Integration of artificial intelligence (AI) in problem-based learning (PBL) for chemistry education

Abstract

This study used the Decision-Making Trial and Evaluation Laboratory (DEMATEL) method to examine the cause-and-effect relationships among key factors influencing the integration of Artificial Intelligence (AI) in Problem-Based Learning (PBL) for chemistry education. Four main factors were identified and analysed: AI guidance quality, AI adaptability to student mastery levels, AI support for scientific reasoning, and student engagement and motivation. Expert evaluation and cause-effect analysis revealed that AI guidance quality (D–R = 1.378) and AI adaptability to student levels (D–R = 0.348) are the main driving factors in this system. This means both factors exert strong influence on the other factors. In contrast, AI support for scientific reasoning (D–R = –0.477) and student engagement and motivation (D–R = –1.249) function primarily as effect factors—meaning they are influenced by other AI qualities and functions. Priority analysis based on D+R values showed that AI support for scientific reasoning is the most important factor in the overall system (6.227), followed by AI guidance quality (6.163) and AI adaptability (6.045). Overall, these findings emphasize that effective use of AI in chemistry PBL depends more on pedagogy-based AI design rather than technological features alone. AI guidance quality and its ability to adapt to student levels serve as essential foundations for enhancing students' scientific thinking abilities while also increasing their engagement and motivation in learning.