The Effect of Focused Versus Unfocused Written Corrective Feedback on the Development of University-Level Learners’ Explicit and Implicit Knowledge in an EFL Context

Abstract

Recent research in the field of second language acquisition (SLA) has started to pay critical attention to the effects of written corrective feedback on learners’ acquisition of explicit and implicit second language (L2) knowledge (e.g., Shintani & Ellis, 2013). Drawing on Gass’ (1997) Computational Model and skill acquisition theories in SLA (e.g., Anderson, 1983, 2000), the study reported in this thesis examined the comparative effects of focused and unfocused metalinguistic corrective feedback on university-level learners’ acquisition of explicit and implicit knowledge of the English regular and irregular past tense forms. The thesis examines whether and the extent to which the effect of feedback is mediated by the type of feedback (i.e., focused versus unfocused) and the linguistic target (i.e., rule-based versus item-based). This thesis also examines EFL learners’ perceptions of the effectiveness of metalinguistic corrective feedback provided to them. The thesis adopted a pretest, posttest and delayed posttest control-group research design (Phakiti, 2014). A total of 58 intermediate-level Thai university learners of English as a Foreign Language (EFL) participated in the study. They were from four intact classes and were randomly assigned to two experimental groups and one control group. The two experimental groups attended three treatment sessions over a period of four weeks where they performed three story rewriting tasks. The first experimental group (N = 19) received metalinguistic corrective feedback on regular and irregular past tense errors only, while the second experimental group (N = 20) was provided with metalinguistic corrective feedback on a wide range of errors including regular and irregular past tense errors. The control group (N = 19) also performed the same writing tasks but received no feedback. Treatment effects were measured by means of an error correction test (designed to measure explicit knowledge) and both a timed grammaticality judgment test and a timed picture description test (both designed to measure implicit knowledge). All three groups completed the tests prior to the first treatment session (pretest), shortly after the last treatment session (immediate posttest), and again four weeks later (delayed posttest). To examine learners’ perceptions of the feedback intervention, a questionnaire survey and interviews were conducted with the learners in the two experimental groups at the end of the data collection period. Statistical analyses revealed that the two experimental groups improved significantly over time in the accurate use of regular past tense forms on each of the three language tests, while no such improvement was found for the control group. Both groups also significantly outperformed the control group in the immediate posttest and the delayed posttest. However, there was no significant improvement in accuracy over time in either experimental group for the irregular past tense. Overall, it is suggested that both focused and unfocused metalinguistic corrective feedback facilitated learners’ acquisition of regular past tense forms by developing both explicit and implicit knowledge, but neither of these two types of feedback demonstrated an effect on irregular past tense forms. The findings lend support to theoretical explanations for the facilitative role of written corrective feedback in L2 development. Regarding the differential effects of the two feedback types, it was found that although the focused feedback group outperformed the unfocused feedback group on all three tests for the regular past tense, the differences between the two groups did not reach statistical significance. These results suggest that there was no advantage in receiving focused metalinguistic corrective feedback instead of unfocused metalinguistic corrective feedback. In addition, the results from questionnaires and interviews indicate that the learners were generally engaged in the intervention and perceived it as useful for successful correction of regular past tense errors, despite a seeming tendency to prefer an unfocused approach to error treatment. The implications of the research findings are discussed from theoretical, methodological and pedagogical perspectives. Limitations of the thesis and suggestions for future research are also offered.