Information Overload,Similarity, and Redundancy: Unsubscribing Information Sources on Twitter

The emergence of social media has changed individuals' information consumption patterns. The purpose of this study is to explore the role of information overload, similarity, and redundancy in unsubscribing information sources from users' information repertoires. In doing so, we randomly selected nearly 7,500 ego networks on Twitter and tracked their activities in 2 waves. A multilevel logistic regression model was deployed to test our hypotheses. Results revealed that individuals (egos) obtain information by following a group of stable users (alters). An ego's likelihood of unfollowing alters is negatively associated with their information similarity, but is positively associated with both information overload and redundancy. Furthermore, relational factors can modify the impact of information redundancy on unfollowing.