The Continued Influence Effect (CIE) - the persistent belief in misinformation despite corrections – is a challenge to interventions aimed at reducing the risk of harmful content, with social media playing a crucial role in spreading and correcting misinformation. This study explores the effectiveness of community notes, a crowdsourced fact-checking tool by X, in mitigating the CIE across neutral (e.g., food recall) and socially relevant content (e.g., gender quota). A German sample (N = 542) read narratives presented in tweets with one tweet containing misinformation. This tweet was later presented with community notes offering either a simple retraction or a correction with alternative explanation. For neutral content, both retraction types reduced the CIE, with alternative explanations being more effective. For socially relevant content, only corrections with alternative explanations reduced misinformation use. These results suggest community notes can counter misinformation but may be moderated by ideological factors in socially relevant contexts.

Believing misinformation can harm society, yet, what causes the formation and maintenance of these fringe beliefs is unclear. One reason may be people’s reliance on different information sources. In Study 1 (N=914), we examined the sources people self-reported valuing. Believers of misinformation, like climate change is a hoax (fringe), perceived personal investigations, case-control studies, case reports, and experts as high-value, while news, government, pundits, and social media were low-value. Conversely, non-believers of misinformation (mainstream) valued experts, systematic reviews, case-control, and cohort studies, whereas social media, pundits, friends, and news were low-value. Study 2 (N=852) examined whether both groups were more persuaded by arguments from their valued sources. We found fringe and mainstream believers were more persuaded and more likely to act according to sources deemed high-value than low-value. These results highlight the accuracy of self-report data on source persuasiveness and the potential influence of information sources in forming fringe beliefs.

Recent concerns about false memories or harmful beliefs related to fake news have prompted interventions to reduce its impact. One such intervention — fake content warnings —intuitively seems helpful, but has shown minimal effectiveness in research. To investigate why, two studies (each using brain scanning (fNIRS) and eye-tracking) with 36 and 40 participants respectively were conducted. Participants viewed real and fake articles with no warning, a fake content warning, or a similarly constructed irrelevant distractor text. Results revealed no impact of warning activity in the prefrontal cortex or temporoparietal junction, nor did warnings prompt participants to revisit the article. However, when analyzing fake news separately, warnings activated the right temporoparietal junction, indicating heightened suspicion. Despite this activation, participants' memory responses were unaffected by the warning. Thus, while warnings may trigger cognitive awareness, they appear insufficient to counteract false memory formation by themselves. Implications are discussed.

This paper explores belief updating in the context of misinformation, emphasising the influence of source reliability. Across four experiments, we examined how individuals revise beliefs when faced with retracted misinformation and varying source credibility. Experiment 1 revealed that participants discounted retracted misinformation and reverted to prior beliefs, challenging the Continued Influence Effect. Experiment 2 demonstrated that reliable sources facilitated greater rejection of misinformation than unreliable ones. Experiments 3 and 4 showed that participants appropriately weighted source reliability and penalised unreliable sources, regardless of the contradictor’s reliability. These findings indicate that belief updating can be rational and context-dependent when source reliability is emphasised. The results underscore the importance of integrating source credibility into misinformation correction strategies and have broad implications for countering misinformation in political, medical, and other applied contexts. This research further highlights the need for evidence-based approaches to enhance the effectiveness of misinformation correction efforts.