Teams are often introduced with the objective of improving the effectiveness of the organization, as well as the quality of working life for employees. This immediately makes clear that constructs in teamwork research may refer to completely different variables depending on the level that is considered (e.g. organizational effectiveness vs quality of working life). This paper aims to highlight the multilevel nature of team research.
In this paper three related multilevel issues are discussed: a conceptual and theoretical perspective of a multilevel approach to teamwork; measurement issues; the analytical side of a multilevel approach. The paper ends with two examples.
Following Klein and Kozlowski, three types of constructs are distinguished. Global constructs characterize the entire team (e.g. team size). Shared constructs are attributes that stem from individual workers but it is supposed that team members share these perceptions (e.g. interpersonal trust or group cohesion). Compositional properties also stem from individual team member; however, there is no condition of within‐unit agreement (e.g. age, skills or personality traits).
This paper contributes to the multilevel approach to teamwork.
