DIFFERENTIAL EFFECTS OF PROMPTS AND RECASTS IN FORM-FOCUSED INSTRUCTION

Abstract

Four teachers and their eight classes of 179 fifth-grade (10–11-year-old) students participated in this quasi-experimental classroom study, which investigated the effects of form-focused instruction (FFI) and corrective feedback on immersion students' ability to accurately assign grammatical gender in French. The FFI treatment, designed to draw attention to selected noun endings that reliably predict grammatical gender and to provide opportunities for practice in associating these endings with gender attribution, was implemented in the context of regular subject-matter instruction by three of the four teachers, each with two classes, for approximately 9 hours during a 5-week period, while the fourth teacher taught the same subject matter without FFI to two comparison classes. Additionally, each of the three FFI teachers implemented a different feedback treatment: recasts, prompts, or no feedback. Analyses of pretest, immediate-posttest, and delayed-posttest results showed a significant increase in the ability of students exposed to FFI to correctly assign grammatical gender. Results of the written tasks in particular, and to a lesser degree the oral tasks, revealed that FFI is more effective when combined with prompts than with recasts or no feedback, as a means of enabling L2 learners to acquire rule-based representations of grammatical gender and to proceduralize their knowledge of these emerging forms.This study was funded by the Social Sciences and Humanities Research Council of Canada (nos. 410-98-0175 and 410-2002-0988). Parts of this study were presented at the annual meeting of the American Association for Applied Linguistics in Salt Lake City on April 7, 2002; the Congress of the Social Sciences and Humanities in Toronto on May 26, 2002; and at the Congress of the International Association for Applied Linguistics in Singapore on December 12, 2002. I am grateful to the participating teachers and their students, to Lucy Fazio for her role as research associate in the data collection, to José Correa for his assistance with the statistical analyses, and to the following research assistants for contributions to various phases of this research: Susan Ballinger, Kristina Eisenhower, Andréanne Gagné, Sophie Beaudoin, Laura-Annie Bouffard, France Bourassa, Sophie Bourgeois, Elisa David, Mélanie Mathieu, Sophie Prince, Andrea Sterzuk, and David Syncox. I gratefully acknowledge Leila Ranta, Iliana Panova, and four anonymous SSLA reviewers for their helpful comments on an earlier version of this paper.