The role of negative feedback in second language instruction

The present study measures the effectiveness of negative feedback, more specifically the effectiveness of negative feedback in two forms, explicit corrections and corrective recasts. This was done by determining the extent to which negative feedback affected the ability of students learning Spanish to improve their production of the preterit and imperfect forms in the long-term. The study also focused on the gender issue of whether female students were more receptive to feedback than males.; It was hypothesized that learners who participated in interactions with negative feedback would show a greater increase in the accurate production of the target structures than learners who participated in interactions without negative feedback. Second, it was also hypothesized that learners who participated in interactions with explicit corrections would show a greater increase in the accurate production of the target structures than learners who participated in interactions with recasts. Finally, it was hypothesized that females regardless of treatment would outperform males under similar conditions.; Participants in the study included 97 undergraduate college students, enrolled in a second semester beginning Spanish course. The sections participating in the study were randomly assigned to the three treatments—two experimental and one control. Instructional materials were prepared by the researcher and were used by regular instructors previously trained in the use of the specific types of feedback. Two instruments were used to assess participants' performance: a fill-in-the-blank exercise and a composition. Students had been introduced to the use of the preterit and imperfect forms in previous classes as part of their normal curriculum. Classes were tape-recorded and a debriefing questionnaire was given to measure participants' reaction.; The results showed that the participants in the recast group outperformed the participants in the control group in the long-term in the fill-in-the-blank exercise. However, contrary to what was hypothesized, the explicit correction group did not outperform the recast group. With regard to the written compositions, the results indicated that the explicit correction group performed significantly better than the control group in the long-term with the imperfect, but not with the preterit. Gender was found not to be a significant factor.