A Corpus-based Comparative Study On The Use Of Lexical Chunks In My Opinion By Chinese EFL Learners

In recent years, researches on the use of lexical chunks have become a hot topicin the field of SLA, which have aroused more and more attention from both appliedlinguists and corpus linguists. As one of the core approaches, lexical teachingapproach intends to integrate the learning of knowledge with the improvement oflearners’ communication abilities. Lexical chunks are the basic units both forlanguage teaching and learning. Through literature review of the previous researcheson lexical chunks, it is found that previous researches mostly adopted traditionalresearch methods, such as speculative methods, text analyzing methods andexperimental research in controlled classes; however, in recent years, with the rapiddevelopment of the new independent discipline--corpus linguistics, research methodson the use of lexical chunks have changed gradually, which have acquired certainachievements based on corpus methods both in depth and breadth; yet few researchesare found on the use of the three-word lexical chunks in my opinion, which isfrequently used as the discourse marker by Chinese EFL learners and has greatsignificance both on function and discourse structure in the process of learners’interlanguage development.Therefore, based on corpus methods, a comparative study is made, which aims toinvestigate the use of the high frequency lexical chunks in my opinion by ChineseEFL learners in college English writing. Consequently, there are four researchquestions that need to be addressed in this study as follows:(1) What are the generalcharacteristics on the use of chunks in my opinion by Chinese EFL learners?(2) Whatare the significant differences between EFL learners and native speakers?(3) What are the differences among Chinese EFL learners at different levels?(4) What are thereasons that may cause these differences?In order to seek answers to the four research questions, this study adoptedWECCL2.0as the main corpus, meanwhile the sub-corpora involving students’composition texts of Grade1-4are generated; and other two corpora used forcontrastive analysis are ICLE and LOCNESS) The research procedures in this studyto deal with the corpora and data analysis are respectively: Sub-corpus Generator isfirstly used to generate the four sub-corpora separately for Grade1-4; then AntConcTool is adopted to generate a word list including the total number of word types andword tokens, thus the type-token ration of each corpus can be calculated; afterwardsAntConc Tool is employed to calculate the frequency of the lexical chunks in myopinion in each corpus, by means of extracting the concordance hits that contain thelexical chunks in my opinion, and then calculate the normalized frequency taking thetimes of occurrence per10,000words as a unit; finally Chi-square Calculator is usedto verify whether there are significant differences between EFL learners and nativespeakers, as well as differences among Chinese EFL learners at different levels.Based on the above research methods, major findings of the present study can besummarized as follows:(1) The general characteristics can be reported as ChineseEFL learners tend to overuse the high frequency lexical chunks in my opinion withsingle variety;(2) From chi-square test comparisons point of view, it is found thatthere exist significant differences between EFL learners and native speakers;(3) Bymeans of the statistical analysis of the type-token ratio, the normalized frequency andchi-square test comparisons, it is found that Chinese learners at different levels haveextremely significant differences. Low level learners tend to overuse the lexicalchunks in my opinion in college English writing, which is exceedingly far more thanthat of high level learners;(4) The possible reasons contributing to the significantdifferences have been proposed that the overuse of lexical chunks in my opinion lie inthe insufficient mastery of lexical chunks, intentionally avoiding mistakes, negativetransfer of L1, learners’ limited language ability, so on and so forth.Based on data analysis and the findings of the present study, several strong pedagogical implications are provided as follows: English teachers are suggested toadopt corpus in language teaching. Through the help of concordance lines in corpus,teachers can demonstrate various usages of concordance or collocations to studentsprovided by the corpus materials, which is helpful for the learners to strengthen theirconsciousness of using authentic lexical chunks as native speakers; meanwhile,English teachers are supposed to shift much more focus on training students’awareness to use lexical chunks in their writing and teach students how to effectivelyrecognize lexical chunks, in which way can students increase the fluency andaccuracy with a large number of lexical chunks. Finally, some recommendations forfurther research are presented.