A Study On The Effects Of Model-imitation Practice In The Teaching Of Writing To Chinese Non-English Majors

In a long term, both foreign and domestic researchers disagree with each other on the effectiveness of the model-imitation practice in writing teaching. This thesis reports an empirical study on the effects of model-imitation practice in the teaching of writing to 72 Chinese non-English majors in their second semester at a university. The goal of the research was to examine the role of model-imitation practice in the teaching of writing to Chinese non-English majors. To be specific, the study attempted to find out an answer to the doubt: Is model-imitation practice a more effective way of writing teaching than the non-model-imitation one?The arguments in the past had focused on whether model-imitation was an effective way of teaching writing or not. Theoretically, behaviorism, mentalism and procedural approach differ from each other on the views about the role of imitation in the process of language learning. Practically, many people find advantages of using models in the teaching of writing, while others see only problems. From an empirical study perspective, few former researches have taken Chinese non-English major freshmen as subjects or undertaken a mixed design of experimental study to solve the issue; instead, most of previous researches are only personal intuition and conclusions.In view of the existing conflicts and the drawbacks of relative empirical studies, three research questions were addressed in the present study:(1). Is model-imitation practice more effective than non-model-imitation one in the teaching of writing to non-English majors?(2). Can gender influence the impact of model-imitation practice upon subjects' test performance? That is to say, do boys and girls make different progress after the model-imitation practice?high-score writers and low-score writers perform the same after practice model-imitation way of writing learning?To answer the above questions, the present study was about to testify to three following hypotheses:(1). The scores of the two groups make no difference after they take the way of model-imitation practice and the non-model-imitation one, i.e., H0: μA-μB= 0;(2). Gender cannot affect the relationship between model-imitation practice and subjects' writing performance, H0: μM-μF= 0;(3). High-score writers and low-score writers are the same at imitating ability,H0: μH-μL= 0.In order to test the above hypotheses, the study undertook a Pretest/Posttest nonequivalent groups design and used a convenience sample, which was two intact classes from an intact university?the history class and the Chinese class as respectively the experimental group (EG) and the control group (CG). There were a total of 72 subjects participating in the experiment. However, due to the absence of some subjects for various reasons, there were only 64 of them entering the final dataanalysis.The following research tools were used in the study: (1) an open-ended interview before the experiment; (2) a pretest and posttest; (3) a after-course questionnaire. The interview was undertaken with the writing teachers to find out how the teachers went about writing teaching; the pretest and posttest were about to measure the subjects' English writing performance; the questionnaire, to investigate the subjects' attitudes towards model-imitation practice.In the experiment, the independent variable was the model-imitation practice, with the gain scores of the subjects on the pretest and posttest as the dependent variable. Gender and the types of writers from the EG would be taken as moderator variables. For the experiment, the EG adopted the model-imitation practice, while the CG adopted the non-model-imitation one. The same teacher taught the two groups, with the schedule and settings of teaching for the two groups being keptthe CG adopted the non-model-imitation one. The same teacher taught the two groups, with the schedule and settings of teaching for the two groups being kept nearly the same.The data from the interview and the questionna...