Search

Prior work has found that early corrections are often more effective than corrections encountered sometime after exposure to misinformation. However, these studies have generally considered only brief delays between misinformation exposure and correction, and do not explore processing style as a potential moderator of correction timing. We conducted a two-wave online experiment randomly assigning participants to receive corrections either shortly after exposure or approximately one week later. We find that both immediate and delayed corrections influenced belief accuracy and policy support, but immediate corrections were no more influential than delayed corrections, contrary to earlier findings. In addition, processing style had no moderating effect on the influence of correction timing. [ABSTRACT FROM AUTHOR]

In a field experiment conducted during the 2012 general elections in the U.S. Nyhan and Reifler found that the threat of fact-checking deterred state legislators from making false or misleading statements. The current study presents a conceptual replication and extension of this influential study by utilizing a similar treatment that leverages a recent partnership between local media outlets and fact-checking organizations, assessing the effects of the treatment on the accuracy of legislators' statements on Twitter around the first impeachment trial of Donald Trump. Results provide limited evidence of the effects of our treatment on the accuracy of legislators' posts, even among legislators within media markets directly affected by this partnership. We conclude with a discussion of the theoretical and practical relevance of these results and avenues for future research. [ABSTRACT FROM AUTHOR]

The rise of negative partisanship raises the possibility that perceptions of what the partisan out-group believes on a factual matter could serve as a cue for one's own factual beliefs. In the current paper, we present the results of an online experiment using a sample of self-identified conservatives and liberals on Amazon's Mechanical Turk platform. Across several statements on various political issues, participants were randomly assigned to receive a corrective message, polling information about the factual beliefs of members from the partisan out-group, or both. We find that while the corrective message improved belief accuracy, information about the out-group did not influence belief accuracy either directly or by moderating the influence of corrections – even among those with the strongest antipathy toward the out-party. We discuss the implications of the results for the role of negative partisanship for misinformation and corrective messages. [ABSTRACT FROM AUTHOR]