Generative AI Empowering College English Writing Teaching Model Construction, Practical Path, and Effect Evaluation

Abstract

To address persistent problems in college English writing teaching, including low student motivation, delayed teacher feedback, insufficient individualized guidance, and single evaluation methods, this study develops a generative AI-enabled “inspiration-collaboration-internalization” teaching model. The model designs human-computer collaboration across the writing process, including brainstorming, drafting, revising, and reflection. Generative AI is used to provide personalized learning materials, alternative writing perspectives, and timely language feedback, while teachers focus on higher-order guidance involving argument depth, logical coherence, and structural organization. The model further encourages students to gradually reduce dependence on AI support and develop independent writing ability. A multilevel evaluation mechanism combining AI-assisted initial assessment and teacher feedback is established. A one-semester quasi-experiment and mixed-methods evaluation show that the experimental group’s writing score increased more than the control group’s, with gains of 1.65 versus 0.72 points and an effect size of 0.82. Students reported high satisfaction, with M = 4.28, and AI-use frequency decreased as writing competence improved.