The Effectiveness of Grammar Instruction

Abstract

Norris和Ortega針對文法教學所作廣泛的整點性分析研究，發現文法教學成效很大，也因此看起來強力支持文法教學在外語教學中的應用。本篇文章旨 在說明這些結果事實上指向相反的方向。文法教學能產生如此深刻而普遍的效應，其實是因為實驗者挑選的研究，大多是受極度人工化的測驗技術所支配(此領域的 研究普遍如此)，但是有些測量學生在實際使用外語能力(文法教學的效能)的研究卻發現其效應值遠低於前者所言，甚至是近乎零的效應。這些看似中肯的結果， 卻多半(或甚至全部)被與本全無關的因素而強化了。

Norris and Ortega’s extensive meta-analysis of research on grammar instruction found that the practice has strong effects, apparently providing strong support for its use in the foreign language classroom. The thesis of this paper is that the findings actually point in the opposite direction. The large overall effect size was due to the predominance in the sample (as in the field in general) of studies that relied on highly artificial testing techniques. Research that measured effects on learners’ practical ability to use the target language (the effectiveness of the instruction) produced much less impressive effect sizes, not reliably different from zero. These modest results owe much (possibly all) of their strength to extraneous factors.