Latent profile analyses of explanations about the origins of SARS-CoV-2 within and across countries

Abstract

Research on belief endorsement typically estimates an average mean value to describe all individuals even though groups of individuals can have coexisting beliefs within the mind. This study employed a person-centered approach based on latent profile analysis (LPA) to investigate within- and between-population variation in the endorsement patterns of multiple explanations for the origin of SARS-CoV-2, including zoonotic, human-made, and supernatural explanations, across eleven countries. Results revealed seven latent profiles that varied in the number and type of explanations endorsed, as well as the degree of certainty associated with each. Among demographic characteristics, political orientation and religious affiliation predict whether particular types of explanations are endorsed or rejected. Latent profiles characterized by explanatory uncertainty and endorsement of conspiratorial explanations were less likely to wear masks and get vaccinated. We discuss the implications of person-centered analytical approaches for documenting variations in explanation endorsement and as predictors of behavioral health outcomes.