A Study on the Construction of a “Teacher-Led, GenAI-Assisted” Dual-Layer Model for Senior High School English Writing Feedback

Abstract

In high school English writing instruction, GenAI tools (such as ChatGPT, DeepSeek, Kimi, etc.) possess the capability to provide students with immediate, personalized feedback. However, they also pose risks such as over-reliance and homogenization of thinking. Targeting the cognitive maturity and autonomous learning abilities of senior high school students in China, this study constructs a “teacher-led, GenAI-assisted” dual-layer model. The first layer involves the teacher in setting goals, guiding strategies, and fostering critical judgment. The second layer utilizes GenAI to provide feedback on linguistic forms, pattern recognition, and expression suggestions.