Implicit corrective feedback in computer-guided interaction: Does mode matter?

摘要

Interaction research on recasts has been largely confined to examining the effects of recasts delivered in an oral, face-to-face modality. The challenge of creating task conditions in a written interactive mode, in which the dynamism and immediacy of recasts can be brought to bear on L2 development, has been a formidable obstacle to advancing the exploration of written recasts. This study examines the developmental effects of recast-intensive interaction on ESL question formation and morphosyntactic accuracy. It compares the effects of recasts delivered in oral (face-to-face) tasks with recasts delivered in analogous written (computer-guided) tasks. In a pretest/posttest design, 56 high school ESL learners engaged a series of communicative tasks with either a native English speaker or an intelligent virtual interlocutor. The results indicated that participation in recast-intensive interaction was a significant predictor of both ESL question development and improvement in morphosyntactic accuracy. The mode of interaction---oral or written---had no effect on development.