|dc.description.abstract||Written CF is a common practice in L2 teaching despite the theoretical controversies about its contribution to L2 development (Truscott, 1996; Bitchener & Ferris, 2012). Empirical research into this issue has been dominated by quasi-experiments in the cognitive framework. By inferring L2 development from the accurate development in the written output, these quasi-experiments have generated somewhat inconclusive results. Framed in the perspectives of both the cognitive theories and the Dynamic Systems Theory, this project examined the L2 learning potential of written CF (enhanced by revision) when accurate development was not achieved. It also examined the possible moderating effects of revision type and L2 motivation. The possible causes of the different extents to which the students benefited from written CF was explored in two follow-up case studies. To accommodate the difference in research aims (i.e. descriptive vs. exploratory), a mixed-method approach was adopted: a quasi-experiment was followed by a multi-case study.
Examining the efficacy of two written CF types (i.e. direct feedback and metalinguisitic explanation) on the development of the English passive voice in comparison with writing practice, the quasi-experiment was conducted among 87 vocational college students in China. It involved three writing tasks, one revision task (feedback groups only), and an L2 motivation questionnaire survey. The results showed that, compared with writing practice, written CF did not significantly contribute to the accurate development of the target feature. However, compared with writing practice, written CF significantly contributed to the partial development of this feature. Moreover, one L2 motivational variable, Ideal L2 self, significantly moderated the effect of one written CF type, direct feedback.
To explore the possible causes of the different extents to which the students benefited from written CF, the follow-up multi-case study focused on two students who had participated in the quasi-experiment and differed significantly in the accurate development of the target feature. Data were collected via three writing tasks, one revision task and two stimulated recall interviews with each participant. The analyses of stimulated recall and texts revealed a link between strategy use in the cognitive processing of written CF and the retention of written CF over time.
Hence, this project revealed that, in the EFL context in China, written CF may contribute to L2 development even when accurate development is not achieved. Moreover, the efficacy of written CF can be influenced by L2 motivation and strategy use in the cognitive processing of written CF. Therefore, it is suggested that written CF can be a useful tool in the EFL teaching in China, and learner differences, such as L2 motivation and learning strategies, need to be considered while providing written CF in this context. Replicative research is suggested in other L1 contexts to test the generalizability of the present findings.||en_NZ