Let's Get United and #ClearTheShelters: The Factors Contributing to Users' Network Centrality in Online Social Networks

This study explores the factors contributing to online users’ network centrality in a network on Twitter in the context of a social movement about the “clear the shelters” campaign across the United States. We performed a social network analysis on a network including 13,270 Twitter users and 24,354 relationships to reveal users’ betweenness, closeness, eigenvector, in-degree, and out-degree centralities before hypothesis testing. We applied a path analysis including users’ centrality measures and their user-related features. The path analysis discovered that the factors of the number of people a user follows, the number of followers a user has, and the number of years since a user had his account increased a user’s in-degree connections in the network. Together with the user’s out-degree connections along with in-degree links pushed a user to have a strategic place in the network. We also implemented a multi-group analysis to find whether the impact of these factors showed differences specifically in replies to, mentions, and retweets networks.