Error Analysis Of English Writing In College English Placement Test Of Hunan University

Writing competence has been playing a significant role in evaluating Chinese English learners'proficiency, while writing performance of students is far from satisfactory according to the statistics declared officially from the examination center of CET-4 &CET-6. Research on writing in CET-4 displays that the average score of writing is far lower than the other items ( listening, reading,etc.) (Yu Keying,2002). Aims to foreign language writing teaching, the researchers prone to analyze the errors occurred in which with relevant theories. However, relatively few empirical studies have been conducted in investigating the errors in writing at each level.Based on the error analysis of the first-year non-English major students'writings on computer-based College English Placement Test (CEPT) in Hunan University (HNU), this dissertation attempts to find out the distribution and frequency of errors in writing at each level so as to help language teachers adopt appropriate teaching strategies and methods of error correction, then to promote their profession of writing teaching for English learners with the same proficiency in writing.The experiment of this study is composed of six procedures: 1. Testing participants. 4785 first-year non-English-major undergraduates of Hunan University participated in CEPT. 2. Scoring. Each of the compositions will be scored at least by 2 raters, who are well-trained raters with rich scoring experience, all the raters were selected for participants in the CEPT program are experienced in English teaching in colleges and familiar with writing scoring criteria. 3. Questionnaire for testees. Shortly after the CEPT, an after-activity questionnaire was administered to gather information about the students'backgrounds. 4. Samples collection. 150 writing samples( 30 compositions at each level×5 levels) are collected from the first year non-English major students'compositions ( 4785 compositions in all) in the CEPT of Hunan University. 5. Errors analysis. Based on the classification of the errors by reference to three levels of language: the levels of substance, text and discourse which specified in Exploring Error Analysis written by Carl James(2001), the writing samples chosen have been conducted with the theory of Error Analysis, including description of errors, explanation of errors and evaluation of errors. 6. Data collection and analysis. The datas were input into SPSS13.0, then the normal distribution hypothesis was tested, the correlation between writing level and error, the correlation among errors were analyzed with SPSS. The results of the study are following: 1. The students differently proficient in writing commit errors variously. The 010 level students are easy to have errors in tense and topical coherence at the level of text and discourse respectively. For the 020 level students, the errors focus on distortion (including omission, misselection, blend,etc.) and topical coherence. In level 030, subject-verb agreement and third person singular verbs accounts the highest percentage at the level of text while the relational coherence at the level of discourse. As for the level 040, the most frequent error types are the same as the 030 level, but the proportion of the errors obviously decreased. Compared with the above levels, the most frequent errors for the highest level 050 are errors in clause (including superfluity, omission, misorder and misselection) and relational coherence. 2. The data obtained from the experiment have indicated that grammar is regarded as the biggest obstacle in writing. According to the overall distribution of total errors in terms of error causes, intra-lingual interference plays the most dominant role in contributing to students'poor writing. 3. There is significant negative correlation between writing level and grammatical errors as well as discourse errors. Lexical errors and grammatical errors are positive correlation.In the end, the paper shows the implications for EFL writing teaching and limitations of the research as well as follow-up research.