Investigating Explainable Artificial Intelligence to Enhance Equity and Inclusivity in Adaptive Learning: Empirical Insights from Under-Resourced Environments

Abstract

This study investigates the integration of Explainable Artificial Intelligence (XAI) into
adaptive learning systems, drawing on Human–AI Complementarity Theory and Epistemic
Justice Theory to examine how explainability supports equity and inclusivity in resourceconstrained
settings. Using a mixed-methods design, the study combined pre- and post-test data
with qualitative insights to assess learner confidence in the system, engagement, and
performance. Findings show that learners using XAI-supported systems achieved significantly
higher post-test scores (F₁,₁₁₁ = 24.78, p < .001, η2 = .08) and reported greater confidence
in the system due to clearer, contextually relevant explanations. The results indicate that XAI
reduces confusion, strengthens fairness perceptions, and enhances learner autonomy. The study
concludes that explainability is essential for equitable AI-mediated learning, particularly where
resource constraints and diverse learner needs amplify the risks of opaque systems. It
recommends the adoption of XAI-enhanced platforms in resource-constrained settings to
improve transparency, trust, and academic achievement.