ChatGPT, Achievement Motivation, and Problem-Solving: Evidence from Thai and Cambodian Undergraduates

Abstract

Background. The rapid global adoption of ChatGPT has raised growing interest in its role in supporting student learning, yet the effectiveness of generative AI tools in enhancing higher-order cognitive skills such as problem-solving may depend on individual learner characteristics that remain underexplored. Purpose. This quantitative study aimed to examine the relationship between ChatGPT usage and problem-solving ability among students enrolled in a Media Design course, and to test whether this relationship is moderated by achievement motivation. Method. The study involved 115 university students from four intact classes across two higher education institutions in Thailand and Cambodia. Data were collected through a survey questionnaire and analysed using Partial Least Squares Structural Equation Modeling (PLS-SEM), preceded by a measurement invariance test (MICOM) to assess the appropriateness of pooling data across the two national samples. Results. The findings indicate that ChatGPT usage was positively and significantly associated with problem-solving ability (? = 0.386, p < 0.001), as was achievement motivation (? = 0.291, p < 0.01). The interaction between ChatGPT usage and achievement motivation also showed a significant positive effect (? = 0.157, p < 0.05), indicating that achievement motivation strengthens the relationship between ChatGPT usage and problem-solving ability. The model explained 44.6% of the variance in problem-solving ability (R² = 0.446). Conclusion. This study has significant implications for instructional practices in higher education settings, particularly for lecturers designing AI-integrated learning activities. By understanding how achievement motivation shapes students’ benefit from ChatGPT, educators can tailor instructional strategies that support both technology adoption and the development of independent problem solving skills.