Social structure, knowledge sharing, and project performance in open source software development

Open Source Software (OSS) development has attracted considerable attention from both researchers and practitioners. A successful OSS project is created through the growth of a community of individuals with shared interests. However, little is known about this virtual community and how its network structure influences project performance in OSS teams.; Drawing on prior literature and social network theories, this study examines the associations among network structure, knowledge sharing, and project performance. The study hypothesizes that two properties of social network structure (i.e., degree of centralization and core/periphery fitness) affect project performances in OSS teams with knowledge sharing as a mediator. Project samples were selected from SourceForge, the world's largest website hosting OSS projects. Multiple research methods such as social network analysis and content analysis have been used to measure the focal variables. The results indicate that both network structure and knowledge sharing significantly impact project performance. In addition, the quantity of knowledge sharing mediates the relationship between social network and project performance.; This research contributes to an increased understanding of social networks and knowledge collaboration in virtual teams. Specifically, it provides empirical evidence relating to how social capital influences collective actions and productivities in OSS teams. The findings also have significant implications for practice as they will be helpful to OSS developers by providing guidelines and suggestions for constructing an efficient social network structure to enhance the performance of OSS teams.