What have we learned, and what is yet to be learned about social media populism? A scoping review and meta-research

Abstract

The rise of populist communication on social media has attracted increasing scholarly attention since the late 2010s. However, the global research landscape on this phenomenon has not yet been comprehensively mapped, and there is also an absence of meta-research in the field to identify the quantitative design challenges scholars face. This article fills those gaps and suggests that the field should refine research designs. The study employs a two-stage empirical strategy. First, it conducts a scoping review of 122 peer-reviewed quantitative studies to synthesize major findings on populist communication on social media platforms. Second, it examines methodological and contextual factors associated with confirming, partially supporting, or rejecting hypotheses in a subset of 62 studies with Quasi-Poisson regressions. The results suggest that none of the examined variables affect the number of supported or partially supported hypotheses per study. However, multicountry designs, the inclusion of noncampaign periods, and longer timeframes are associated with a higher likelihood of rejecting hypotheses. These patterns point to an asymmetry in the evidence base: research design features predict hypothesis rejection more consistently than they predict hypothesis support. Consequently, we outline recommendations that can address these issues in future research.