(Don't fear) the factors: An item-level meta-analysis of the fear of COVID-19 Scale's factor structure and measurement invariance.

Abstract

The global COVID-19 pandemic saw marked research and clinical interest in evaluating pandemic-related distress, namely fear and anxiety regarding infection and death. The most widely used and earliest developed measure of COVID-19 distress is Ahorsu et al. (2022) seven-item Fear of COVID-19 Scale (FCV-19S). To investigate the factor structure and measurement equivalence of the FCV-19S, we conducted an item-level meta-analysis synthesizing 1155 effect sizes across k = 55 independent samples comprising N = 71,161 individuals. We found that a two-factor measurement model comprising a four-item Emotional factor and a three-item Psychosomatic factor exhibits better fit than the originally proposed single-factor measurement model. Moreover, the bidimensional FCV-19S exhibits partial scalar/strong invariance across the general population, healthcare workers, schoolteachers, and university students as well as partial metric/weak invariance across samples from Bangladesh, China, Japan, Pakistan, Poland, and Portugal. Despite the theoretical and practical implications of these findings, more primary research across a wider range of sample types and countries is undoubtedly needed for further evaluation of the FCV-19S's psychometric properties and generalizability.