文法糾正對第二語言寫作效益之後設分析
No Thumbnail Available
Date
2014-09-??
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
英語學系
Department of English, NTNU
Department of English, NTNU
Abstract
本研究透過後設分析將文法糾正對第二語言寫作成效之量化研究進行統整。雖然過去已有不少後設分析研究探討文法糾正之效益為何，但是Truscott (2007) 是過去後設分析研究裡唯一採用較嚴謹的設計條件進行分析的研究，而他宣稱文法糾正對於學生未來的寫作表現有小的負面影響。然而，他的後設分析卻只分析5 篇研究，並且沒有考量相關的中介變項。本後設分析研究之目的為（一）增加符合設計條件的研究，即擴增Truscott (2007) 的分析範疇，從5 篇增至26 篇研究，以及（二）考量中介變項的分析來檢視Truscott 的結論。本研究主要結果發現中介變項裡的回饋範圍（集中與非集中回饋）對學生的文法正確性有顯著的影響。不論集中與非集中回饋研究進行分析時，結果呈現小而正面的效果量 (d = 0.444)。而只針對集中回饋研究時，則呈現大且正面的效果量 (d = 0.851)。根據本研究結果，我們建議（一）與其混合所有文法糾正的種類，來回答文法糾正是否有效這種廣泛的問題，不如承認現今研究的確證實有些方法或狀況是有效的；（二）在未來的研究裡致力於探討文法糾正效益之關鍵因素為何。
This meta-analysis synthesizes quantitative studies on the effectiveness of corrective grammar feedback on L2 writing. Although previous meta-analyses have been conducted to explore correction effects across studies, Truscott (2007) is the only previous meta-analysis on grammar correction that adopts more strict inclusion criteria, and he found that grammar correction had a small negative effect on students’ future writing. However, that meta-analysis covered only five studies and considered no potential moderating variables. The two purposes of the present meta-analysis are (1) to examine the degree of effect of grammar correction on L2 writing by expanding the pool of qualified studies included from the five studies of Truscott (2007) to 26 studies, and (2) to examine the effects of moderating variables on the effectiveness of grammar correction to see if Truscott’s conclusions still stand. Our most important finding is the striking influence of the moderating variable of the scope of feedback. Studies using focused feedback showed a large positive effect size (d = 0.851). This fell to a small positive effect (d = 0.444) when we conflated focused feedback studies with studies that used unfocused feedback. Our results suggest that rather than conflating all sorts of grammar correction practices to seek an answer to the broad question of whether grammar correction works, it is more useful to acknowledge the evidence now that some methods or conditions work while others do not, and to concentrate future research on learning more about the difference.
This meta-analysis synthesizes quantitative studies on the effectiveness of corrective grammar feedback on L2 writing. Although previous meta-analyses have been conducted to explore correction effects across studies, Truscott (2007) is the only previous meta-analysis on grammar correction that adopts more strict inclusion criteria, and he found that grammar correction had a small negative effect on students’ future writing. However, that meta-analysis covered only five studies and considered no potential moderating variables. The two purposes of the present meta-analysis are (1) to examine the degree of effect of grammar correction on L2 writing by expanding the pool of qualified studies included from the five studies of Truscott (2007) to 26 studies, and (2) to examine the effects of moderating variables on the effectiveness of grammar correction to see if Truscott’s conclusions still stand. Our most important finding is the striking influence of the moderating variable of the scope of feedback. Studies using focused feedback showed a large positive effect size (d = 0.851). This fell to a small positive effect (d = 0.444) when we conflated focused feedback studies with studies that used unfocused feedback. Our results suggest that rather than conflating all sorts of grammar correction practices to seek an answer to the broad question of whether grammar correction works, it is more useful to acknowledge the evidence now that some methods or conditions work while others do not, and to concentrate future research on learning more about the difference.