Mike D Ferguson

Mike D Ferguson is a Industrial Organizational Psychologist based in Southern California, specializing in People Analytics and Employee Research. His intellectual passions center on productivity, teamwork, training and development, and workplace technology.

Which Differences Make a Difference? Comparing Diversity Metrics and their Relationships to Outcomes in Virtual Teams

Full Thesis

Proposal Presentation

Defense Presentation

Summary: Inconsistent conceptual definitions and operationalizations may account for conflicting effects of diversity on outcomes. Further, although evidence suggests that contextual moderators are particularly relevant for understanding the association between diversity and outcomes, contextual effects are not consistently accounted for in the literature. Given that most modern work teams collaborate virtually to some extent, the context of team virtuality may be a key omitted moderator that could affect the relationship between team diversity and outcomes.

My thesis explored these questions by first exploring whether the measurement and mathematical operationalization of diversity differentially predicts team processes and outcomes. I then examined contextual factors that may moderate the relationship between diversity and team processes and outcomes, to ultimately determine which measures of diversity make a difference to teams with differing levels of virtuality

Participants: I used archival data from nine different MBA graduate and upper-level undergraduate business courses across four universities. The final sample used for the study included 101 teams with three to six members per and a total of 451 participants.

Methodology: I adopted an inductive approach to explore the effects of diversity measurement on team outcomes. Specifically, I compared diversity operationalizations in terms of convergence using correlations, investigated computational bias in diversity metrics and their proposed corrections using a series of regressions, and contrasted simple proportional measures with other metrics in terms of their predictive validity for team outcomes using a series of structural equation models (SEM). I then assessed the independent effects of diversity on two attributes that vary in terms of their job-relatedness using a SEM. Finally, I tested moderating effects of virtuality on the association between diversity and outcomes using another SEM.

Results: Findings from the inductive analysis showed that diversity metrics demonstrated a high degree of convergence and revealed a small but significant team size bias in the uncorrected versions of diversity metrics. Diversity metrics showed incremental predictive validity over simple proportional measures for only gender, but not for any of the four other attributes. Neither job-related educational specialty diversity nor not-related ethnicity diversity showed any significant main effects on team processes and outcomes. Finally, team virtuality showed an unhypothesized main effect on team processes and outcomes, and was found to moderate the relationship between educational specialty diversity and task-based outcomes, but not for ethnicity diversity.

Conclusions: Results suggest that some measurement effects may exist, but are likely small and may not pose as substantial threat to validity of findings in diversity studies. Further, team virtuality likely plays a substantial role in determining team outcomes and may impact the effect of diversity through making differences more or less salient.