|
Jason Reifler
University of Southhampton, UK @jasonreifler Online Consumption of Voter Fraud Misinformation and Approaches to Correcting Voter Fraud Misperceptions Unreliable information about election fraud is abundant online and threatens to undermine confidence in American elections. For a representative sample of Americans, we measure online exposure to content about fraud and about the January 2021 insurrection at the U.S. Capitol from both trustworthy and untrustworty sources, comparing exposure to both types of content across those who supported Joe Biden versus Donald Trump in 2020 and according to respondents' conspiracy predispositions. We evaluate the correlations between information flows and beliefs about specific types of election fraud, and we measure how those beliefs changed during and after the 2020 election. Finally, we report results from a survey experiment conducted after the 2020 election showing that corrections debunking claims of widespread voter fraud can reduce misperceptions. However, our data on media consumption show that exposure to corrections was relatively rare, especially among those who most overestimate the prevalence of election fraud. The results underscore the importance of increasing exposure to trustworthy online news content and fact-checking among those most likely to hold misperceptions about the integrity of American elections. |
|
Onur Varol
Sabanci University, Turkey @onurvarol Analyzing User and Group Behavior during Elections: Coordination, Bots and Misinformation The unprecedented growth in social media use and large-scale information collection pose new threats, but also offer new opportunities. Modeling and managing complex interactive systems requires the analysis of social and technological signals to gain new insights into human society and individual behavior. Online social networks play an essential role in our access to information and are a good proxy for studying population-level behavior patterns and individual-level predictions. In this talk, I will present my research analyzing various account behaviors, from social bots spreading misinformation to coordinated activities during political campaigns and elections. I will also present the dataset we collected and the system we developed for the #Secim2023 project, which we used to study the recent Turkish presidential elections. |
|
Joana G Sá
University of Lisbon, Portugal @mjoanasa Who Believes in Disinformation? Complementary Approaches to Studying Human and Algorithmic Bias The unprecedented growth in social media use and large-scale information collection pose new threats, but also offer new opportunities. Modeling and managing complex interactive systems requires the analysis of social and technological signals to gain new insights into human society and individual behavior. Online social networks play an essential role in our access to information and are a good proxy for studying population-level behavior patterns and individual-level predictions. In this talk, I will present my research analyzing various account behaviors, from social bots spreading misinformation to coordinated activities during political campaigns and elections. I will also present the dataset we collected and the system we developed for the #Secim2023 project, which we used to study the recent Turkish presidential elections. |
