Evaluating the Impact of Attempts to Correct Health Misinformation on Social Media: A Meta-Analysis

ABSTRACT

Social media poses a threat to public health by facilitating the spread of misinformation. At the same time, however, social media offers a promising avenue to stem the distribution of false claims – as evidenced by real-time corrections, crowdsourced fact-checking, and algorithmic tagging. Despite the growing attempts to correct misinformation on social media, there is still considerable ambiguity regarding the ability to effectively ameliorate the negative impact of false messages. To address this gap, the current study uses a meta-analysis to evaluate the relative impact of social media interventions designed to correct health-related misinformation (k = 24; N = 6,086). Additionally, the meta-analysis introduces theory-driven moderators that help delineate the effectiveness of social media interventions. The mean effect size of attempts to correct misinformation on social media was positive and significant (d = 0.40, 95% CI [0.25, 0.55], p =.0005) and a publication bias could not be excluded. Interventions were more effective in cases where participants were involved with the health topic, as well as when misinformation was distributed by news organizations (vs. peers) and debunked by experts (vs. non-experts). The findings of this meta-analysis can be used not only to depict the current state of the literature but also to prescribe specific recommendations to better address the proliferation of health misinformation on social media.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.

Notes

1. Rather than directly comparing exposure to correction of misinformation with a misinformation-only control condition, three studies in the sample (Bode & Vraga, 2018; Vraga & Bode, 2017; 2018) compared the effect of exposure to correction of misinformation with a neutral control condition. Similarly to fact-checking messages, these studies introduced the false statements and then immediately debunked them. When studies did not include a no-correction condition, the contrast between exposure to misinformation and exposure to correction was assessed with within-subject effects focusing on participants attitudes/intentions/behavior after exposure to misinformation and then following its correction (e.g., Study 3 in Kim, 2019; Kim et al., 2017; Study 2 in Lee, 2019).

2. When the source of misinformation was a news agency (e.g., Washington Post) but the message was shared by a private social media user (e.g., random person’s Newsfeed), the misinformation was attributed to the news agency (e.g., Bode & Vraga, 2015). This decision was based on the fact that (a) the true source of the message was the news agency and (b) the social media user was unfamiliar to the participants in the study.