【作 者】：
李海峰， 王 炜
【关 键 词 】：
生成式人工智能； 高阶思维； 教学模式； 人机协同； 智能教育
【栏 目】：
课程与教学
【中文摘要】：
大量学生利用生成式人工智能生成作业，导致了学生高阶思维能力难以有效培养。在人机会话中嵌入“争论噪音”，将是解决这一问题的可能途径。研究采用准实验研究设计，基于高阶思维、建构主义和争论学习理论，利用文心一言、学习分析和腾讯QQ等技术工具，开发了争论式智能会话机器人，构建了人机争论探究法教学模式。结果表明，实验组的学习成绩、批判性思维能力、问题解决能力和学习态度显著优于对照组，但是创新能力效果不显著。为提高人机争论学习效果，教师需要构建特定的高阶思维学习支架，优化人机争论学习的算法机制，培养学生的人机争论素养，研制师生机三元协同争论机制。
【英文摘要】：
A large number of students use generative artificial intelligence to generate homework, which makes it difficult to effectively cultivate students' higher-order thinking ability. Embedding "argument noise" in human-machine conversations would be a possible way to solve this problem. Based on higher-order thinking, constructivism and argumentative learning theory, the study adopts quasi-experimental research design and uses Ernie Bot, Learning Analytics, Tencent QQ and other technical tools to develop argumentative intelligent chatbots and construct a teaching model of human-machine arguing inquiry method. The results show that the experimental group's academic performance, critical thinking ability, problem solving ability and learning attitude are significantly better than those of the control group, but the effect of innovation ability is not significant. In order to improve the effectiveness of human-machine argument learning, teachers need to build a specific learning scaffold for higher-order thinking, optimize the algorithmic mechanism of human-machine argument learning, cultivate students' human-machine argument literacy, and develop a teacher-student-machine ternary collaborative argument mechanism.