Social media’s contribution to political misperceptions in U.S. Presidential elections

There is considerable concern about the role that social media, such as Facebook and Twitter, play in promoting misperceptions during political campaigns. These technologies are widely used, and inaccurate information flowing across them has a high profile. This research uses three-wave panel surveys conducted with representative samples of Americans during both the 2012 and 2016 U.S. Presidential elections to assess whether use of social media for political information promoted endorsement of falsehoods about major party candidates or important campaign issues. Fixed effects regression helps ensure that observed effects are not due to individual differences. Results indicate that social media use had a small but significant influence on misperceptions about President Obama in the 2012 election, and that this effect was most pronounced among strong partisans. Social media had no effect on belief accuracy about the Republican candidate in that election. The 2016 survey focused on campaign issues. There is no evidence that social media use influenced belief accuracy about these topics in aggregate, but Facebook users were unique. Social media use by this group reduced issue misperceptions relative to those who only used other social media. These results demonstrate that social media can alter citizens’ willingness to endorse falsehoods during an election, but that the effects are often small.


Appendix A. Question wording and descriptives for 2012 election study
Social media use was assessed in a two ways. First, a question asked "How often did you use the following strategies to get to the news sites that you used in the past month?" Several behaviors were listed, including "Followed a link posted on a social network, such as Facebook or Twitter." Second, respondents who answered in the affirmative when asked if they had ever used "an online social networking site, such as Facebook, Twitter, or LinkedIn" were asked "How often did you use social networking sites to do the following in the past month? (a) Read news headlines or short news summaries about the Presidential candidates or the campaign; (b) read messages from, or profiles of, the presidential candidates or the campaign; (c) read political opinions about the presidential candidates or the campaign; (d) look at a video or image about the Presidential candidates or the campaign." Descriptives are reported in the main text of the manuscript.
Offline news use was assessed by asking respondents how often they get news or information from five offline sources in the past month: print newspapers, television, print magazines, radio news, and political talk radio. Responses are given on a five-point scale, from "never" to "every day or almost every day", and are recoded so that higher values correspond to more frequent use. The items were combined to form a single composite measure (M = 3.432, SD = .862, alpha = .716) There were several measures of online news use. Partisan site use was assessed by asking respondents how often they got information about the election in the past month from the website of a major national news organization frequently characterized as favoring liberal (conservative) positions and the websites of a politically liberal (conservative) online news organization or blog. Measures of non-partisan site use asked about major news organizations not frequently characterized as favoring a particular party of ideology, and nonpartisan online news organizations and blogs. Responses are given on a five-point scale, from "never" to "every day or almost every day", and are recoded so that higher values correspond to more frequent use. The items were combined to form a single composite measure (M = 1.527, SD = .688, alpha = 838).
Email use was assessed by asking respondents how often they read political emails from three sources in the past month: friends and family; political candidates, parties, or other groups; and news organizations. Responses are given on a five-point scale, from "never" to "every day or almost every day", and are recoded so that higher values correspond to more frequent use. The items were combined to form a single composite measure (M = 1.715, SD = .888, alpha = .804).

Appendix B. Question wording and descriptives for 2016 election study
Social media use was assessed in two stages. First, respondents indicated which of 14 different social media sites they had used at least once in the past month (listed in the Online news use subsection, below). Second, respondents who reported using any of those sites were presented with the following message: "You indicated that you use some type of social media. Though not always intentional, people do sometimes encounter news and information about politics, including the election, using these services. How about you? How often did you use social media to do the following in the past month? (a) Read political news headlines or short news summaries; (b) Access and read political news stories (not just headlines or summaries); (c) Read messages from, or profiles of, political candidates, leaders, or organizations; (d) Read other users' comments about political news stories; (e) Read other users' opinion about politics; (f) Look at a political photo, meme, video, gif, etc." Descriptives are reported in the main text of the manuscript.
Offline news use was measured as it was in 2012. The items were combined to form a single composite measure (M = 2.625, SD = .877, alpha = .760).
Online news use was based on the program list technique (Dilliplane, Goldman, & Mutz, 2013). Respondents first indicated whether they had learned about politics in the prior month from online sources, including news sites, blogs, news app, or from social media sites or apps, such as Facebook, YouTube, or Twitter. Those who had were presented with a list of online sources, in random order, that included 65 frequently used news sites of varying degrees of partisanship. Use was measured as a count of sites (M = 4.271, SD = 5.133). A complete list of news sites follows: Google News (news.google.com), Yahoo News (news.yahoo.com), CNN (cnn.com), Buzzfeed (buzzfeed.com), Washington Post (washingtonpost.com), USA Today (usatoday.com), Vice (vice.com), Daily Mail (dailymail.co.uk), Business Insider (businessinsider.com), BBC (bbc.com), The Guardian (theguardian.com), NBC News (nbcnews.com), NPR (npr.org), ABC News (abcnews.go.com), CBS News ( Furthermore, for respondents who reported getting news from social media, fourteen social media platforms were included in the list of online news sources, in random order. These platforms were: Facebook (www.facebook.com), Google Plus (plus.google.com), LinkedIn (www.linkedin.com), Twitter (www.twitter.com), Tumblr (tumblr.com), Reddit (www.reddit.com), Pinterest (www.pinterest.com), YouTube (www.youtube.com), Vine (vine.co), Instagram (www.instagram.com), Snapchat (www.snapchat.com), Periscope (www.periscope.tv), Meerkat (meerkatapp.co) , Yik Yak (www.yikyak.com) Email use was assessed by asking respondents how often they read political emails in the past month from either their friends and family, or anyone else. The items used the same fivepoint response scales as were used in 2012, from "never" to "every day or almost every day", and were coded so that higher values correspond to greater frequency. The items were combined to form a single composite measure (M = 1.544, SD = .731, r = .665).           Change in exposure to political information on social media since prior wave (1)

References
(2) Note. Top figure (1) shows cross-sectional relationship, bottom figure (2) shows relationship between changes in social media use and belief accuracy.   Change in exposure to political information on social media since prior wave Graphs by Wave