From Dependence to Development: The Role of AI in Enhancing Writing Proficiency Among Chinese English Majors

Abstract

The integration of artificial intelligence (AI) tools into higher education writing instruction is reshaping traditional pedagogical practices, yet empirical evidence on their effectiveness remains limited among English majors. This study aimed to examine the impact of human-AI collaborative instruction on the writing proficiency of English majors in a Chinese university. Using a quasi-experimental design, 80 third-year English majors were assigned to either an experimental group (AI-assisted instruction with teacher guidance) or a control group (traditional instruction). Writing proficiency was measured through pre- and post-intervention IELTS writing tasks, scored by three trained raters using analytic band descriptors. Findings revealed that the experimental group achieved significantly higher gains in overall writing performance compared to the control group (p < .001), with particularly large improvements in lexical resource and grammatical accuracy. ANCOVA results confirmed the instructional approach as a significant predictor of post-test performance, independent of baseline scores. The findings highlight the pedagogical potential of human-AI collaboration in improving writing outcomes and offer empirical support for its integration into curriculum design of English major.