The effect of embodied learning on students' learning performance: A meta-analysis

Embodied learning has attracted considerable attention in recent years. However, there is no academic consensus on whether embodied learning effectively enhances students' learning performance.

This study aims to examine the overall effect of embodied learning on students' learning performance through a meta-analysis. It also seeks to explore variations based on moderators such as discipline, educational level, experiment period, sample size, region, learning approach, embodied level and type.

A meta-analysis was conducted on 46 studies (66 effect sizes) published between 2010 and 2025. These studies were analyzed to calculate the overall effect size (Hedges' g) and explore potential moderating variables.

The results found that: 1) Embodied learning has a moderately positive effect on students' learning performance (g = 0.406, 95%CI [0.264,0.548]), with no significant differences across regions; 2) The effect of embodied learning is greater in the humanities compared to other disciplines (e.g., math); 3) Compared to other educational levels, embodied learning has the greatest impact on high school students' learning performance; 4) The impact of embodied learning is significantly greater during a one-term experiment period than other periods; 5) Compared with other sample sizes, the embodied intervention group with more than 50 participants has the best effect on their learning performance. 6) Embodied learning in small groups has a greater effect on students' learning performance than other learning approaches; 7) High-level embodied learning has a more significant effect on students' learning performance than low-level embodied learning; and 8) Active embodied learning has a greater effect on students' learning performance than passive embodied learning. These findings provide valuable insights for future practice and research on embodied learning.