Effects of teaching games for understanding on students’ learning outcomes in physical education: a hierarchical meta-analysis of domain-specific effects

The purpose of this meta-analysis was twofold. First, to examine the effects of Teaching Games for Understanding (TGfU), compared with traditional instructional methods, on physical education (PE) students' learning outcomes in the cognitive, social-emotional, and physical domains. Second, to explore hierarchical trends spanning the three domains. The 20 primary studies (30 reports) involving 1478 participants showed heterogeneity among effect sizes (ESs). The random-effects model results showed a positive overall effect (g = 0.85, 95% CI = [0.50, 1.20], p < .001) on learning outcomes in the cognitive, social-emotional, and physical domains, suggesting that students learned substantially more via TGfU than traditional instruction. Additionally, initial analyses indicated a preliminary trend suggesting that the social-emotional domain may demonstrate a strong positive effect (g = 1.27, 95% CI [0.29, 2.27], p < .05), though this requires further large-scale validation. This was followed by the cognitive domain (g = 0.92, 95% CI [0.32, 1.52], p < .01), and the physical domain (g = 0.57, 95% CI [0.20, 0.94], p < .01). The overall effect and the hierarchy of domain-specific effects demonstrate TGfU's superiority over traditional instruction and offer clear guidance for its strategic implementation in diverse PE settings, supporting both holistic and domain-targeted student development.