Do diverse groups cheat less than homogenous groups?

Hurvitz, Grant
Journal Title
Journal ISSN
Volume Title
This thesis provides empirical evidence to test the hypothesis that diverse groups cheat less than homogenous groups. The thesis explores both whether or not diversity provides more than just varied perspectives and the benefit of equal opportunity. The research is carried out in an academic setting and is based on an experiment with 688 Second and Third year Auditing students. It uses an experimental task designed to identify instances of dishonesty by individuals in different groups through the completion of a short mathematical assessment. Social psychological theories which are relevant for understanding group composition, social norms, conformity and self-awareness are considered as possible explanations for the manner in which diversity might affect the ethicality of an individual’s decisions within a group. Using a two-way ANOVA the results show that variations in the composition of a group (in terms of diversity or homogeneity) have a significant effect on the frequency of dishonest behaviour by individuals within the respective group. In particular, when students were observed those who were members of diverse groups had test scores which were not significantly different from students writing in homogenous groups. Conversely, unobserved students in homogenous groups had test scores which were significantly higher than in unobserved diverse groups. These results should be relevant for accounting practitioners and corporate governance policymakers - given the emphasis placed on diversity of management groups in terms of codes of best practice - as well as for academics interested in better understanding the dynamics of student groups.