The Effect of Accuracy Instructions on Coronavirus-Related Belief Change Following Conversational Interactions

In a high-risk environment, such as during an epidemic, people are exposed to a large amount of information, both accurate and inaccurate. Following exposure, they typically discuss the information with each other. Here, we assess the effects of such conversations on beliefs. A sample of 126 M-Turk participants rated the accuracy of a set of COVID-19 statements (pre-test). They were then paired and asked to discuss these statements (low epistemic condition) or to discuss only the statements they thought were accurate (high epistemic condition). Finally, they rated the accuracy of the initial statements again (post- test). We do not find an effect of the epistemic condition on belief change. However, we find that individuals are sensitive to their conversational partners and change their beliefs according to their partners’ conveyed beliefs. This influence is strongest for initially moderately held beliefs. In exploratory analyses, we find that COVID-19 knowledge is predicted by trusting Doctor Fauci, not trusting President Trump, and feeling threatened by COVID-19, whereas believing COVID-19 conspiracies is predicted by trusting President Trump, not trusting Doctor Fauci, news media consumption, social media usage, and political orientation. Finally, we find that news media consumption positively predicts believing COVID-19 conspiracies, even when controlling for demographic variables including political ideology, and that this effect is not driven by a particular news network, but instead it is a general effect of news media consumption.