Integrating the person-centered approach with the study of vaccine hesitancy: Applying latent profile analysis to identify vaccine hesitancy subpopulations and assess their relations with correlates and vaccination outcomes

Abstract

In scholarly and popular discussions of vaccine hesitancy, authors have repeatedly referred to different “types” of vaccine hesitant individuals; however, almost all modern research on vaccine hesitancy utilizes variable-centered approaches to identify the relation of variables rather than a person-centered approach to identify subpopulations, which suggests that a discrepancy exists between conceptual discussions and empirical research on vaccine hesitancy. For this reason, the current article conducts a latent profile analysis (LPA) on the dimensions of a well-supported vaccine hesitancy measure, which assess hesitancy towards vaccines in general. We also assess the relations of the resultant profiles (e.g., subpopulations) with relevant self-reported outcomes and correlates, wherein most of our outcomes are associated with COVID-19 and flu vaccines. Our LPA results support the existence of eight vaccine hesitancy profiles. The profile with the most unfavorable vaccination outcomes (e.g., willingness, receipt, and word-of-mouth) was associated with greater perceptions that vaccines cause health risks and unneeded when healthy; the profile with the most favorable vaccination outcomes was associated with low levels of all vaccine hesitancy dimensions. The other profiles produced a clear gradient between these two extremes. The profiles also differed regarding their standing on correlates, but the clearest difference was their relation with political orientation. Profiles with more unfavorable vaccination outcomes were associated with conservatism, whereas profiles with more favorable vaccinations outcomes were associated with liberalism. These results provide a new perspective for current understandings of vaccine hesitancy and open several avenues for future research.