Student engagement in ChatGPT-assisted mathematics learning: A phenomenological study

This study explored the lived experiences of high school students using ChatGPT in mathematics learning through a student engagement framework and the interpretive lens of Self-Determination Theory (SDT). Using a descriptive phenomenological design, data were gathered from 58 students through eight focus group discussions and analyzed using Colaizzi's method. The findings revealed that participants described ChatGPT as shaping their engagement in multiple ways. Students valued its accessibility, immediate feedback, and step-by-step explanations, which they described as supporting their participation, conceptual understanding, and confidence in learning mathematics. Emotional engagement was reflected in students' experiences of psychological safety and perceived support, while cognitive engagement was reflected in participants' perceptions that ChatGPT supported their understanding and critical evaluation of AI-generated responses. However, participants also reported disengaging experiences, including overreliance on ChatGPT, reduced interaction with teachers and peers, challenges in using mathematical symbols, and concerns regarding inaccurate responses. These findings suggest that participants perceived ChatGPT as creating both opportunities and constraints for meaningful mathematics learning. Interpreted through SDT, participants' accounts reflected experiences that may be interpreted as supporting or frustrating psychological needs related to autonomy, competence, and relatedness. This study contributes to the growing literature on AI-assisted learning by integrating student engagement and SDT to provide a nuanced understanding of students' experiences. The findings underscore the importance of guided AI use, critical evaluation, teacher support, and collaborative learning to promote meaningful and sustainable engagement in mathematics.