This study aims to examine the association between perceived artificial intelligence (AI)-adaptive learning and the hybrid-flexible (HyFlex) modality with personalized learning, inclusive education, and perceived equitable access in digitally constrained higher education settings.
An explanatory sequential mixed-methods design combined a cross-sectional survey of 469 students from three Indonesian universities with structured interviews with ten students. Partial least squares structural equation modeling (PLS-SEM) with 5,000 bootstrap subsamples tested six direct and four specific indirect associations, while thematic analysis explained statistical patterns and identified relevant implementation conditions.
All six direct and four specific indirect associations were significant. Inclusive education had the strongest direct association with perceived equitable access (β = 0.606), whereas the specific indirect association between the HyFlex modality and perceived equitable access through inclusive education was the largest (β = 0.329). The interviews identified adaptive explanations and flexible participation, alongside constraints related to connectivity, device availability, digital literacy and lecturer preparedness.
Cross-sectional self-reports, eligibility-based course-cluster sampling, same-source measurement, a male-dominated sample, and a small qualitative subsample limit causal and generalizable claims of this study. Longitudinal mixed-methods research is required to address these concerns.
Universities should combine low-bandwidth alternatives, accessible assessment routes, device and connectivity support, responsible AI governance, and lecturer development.
This study jointly models students' perceptions of AI adaptation and HyFlex as complementary affordances and compares personalization and inclusive education as specific indirect pathways to perceived equitable access. This contribution concerns students' perceptions and does not provide objective evidence of institutional equity.
