Undergraduate English Majors' Views on ChatGPT in Academic Writing: Perceived Vocabulary and Grammar Improvement

Abstract

This study investigates undergraduate English majors' perceptions of ChatGPT in enhancing vocabulary and grammar in academic writing. Utilizing a mixed-methods convergent design, data were collected from 31 students via pre- and post-survey questionnaires, 20 participants’ reflective journals, and semi-structured interviews with 10 volunteers. Quantitative findings revealed significant improvements in students’ perceptions of vocabulary accuracy, relevance, and depth. Thematic analysis of qualitative data identified benefits such as enriched vocabulary, improved grammatical accuracy, and increased confidence in academic writing. Challenges included overdependence, difficulty interpreting feedback, and a lack of originality in AI-generated suggestions. Students employed strategies to optimize ChatGPT use, such as asking specific questions, selectively applying feedback, and balancing AI input with personal judgment. The study highlights ChatGPT's ability to provide tailored feedback, foster confidence, and support vocabulary development while underscoring the importance of responsible use to mitigate overreliance and maintain originality. The findings underscore ChatGPT's potential to enhance academic writing skills when integrated thoughtfully into curricula. However, overuse risks shallow learning, suggesting a need for instructional strategies that can promote rigorous analysis with AI tools. Future research should explore long-term impacts, comparisons with other AI tools, and strategies for ethical and effective integration of ChatGPT in higher education.