Working Memory and Corrective Feedback in Interaction = Working Memory and Corrective Feedback in Interaction

다국어 초록 (Multilingual Abstract)

The present study is designed (a) to examine the relative efficacy of one feedback move over another (recasts vs. metalinguistic feedback) on the acquisition of the English past unreal conditional and (b) to explore whether working memory (WM) mediates the extent to which L2 learners benefit from the two feedback moves. Eighty-three Korean EFL learners were randomly assigned to one of three groups: recast (n=24), metalinguistic (n=27), and comparison (n=32). Each individual learner carried out a series of tasks and tests: (a) one WM span task (operation span), (b) one dyadic intensive oral treatment session, (c) oral production tests, and (d) grammaticality judgment tests (GJTs). Results showed that recasts were slightly more effective than metalinguistic feedback at precipitating development of the English past unreal conditional as evidenced in learners` GJT performance. However, no significant or noticeable difference was observed between the two feedback conditions in learners` oral performance. WM was significantly predictive of the oral pre-to-post development of the metalinguistic feedback condition only. (Gwangju National University of Education)