The Effect Of Students' Preferences On The Result Of Corrective Feedback In High School English Classrooms

Corrective feedback, as an important component of English classroom teaching,has been hotly studied since 1980s. On account of the fact that languages, especially foreign languages, are usually learned in the classroom, it is of great importance to pay attention to students' and teachers' attitudes toward and preferences of corrective feedback. By investigative research, or semi-structured interviews, many researchers have explored the differences between teachers and students on their preferences of corrective feedback, as well as teachers' actual corrective feedback moves in the classroom. Nevertheless, few empirical studies of corrective feedback in high school English classrooms have been conducted, let alone studies of the relations between students' preferences of corrective feedback and the repair rates of different types of corrective feedback. Therefore, the present thesis explored students' and teachers'preferences of corrective feedback and the impact of students' preferences on the results of teachers' corrective feedback moves, employing questionnaire investigation and its consequent action research. The research questions of the present study are as follows: 1) Are there any differences between types of corrective feedback favored by students and teachers? If yes, what are the differences? 2) What types of corrective feedback are adopted by teachers in real classroom interactions? Are they consistent with teachers' preferences? 3) Do corrective feedback moves favored by students yield higher repair rate? If yes, what's the reason?170 senior one students and 22 high school English teachers participated in the questionnaire investigation. All the 170 students come from a four stared high school in Nanjing, Jiangsu province, while the 22 teachers are working at schools in Nanjing,Shanxi and Chongqing. One class in this four stared high school was chosen to have the classroom interactions recorded in the consequent action research, which involved a pre-test and a post-test. In the pre-test, the teacher followed her usual teaching method without any intervention from the researcher, while in the post-test, based on the findings of the questionnaire investigation, the researcher requested the teacher to use the types of corrective feedback favored by the students i.e., meta-linguistic feedback, recast and elicitation, more frequently. Altogether 29 classes were recorded and 22 of them (more than 18 hours) were transcribed into written texts for later analysis.Through the quantitative and qualitative analysis of the questionnaire data, the distribution of errors and corrective feedbacks in classroom interactions, and the repair rate of each type of corrective feedback in the pre- and post-test, the following conclusions are drawn: 1) Students' preferences of corrective feedbacks are:meta-linguistic feedback> recast>elicitation>repetition>clarification request>explicit correction. Teachers' preferences of corrective feedbacks are: recast> elicitation=repetition> meta-linguistic feedback> clarification request> explicit correction. 2) In the classroom interactions, the actual frequency of each type of corrective feedback used by teachers is: recast> elicitation= repetition> meta-linguistic feedback> recast=clarification request. 3) Corrective feedbacks such as meta-linguistic feedback and recast favored by the students can elicit more repairs. However, Spearman correlation analysis indicates that there is no significant correlation between students' preference and the repair rates of corrective feedbacks.Based on the above findings, the following implications for teaching are proposed. Teachers should strengthen their own theoretical knowledge and apply theory into practice. In this way, not only can students' errors get corrected, but the teachers themselves can improve their academic ability. In classroom interactions,teachers are suggested to use multiple types of corrective feedbacks at the same time to boost more repairs instead of just using one type of corrective feedback on one error.