Validation of Self-Regulated Writing Strategies for Advanced EFL Learners in China: A Structural Equation Modeling Analysis.

Abstract

This study aims to validate self-regulated writing strategies for advanced EFL learners through a structural equation modeling analysis. Two sets of advanced, university-level EFL learners in China were recruited on the basis of results from a nationwide standardized English test. Sample 1 consisted of 214 advanced learners and served mainly as a data source for exploratory factor analysis. Sample 2 consisted of 303 advanced learners; data from this group were used to conduct confirmatory factor analyses. The results confirmed the goodness of fit of the hierarchical, multidimensional structure of self-regulated writing strategies. This hierarchic model has the higher order of self-regulation and the second order of nine self-regulated writing strategies that belong to four dimensions. In terms of model comparisons, the indices of Model 1 (nine-factor correlated model of EFL writing strategies for SRL) and Model 2 (four-factor second-order model of EFL writing strategies for SRL) mark significant improvements in terms of fit over the indices of Model 3 (one-factor second-order model of EFL writing strategies for SRL). This means the four-factor model (cognition, metacognition, social behavior, and motivational regulation) offered a better explanation for advanced EFL learners than the model treating self-regulated writing strategies as a single convergent factor. These findings, in some ways, differ from the results of earlier research on EFL learners' self-regulated writing strategies, and the findings of this study have certain implications for L2 writing teaching and learning.