(De)construction on the road ahead: Item response theory and psychometric analysis of self-report measures of active and passive social media use

Abstract Active social media (SM) use involves engaging with other users (e.g., posting content, commenting), while passive social media use does not involve engaging with other users (e.g., scrolling). These constructs are fraught with measurement issues. This manuscript provides item response theory and updated psychometric analyses of two commonly used self-report measures developed nearly a decade ago: the Passive and Active Use Measure and the Multidimensional Scale of Facebook Use. This updated analysis provides a much-needed snapshot of their performance in a contemporary SM landscape. Item response theory provides insights into how these measures perform across the latent trait continuum. Findings indicate that existing measures are unreliable and provide inadequate information at all levels of active and passive use. Multidimensional scaling suggests that these issues may emerge from attempts to impose a latent trait framework (i.e., viewing active and passive use as deep, underlying traits) on constructs that are better conceptualized as a set of loosely related objective behaviours. In line with this, analyzing SM behaviours separately reveals disparate associations with wellbeing. We provide recommendations for researchers to increase reliability and validity in existing approaches to measuring active and passive use, as well as for researchers seeking to develop new measures.