Article summary
Abstract
The rapid development of artificial intellengence (AI) has created new opportunities
for improving mathematics teaching and learning. chatGPT, as a generative AI-based
conversational learning tool, can provide students with immediate feedback,
explanations, examples and alternative solution strategies. This study examines the
effects of ChatGPT- assisted learning on student’s mathematical reasoning, problem
solving ability and mathematical achievement in Bagmati Provience, Nepal. A quasi
experimental pre- test- post- test control group design was proposed for the study. The
study used the sample of 300 secondary- level students, comprising 150 in students
in the experimental group and 150 in the control group. The experimental group
received mathematics instruction supported by chatGPT, whereas the control group
received conventional instruction. Mathematical reasoning, problem solving and
achievement were measured through structured tests administered before and after
the intervention. The stimulated findings indicated that the experimental group
demonstrated greater improvement in all three outcomes. Mathematical reasoning
increased from a pre –test mean 24.80 to 36.70 , problem solving performance from
23.90 to 35.90 and mathematical achievement from 52.10 to 70.80. the corresponding
post- test mean for the contro group were 28.90, 28.40 and 61.50 respectively.
Independent- sample t- tests indicated statistically significant differences between
the groups for mathematical reasoning, t(197) = 16.02, P < 0.001, problem solving,
t(198) = 15.15, p < o.oo1 and achievement, t(196) = 16.51, P < o.oo1. The findings
suggest that chatGPT- assisted learning has potential as a supplementry approach
for enhancing mathematics learning. However, effective teacher guidance verification
of AI generated responses and responsible use of ChatGPT are essential to avoid
overdependence on AI. The study provides a basis for further empirical investigation
of generative AI supported mathematics education in Nepal.