Play to learn: A meta-analysis on the impact of game-based learning on the language learning outcomes of young and middle-aged adults

Abstract

Background: One emerging teaching methodology is Game-Based Learning (GBL), which utilizes game mechanics to enhance individual learning experiences. This meta-analysis examined the differential impact of game-based learning compared to traditional learning methods among young and middle-aged adults.

Methods: Comparative studies that explores the difference between traditional and GBL among young and middle adults were obtained from three different databases. Literature search before April 2025 yielded a total of 1906 citations and study selection brought down that total number of studies to seven which were then subjected to meta-analysis. Risk of bias assessment (ROBINS-I and Rob 2.0) and sensitivity analysis were done to further assess the robustness of the model.

Results and Discussion: Meta-analysis using a random-effects model showed that improvements in the language learning outcomes favor GBL compared to traditional learning methods. A significant moderate effect size (Cohen’s d = 0.79, p < 0.00001) and minimal to moderate heterogeneity (I² = 37%) but non-significant (P = 0.13) variability (Cochran’s Q = 11.10) among the included studies were obtained. Risk of bias assessment and sensitivity analysis showed low to moderate concerns, which supports the robustness of the methodology and validating the comparison between the post-test scores of the selected studies.

Conclusion: The findings of the study suggest that GBL is more effective than traditional learning methods in improving the language proficiency of an individual, as indicated by meta-analysis results. This study highlights the effectiveness of GBL and its potential for real-world application in education settings.