Collaboration and creativity: effects of tie strength

Abstract

This dissertation studies the relationship between collaboration networks and scientific creativity. It finds significant knowledge spillover from new collaborations to repeated collaborations, and proposes a network approach to understand scientific creativity at the egocentric network level beyond the boundary of teams. To understand the network effect (specifically, effects of tie strength) on creativity, it integrates literature on small groups and social networks and adopts a creative-process model. An inverted U-shaped relationship between tie strength and creativity is observed, because of the mixed impacts of tie strength at different stages of the creative process. Furthermore, it explores the effect of tie configurations and finds that the skewness of tie strength distribution moderates the effect of tie strength. In addition, it also tests two competing explanations for the association between strong tie and low creativity: creativity-decline hypothesis versus cost-reduction hypothesis. Finally, there is no evidence that collaboration networks would raise the visibility of previously published papers, but there is a significant prestige effect in gaining citations.