An AI‐Supported Adaptive Learning Environment for Arabic Reading Development: A Quasi‐Experimental Study in Primary Education

ABSTRACT

This study evaluated the effectiveness of an artificial intelligence (AI)‐supported adaptive learning environment in developing Arabic reading skills among primary school pupils. Arabic reading presents distinctive challenges for beginning learners because letter forms vary according to position, short vowels are frequently omitted and the script is written from right to left. Conventional whole‐class instruction may not adequately accommodate differences in prior knowledge, learning pace and cognitive readiness. A quasi‐experimental pre‐test–post‐test control‐group design was employed with 72 pupils from two intact classes in one Malaysian primary school. The experimental group used an adaptive environment that adjusted task difficulty and learning pathways according to response accuracy and provided immediate corrective, explanatory and motivational feedback. The control group received conventional teacher‐led instruction using the same curricular content and instructional duration. Both groups completed an expert‐validated Arabic reading test before and after the 4‐week intervention, while pronunciation and fluency were assessed using a validated rubric. The experimental group demonstrated substantially greater improvement than the control group, with a statistically significant post‐test difference and a very large effect size (Cohen's d  = 1.91). Higher pronunciation accuracy, reading fluency and learner engagement were also observed in the experimental group. The findings suggest that AI‐supported adaptation and timely feedback can provide differentiated support for beginning Arabic readers. Although the single‐school sample limits broad generalisability, the study offers practical evidence for carefully supervised integration of adaptive learning into primary Arabic language instruction.