How to prompt ethical use? Investigating students’ intention to use ChatGPT for graded assignments

This study investigates the factors influencing university students’ intention to use ChatGPT for schoolwork, as well as how they determine when such use is acceptable. Focusing on university students in Singapore, we employed a mixed-methods design comprising two phases. Study 1 involved an online survey (n = 540) guided by the Theory of Planned Behavior, which found that injunctive norms, perceived behavioral control, and outcome expectancy significantly predicted students’ intention to use ChatGPT for graded writing assignments. Study 2 consisted of in-depth online interviews (n = 20), which provided insight into how students define acceptable versus unacceptable use of ChatGPT. The interviews found that students’ ethical boundaries are shaped by both pragmatic and normative considerations, with grade weightage and perceived impact on learning serving as key reference points in determining ethicality.

本研究考察了影响大学生将ChatGPT用于学业的使用意愿，以及他们如何判定此类使用行为的可接受性。本研究以新加坡大学生为研究对象，采用包含两个阶段的混合研究设计。研究1为基于计划行为理论（Theory of Planned Behavior）的线上问卷调查，有效样本量为540份，结果显示指令性规范、感知行为控制以及结果期望均能显著预测大学生在评分写作任务中使用ChatGPT的意愿。研究2为深度线上访谈，受访样本量为20人，该研究揭示了大学生如何界定ChatGPT使用的可接受与不可接受边界。访谈结果表明，学生的伦理边界同时受实用主义考量与规范性考量的共同塑造，其中成绩占比以及对学习的感知影响，是判定使用行为伦理合理性的核心参考依据。