Emotions in Microblogs and Information Diffusion: Evidence of a Curvilinear Relationship

Department of Decision Sciences and Managerial Economics

How do emotions embedded in Social Media (SM) content impact information diffusion? Prior research provides inconsistent answers to this question, with one stream of studies suggesting that strong emotions can increase the likelihood of content diffusion and another suggesting that strong emotions can lead to users self-censoring themselves. Our research combines these contradictory arguments to predict that content emotions relate to ID in an inverse U-shaped manner. We further posit that the relationship is heterogeneous depending on a) the context of SM communication and b) the valence of embedded emotions. Empirical analyses of tweets collected from multiple settings lend support to our hypotheses. Our theorizing and results a) contribute to Information Systems (IS) literature by presenting a nuanced account of content-sharing and b) provide practical guidance to content creators who want to maximize the diffusion of their content.