Misconceptions about mental health can have negative effects on therapy, education, and social interactions. Misconceptions about mental health can be formed through misinformation being spread online from a variety of sources. The current study manipulates and examines the effects of social media users’ justification for knowing on participants’ perceived credibility and knowledge revision. Justification for evidence was manipulated within subjects. There were 3 types of justifications: personal experience, professional experience, or no justification. To test the effects of evidence justification, we used two dependent variables: perceived credibility and knowledge revision. MTurk participants (n = 111) completed pretest assessments regarding mental health and general science knowledge. They then read 11 experimenter-derived Twitter threads, each containing a misconception, two tweets with a refutation, and a statement of justification for the refutation. After each Twitter thread, participants were asked to rate the perceived credibility of the refutation texts. Participants were later given a posttest to measure knowledge revision as well as a series of questions that measured epistemic belief systems. We hypothesized that participants would be more likely to revise their misconceptions when the justification was personal expertise compared to when the justification was professional expertise or no justification is given. The findings did not support these hypotheses, instead indicating that the highest perceived credibility rankings came from professional expertise while knowledge revision occurred in all conditions.

I studied how hostile and benevolent language influences one’s ability to change their misconceptions. Participants were less likely to revise their misconceptions when reading tweets with hostile language than those exposed to benevolent language, which stresses adopting a neutral or benevolent tone to increase the likelihood of successful revision. This may be due to a shift of memory resources from the less engaging Tweet information to the more engaging, evocative hostile language.