Association for Public Policy Analysis & Management

There is too much false information circulating on social media convincing users of falsehoods, especially during elections. This research addresses the question of how to help social media users better resist false information on social media platforms, particularly around election seasons. Its principle audience are social media platform policy-makers interested in securing their platform and users from false information, but it also has application for educational, non-profit, and media audiences. Falsehoods shared online have created greater discord and confusion during elections, deepened social-political fissures, contributed to a trend of abandoning facts, and may erode democratic foundations. I will present on how social media platforms may help their users better discern and resist false information on their platforms. Three lines of effort focus my research. 1) I present a curated research guide, written for social media policy makers, applying social science findings to helping users resist persuasion from false information. I also carry out two experiments based on the most promising of these findings. 2) I perform a network and text analysis on Twitter data to find communities that share the greatest amount of false information. This proof-of-concept is designed to scale and allow social media platforms to identify vulnerable networks on which to triage their interventions. Finally, 3) I conduct a survey experiment testing inoculation, which inoculates against false information by forewarning of specific tactics or arguments in false messages. I test if inoculation also works against messages intended to produce over-whelming emotional over-responses that can cloud or overpower cognitive faculties. I will present interim findings on these lines of effort and give tentative suggestions on how social media companies may help social media users better resist false information, most especially around elections.