Enhancing Students’ Learning Outcomes in Mathematics through Intelligent Tutoring Systems Based on Real-Time Feedback

Abstract

This study examined the effectiveness of an intelligent tutoring system (ITS) driven by real-time feedback in enhancing students’ mathematical learning outcomes, as defined by the Structure of Observed Learning Outcomes. Conducted within the domains of Measurement and Geometry, the study employed a randomized controlled trial involving 120 students from the Mathematics Program for Gifted Students at Khon Kaen University Demonstration School. Of these, 78 were selected through systematic random sampling and assigned to experimental and control groups. The experimental group engaged with the ITS featuring automated, real-time feedback, while the control group used the same system without feedback. Both groups utilized the system through adaptive diagnostic assessments. The diagnostic test demonstrated strong psychometric properties (Rasch model difficulty range = -2.34 to 1.71; Cronbach’s α = 0.81; IRT reliability = 0.89). Statistical analyses—including independent t-tests, repeated measures ANOVA, and relative gain scores—revealed significant learning gains in the experimental group (p < .05). Mean scores increased from 1.90 to 3.54 in the experimental group and from 1.90 to 2.85 in the control group. The experimental group reached an advanced level (M = 84.83%), significantly outperforming the control group (M = 61.54%), with an effect size of 0.73. These findings highlight the potential of ITS with diagnostic feedback to foster deep, structured mathematical understanding and offer valuable insights for personalized digital learning.