Keywords: misinformation, social media, affective sciences, social psychology
The spread of misinformation online continues to have negative social and political consequences and efforts to contain it have had limited success. In this research we address a gap in the study of misinformation by focusing on the specific affective antecedents of engagement with misinformation. Across two studies using content posted to Twitter (N = 90,237 URLS, 18,436 responses) and Facebook (N = 9,889 URLS) during a period of heightened misinformation spread (the 2016 US presidential election), we find evidence that misinformation news links were consistently associated with more moral outrage evocation than factually-accurate URLs. We also found outrage-evoking misinformation was associated with more engagement (likes, shares, replies) than both misinformation that is less outrage-evoking and outrage-evoking factually-accurate news. On Facebook, misinformation was more strongly associated with outrage than with any other affective reaction, which was not the case for factually-accurate news. Outrage-evoking misinformation was also associated with more frequent “mindless” sharing; or sharing content without clicking into or reading it. Our results suggest that moral outrage is a key component of the spread of misinformation online. Across social media platforms expressions of moral outrage may act as a behavioral marker (above and beyond other emotions) for misinformation with the potential to spread. Tracking expressions of outrage at scale my therefore enable faster and more generalized algorithms for detecting misinformation before it spreads: while the content of misinformation changes practically by the hour, the moral expressions associated with spreading may be a stable and predictable feature of human psychology.