Examining the role of social influence, learning value and habit on students’ intention to use ChatGPT: the moderating effect of information accuracy in the UTAUT2 model

This study examined factors influencing students’ intention to use ChatGPT using UTAUT2 model. The cross-sectional study is based on responses collected from 578 students selected through convenience sampling at a university in Nepal through a structured questionnaire. Using PLS-SEM, the study found that habit (β = 0.315, p < 0.001), learning value (β = 0.254, p < 0.001), and social influence (β = 0.127, p < 0.001) were the significant factors positively affecting students’ intention to use ChatGPT. In contrary, effort expectancy, hedonic motivation, facilitating conditions, and performance expectancy did not significantly influence intention to use. The study extended the UTAUT2 model by incorporating information accuracy as a moderating variable. However, information accuracy of ChatGPT did not moderate the hypothesized relationships. This study provides valuable insights for intention to use ChatGPT in education, with a unique focus on the moderating role of information accuracy. The findings offer practical implications for enhancing the adoption and effectiveness of AI tools in education, contributing significantly to the literature on educational technology.

本研究基于统一技术接受与使用理论2模型（UTAUT2），探究影响学生使用ChatGPT意愿的相关因素。本项横断面研究依托结构化问卷收集数据，研究样本为通过便利抽样法选取的尼泊尔某高校578名学生。本研究采用偏最小二乘结构方程模型（PLS-SEM）开展分析，结果显示：使用习惯（β = 0.315, p < 0.001）、学习价值（β = 0.254, p < 0.001）与社会影响（β = 0.127, p < 0.001）为显著正向影响学生使用ChatGPT意愿的因素；与之相反，努力期望、享乐动机、便利条件与绩效期望对学生的使用意愿无显著影响。本研究将信息准确性作为调节变量纳入UTAUT2模型以拓展其应用场景，但研究并未发现ChatGPT的信息准确性对上述假设的变量关系存在调节作用。本研究聚焦信息准确性的调节作用，为教育场景下ChatGPT使用意愿的相关研究提供了极具价值的理论洞见。研究结果可为提升教育领域人工智能工具的采纳率与应用效能提供实践参考，同时为教育技术相关学术文献作出了重要贡献。