<![CDATA[Purpose – This meta-analysis systematically examines the association between artificial intelligence (AI) interventions and academic achievement among Nigerian university students, synthesizing empirical evidence from 2022 to 2025.Methods – Following PRISMA 2020 guidelines, 47 primary studies meeting rigorous eligibility criteria were identified, yielding a combined meta-analytic sample of 8,234 undergraduate and postgraduate students from federal and state universities in Nigeria. A random-effects model with Hedges' g as the primary effect size metric produced an overall pooled estimate of g = 0.68 (95% CI [0.54, 0.82]), indicating a moderate-to-large positive association between AI integration and academic performance.Findings – However, substantial heterogeneity (I² = 86.5%) indicates that this overall estimate masks considerable variation across implementation contexts and should be interpreted with caution rather than as a stable, universally applicable effect. Moderator analysis identified significant variations across learning strategy, subject area, AI role and type, intervention duration, and sample size. Intelligent tutoring systems delivering individualized instruction in STEM disciplines over sustained periods yielded the largest effects. Infrastructural deficits, limited financial resources, and insufficient faculty AI competency were the most prevalent implementation barriers.Research implications – This study provides an empirically grounded synthesis of AI's educational associations within a resource-constrained developing country context while acknowledging the methodological limitations that constrain the certainty of the conclusions.Originality – The Implications for evidence-based policy and institutional practice in Nigerian higher education are discussed.]]>
Copyrights © 2026
