The influence of social media on body image has been extensively studied over recent years, yet findings are often inconsistent, with many studies reporting small or negligible effects. Such inconsistencies have raised important theoretical and methodological questions about what null effects truly mean. Using social media and body image research as an exemplar, this paper explores the meaning of null effects in the association between these constructs. We outline how opposing mediators — such as receiving social validation versus negative feedback — may cancel one another out, producing null statistical effects despite underlying causal processes. We propose an outcomes balance perspective of risk, whereby the aggregate balance of positive and negative mediator effects is needed to fully understand underlying causal relationships. We present four key recommendations for researchers to apply this outcomes balance perspective : (1) reconceptualise risk as a balance of positive and negative mediating effects, with mechanisms of change reflected in shifts in this risk balance over time; (2) routinely test for mediation even when total effects are small; (3) employ within-person and intensive longitudinal designs to assess individual differences in risk balance; and (4) triangulate findings across study design types to better understand boundary conditions for the effect one is interested in exploring. This outcomes balance perspective may encourage re-examination of established lines of argument in the body image literature. Indeed, arguments raised here are generalisable across psychological research and should thus interest all readers. • Social media-body image links are often small and inconsistent. • Opposing mediators may cancel out true effects. • We propose an outcomes balance perspective of risk. • Mediation testing may be needed even when direct effects are small. • We offer key recommendations to advance theory and research design.
Jarman et al. (2026) studied this question.