MACHINE LEARNING FOR PERSONALIZED EDUCATION: A STUDY ON ADAPTIVE LEARNING SYSTEMS

Abstract

This research explores the application of machine learning-driven adaptive learning technologies in higher education and their effectiveness in Pakistani context using mixed-methods approach. For a balanced representation of disciplines and socioeconomic status, the researchers implemented stratified random sampling to select 250 undergraduate students from five universities located in Lahore and Karachi. For six months, the research team utilized a machine learning-based adaptive learning system which personalized modules according to learners’ patterns, performances, and knowledge gaps. The research design encompassed both qualitative and quantitative methodologies which included the analysis of pre and post-test scores, learning analytics generated by the adaptive learning system, semi structured interviews conducted with 30 students and 15 faculty members, and institutional technology-challenge surveys. The quantitative findings, analyzed using t-test and ANOVA, posited statistically significant learning improvements for students using adaptive learning systems in comparison to those in traditional learning setups. Qualitative thematic analysis surfaced five primary themes which included student engagement, personalized learning, technological challenges, faculty resistance, and a call for institutional commitment. The technology-adaptive learning system markedly enhanced student academic performance and engagement, demonstrating the value of adaptive learning technologies in Pakistani higher education, even with systemic infrastructural challenges. The research concluded that machine learning-based personalized education has a lot of potential to change positively higher education in Pakistan. However, the implementation of this change hinges on overcoming technological barriers, faculty training, and institutional support.