Dynamic Models of Communication in an Online Friendship Network

Abstract

In this article, we argue for the usefulness of relational event network analysis to study online communication networks. Unlike other network analytic techniques that require online communication data to be summarized prior to analysis, relational event network analysis uses un-summarized time-stamped data to track the dynamic evolution of communication networks. To illustrate, we use relational event network analysis to analyze the evolution of a communication network within the virtual world Second Life. Results suggest that there are different patterns of communication among nonfriends and friends within the network. Nonfriends tend to communicate with those they have communicated with in the past, reciprocate communication, and close communication triads. Friends tend not to communicate with those they have communicated with in the past, instead preferring to reciprocate communication and close triads. We discuss implications for the study of online communication and identify directions for future research using relational event network analysis.