Individual Submission Summary
Share...

Direct link:

How Online Comments Affect Polarization and Incivility

Thu, August 29, 4:00 to 5:30pm, Hilton, Columbia 4

Abstract

Though studies have measured the content and civility of online comments about politics, less is known about who comments online or what effects these messages have on the behavior of others. We make two contributions. First, prior research and reporting suggests that a non-representative subset of Americans tend to self-select into online discussion, but the extent to which these differences distort the comments that people see is unknown. Second, scholars have not considered whether this seemingly distorted picture of the political views of the American public itself warps the perceptions of others. We therefore test whether people exposed to extreme/uncivil comments (which are often popular and thus displayed by social media algorithms) fail to account for their unrepresentativeness, especially for opposition partisans, and respond with increased negative affect toward politics and the opposition party. We conduct three studies to test these hypotheses. First, we draw on ANES 2016 and Pew American Trends Panel data to examine the differences between frequent online commenters and the general public. Second, we use a novel design to compare levels of incivility and extremity among comments scraped from media outlet Facebook pages with comments we solicit on the same articles from respondents in a nationally representative survey experiment. In this study, we also randomize exposure to highly visible Facebook comments to assess whether they increase expressed levels of incivility and extremity in survey respondents’ comments. Finally, we conduct a follow-up survey experiment to examine whether the comments we drew from Facebook have stronger negative effects than comments from the representative sample.

Authors