Effects of form-focused instruction, corrective feedback, and individual differences on the acquisition of Chinese wh-questions and classifiers

Abstract

This dissertation addresses the differential effects of two oral corrective feedback strategies, recasts and metalinguistic prompts, on the acquisition of Chinese wh-questions and classifiers, while examining how individual differences (i.e. language analytical abilities, attitudes, and anxiety) would moderate the effects of CF. Two beginning Chinese classes were randomly assigned to the recast or metalinguistic prompts group. In a span of 5 weeks, a total of 4 treatment sessions took place. Students were tested with an oral production task and a written error correction task before, immediately after, and two weeks after the treatment. Mixed-method ANOVAs were used to analyze the differential effects of the two CF strategies on the acquisition of wh-questions and classifiers. In addition, students also completed two questionnaires, with one testing their language analytical ability, and a combined questionnaire measuring their attitudes and anxiety. Multiple regressions were used to analyze the relationship between individual differences and students’ learning outcome. The results showed that the metalinguistic prompts group had significant gains in accuracy in all measures, regardless of testing time (posttests or delayed posttests), target forms (wh-questions or classifiers), and testing mode (oral production or written error correction tests). The recast group showed significant gains in the written tests for wh-questions and classifiers, but only achieved significant short-term gains for wh-questions in the oral test. Regarding individual differences, we found that learners’ language analytical abilities and attitudes were important in predicting their test performance, while anxiety did not affect the learning outcome. Results were discussed within the Interactional Cognitive Framework. Form-focused instruction, along with metalinguistic prompts, which were consistent, output-pushing, and reminded students of previous learned declarative knowledge, worked better than input-providing CF (recasts) for both syntactic and lexical features. Metalinguistic prompts withheld the target L2 forms, provided metalinguistic comments, and pushed for modified output, which may have increased the likelihood for learners to close the gap between their existing knowledge and the target L2 forms, and convert declarative knowledge into procedural knowledge. The findings also suggested that CF could be delivered without raising students’ anxiety, and helping students maintain positive attitude was important for their language development.