Corrections of Scientific Misinformation are Impacted by Source Traits and More Effective When Labelled as a Fact-Check

Misinformation corrections are generally effective, but correction efficacy varies depending on different factors. In a pre-registered experiment, we tested whether framing (fact-check, scientific disagreement), source traits (integrity, expertise, benevolence), and/or the political valence of scientific claims (neutral, political) impacted beliefs and trust. Specifically, we examined belief in the original misinformation claim, belief in the correction, and trust in the original claimant and the refuter, along with broader trust in science. Correction source traits, particularly expertise and integrity, led to greater belief in the correction and trust in the correction source, and reduced belief and trust for the misinformation and misinformation source. Participants were also sensitive to the framing of the correction, with lower misinformation belief and greater correction belief when corrections were labelled as a ‘fact-check’ rather than a scientific disagreement. Political valence was also impactful, leading to reduced belief updating and lower trust in the refuter when misinformation was politically congruent. Participants were also less responsive to corrections for political than neutral claims. Overall, our results demonstrate that correction efficacy may be enhanced when corrections are made by experts without vested interests. Explicitly labelling corrections as fact-checks is also beneficial. Importantly, although political congruence impacted the efficacy of corrections, corrections were still effective for political claims. To enhance efficacy, we recommend labelling corrections as ‘fact-checks’ where appropriate, and to highlight the expertise and integrity of the correction source.