Understanding and predicting team performance: a collaboration perspective

Collaboration is an important way to pool resources and expertise from a team and achieve goals that are difficult for individuals. As human beings tackle more complex problems, collaboration is becoming more prevalent and important. Thus, better understanding and predicting team performance from collaboration patterns can aid team management and decision-making on personnel selections. Specifically, team members use their skills, knowledge, expertise, and tactics learned from their previous collaboration experiences for subsequent collaborations. In order to connect individuals’ previous collaboration experiences with future team performance, this dissertation adopts data-driven analyses to study team performance in research and sports teams. Using large-scale scholarly data, the connection between the expertise of scholars obtained from their research experiences and research performance in emerging and established areas is explored. Additionally, I predict the performance of sports teams using sports coaches’ hierarchical structures. Implications of this dissertation are not limited to team management in research and sports but can be generalized to other domains where collaborations are an essential part of predicting team performance.