Factbait: Emotionality of Fact-Checking Tweets and Users’ Engagement during the 2020 U.S. Presidential Election and the COVID-19 Pandemic

Abstract

Given the importance of fact-checking in reducing the spread of false information on social media, prior research has examined effective fact-checking strategies. The current study addresses this question by conducting a computational analysis of actual fact-checking tweets of three representative fact-checking organizations in the United States (Factcheck.org, PolitiFact, Snopes), replies, and retweets (N = 166,526) on Twitter made from September 29 to November 3, 2020, when the 2020 U.S. presidential election and the COVID-19 crisis co-occurred. The results show that fact-checking tweets with a greater degree of anxiety and anger generally receive more replies, but those exhibiting sadness are retweeted less. Additionally, fact-checking tweets with heightened levels of anxiety, anger, sadness, or negativity, in general, tend to elicit replies featuring a degree of anxiety, anger, sadness, or negativity, respectively. Our findings suggest that emotions can be utilized as drivers of engagement in fact-checking tweets, meaning that the emotional impetus can potentially serve as an important strategy to make fact-checking efforts more impactful in uncertain situations. However, fact-checking organizations should be aware that emotional appeals in fact-checking posts catalyze correspondingly emotional responses from their audiences, which reflects the emotional contagion process.