Impact of AI-Based Systems on student engagement and mathematics achievement
An intelligent tutoring and an adaptive learning system
Students from socioeconomically disadvantaged backgrounds tend to struggle with mathematics, scoring significantly below the state and national average. Students at the high school, located in South Georgia, United States, have shown poor mathematics achievement. To narrow the learning gap, the school implemented the ALEKS, an Intelligent Tutoring System, and Edmentum Exact Path, an Adaptive Learning System. I wanted to inquire if these AI-based systems enhance student engagement and mathematics achievement when used as a supplemental tool.
Teacher versus System
This study utilized a quasi-experimental design comparing teacher-led instruction with Edmentum Exact-Path-led instruction when used as a supplemental tool. The results of this study indicated that AI-based systems have the potential to enhance the mathematics achievement of underachieving students in a rural context. Students who received traditional teacher-led instruction demonstrated higher levels of both affective and cognitive engagement, whereas students who received Edmentum-led instruction showed only a positive impact on cognitive engagement, not affective engagement. This highlights the important aspect of keeping humans in the loop, as AI-based systems cannot provide emotional support to students.
Supplemental tool
This thesis highlights the importance of AI-based systems in enhancing mathematics achievement and student engagement. These systems are effective when used as a supplemental tool alongside teacher-led instruction. The findings from this study suggest that AI-based systems can provide adaptive and personalized instructions to the students from socioeconomically disadvantaged populations, especially in rural settings where limited resources and qualified teachers in subjects like mathematics are scarce. In under-resourced schools, AI-based systems have the potential to bridge equity gaps by addressing individual needs.
Teacher-student interaction is critical
Although students who used AI-based systems showed improvement in mathematics achievement, students in the teacher-led instruction group demonstrated greater gains in mathematics achievement and higher levels of cognitive achievement. This highlights the critical role of teacher-student interaction and supports the notion that AI-based systems should augment rather than replace traditional instructions.
About Rashmi Khazanchi
Rashmi Khazanchi has been a K–12 educator for 25 years. She earned a master’s degree in organic chemistry from Maharshi Dayanand University, India, in 1995, and a postgraduate diploma in health psychology and behavior modification from Amity University, New Delhi, India, in 2003. She later completed an Educational Specialist (Ed.S.) degree in Curriculum and Instruction from Lincoln Memorial University in Tennessee, United States, in 2013.
On Friday, May 8, 2026, at 13:30 (CET), she will defend her dissertation, Artificial Intelligence in Education: Impact of AI-Based Systems on Mathematics Achievement, in an online defense organized by the Faculty of Behavioural and Health Sciences at the Open University in Heerlen.
Promotor is Prof. Dr. Hendrik Drachsler (Open Universiteit), and co-promotor is Prof. Dr. Daniele Di Mitri (German University of Digital Science). You can follow the thesis defense live: www.ou.nl/live.
