Enhancing Critical Thinking in EFL Writing Through an AI-Supported Blended Learning Model

Abstract

This study introduces a novel, critical thinking-oriented blended learning model for English as a Foreign Language (EFL) writing instruction. It addresses the gap in the lack of explicit cognitive guidance for critical thinking in many blended learning models, which often prioritize technology over pedagogy. Grounded in the Paul-Elder Triadic Model and the process-genre approach, this model integrates online, face-to-face, and AI-enhanced learning to cultivate critical thinking within writing development. The model features three hierarchical layers: Goal (critical thinking objectives), Process (pre/in/post-class activities), and Environment (integrating online, offline, and AI components). A quasi-experimental study (N=52 non-English majors, Chinese technical university) compared the model to traditional instruction. Results from the International Critical Thinking Test and TOEFL-based writing assessments indicated significantly improved critical thinking skills in the experimental group, specifically in identifying assumptions, concepts, and inferences, and applying intellectual standards (clarity, relevance, breadth). Writing proficiency improved in both groups, but without significant between-group differences, possibly due to intervention length and the importance of linguistic ability in EFL writing. This study provides empirical evidence for the effectiveness of a theory-driven, AI-enhanced blended learning model specifically designed for critical thinking in EFL writing, extending beyond general approaches by strategically integrating AI to support distinct writing process stages and critical thinking development.