重探英語廣泛閱讀對第二語言習得之成效：後設分析研究

Abstract

在過去數十年中，廣泛閱讀時常被用來作為一種提升第二語言學習能力的教學方法。現今已有許多實徵研究致力於探討廣泛閱讀對於第二語言習得不同面向之成效，如學習者的閱讀能力、單字量以及寫作表現。為了統合眾多研究的結果，本研究對1983年至2020年的53篇實徵研究進行後設分析，總計萃取了248個效果量，且研究規模達6,520名參與者。此研究採用了隨機效果模式，結果顯示廣泛閱讀對於第二語言學習者的語言表現有一個輕度的效果量 (g = 0.407)。此外，調節變項的分析結果顯示，在十四個潛在調節變項中，有九個是重要的顯著調節因子，其中包含了學習者的母語、第二語言程度、教育程度、出版形式、廣泛閱讀執行時間長短、廣泛閱讀任務多寡、成效測量種類、書目多寡、閱讀媒材種類及閱讀量多寡。本研究旨在探究廣泛閱讀對於第二語言習得多面向之成效、提供教育現場及學者對於廣泛閱讀更全面性的洞見、指認出影響廣泛閱讀之關鍵調節變量，並進一步發展出更適合第二語言學習者的廣泛閱讀模式。最後，本研究亦指出一些研究限制。Over the past few decades, extensive reading (ER) has implemented as a teaching instruction to facilitate better second language (L2) learning outcomes. There have been numerous empirical studies investigating the effects of ER on different aspects of second language acquisition (SLA) such as reading, vocabulary acquisition and writing performance. To integrate the findings of existing ER studies, a total of 53 primary studies published between 1983 and 2020 were included and 248 effect sizes extracted with 6,520 participants were meta-analyzed. This study applied a random effects model to justify the effect sizes in the primary studies of ER research. The random-effects model demonstrated an overall robust effect size (g = 0.407), indicating that ER intervention is indeed effective. Nine out of the fourteen potential moderators were found to significantly relate to the variability in effects, including leaners’ L1, L2 proficiency level, education level, publication, length of ER intervention, accountability, outcome measure, library size, material type and number of words read. The purpose of this study is to explore the effectiveness of ER instructions in order to assist L2 practitioners and researchers in gaining more insights into ER, identifying several key factors that determine the effectiveness of ER, and developing a suitable teaching approach for L2 learners. In addition, based on the findings, the limitations of current study are also provided.