Effects of SRSD on Students’ Metacognitive Knowledge, Self-efficacy, Text Revision, and Text Quality in L2 Writing

Abstract

Revision has been widely accepted as a crucial part of writing processes, outcomes, and development. Researchers see the need for inexperienced writers to learn how to revise effectively and proposed various factors that account for the different ways experienced and less experienced writers approach revising. As a complex, cognitivelyconsuming activity, revising necessitates substantial metacognition and self-regulation. For revision to be effective, writers also need to possess high levels of self-efficacy to engage in metacognitive monitoring of the sub-processes underlying revision. Furthermore, writing in second language (L2) poses additional constraints on writers’ revision process, especially for those of low L2 proficiency, rendering metacognition and self-efficacy more crucial. Comprising two studies, this mixed-methods research explored the effectiveness of a self-regulated strategy development (SRSD) model, a multicomponent strategy instruction approach that has shown to be effective in influencing students’ metacognitive knowledge and self-efficacy in the first language (L1) context, in an English-as-a-foreign-language (EFL) context. Study One, the preparatory stage of this research, reported the development and validation of two self-report instruments designed to measure students’ metacognitive knowledge about revision and selfefficacy for text revision in EFL writing. Data were collected from 446 undergraduate students from four universities in Central China via convenience sampling. Their responses were subjected to factor analyses and reliability tests to check the validity and reliability of the two newly-developed questionnaires. Study Two, the main study, explored the effects of the SRSD model on students’ metacognitive knowledge, self-efficacy for text revision, text revision performance and written text quality in EFL writing. Instructional effects of genre-specific writing evaluation criteria and general writing evaluation criteria taught within the SRSD model were compared. A quasi-experimental design with three instructional conditions, namely, SRSD with genre-specific criteria instruction (GSCI), SRSD with general writing criteria instruction (GCI), and regular writing instruction without SRSD, was carried out. Participants (n = 102) were recruited from three intact classes in a university in Central China; each of the three participating groups was assigned to one of the instructional conditions. The instructional effects were examined via the self-report instruments developed in Study One, writing and revising tests, and semi-structured interviews. Quantitative and qualitative results together pointed to the effectiveness of the SRSD instruction in improving students’ metacognitive knowledge and influencing their self-efficacy for text revision. Both SRSD groups reportedly became more knowledgeable of themselves as revisers with increased awareness of the strengths and weaknesses of their revision abilities. They also reported having gained a more comprehensive understanding of task schema for revision and more extensive knowledge of revision strategies after instruction. Positive changes in their self-efficacy for text revision emerged following instruction; both SRSD groups reported greater willingness to attend to problems involving higher-level features of a text, an activity they previously found demanding and thus avoided performing. Both SRSD groups made more revisions involving larger segments of texts and produced more meaningchanging revisions, contributing to improvements in overall text quality. Overall, this research provides empirical evidence to support claims of the effectiveness of both SRSD conditions in facilitating the development of students’ abilities to revise in an EFL context.