The network as dependent variable : antecedents of online social network formation.


Access rights

Worldwide access

Journal Title

Journal ISSN

Volume Title



Although social networks change over time, network dynamics remains an understudied area of social network research. The detailed logs that are a feature of online social networks allow for easier observation and capturing of social network evolution. This study examines the evolution of an online social network. The network is proposed to evolve with varying tendencies of density, reciprocity, centralization, and demographic influences under two competing theories – social capital theory and social comparison theory. Using a longitudinal data set of email communications that span a period of twenty-six months, the study examines the dynamics of the email communication network of a technology company. Prior social network research is dominated by descriptive studies. This study is different because it not only tests the statistical significance of the influence of structural and demographic factors on network evolution, but also their relative strengths. Further, the study also examines how online social networks differ from offline social networks. It examines the differences between the (online) email communication network and the (offline) advice, brainstorming, and working networks within the same organization. The study finds that the email network evolves with tendencies towards high density, high centralization, and high reciprocity. Demographic influences are inconsistent over time; they are salient only in certain months. The study also finds that online networks exhibit higher levels of density, centralization, and reciprocity compared to offline networks. As such, the study contributes to theory by finding (I) antecedents of social network evolution and (II) differences between online and offline networks.



Social network analysis. Social network formation. Exponential random graph modeling. Online networks.