中国电化教育

生成式人工智能的广泛应用正在重塑大学生的学习行为方式。基于问卷调查数据,采用结构方程模型,实证分析生成式人工智能介入高校学习情境中大学生学习行为的作用机制。研究发现,感知易用性与感知有用性显著推动学习行为由传统模式向技术辅助模式转向,且感知易用性会通过感知有用性的中介效应进而发挥更为显著的作用;近端社会规范和制度情景对学习行为转向产生积极的作用,且通过感知有用性的中介效应,对学习行为转向产生更为显著的作用。相较之下,风险认知对学习行为转向的影响不明显,表明在当前高校学习情境中,技术带来的即时效率收益在行为决策中占据主导地位,而相关规范与风险治理尚未充分内化为稳定的学习行为规则。研究结果从学习行为视角拓展了技术接受模型的解释边界,为理解生成式人工智能嵌入高等教育学习实践的行为机制提供了证据。

The widespread application of generative artificial intelligence(GenAI) is reshaping university students' learning behaviors. Based on questionnaire survey data, this study employs structural equation modeling to empirically examine the mechanisms through which Gen AI influences learning behaviors in higher education contexts. The results indicate that perceived ease of use and perceived usefulness significantly promote the shift of learning behaviors from traditional modes toward technology-assisted patterns, with perceived ease of use playing a more prominent role. Interpersonal relationships and the social environment exert significant effects on the synchronized migration of learning behaviors by providing normative signals and legitimacy support. Moreover, social environment and perceived ease of use indirectly reinforce the sustained transition of learning behaviors by strengthening students' perceptions of the learning value of Gen AI. In contrast, risk perception does not exert a significant inhibitory effect on learning behavior transition, suggesting that in current higher education settings, immediate efficiency gains brought by GenAI dominate behavioral decision-making, while norms and risk governance related to GenAI have not yet been fully internalized as stable behavioral boundaries. By adopting a learning behavior perspective, this study extends the explanatory scope of technology acceptance research and provides empirical evidence for understanding the behavioral mechanisms through which generative artificial intelligence becomes embedded in higher education learning practices.