大学生对英语写作课程中人工智能工具的感知研究——以贺州学院为例

Research on University Students’ Perceptions of Artificial Intelligence Tools in English Writing Courses—A Case Study of Hezhou University

摘要: 文章以贺州学院为例，探讨了地方应用型本科院校大学生对人工智能英语写作工具的感知与使用现状。通过对300名在校大学生的问卷调查与焦点小组访谈，文章系统分析了学生对机器翻译工具(如谷歌翻译)与生成式智能写作助手(如Grammarly)的使用偏好、效果评价及影响因素。研究发现，超过75%的学生更倾向于使用机器翻译工具进行语言转换，而智能写作助手主要用于文本润色与风格优化。研究进一步揭示了英语水平差异对工具使用模式的显著影响：高水平学习者能够批判性地将工具作为辅助资源，而低水平学习者则易形成依赖性使用习惯。该研究发现人工智能工具在英语写作教学中具有双重效应：一方面能有效缓解学生的写作焦虑，提供即时语言支持；另一方面过度依赖可能削弱学生的基础写作能力发展。工具的实际效用受到学生英语水平、使用目的及工具功能特性的共同影响。基于研究结果，文章提出构建“教师主导–技术辅助”的协同教学模式，并建议针对不同水平学习者实施分级引导策略。最后，研究对人工智能工具的教育适用性优化及未来研究方向提出了具体建议，为地方院校英语写作教学改革提供参考依据。

Abstract: This study examines the perceptions and usage patterns of artificial intelligence-powered English writing tools among undergraduates at Hezhou University, a regional applied undergraduate institution. Through questionnaire surveys and focus group interviews with 300 enrolled students, it systematically analyses their preferences, efficacy assessments, and influencing factors regarding machine translation tools (e.g., Google Translate) and generative AI writing assistants (e.g., Grammarly). Findings reveal that over 75% of students favour machine translation tools for language conversion, while intelligent writing assistants are primarily employed for text refinement and stylistic optimisation. The study further highlights the significant impact of varying English proficiency levels on tool usage patterns: high-level learners critically utilise tools as supplementary resources, whereas lower-level learners tend to develop dependent usage habits. This study identifies a dual effect of AI tools in English writing instruction: while they effectively alleviate writing anxiety by providing instant language support, overreliance may undermine the development of foundational writing skills. The practical efficacy of these tools is jointly influenced by learners’ English proficiency, usage objectives, and tool functionalities. Based on these findings, the paper proposes establishing a collaborative teaching model centred on ‘teacher-led instruction with technology-assisted support’ and recommends implementing tiered guidance strategies tailored to learners of varying proficiency levels. Finally, this study offers specific recommendations for optimising the educational applicability of AI tools and future research directions, providing a reference framework for reforming English writing instruction in local institutions.