Searching Online To Evaluate Misinformation Can Increase Its Perceived Veracity

• In Event: Correction! Can We Mitigate Misperceptions and Incivility?

Abstract

With misinformation introducing challenges in domains ranging from public health to democratic governance, significant attention has been paid to understanding the spread of and belief in online misinformation, with a particular focus on social media platforms. Comparatively little focus has been placed on other central features of the digital information ecosystem, such as online search engines. As the cost of producing and distributing information online has fallen and the sheer volume of information on the internet has risen, reliance on traditional gatekeepers has been significantly reduced, leaving search engines to fill the role of 21st-century gatekeepers by sorting and validating online content for the public. In this new role, search engines have become influential in users' political knowledge and public opinion. The growing reliance on search engines for information verification has been encouraged by social media companies, civil society, and government agencies, all of which have invested in campaigns to encourage online users to research news they believe may be suspect through online search engines, such as Google, with the goal of reducing belief in misinformation. Although search engines play a key role in how people evaluate information online, we know little about how searching online to evaluate misinformation (SOEM) impacts belief in misinformation.

While conventional wisdom suggests that searching online when evaluating the veracity of misinformation would reduce belief in it, there is little empirical evidence with which to evaluate this claim. Across five pre-registered experiments, we present consistent evidence that online search to evaluate the truthfulness of false news articles increases the probability of believing misinformation in a multitude of different contexts and experimental designs. In one between-subjects study we tested the effect of SOEM on belief in highly popular false/misleading directly after an article's publication (within 48 hours) using a randomized controlled trial where participants were randomly assigned to one of two groups: the treatment group, in which respondents were encouraged to search online to help them evaluate randomly assigned news articles; or the control group, in which they were not encouraged to search online. In a second within-subjects study, we tested whether the effect of SOEM can change an individual's evaluation after they had already assessed the veracity of a news story and anchored themselves to an answer. Our third study measured the effect of SOEM months after publication, rather than directly after publication, while Study 4 measured the effect of SOEM on recent news about a salient topic with significant news coverage (in our case, the Covid-19 pandemic). Our final study was similar to the first study, but we monitored how individuals search online to evaluate misinformation. In all of these studies, we found that encouraging individuals to search online increased the number of individuals who rated false news articles as true. In some of these studies, we found this effect to be up to 26%.

To shed light on this relationship, we combined survey and digital trace data, collected using a custom browser extension, to investigate the mechanism. We find that the search effect is concentrated among individuals for whom search engines return low-quality information. Our results demonstrate that those who search online to evaluate misinformation risk falling into data voids, or informational spaces where there is plenty of corroborating evidence from low-quality sources. Analysis of online search strategies showed that those who used the headline of the false article being evaluated or distinct terms used by the false article as a search query were much more likely to fall into these data voids and be exposed to unreliable information by search engines.

Finally, to fully evaluate the effect of recommending individuals to go search online, we also measured the effect of searching online to evaluate true news on belief in true news. We find that the search effect on belief in true news is similar to the search effect on belief in false/misleading news, but when we subset the search effect by the quality of the source, we find a strong search effect on belief in true news from low-quality sources, but no search effect on belief in true news from mainstream sources.

All of the findings in this paper highlight the need for media literacy programs to ground their recommendations in empirically tested interventions and search engines to invest in solutions to the challenges identified here. Building off our findings, media literacy programs should supplement a general recommendation to search online with more targeted techniques that teach individuals how to use proper search terms and identify quality news sources.

Authors

• Kevin Aslett, University of Central Florida

• Zeve Sanderson, New York University

• William Godel, New York University

• Nathaniel Persily, Stanford University

• Jonathan Nagler, New York University

• Joshua A. Tucker, New York University