The result of Analysis of Factors Influencing the Sustainability of Generative AI Use in the Learning Process Among Jabodetabek University Students.xlsx

This research analyzes the factors that affect university students in the Jabodetabek area to continue using ChatGPT in their learning process. With the fast adoption of generative AI tools in education, ChatGPT is one of several platforms that has become widely used in the world for academic assistance. For the sustainability of ChatGPT usage not only gets affected by technical capabilities but also by psychological and social factors. This research used extended Technology Acceptance Model (TAM) by combining mediating variables like Perceived Trust and Satisfaction, and external variables like Competence and Social Influence. 415 valid respondents who are active university students and ChatGPT users collected with quantitative methods. SMART Partial Least Squares Structural Equation Modeling (PLS-SEM), SmartPLS 4 were used to analyze the data. All hypotheses proposed are significant statistically, with the most influential factors are Perceived Usefulness and Competence. Satisfaction and Trust serve as critical mediators, and Continuous Usage Behavior is also affected by Social Influence. The results found that the importance of improving digital competence of the students, trust, and fostering ethical AI usage in academic policies and support systems, provides valuable implications for future studies and academic exploration.

本研究旨在探讨雅加达大都会区（Jabodetabek）大学生在学习过程中持续使用ChatGPT的影响因素。随着生成式AI（Generative AI）工具在教育领域的快速普及，ChatGPT已成为全球范围内广泛应用于学术辅助的平台之一。ChatGPT的使用持续性不仅受技术性能的影响，同时也与心理、社会因素密切相关。本研究对技术接受模型（Technology Acceptance Model, TAM）进行扩展，纳入感知信任与满意度作为中介变量，并引入胜任力、社会影响作为外部变量。研究通过定量研究方法共收集到415份有效问卷，受访者均为在读大学生且为ChatGPT使用者。本研究采用偏最小二乘结构方程模型（Partial Least Squares Structural Equation Modeling, PLS-SEM）结合SmartPLS 4工具开展数据分析。所有提出的研究假设均具有统计学显著性，其中影响力最强的因素为感知有用性与胜任力；满意度与信任发挥关键中介作用，持续使用行为同时受社会影响的显著正向作用。研究结果显示，提升学生数字胜任力、构建信任体系，以及在学术政策与支持系统中推广人工智能伦理使用，可为后续相关研究与学术探索提供重要的参考价值。