Adaptive and Personalized Learning in Higher Education: An Artificial Intelligence-Based Approach

Abstract

The integration of Artificial Intelligence (AI) in higher education offers a potential solution to the scalability of personalized learning, yet empirical frameworks connecting diagnostic data with teacher-mediated interventions remain limited in developing contexts. This study adopts a sequential multi-phase research design to address this gap. Phase 1 comprised a diagnostic quantitative analysis of the National Survey on Access and Permanence in Education (ENAPE 2021), involving a representative sample of 3422 Mexican undergraduate students. Using Exploratory Factor Analysis (KMO = 0.96) and Pearson correlations, the study established a structural baseline. Phase 2 implemented a quasi-experimental exploratory pilot (N = 23) across two academic clusters (Civil Engineering and Nutrition) using “ActivAI”, a custom GPT configured with Retrieval-Augmented Generation (RAG). Results from Phase 1 revealed a strong, statistically significant correlation (r=0.72, p<0.01) between the perceived impact of education on daily life and the perception of equity, identifying “relevance” as a key driver of accessibility. Phase 2 results demonstrated high student satisfaction with AI-driven personalization (M = 4.49, SD = 0.64), although disciplinary variations in engagement were observed (SD = 0.85 in Nutrition versus 0.45 in Engineering). The study concludes by proposing the Dynamic Integration Model, which leverages AI not as a replacement for instruction but as a scalability toolkit for teacher-led orchestration, ensuring that personalization addresses dynamic student needs rather than static learning styles.

Ähnliche Arbeiten