A Contrastive Study Of Validity In CET Controlled Writing Task And Free Style Writing Task

Writing, as one of the major language skills, is considered a valid and expressive form to reflect how well a student of English, particularly a student who learns English as a foreign language, has mastered that language. Nowadays, the pattern being used in CET writing sub-test is the simple single style of controlled writing task, an argumentative essay with a given title and outline, which has dominated the CET writing sub-test for a long time. Many scholars and teachers have been questioning this rigid test pattern. In opposition to controlled writing task, free style writing task provides candidates more space to create their own compositions without limitation of outline and fixed structure. This thesis intends to investigate the test validity of the two writing tasks contrastively. The general research question is:Between controlled writing task and free style writing task, which testing method can better assess students'EFL writing ability in terms of test validity, and to what extent is this method more valid? To answer this question, the author employs three instruments:questionnaires, interviews and writing tests.The results from questionnaires provide evidences that most of subjects hold favorable opinions of free style writing task that offers them sufficient space to think independently, generate and organize their own thoughts. Furthermore, their ability to organize and think creatively will be improved by sequencing their ideas and organizing a composition in a logical and coherent way completely on their own. By contrast, many subjects regard the controlled writing task as rigid, stereotyped and bored, preventing them from creating their own ideas and organizing compositions in their own way, thus depriving them of the opportunities to show their best level of writing competence due to the limitation of provided outline or topic sentences.The interview survey explores the detailed process of subjects'completion of writing and their specific feelings about these two different writing tasks, which reveals that when dealing with free style writing task, most of interviewees feel that it is tougher than controlled writing task, because without guidance of outline or topic sentences in advance, they must fulfill task solitarily. However, some of them think that accomplishment of free style writing task is a process of revealing their true writing ability, thus they have chance to perform at their best due to solitariness. Finally, the interview results also indicate that whatever writing tasks the interviewees handle, same problem occurs that they are incapable of picking up right words to express their thoughts exactly. The crux of the problem lies in less reading and limited vocabulary. Based on correlation coefficient computed with Spss, both controlled writing task and free style writing task are highly correlated with teacher's assessment, which means both of them are valid methods to assess students'writing ability. However, compared with controlled writing task, free style writing task is more valid in evaluating students' writing competence according to correlation coefficient. In accordance with Mean in Paired-sample T-test procedures, it is revealed that free writing task is more severe than controlled writing task to the subjects.Therefore, a conclusion can be drawn that, compared with controlled writing task, free style writing task is a more valid testing method to better assess students' EFL writing ability and exert a beneficial effect on college English writing teaching and learning. It is hoped that the findings of the current research may provide instructive implications and recommendations on the development of CET writing sub-test research and English writing teaching and learning in universities.