AI-Driven Diagnostic and Tutoring Systems in Mathematics Classrooms: A User Experience Perspective

This study employed a user experience (UX) research approach to investigate students' interactions with an AI-driven diagnostic and tutoring system in real mathematics classrooms across six regions of Thailand. The system featured a five-level diagnostic structure and personalized learning interventions tailored to diverse performance levels. Using a stratified random sampling method, 294 Grade 7 students from 14 schools participated in the study. Data was collected using a validated Learner Feedback and Usability Questionnaire to capture perceptions of appropriateness, usability, and satisfaction. Results showed that 98% of students agreed with their assigned diagnostic level, and 81% expressed confidence in the system’s feedback and recommendations. While the average satisfaction rating was 3.5 out of 5, 22% of students did not engage with supplementary materials, and 15.92% reported dissatisfaction due to usability challenges. The findings provide practical insights into integrating AI tools in education, highlighting the system’s potential to enhance personalized learning and areas for UX-based improvement, including more precise navigation and adaptive content delivery.