The Effects of Digital Technologies on Physical Education Outcomes: A Systematic Review and Meta-Analysis

Background. Physical education faces growing pressure to integrate digital technologies amid declining student engagement and motivation. Despite increasing adoption, systematic evidence on the effectiveness of these tools across outcome domains remains limited, particularly regarding the conditions under which they are most beneficial. Objectives. This systematic review and meta-analysis examined the effects of digital technologies on physical education outcomes, investigating the overall magnitude of effects, domain-specific effects (physical, motivational-affective, cognitive), and moderators of heterogeneity. Materials and Methods. Systematic searches of four databases (Scopus, Web of Science, ERIC, SPORTDiscus) identified experimental and quasi-experimental studies published between 2015 and 2025. Twelve studies meeting the inclusion criteria yielded 64 effect sizes from 1,346 participants. Random-effects meta-analyses with REML estimation were conducted to calculate pooled Hedges' g effect sizes, supplemented by prediction intervals to account for extreme heterogeneity. Results. The overall pooled effect was g = 0.745 (95% CI [0.590, 0.900], p < .001), with domain-specific estimates of g = 0.854 (physical), g = 0.616 (motivational-affective), and g = 0.936 (cognitive). However, extreme heterogeneity across all models (I² = 99.98%, τ² = 0.373) renders these pooled estimates unstable and non-generalizable. The 95% prediction interval (PI) [−0.471, 1.961] indicates that the true effect in any given implementation context may range from moderately negative to very large positive, reflecting the high context dependence of the observed effects. Conclusions. Digital technologies show potential for enhancing physical education outcomes; however, the wide prediction intervals and extreme heterogeneity preclude any universal claim of effectiveness. The practical impact of a specific implementation critically depends on technology type, pedagogical integration, teacher expertise, and educational context. Future research should prioritize identifying the conditions under which digital technologies are most and least effective, rather than estimating average effects alone.