The effects of repair techniques on L2 learning as a product and as a process : CA-for-SLA investigation of classroom interaction

Abstract

The link between L2 classroom learning and repair of syntactical errors has not yet been sufficiently explicated. Moreover, previous findings on the effectiveness of different repair types have been inconclusive. Therefore, this thesis aims to investigate the effects of different types of repair technique (RT), namely recasts, prompts and explicit correction, on L2 classroom learning of the English passive voice as both a product and a process, using a methodology that has not previously been employed with this perspective by any other study. For this purpose, the focus is on establishing whether a relationship exists between types of RT and learners' test scores (i.e., classroom product) on the one hand and between these types and classroom interaction (i.e., classroom learning processes) on the other, in order to show how interaction and ( learning ) are bound up together. To achieve this goal, two methods were used to collect and analyse data obtained from two sources. First, a quantitative experimental research design consisting of an intervention using a pre-test/post-test measure was employed to quantify the relationship between learners' pre-test/post-test scores and the types of RT, in order to assess their effectiveness in promoting learning of the English passive voice. Five groups of EAP learners participated in the study during their focus-on-forms (FonFs) classes at King Faisal University in Saudi Arabia. Three groups received one repair type each in response to their errors. In the fourth group, learners' errors were not corrected at all, while learners in the fifth group, where no intervention conditions were implemented, received their teacher's normal corrective behaviour. The quantitative analysis revealed a statistically significant positive effect of all types of RT on language learning. with corrective recasts being the most beneficial, followed by prompts and then explicit correction. Additionally, ignoring learners' errors proved to have a slight - though not significant - negative effect on learning. Second, a qualitative CA approach was adopted to analyse teacher-learner interaction during the intervention. Sixteen hours of audio- and video-recorded classroom interaction was transcribed and analysed using CA tools and methods. CA is not employed as a learning theory or as a tool to measure learning, but as part of a strategy to explore learning opportunities and the types of behaviour that could be associated with them. The results of the qualitative analysis of the interaction uncovered aspects of the interaction-learning relationship by demonstrating how learning processes occurred in the intersubjective spaces between participants and by portraying why the different RT types varied in their effectiveness in promoting opportunities for language development and learning of the passive voice. By applying a product-process perspective to the interaction-learning relationship, this thesis leads to a rethinking of the close and complex relationship that exists between repair and classroom learning. Hence, the present study attempts to bridge the gap between the cognitive and social-interactional schools of SLA and seeks to inform current perspectives on theory and practice.