Guo Fei, Zhang Lanwen, Shi Tianle, Coates Hamish
School of Education, Tsinghua University, Haidian District, Beijing 100084, China.
Policy Research and Planning Office, Tsinghua University, Haidian District, Beijing 100084, China.
Behav Sci (Basel). 2025 Jul 25;15(8):1011. doi: 10.3390/bs15081011.
Generative AI (GenAI) technologies have been widely adopted by college students since the launch of ChatGPT in late 2022. While the debate about GenAI's role in higher education continues, there is a lack of empirical evidence regarding whether and when these technologies can improve the learning experience for college students. This study utilizes data from a survey of 72,615 undergraduate students across 25 universities and colleges in China to explore the relationships between GenAI use and student learning engagement in different learning environments. The findings reveal that over sixty percent of Chinese college students use GenAI technologies in Academic Year 2023-2024, with academic use exceeding daily use. GenAI use in academic tasks is related to more cognitive and emotional engagement, though it may also reduce active learning activities and learning motivation. Furthermore, this study highlights that the role of GenAI varies across learning environments. The positive associations of GenAI and student engagement are most prominent for students in "high-challenge and high-support" learning contexts, while GenAI use is mostly negatively associated with student engagement in "low-challenge, high-support" courses. These findings suggest that while GenAI plays a valuable role in the learning process for college students, its effectiveness is fundamentally conditioned by the instructional design of human teachers.
自2022年底ChatGPT推出以来,生成式人工智能(GenAI)技术已被大学生广泛采用。虽然关于GenAI在高等教育中作用的争论仍在继续,但缺乏关于这些技术是否以及何时能够改善大学生学习体验的实证证据。本研究利用对中国25所大学和学院的72,615名本科生进行的一项调查数据,探讨GenAI的使用与不同学习环境中学生学习参与度之间的关系。研究结果显示,超过60%的中国大学生在2023 - 2024学年使用GenAI技术,学术用途超过日常用途。在学术任务中使用GenAI与更多的认知和情感参与有关,尽管它也可能减少主动学习活动和学习动机。此外,本研究强调GenAI的作用在不同学习环境中有所不同。GenAI与学生参与度的积极关联在“高挑战和高支持”学习环境中的学生中最为显著,而在“低挑战、高支持”课程中,GenAI的使用大多与学生参与度呈负相关。这些发现表明,虽然GenAI在大学生的学习过程中发挥着重要作用,但其有效性从根本上受到人类教师教学设计的制约。