Individual differences in vulnerability to misinformation when forming impressions of political candidates

Previous work has shown that false information continues to affect decision making even after being corrected, a phenomenon known as “continued influence effects” (CIEs). Here we demonstrate that vulnerability to CIEs varies systematically between individuals as a function of demographic and psychometric variables. We developed a set of mock social media accusations, refutations, and control stimuli targeting fictional political candidates. We observe robust within-participant CIEs: candidates targeted by corrected accusations evoke lower feeling thermometer ratings than candidates not targeted by accusations. Individuals who rely more on intuitive feelings show larger CIEs, whereas those who score higher on digital literacy show reduced CIEs. These results suggest that analytic thinking plays a role in countering the continued influence of corrected misinformation. Interestingly, older adults appear less vulnerable to CIEs than their younger counterparts, a counterpoint to prior findings that older adults share more false content on social media. We find no effect of political orientation on CIEs despite its influence on explicit identification of misinformation. Finally, after a two-day delay, accusation stimuli are remembered better than refutations, suggesting that accusations stimulate higher-priority processing than refutations, potentially due to stronger emotional arousal. Our results suggest that analytic thinking and digital literacy could be protective when people must judge political candidates targeted by refuted false information.