A Comparative Study Of The Effects Of Recasts On Learners' L2 Acquisition In FTF And CMC Instructional Settings

Recasts, as an implicit corrective feedback strategy, have been the subject of considerable Second Language Acquisition (SLA) research. Quite a number of research studies have indicated that recasting is the most common form of corrective feedback used in classroom interaction (eg. Lyster & Ranta, 1997; Lyster 1998; Ellis et al, 2000). However, no consensus has yet been reached about its efficacy. Although recasts have been the focus of many empirical studies, in both the laboratory and the classroom, few of them have compared and contrasted their performance in different instructional settings. Aiming to fill this gap, this study uses the theoretical framework of Long's Interaction Theory(Long, 1996) to compare and contrast the efficacy of recasts on L2 development in different instructional settings, while exploring the factors that cause the differences. In this way, it sheds light on the nature of recasts, the role of instructional setting, and their respective effects on L2 development.This study adopted a pretest, posttest and delayed posttest design to investigate whether recasts can enhance learners'L2 acquisition and to what extent the instructional setting affects L2 development. Both oral production tests and written grammacality tests were administered to learners to examine recasts'effect on different aspects of L2 knowledge development.In this experiment, recasts were operationalized as either partial or full reformulations of the learners'ill-formed utterances (Sheen, 2006) . The learner's L2 development was operationalized as the development of their grammatical competence, defined by Canale and Swain in 1980 as"the knowledge of lexical items and of rules of morphology, syntax, sentence-grammar semantics, and phonology"(Canale and Swain, 1980: 29). Meanwhile, a specific target structure--the hypothetical conditional was also chosen to examine recasts'effect on learners'acquisition of a rather complex morphosyntactic structure. Two instructional settingswere employed: face-to-face classroom and computer-mediated communication environment using Qingyang video conferencing software. The following are the research questions the study strives to answer:1. Do recasts have significant influence on learner's L2 development of grammatical competence?2. Does the instructional setting, either Face to Face (FTF) or Computer Mediated Communication (CMC), affect the impact of recasts on learners'L2 development of grammatical competence?3. Is there any interaction between recasts and the instructional setting?4. 1) Do recasts function differently under different instruction settings (FTF or CMC)?2) Do instructional settings (FTF or CMC) effect the functioning of recasts ?3) If instructional settings do have different effects on recasts, what are the possible causes of these differences?4) How do recasts affect learners'L2 development in this experiment?5. Do recasts have any significant effect on the acquisition or retention of a specific target structure—the hypothetical conditional? Why or why not?The present study is a quasi-experiment conducted with 200 students who were randomly assigned into four conditions: face-to-face control group, face-to-face experimental group, computer-mediated communication control group and computer-mediated communication experimental group. The experiment was designed to compare and contrast the different efficacies of recasts under FTF and CMC settings and also to disclose the factors that lead to the differences.First, all the four groups took the pretest to check if their primary English level was at the same level. The two experimental groups in the FTF and CMC settings received the same amount of treatment (two treatments, 50 minutes each) respectively, and the treatment in this experiment refers to the teacher's feedback to students'errors in the form of recasts. The two control groups performed the discussion twice (50 minutes each) on the same topic in class without receiving any feedback on their errors. After two treatments, the post-tests (oral and written ) were conducted in order to find out the short-term effect of recasts on learners'proficiency. Three weeks later, the delayed post-tests (oral and written) were administered to investigate the long-term effect of recasts. The experiment spanned seven weeks.Then quantitative data were collected and were processed with SPSS and SAS software using T- test and one-way and two-way ANOVAs. According to the results of the statistical analysis, the major findings are as follows:After the treatment, the two experimental groups, in FTF and in CMC, taken together, showed no overall superiority in learning over the two control groups taken together, so there was no obvious relationship between recasts and L2 development, but in specific instructional settings recasts'function was amplified and did influence L2 learning significantly. However, the result of the comparison between the entireties of the two groups in each instructional setting demonstrates that instructional setting can affect L2 learning and this is more evident when instructional setting is combined with feedback to demonstrate its function. Thus, recasts and instructional settings interact variably depending on the different instructional settings and the amount of recasts.Further, there are significant differences between the functioning of recasts under FTF and CMC, which may be caused by the unique features of CMC, such as its more equitable and non–threatening context, written modality, retention of the chatscript, saliency, longer processing time, enhanced noticing and special classroom dynamics.In this study, recasts exerted effects on the learners'general L2 development rather than their development of the target structure. A variety of reasons may contribute to this result, but the primary one is that feedback on different linguistic structures is perceived by learners differently, especially when the target structure is the hypothetical conditional --- a rather difficult morphosyntactic structure for Chinese learners .This study has provided support for the validity of Long's Interactional Theory by demonstrating the successful application of interaction in different instructional settings. It has revealed the subtle relationship between recasts and L2 development, thus uncovering the diverse nature of recasts. It also reveals the importance of the instructional setting and draws attention to the interaction between recasts and the instructional setting. By comparing and contrasting the function of recasts in FTF and CMC, the study has presented the specific characteristics of each instructional setting and explored how they influence the efficacy of recasts. Finally, this study has special significance for the development of Chinese EFL teaching as it can assist Chinese EFL teachers to find more effective pedagogical tools and construct more effective instructional settings for foreign language teaching.