Online Information Seeking during the 2018 U.S. Congressional Elections

• In Event: Susceptibility to Fake News And Misinformation: Who and Why?

Abstract

Research addressing the public’s exposure to and engagement with partisan and fake news in online environments has proliferated in recent years, but has largely focused on social media. To better understand how people seek out information online, we recruited a digital panel participants — oversampling strong partisans — to install a browser extension that collected their web searches (n = 265) and browsing history (n = 333) during the 2018 U.S. Congressional elections. Participants' had the extension installed on their desktop computers from October 18, 2018 to January 1, 2019, and completed a survey containing a battery of questions about their demographics and political preferences both at the beginning and end of the study.
Using recently developed datasets to quantify the political partisanship of web domains and classify web domains as producers of fake news, we analyzed and compared how our participants surfed and searched the web. We found that (1) Republicans browsed and conducted searches that returned significantly more right-leaning domains than Democrats, and (2) exposure to fake news was more concentrated in browsing, where 8.4% of participants accounted for 90% of fake news exposures, than in searching, where 28.7% of participants accounted for 90% of fake news exposures. We also found that Republicans browsed significantly more fake news than Democrats, but found no significant differences in exposure to fake news by political party ID in participants’ searches. Similar to prior work on social media, we found that exposure to fake news, in both searching and browsing, was concentrated among older Republicans. Within web search, we found that 1.2% of all participant searches returned a fake news domain, and 4.3% of news-related searches returned a fake news domain. Our findings highlight the cross-platform pattern of concentrated exposure to fake news, and shed light on the extent to which web search is a vector for exposure to fake news.

Authors

• David Lazer, Northeastern University

• Christo Wilson, Northeastern University

• Ronald Edward Robertson, Northeastern University