Empirical research has been shown that team mental models are the basis for team learning. Team mental models correspond to a mental representation of the knowledge team members share regarding relevant task and team aspects and the environment in which they operate. For team members to be able to effectively engage in learning behaviors they need to share an understanding about different aspects of work. Therefore, without team mental models, learning and performance will not be as beneficial as with team mental models. In this study we tested the notion that teams need to start from a basic common ground in order to effectively learn from previous experiences. Team learning has traditionally been measured in two different ways: (1) as a process consisting of team members behaviors, and (2) as an outcome operationalized as the slope of performance over time. We integrate both approaches and predict that the extent to which team learning behaviors actually result in performance increase depends on the similarity and accuracy of team members’ task, team, and temporal mental models. We tested our longitudinal model in a sample of 67 teams performing in a management simulation over five consecutive time periods. We estimated random coefficient growth models following the six-step model estimation procedure of Bliese and Ployhart (2002). Our findings suggest that although team learning does not have a direct effect on performance growth, particularly teams’ temporal mental model similarity is crucial for the translation of learning behaviors into performance improvement. When teams have low temporal mental model similarity, team learning is even detrimental at the beginning of task performance without leading to high performance levels. When teams have high temporal mental model similarity, team learning behaviors lead to initial negative performance outcomes but after this initial negative effect it leads to a rapid growth in performance. Temporal mental model accuracy mainly impacts the initial state of performance and not the slope of performance. Our longitudinal study allow us to understand not only that mental model similarity is important for the translation of team learning into performance growth, but also allow us to understand when mental model similarity might be most or least effective.

--

team learning; shared mental models; team performance; longitudinal analysis