Widespread misinformation about the COVID-19 pandemic has presented challenges for communicating public health recommendations. Should campaigns to promote protective behaviors focus on debunking misinformation or targeting behavior-specific beliefs? To address this question, we examine whether belief in COVID-19 misinformation is directly associated with two behaviors (face mask wearing and social distancing), and whether behavior-specific beliefs can account for this association and better predict behavior, consistent with behavior-change theory. We conducted a nationally representative two-wave survey of U.S. adults from 5/26/20-6/12/20 (n = 1074) and 7/15/20-7/21//20 (n = 889; follow-up response 83%). Scales were developed and validated for COVID-19 related misinformation beliefs, social distancing and face mask wearing, and beliefs about the consequences of both behaviors. Cross-lagged panel linear regression models assessed relationships among the variables. While belief in misinformation was negatively associated with both face mask wearing (B = −.27, SE =.06) and social-distancing behaviors (B = −.46, SE =.08) measured at the same time, misinformation did not predict concurrent or lagged behavior when the behavior-specific beliefs were incorporated in the models. Beliefs about behavioral outcomes accounted for face mask wearing and social distancing, both cross-sectionally (B =.43, SE =.05; B =.63, SE =.09) and lagged over time (B =.20, SE = 04; B =.30, SE =.08). In conclusion, belief in COVID-19related misinformation is less relevant to protective behaviors, but beliefs about the consequences of these behaviors are important predictors. With regard to misinformation, we recommend health campaigns aimed at promoting protective behaviors emphasize the benefits of these behaviors, rather than debunking unrelated false claims. Past studies Concern about misinformation predates the COVID-19 pandemic. Much attention has been paid to the proliferation of political misinformation, particularly in the wake of the 2016 U.S. presidential election (Guess et al., 2020), and how to effectively counter it (Cook et al., 2015). Misinformation is comparatively understudied in the health domain (Kreps & Kriner, 2020; Southwell et al., 2019). There is some evidence that belief in specific false health information is associated with undesirable outcomes, including lowered vaccination rates (Jolley & Douglas, 2014; Oliver & Wood, 2014), reduced contraceptive use (Thorburn & Bogart, 2005), and nonadherence to antiretroviral treatment (Bogart et al., 2010), although overall, the research is limited and the findings mixed (Nan et al., in press).
