A Correlational Study Of English Majors’ Modified Output And Working Memory Capacity

Inspired by Long’s statement on the important role of internal factors in SLA, this study hypothesizes that there may likely be some relationship between modified output and L2learners’internal factors. The study chooses to focus on the relationship between English majors’ modified output and their working memory capacities in the task-based interactional activities, and seeks to disclose learners’inner cognitive mechanism of working memory through their modified output. To be specific, this study addresses two research questions:1) What is the frequency and distribution of responses in English majors to the provided feedbacks in the context of a task-based interaction?2) What is the correlation between modified output and working memory capacity?The subjects are Year Three English majors, who were asked to take a test and complete a task. Two instruments were used in the current study:One is MindTest2.0, a working memory span test software, designed to measure the working memory capacity of the selected subjects. The other is a story-completion task, in which the subjects were asked to complete a picture story in English through interaction with the interlocutor, and the task was designed to investigate students’modified output in the process of conversational interaction. The MindTest2.0generated the results, and the interaction was tape-recorded. The data was coded or transcribed and then analyzed with the assistance of SPSS16.0. Detailed analysis yields the following findings:Firstly, of all the responses to the clarification requests provided, three categories were identified:modified output, repeated responses, and other types of possible response, which account for66.2%,18.6%, and15.3%respectively. Obviously, the most frequently adopted response is modified output.Secondly, the correlational analysis reveals that modified output is positively correlated to working memory capacity. Learners’working memory capacities were found to affect their production of modified output in the task-based interaction. The results indicate that learners with higher working memory capacities tend to produce more modified output in the interaction, while learners with lower working memory capacities produce less output modifications. What’s more, the results of the linear regression analysis indicate that learners’working memory test scores enjoy a strong predicting power to their production of modified output. The results support the hypothesis proposed in this study:Learners with higher working memory capacities are more likely to modify their output in interaction than those with lower working memory capacities.The findings may have some pedagogical implications. To language teachers, modified output may have practical value in the sense that teachers may resort to different corrective feedbacks to create as many opportunities as possible for students to modify and improve their output to the most satisfactory degree. To learners, working memory could be trained and integrated into output modifications so as to develop their learning strategies.