We investigate the effects of sentiment and issue salience on emotionally labeled responses to posts from political actors on Facebook (i.e., Reactions). We use an automated content analysis of Facebook posts and voter survey data in a multilevel negative binomial regression approach. Findings show that the sentiment of a post relates to the number of "Love" and "Angry" Reactions. Furthermore, if a post addresses an issue that constituents perceive as salient, this positively influences the number of "Angry" Reactions only. We also find that the effect of sentiment on "Angry" Reactions is highest when issue salience is low.
Why & How I Found This
My Professor suggested reading this paper for our project.