生成式人工智能支持大学生自主学习的教学引导

Instructional Guidance for University Students’ Self-Regulated Learning Supported by Generative Artificial Intelligence

摘要: 生成式人工智能提高了大学生课后获取解释与反馈的便利性，但借助工具完成任务并不必然意味着形成理解。本文采用文献分析与教学情境讨论，围绕课程阅读、例题学习和报告修改，分析求助目标不清、生成内容判断依据不足及反馈未转化为后续行动等问题。依据自我调节学习的前瞻、表现和反思三个阶段，结合自我解释研究与反馈素养框架，提出相应的教学引导：在前瞻阶段根据已有知识定位困难并确定求助目标，在表现阶段依据课程内容监控理解并调整帮助，在反思阶段通过理由解释和有依据的修订评价进展、安排后续学习。教学引导应依据学生在具体任务中的知识准备和困难类型调整帮助，对先修知识不足的学生提供必要讲解，对已有思路的学生保留判断和解释的机会。上述建议旨在将工具使用与既有作业、课堂反馈相衔接，其适用性与实际效果仍需在真实课程中检验。

Abstract: Generative artificial intelligence makes explanations and feedback more accessible to university students, but completing a task with assistance does not necessarily indicate understanding. Drawing on a literature-based analysis and illustrative instructional situations involving course reading, worked examples, and report revision, this paper examines unclear help-seeking goals, insufficient criteria for judging generated content, and failure to translate feedback into subsequent action. The forethought, performance, and self-reflection phases of self-regulated learning are considered alongside self-explanation research and the student feedback literacy framework. The proposed guidance involves identifying difficulties and setting help-seeking goals during forethought, using course-based criteria to monitor understanding and adjust assistance during performance, and evaluating progress through reasoned explanations and revisions during self-reflection. Guidance should be adapted to students’ prior knowledge and difficulties in specific tasks. Necessary explanations should be provided when prerequisite knowledge is insufficient, while opportunities for judgment and explanation should be retained when students have an initial approach. These proposals connect tool use with existing assignments and classroom feedback; their applicability and effects require investigation in authentic courses.