초록 열기/닫기 버튼
This study explored which corrective feedback (didactic recasts, metalinguistic clues, and repetition) enabled students to repair the form more than when they were exposed to the same opportunities to uptake the form. The study was conducted using 48 Korean university students. The students were asked to write down the correct past forms of the irregular verbs during the pre-test, the immediate post-test, and a two-weeks-delayed post-test. Based on the pre-test, the three verbs that students answered incorrectly were selected for different feedback treatments. The results from the analysis with the recasts group showed higher rates of successful repair for the immediate pushed output. For the analysis of the two-weeks-delayed post-test, among linear mixed-effect models, the random intercept model was considered to be the best fitted model. This model illustrated that didactic recasts and metalinguistic clues were significantly efficient, even though the effectiveness was much greater on recasts. These results can be applied to teaching, in that recasts can be effective in enabling students to notice phonological and lexical errors; providing explicit corrective feedback might help improve the students’ accuracy when learning irregular morpho-syntactic items.
