ANALYZING THE IMPACT OF CHATGPT USAGE ON LEARNING BEHAVIOUR AND ACADEMIC PERFORMANCE OF MCA AND MMS STUDENTS

Abstract

Artificial Intelligence (AI) tools, particularly ChatGPT, have rapidly emerged as widely adopted academic aids among university students. This study examines the effect of ChatGPT usage on the learning behavior and academic performance of postgraduate students pursuing MCA (Master of Computer Applications) and MMS (Master of Management Studies) programs at an Institute of Technology.

A structured questionnaire was administered to 80 students, capturing data on frequency of usage, purpose of use, average daily time spent, learning behavior through Likert-scale statements, and last semester SGPI. Three analytical techniques—K-Means Clustering, Pearson Correlation Coefficient, and Multiple Linear Regression—were applied to uncover behavioral groupings and performance relationships.

The findings reveal that 80% of students use ChatGPT frequently or always, primarily for concept understanding (42.5%) and exam preparation (26.25%). K-Means analysis identified three distinct user clusters. Pearson correlation indicated weak but directionally meaningful associations between usage frequency and SGPI.

The regression model identified usage frequency and answer verification as the most positive predictors of academic performance. The study concludes with implications for the responsible integration of AI tools in higher education.