The Effects of Digital Game-based Learning Task in English as a Foreign Language Contexts: A Meta-analysis

英語為外語環境下數位遊戲式學習任務效益之後設分析

本研究採用後設分析，探討數位遊戲式學習在英語為外語環境下之效益為何。針對25篇研究進行分析後，發現中度且正面的效果量（固定效果模式：d = 0.695， p < .05；隨機效果模式：d = 0.777， p < .05）。其結果顯示，在英語為外語環境裏，數位遊戲式學習相較於傳統教學（如：文法翻譯法或聽說教學法）有效。本研究根據任務導向語言學習與數位遊戲式學習的關係發展一套編碼系統，進一步分析數位遊戲式學習的效益為何。該編碼系統包含三種任務特徵，而這三種特徵共有六項潛在調節變項。在這些潛在調節變項中，任務投入特徵裏的實驗時間及語言知識對數位遊戲式學習最有顯著效果。根據結果，本後設分析研究建議，以英語為外語學習者接受較長時間的數位遊戲式學習，對其程序性語言知識的發展助益較大。未來研究應探討其他潛在調節變項對數位遊戲式學習有何效益，得以更深入了解該領域的全貌。

A meta-analysis was conducted to investigate the effects of digital game-based learning (DGBL) in English as a foreign language (EFL) contexts. Twenty-five studies were included and they yielded a medium positive effect size (d = 0.695， p < .05 under the fixed-effect model; d = 0.777， p < .05 under the random-effects model)， suggesting DGBL to be more effective than traditional instruction such as grammar translation methods or audio-lingual methods in EFL contexts. A coding scheme was developed based on the relationship between task-based language learning and DGBL. Six potential moderating variables in three task characteristics were analyzed to further investigate the DGBL effectiveness. Among these potential moderating variables， the treatment duration and linguistic knowledge in task involvement characteristics were found to be influential in DGBL effectiveness. This meta-analysis suggests that digital games should be effective for the acquisition of procedural knowledge of EFL learners who receive the long treatment duration of DGBL. Future research should explore further distinctions in other potential moderating variables in order to provide a more detailed picture of what sorts of variables are influential and what sorts are not.