Insight into the accuracy of COVID-19 beliefs predicts behavior during the pandemic

Abstract

Susceptibility to COVID-19 misinformation--believing false statements to be true--negatively relates to compliance with public health measures. Here, we make the prediction that metacognitive insight into the varying accuracy of own beliefs predicts compliance with recommended health behaviors, above and beyond the accuracy of these beliefs. In a national sample of German citizens, we investigate metacognitive sensitivity, the degree to which confidence differentiates correct from incorrect beliefs. Bayesian and frequentist analyses show that citizens with higher metacognitive sensitivity were more likely to adopt recommended public health measures. Importantly, this benefit of metacognitive introspection into own beliefs held controlling for the accuracy of the beliefs. The present research highlights that insight into the varying accuracy of beliefs, rather than only the beliefs themselves, relate to citizens’ behavior during the pandemic

Navigating the COVID-19 Infodemic: The Influence of Metacognitive Efficiency on Health Behaviors and Policy Attitudes

“…Importantly, we also found that individual differences in metacognitive insight about COVID-19 knowledge were predictive of attitudes toward restrictions, behaviours during lockdown and vaccination intentions -even after taking into account covariates such as the accuracy of knowledge itself, education, political alignment or income. These findings, which are in agreement with a recent study on a German sample (Fischer et al, 2021), illustrates how metacognitive insight about one's own knowledge can affect decision-making in areas relevant for public health and social measures, in agreement with prior research in other domains (Hadar et al, 2013;Meyer et al, 2013;Parker et al, 2012).…”

Contested science: Individuals with higher metacognitive insight into interpretation of evidence are less likely to polarize

“…However, as both measures are concerned with citizens' reflective ability, one might expect both object-level reasoning (assessed with the CRT or numeracy) and metacognitive sensitivity (meta-d') to be somewhat related. Future research could investigate how object-level and metacognitive reflective abilities jointly shape citizens' reasoning about, or behavior in relation to, contested science (Fischer, Huff & Said, 2021a).…”

Do people know when they are good at spotting liars? – Metacognitive efficiency in Lie Detection

“…As mentioned above, metacognitive efficiency has been assessed in a wide range of different domains, for instance, in perception tasks: M ratio ~100% (Mazancieux et al, 2020;Palmer et al, 2014), in memory tasks: M ratio ~73% (Palmer et al, 2014) and M ratio ~50% (Mazancieux et al, 2020), as well as in knowledge tasks: M ratio ~100% for science knowledge, and M ratio ~50% for climate change knowledge (Fischer et al, 2019), M ratio ~86 % for COVID-19 knowledge (Fischer et al, 2021). To illustrate, for COVID-19 knowledge, participants did not use 14% of the evidence they used for the knowledge task when making confidence judgments.…”

“…Unlike previously used measures of metacognition, the measure we apply in this article (metacognitive efficiency) investigates the confidence-accuracy puzzle in lie detection for the first time free of response bias and independent of lie detection performance. We use M ratio as a universal measure of metacognitive efficiency, which has already been used in various areas of research such as perception, memory (Bang, J. W., Shekhar, M., & Rahnev, D., 2019;Folke, Ouzia, Bright, Martino, & Filippi, 2016;Hainguerlot, Vergnaud, & Gardelle, 2018;Mazancieux, Fleming, Souchay, & Moulin, 2020;Muthesius et al, 2022;Palmer, David, & Fleming, 2014;Reyes et al, 2020), and knowledge (Fischer, Amelung, & Said, 2019;Fischer, Huff, & Said, 2021). In re-analyses of twelve lie detection studies with N=2817 participants in total, we calculate M ratio by applying a hierarchical Bayesian approach implemented by Fleming (2017).…”