Models and strategies for implementing adaptive learning in elementary school

Abstract

AI-based adaptive learning in elementary school is a revolutionary way to reduce inequalities in learning and improve student achievements. This potential is however not fully harnessed, with infrastructure shortfall, lack of teacher training and issues in resource distribution impeding governments in expanding ESSP-based models at scale. The field is in want of a consolidated repository of empirically evaluated AI techniques that have been designed for the primary education context. The present study fills in this gap using a systematic review of literature (2015–2025) coupled with a descriptive qualitative approach. Four archetypal AL-frameworks applicable in the MI/SD context have been identified by the research: Personalized Learning, Differentiated Instruction, Adaptive Inquiry Learning, and Hooked Adaptivity. To apply the methods effectively, the paper suggests three strategic pillars: using AI for large scale real time adaptation, flexible curriculum to enable differentiated learning and redefining teachers as data driven facilitators through continuous professional development. By reframing the question from tool-use to a data-guided pedagogical milieu, this research offers hope for how we can optimize student mood and teacher efficacy. In short, these findings have important theoretical and practical implications for educators and policy makers who endeavor to negotiate the moral and logistical complexities of contemporary early childhood education.