【作者】 张嘉婧； 郭又铭； 朱神海；

【Author】 ZHANG Jiajing;GUO Youming;ZHU Shenhai;College of Foreign Study,Guangxi Normal University;

【机构】 广西师范大学外国语学院；

【摘要】 本研究在自动作文评分系统(AES)技术发展背景下,以豆包AI为例,对比分析了生成式人工智能(GAI)与中学英语教师在英语写作反馈方面的一致性、态度与策略差异。通过分析50篇学生作文发现,二者的评分结果具有较高一致性,但教师评分相对更宽松。在反馈态度上,豆包AI的反馈效率更高、更积极,尤其擅长内容层面的评价。反馈策略方面,豆包AI偏向于直接纠正,而教师则更多采用元语言引导。研究表明,豆包AI可作为有效的写作反馈辅助工具,但仍需教师介入,以提升反馈的人文性与综合性。本研究为GAI赋能英语写作教学提供了实证依据,并对构建人机协同反馈模式提出了建议。更多还原

【Abstract】 Under the context of Automated Essay Scoring(AES) technology development,this study takes Doubao AI as an example to comparatively analyze the consistency,attitudes,and strategic differences between Generative Artificial Intelligence(GAI) and middle school English teachers in providing feedback on English writing.By analyzing 50 student essays,the findings reveal that their ratings exhibit high consistency,but teacher ratings are relatively more lenient.In terms of feedback attitude,Doubao AI demonstrates higher efficiency and more positive feedback,particularly excelling in evaluating the content dimension.Regarding feedback strategies,Doubao AI tends toward direct correction,whereas teachers more frequently employ metalanguage for guidance.The research indicates that Doubao AI can serve as an effective auxiliary tool for writing feedback,but teacher intervention remains necessary to enhance the humanistic and comprehensive aspects of the feedback.This study provides empirical evidence for GAI-enabled English writing instruction and offers suggestions for constructing a human-computer collaborative feedback model.更多还原