Team composition is an important consideration when forming work teams as it is known to affect team outcomes through team mediators. However, because current recommendations regarding team composition are based primarily on face-to-face teams, they may not be as applicable to virtual teams, a team-type that is becoming increasingly more prevalent. Virtual teams are those whose members are distributed across locations and, consequently, rely primarily on technology to communicate with one another. This virtuality could moderate the effect of team composition on team outcomes because higher levels of virtuality pose challenges that are not apparent at lower levels of virtuality. Thus, compositions that work well for less virtual teams might not be as effective for teams that are more virtual. Therefore, the objective of this meta-analysis was to examine how team virtuality moderates the relationships between deep- and surface-level compositional characteristics and team outcomes. The relationship between deep and surface-level composition and team mediators at different levels of team virtuality was also examined because team composition can affect team outcomes through team mediators. The results indicated that the effects of team composition in highly virtual teams and teams that are low in virtuality did not differ from one another; instead, differences arose when teams were moderately virtual. This could be because moderately virtual teams use communication methods that are rich enough for team member characteristics to be discerned, and communicating electronically initially could prevent surface-level attributes from causing categorization. Second-order sampling error, however, could also be influencing the findings. These results suggest that recommendations regarding team composition are generally applicable across teams of differing levels of virtuality, although moderately virtual teams may require special consideration.
