Abstract
Hostility toward Jewish and Muslim Americans remains a persistent feature of the contemporary United States, intensified by recent geopolitical conflict and amplified through today’s fragmented information environments. Although prior research demonstrates that factual corrections can reduce antisemitic and Islamophobic sentiment, the boundary conditions of these effects remain poorly understood. This project proposes a large-scale experiment to identify when, how, and for whom correction-based interventions meaningfully reduce intergroup hostility. We recruit individuals who express unfavorable attitudes toward Jewish or Muslim Americans and randomly assign them to six conditions: no information, a standard fact-checking intervention, fact-checking preceded by a prompt highlighting potential personal error, fact-checking accompanied by an attack on the source’s credibility, fact-checking accompanied by an attack on the validity of the corrective content, and a compassion-driven perspective-taking video. By systematically comparing these conditions, the study isolates the extent to which epistemic threat, defensive reactions, and credibility disputes undermine the positive effects previously observed in controlled settings. The results will clarify the limits of information-based interventions, adjudicate competing claims about motivated reasoning and belief updating, and provide researchers and practitioners with empirically grounded guidance for reducing antisemitism and Islamophobia in diverse social contexts.
