The Effect Of Written Corrective Feedback On The Overall Quality And Linguistic Accuracy Of Chinese Students' ESL Writing

Researchers are devoted to investigating the effectiveness of written corrective feedback because error correction plays an important role in second language writing.More specifically,researchers are aiming to examine which kind of written corrective feedback could promote more linguistic accuracy.Numerous studies have suggested that written corrective feedback could have a positive effect on improving linguistic accuracy which could facilitate second language acquisition(Ferris&Roberts,2001;Chandler,2003;Sachs&Polio,2007;Beuningen et al.,2012;Wang&Liu,2012;Chen et al.,2013;Li,2013;Shintani&Ellis,2013;Hou,2015).Researchers even conducted studies to test whether there is a certain kind of written corrective feedback or written corrective feedback combination leading to more linguistic accuracy.Shintani and Ellis'(2013)study which focuses on the indefinite article,indicated that metalinguistic explanation was more effective than direct feedback,but its effect was not durable.Shintani et al.(2014)suggested that both metalinguistic explanation and direct corrective feedback could improve the linguistic accuracy of the unreal conditional,however,direct corrective feedback was more effective.Bitchener(2008),Bitchener and Knoch(2008,2009a,2009b,2010),and Jiang and Chen(2014)all carried out studies on articles,and these studies all demonstrated that the effects of metalinguistic explanation were equivalent to that of direct corrective feedback.In addition,researchers have compared the effects of direct corrective feedback with that of indirect corrective feedback.Lalande's(1982)study illustrated that coded corrective feedback led to more linguistic accuracy,which is more effective than direct corrective feedback.Jiang and Chen(2013)investigated the effects of metalinguistic explanation and indirect corrective feedback on the generic NPs,whose results manifested that the effect of metalinguistic explanation was more effective than that of indirect corrective feedback.Hou(2018)examined the effects of metalinguistic explanation on linguistic accuracy.The result showed that error code and metalinguistic explanation were more effective and durable than direct corrective feedback.In addition,metalinguistic explanation had a better effectiveness than error code.Among these comparisons of written corrective feedback,metalinguistic explanation has attracted much attention,however,there were few studies which concern direct corrective feedback,indirect corrective feedback and metalinguistic explanation.What's more,linguistic structures,such as articles,investigated in these studies were not complex enough.Therefore,the present study,focusing on the rule-based errors(articles and third person singular)and item-based errors(prepositions and part of speech),examined the effects of indirect corrective feedback,direct corrective feedback and metalinguistic explanation on the overall writing quality and the linguistic accuracy.The present study aims to investigate the following three research questions:(1)Is written corrective feedback(WCF)more effective in improving the overall writing quality than writing with no feedback?If yes,is there any difference in the effect of direct corrective feedback(DCF),indirect corrective feedback(ICF)and metalinguistic explanation(ME)on learners' improvement of the overall writing quality?(2)Does WCF lead to more accurate use of rule-based structures or item-based structures?If yes,is there any difference in the effect of DCF,ICF and ME on learners5 improvement of linguistic accuracy of rule-based structures or item-based structures?(3)Are different types of WCF effective in promoting the accurate use of articles,third person singular,prepositions and part of speech?If yes,is there any difference in the effect that DCF,ICF and ME have on the longer term effect of the four target structures?In the present study,165 subjects,who were the second-year non-English major students,were assigned into four groups:the control group(n=31),the indirect corrective feedback group(n=44),the direct corrective feedback group(=44),and the metalinguistic explanation feedback group(n=46).The effects of WCF on improving the overall writing quality and linguistic accuracy were investigated by using a"pretest-treatment-immediate post-test-delayed post-test" design,and by employing both qualitative and quantitative methods.Results of the present study suggest the following three major findings:(1)The three experiment groups which receive WCF on grammar,all gain more scores in the two post-tests,which indicates that WCF is effective in improving the o'verall writing quality.At the same time,the overall writing quality of the control group improves significantly in the two post-tests,which suggests that writing practice can help increase scores of learners' writing.However,compared with the three experiment groups,the effect of writing practice is not as effective as the three types of WCF.(2)WCF is effective in promoting the linguistic accuracy of both rule-based structures and item-based structures.More precisely,the present study indicates that for rule-based structures,ME has a more positive and significant effect of helping improve the linguistic accuracy than ICF and DCF in either the short or longer term,which suggests an advantage for ME over ICF and DCF in new pieces of writing.For the item-based structures,ME is the most effective feedback in the short term,while in the longer term,there are no significant differences between ME and DCF.This means ME and DCF have a similar long-term effect of helping improve the linguistic accuracy of item-based structures.(3)WCF has a positive effect on promoting the accurate use of the four target structures.More specifically,for articles,ME has a more significant long-term effect than ICF and DCF on improving the linguistic accuracy.For third person singular,ME has a better effect than ICF and DCF on improving the linguistic accuracy in the short term.But in the longer term,no significant differences are found among the three different types of WCF.For prepositions and part of speech,ME and DCF outperform ICF on improving the linguistic accuracy in either the short or longer term.Based on the findings,the present study provides some empirical evidences on the effectiveness of WCF or WCF combination on improving learners' overall writing quality and linguistic accuracy.In addition,the present study offers some suggestions for future research.Firstly,teachers should provide WCF for learners and they should continue providing feedback in the longer term.Providing WCF helps learners to notice the gap between their interlanguage and the target language,thus promoting acquisition.Secondly,the results of the present study suggest that different linguistic categories should be provided with different WCF.Therefore,when providing WCF,teachers should take the types of structures into consideration and provide corresponding feedback rather than give the same feedback to all structures,thereby making full use of teachers' WCF.Finally,The results also suggest that focused feedback plus metalinguistic explanation is more effective for rule-based structures.Therefore,in actual teaching situation,teachers may consider adopting a feedback combination instead of a single feedback in their teaching process.