accounting groups. I also examine the research use and potential of sociometric badges, a special wearable sensor technology that measures paralinguistic honest signaling by capturing fine-grained data on individuals’ intra-group patterns of vocal, nonverbal interaction. Additionally, I propose several avenues for future research that will contribute to academic literature as well as practice. In the second study, I conduct an experiment to study the effect of collective intelligence and honest signaling on fraud brainstorming effectiveness. I employ a between-subjects viii design in which 41 audit brainstorming groups, comprising 152 upper-level accounting students, are assigned to either an electronic or a face-to-face group condition. I hypothesize and find that both collective intelligence and brainstorming mode influence fraud brainstorming effectiveness, and that electronic groups are more effective than faceto-face groups. This latter effect, however, is contingent on brainstorming mode: high collective intelligence improves the brainstorming performance of electronic groups more than that of face-to-face groups. Neither average member intelligence nor maximum member intelligence is predictive of brainstorming effectiveness. These results imply that process losses associated with face-to-face brainstorming attenuate the countervailing beneficial impact of high collective intelligence. Speaking turn variance, as measured by sociometric badges, is negatively correlated with both collective intelligence and fraud brainstorming effectiveness. These findings suggest that members of higher performing brainstorming groups participate more equally in group conversations and that are better at reading social signals, as compared to members of less effective groups.

Download

Share

COinS