An Experimental Study On Corpus-based Approach In College English Synonym Teaching

The present thesis has been conducted a corpus study from an experimental study. Thepurpose of the study is to test whether the corpus-based approach can help college studentsimprove their level of English synonyms significantly. At the same time, it is to test whetherthis new approach is more feasible and more efficient than the traditional one，which is on thebasis of dictionary or the teachers’ explanation.Bearing the purpose in mind, the author chose65non-English-major first-yearundergraduate students from Da Tong University as experimental subjects. Classes wererandomly assigned as the control class (class1) and the experimental class (class2). In thecontrol class the traditional approach was used, while in the experimental class thecorpus-based approach was used. In this thesis, the author exemplified and introduced indetail three corpus-based synonym discrimination methods. At the very beginning of theexperiment, the author carried out a pre-test to examine whether the students in two classes atthe same level or difference in learning and comprehending synonyms. The results manifestthat no significant difference existed between the two classes. After a semester finished, thestudents were asked to participate the post-test. The results show that in the experiment classthe subjects made great progress in learning English synonyms, while in the control class thestudents made little progress. In addition, the questionnaire and the interview were carried outin the experimental class. The questionnaire included9items and can be divided into fourparts. The interview was designed to investigate the students’ suggestion and evaluation forthe new approach.Besides, the researcher used Statistical Package for the Social Science (SPSS17.0) to analyze the raw data collected from the two tests and questionnaire.1）Independent SamplesT Test was adopted for analyzing the Pre-test or the Post-test between the two classes. In thesame test, the scores of two different classes were compared.2) Paired Samples T Test wasadopted for analyzing CC or EC between the two tests. It was to test whether the students ofthe same class made great progress in two different tests.3) Chi-square Test Frequencies wasadopted for testing significantly different frequencies between questionnaire items. Thereason why the author used the Chi-square goodness-of-fit test is to demonstrate the results ofthis thesis further.The results show that corpus-based approach can help college students improvesignificantly their level of English synonyms. However, each method has its advantages anddisadvantages. The best way is to add to the corpus-based method on the basis of theadvantage of the traditional one.