国内外二语研究元分析新进展(2010-2019)  

Recent developments of meta-analyses in L2 research(2010-2019)

作　　者：申云化[1] 张军 SHEN Yunhua;ZHANG Jun Lawrence

机构地区：[1]吉林大学公共外语教育学院,吉林长春130012 [2]西安交通大学外国语学院,陕西西安710049 [3]奥克兰大学教育学部,奥克兰1023

出　　处：《外语界》2020年第3期70-79,共10页Foreign Language World

基　　金：教育部国家留学基金委青年骨干教师出国研修项目[(2017)3192号];吉林大学哲学社会科学横向科研项目[(2018)社科字497号]的阶段性成果。

摘　　要：元分析是借助统计手段对某研究领域一手研究进行系统性综述的方法,其目的是充分考虑抽样误差和无显著统计学意义的结果,避免情感、记忆等主观因素引发的分析疏漏。本文回顾梳理2010-2019年SSCI和CSSCI来源期刊的二语研究元分析论文,得到以下发现:近十年元分析主题多元,涉及二语阅读、二语词汇、二语语法、二语口语、二语写作、二语策略、计算机辅助语言学习等;数据分析工具主要是CMA;多数研究合并标准均差Cohen’s d效应量,部分研究计算Hedge’s g来纠正小样本偏误,少数研究合并Pearson’s r效应量。本文的分析结果对二语研究元分析发展和外语教学具有一定启示意义。Meta-analysis is a quantitative approach to systematically reviewing primary studies in certain research domains.Its main purpose is to consider sufficiently sampling errors and non-statistically significant results,and to overcome analysis omissions caused by subjective factors such as emotion and memory.This paper sifts through the meta-analytic reviews of L2/FL studies published in SSCI and CSSCI indexed journals from 2010 to 2019.It is found that in the past ten years the themes of meta-analyses encompass a wide range of topics including L2/FL reading,vocabulary,grammar,speaking,writing,strategies and computer-assisted language learning.CMA has been a prevalent tool adopted by meta-analysts.A large proportion of meta-analyses aggregate the standardized mean difference Cohen’s d,some opt for Hedge’s g as effect size indicators to render corrections on the small sampling bias,whereas only a small number of studies address Pearson’s r.The results provide some implications for meta-analyses of L2/FL studies and foreign language teaching and learning.

关 键 词：元分析 二语研究 一手研究 效应量 新进展 

分 类 号：H319[语言文字—英语]