Evaluating the Impact of AI-Driven Personalization in Online Learning Platforms

Abstract

The rapid integration of artificial intelligence (AI) in online learning platforms has emerged as a transformative strategy to personalize education and improve student outcomes. This study employed meta-analysis to synthesize findings from 47 empirical studies, encompassing over 13,000 learners, to evaluate the impact of AI-driven personalization on academic performance, engagement, and retention. The aggregate analysis revealed a moderate-to-strong overall effect (Cohen’s d = 0.55), indicating that AI-enabled adaptive learning systems outperform traditional, non-personalized approaches. Moderator analyses further indicated that the effectiveness of these interventions varies significantly by educational level and the specific type of AI tool used, with higher education settings and intelligent tutoring systems yielding more pronounced benefits. Although analyses of publication bias revealed minimal skew, significant heterogeneity was observed, underscoring the influence of contextual factors such as instructional design and learner demographics. These findings substantiate the potential of AI-driven personalization to tailor learning experiences to individual needs, thereby fostering improved educational outcomes. Simultaneously, the study highlights challenges related to algorithmic transparency and ethical considerations that warrant further investigation. The insights provided by this meta-analysis offer a robust foundation for future research and the strategic implementation of AI technologies in online education, contributing to a more engaging and effective digital learning environment.