Video-game based instruction for vocabulary acquisition with English language learners: A Bayesian meta-analysis

Abstract

This meta-analysis reviewed the literature on the efficacy of game-based learning in English as a second language vocabulary acquisition. A systematic search of the literature produced 19 studies that met inclusion criteria. Using Bayesian methods and 20 standardized-mean-difference effect sizes we assessed 1) overall mean effects and between-studies variability, 2) subgroup analyses (grade level, sex, hardware, game type, intervention length, Foreign Services Institute level, allocation, and publication type), and 3) risk of publication bias. The overall effect was moderately large, indicating favorability to the video-game based learning groups. We found evidence of effect-size heterogeneity with a large between-studies standard deviation, along with large Q and I2 values. Subgroup analyses produced mixed results when partially explaining effect-size variability. Furthermore, all publication bias tests indicated a low risk of publication bias. We found that video-game based learning can make a significant difference in promoting English vocabulary acquisition and that integrating entertainment video games into educational contexts can result in substantial student learning gains.