A study of the effect of corrective feedback on foreign language learning: American students learning Chinese classifiers

Hao-Jan Howard Chen, University of Pennsylvania


Issues regarding the role and contribution of corrective feedback (CF) for language learning have been central to second language acquisition (SLA) theory and pedagogy. To address these issues, this study investigated the effect of different types of CF on the acquisition of Chinese classifiers by 38 American college students. Three research questions were posed: Does the immediate effect of CF on learner accuracy generalize across different tasks? Can the short-term impact of CF be sustained over time? Do different types of CF have different effects? Four groups of subjects performed a computer-mediated exercise on Chinese classifiers. Subjects in three experimental groups received different types of corrective feedback when they made errors: The "metalinguistic group" was given correct answers and explanations; the "explicit rejection group" was told that responses were wrong; and the "modeling group" was given correct answers. The control group received no feedback. All subjects were tested with both an oral and a written posttest. Posttests were administered immediately after treatment and again six weeks later. There were three major findings. First, the positive effect of CF on learner accuracy generalized across different posttesting tasks. Second, the short term effect of CF was not sustained after a six-week interval. Third, both the metalinguistic group and the modeling group outperformed the explicit rejection group in the short-term. However, over the longer term, the metalinguistic group outperformed the explicit rejection group only in the written task. The results of this study lend strong support to the view which considers CF to be facilitative. Nevertheless, the rapid decay of the impact of CF also suggests that follow-up activities might be essential for maintaining any short-term gains. Moreover, the finding that metalinguistic input was most beneficial for learners suggests modifications of current claims that metalinguistic input is not useful for SLA. Despite these findings, limitations in the research design (e.g., the amount of treatment and the length of the interval between tests) suggest that additional research is needed to study the role and contributions of CF for SLA, especially with respect to its long-term impact.

Subject Area

Language arts|Linguistics

Recommended Citation

Chen, Hao-Jan Howard, "A study of the effect of corrective feedback on foreign language learning: American students learning Chinese classifiers" (1996). Dissertations available from ProQuest. AAI9712905.