AI-Based Personalized Learning and Student Performance: Evidence from Indian HigherEducation

Abstract

While the number of AI-powered educational platforms being developed and widely adopted in higher education institutions in India is rapidly increasing, there is not a lot of empirical evidence to determine whether such systems are effective for enhancing student outcomes. The study involved giving a structured quantitative questionnaire of the selected colleges of India to undergraduate and postgraduate students. Out of the 506 responses obtained, there were 497 responses that were retained after informed consent screening. The instrument measured four constructs related to AI-based personalized learning, students' learning engagement, and their satisfaction and performance using a 24-item, five-point Likert-scale. Descriptive statistics, Pearson correlation, multiple regression and bootstrapped mediation analysis were conducted. Personalized learning based on AI, along with engagement and satisfaction, accounted for 75.1 % of the variance in academic performance (R2 = 0.751, p < 0.001), with both mediators shown to be significant. The results provide empirical insights on how to guide institutions, policymakers, and EdTech companies to reinforce learning outcomes through AI for personalisation in Indian higher education.