Trusting the tag: electrophysiological signatures of source credibility in fact-checking

As misinformation spreads rapidly on social media, understanding the efficacy of corrective interventions is a key challenge for cognitive science. We investigated behavioral and neurophysiological responses to fact-checking labels addressing medical and healthcare news. While EEG was recorded, 53 participants rated headlines on trust and willingness to share, both at baseline and after receiving ‘Truth’ or ‘Fake’ labels from either credible or low-credible medical sources. Behaviorally, credible source labels influenced trust and sharing intentions more strongly than low-credible labels. The label ‘Truth’ increased both measures only when attributed to the credible source, whereas the label ‘Fake’ decreased ratings regardless of source credibility. Neurally, we found no evidence for a feedback-related negativity (FRN/RewP) in response to labels. Instead, credible sources elicited a larger early P2-like response independent of label content, and single-trial analyses linked P2 and P3 amplitude to trust and sharing, most reliably when credible sources issued refuting corrections. These findings suggest source credibility is registered early and selectively shapes downstream evaluation of fact-checking labels.