Does Who Helps You Impact Your Behavior? Examining the Effects of Social Interactions on Knowledge Sharing in Online Communities

Abstract

Online communities provide vibrant forums for knowledge sharing and are increasingly being used by individual users and firms to source knowledge and create and capture value. Yet, there is much to learn about how the actions of community members affect other members, particularly new members whose continued participation is important to community vibrancy and growth. The extant literature investigates whether or not receiving a reply increases the likelihood that a new user will post again, finding mixed results. We extend this literature in two ways: we examine the effects of (1) receiving a reply on different types of subsequent participation (asking another question and replying to others’ questions) and (2) receiving a reply from community members with different affiliations (user or employee). Based on data from the Statalist community, our findings show that receiving a reply from another user is positively associated with subsequent questioning, but negatively associated with subsequent replying by new users. Conversely, receiving a reply from an employee is positively associated with subsequent replying, but not with subsequent questioning by new users. To help address endogeneity concerns, we run fixed-effects models, implement an instrumental variable-based approach (using a two-stage Heckman selection model), use a bivariate probit model and coarsened exact matching, and conduct qualitative interviews with Stata employees. Our findings suggest that social interactions play an important, albeit nuanced and sometimes unexpected, role in shaping subsequent participation and highlight the importance of examining the effects of firm participation in online communities.

Keywords: online communities, knowledge development, innovation