Cohesive Features In Chinese College Students' English Compositions

The present study is intended to provide a relatively comprehensive account of the cohesive features in Chinese college students' English compositions. 1) How frequently is each cohesive tie used in the student compositions? 2) What is the relationship between the number of cohesive ties used and the quality of the student compositions? 3) Which cohesive tie(s) is/are the better predictor(s) of the quality of the student compositions? 4) What are some of the common features in the student compositions in the use of each cohesive tie? 5) What are the most common cohesive errors in each cohesive tie in the student compositions? These are the questions probed and discussed in this study.One hundred sample compositions were selected systematically from 4845 compositions written by college non-English majors who took part in the 2004 National Entrance Test of English for M.A./M.S. Candidates and applied for a postgraduate program in Lanzhou University. They were rated holistically by 2 raters and a paired t-test of significance was conducted to compare the difference in their ratings. The average of their scores was the final result for each student composition and was used for data analysis. Halliday and Hasan's (1976/2001) concept of cohesion and taxonomy of cohesive devices were adopted to analyze the samples for their cohesive features. Then descriptive statistics was conducted to settle Research Question 1), Pearson correlation analysis Research Question 2) and multiple regression analysis (stepwise) Research Question 3). With regards to Research Questions 4) and 5), a qualitative analysis was performed on the common features, including errors, in using each cohesive tie by the students.The findings reveal that these Chinese college students employed a variety of cohesive ties in their English compositions, among which lexical category had the highest percentage, followed by the categories of reference and conjunction. Substitution and ellipsis were seldom used. The quantitative analysis indicates that the composition scores were positively co-related with the total number of cohesive tiesand lexical synonymy (LS) and personal reference (RP) among the categories of cohesive ties were better predictors of the quality of the student compositions. However, the results of the quantitative analysis suggest a weak relationship between the total number of the ties used and the composition scores.The qualitative analysis of the cohesive features in the compositions shows that among the three subcategories of reference ties, personal reference had the highest percentage of use, followed by demonstrative reference and comparative reference. The examination of the most frequently used reference devices proves that the first person pronouns had the highest frequency in the occurrence of personal referential cohesion. The major referential cohesive errors involved the shifted use of pronouns, reference ambiguities, the confusion of demonstratives "this" and "that" and difficulty in using the definite article "the" correctly. With regard to conjunction cohesion, the students appeared to have an awareness of three subcategories of it, but there might be an overuse of such cohesive items as "but", "and" and "so". There is also a lack of connector variety. Four types of problems with the use of conjunctive ties were identified: improper addition, inappropriate omission, misuse (logically) of conjunctive expressions and misapplication (grammatically) of them. The most prominent feature of lexical cohesive ties used was the overuse of lexical repetition. In addition, lexical synonymy was not widely used in the sample compositions.This study further documents the previous evidence about the effect of cohesive ties on writing quality. It can also help promote the understanding of the cohesive features in Chinese college students' English compositions and provide insights into the teaching of English, especially English writing, to Chinese college students learning English.