大学化学

上一篇    下一篇

人工智能时代“学生-同伴-AI-教师”协同课堂：费曼学习法驱动的物理化学教学变革

李荣金1,2, 朱荣娇1,2, 陈丽1,2, 冯霞1,2, 刘俊吉1,2   

  1. 1 天津大学理学院化学系，天津 300072;
    2 化学化工国家级实验教学示范中心(天津大学)，天津 300354
  • 收稿日期:2025-11-20 录用日期:2025-12-03
  • 通讯作者: 李荣金 E-mail:lirj@tju.edu.cn Rongjin Li
  • 基金资助:
    天津大学教改项目(DSFZ2025030)

Teaching transformation in physical chemistry through the Feynman technique: building a student–peer–AI–instructor collaborative classroom in the age of artificial intelligence

Rongjin Li1,2, Rongjiao Zhu1,2, Li Chen1,2, Xia Feng1,2, Junji Liu1,2   

  1. 1 Department of Chemistry, School of Science, Tianjin University, Tianjin 300072, China;
    2 National Demonstration Center for Experimental Chemistry and Chemical Engineering Education, Tianjin University, Tianjin 300354, China
  • Received:2025-11-20 Accepted:2025-12-03
  • Contact: Rongjin Li E-mail:lirj@tju.edu.cn

PDF

132

赞

摘要： 物理化学概念抽象、逻辑严密，传统教学模式易导致学生陷入“会计算却不理解、能记忆却难迁移”的困境。生成式人工智能(AI)在提供学习支持的同时，也带来了“认知外包”的风险。如何在利用AI优势的同时保障学生的自主思考能力，成为教学改革的关键。本文基于生成性学习理论，引入费曼学习法，构建了“课前预学→课堂微讲解→同伴互动→板书推演→教师精讲→检索练习→实验拓展”的物理化学教学新范式。费曼学习法的引入能有效揭示并矫正学生的概念理解误区，提升其学科表达与逻辑推理能力。通过与人工智能工具深度融合，构建起“学生-同伴-AI-教师”协同的学习生态，在发挥AI优势的同时有效管控认知外包风险，为人工智能时代的高阶思维能力培养提供了可行方案。

关键词: 教育学理论, 教学方法, 费曼学习法, 讲中学, 生成式人工智能

Abstract: Physical chemistry is characterized by abstract concepts and rigorous logical frameworks, yet conventional teaching methods often leave students trapped in a paradoxical state of “mastering calculations without genuine comprehension” and “memorizing content without achieving knowledge transfer”. While generative artificial intelligence (AI) provides substantial learning support, it simultaneously introduces the risk of “cognitive outsourcing”. The key challenge for contemporary pedagogical reform lies in harnessing AI’s advantages while preserving students’ capacity for independent thinking. Drawing upon generative learning theory, this study incorporates the Feynman Learning Technique to develop an innovative instructional framework for physical chemistry, structured around seven key phases: pre-class preparation, in-class micro-teaching demonstrations, peer interaction, whiteboard derivations, instructor-led clarification, retrieval practice, and experimental extensions. The integration of the Feynman approach effectively identifies and rectifies students’ conceptual misunderstandings while enhancing their disciplinary communication skills and logical reasoning abilities. Through deep integration with AI tools, this model establishes a synergistic learning ecosystem connecting students, peers, AI systems, and instructors. This approach capitalizes on AI’s benefits while effectively mitigating cognitive outsourcing risks, offering a practical solution for cultivating higher-order thinking skills in the artificial intelligence era.

Key words: Educational theory, Instructional methods, Feynman technique, Learning-by-teaching, Generative artificial intelligence