Artificial Intelligence on the Personalization of UniversityLearning

Abstract

The present study identifies the effect of artificial intelligence on individualized learning in higher education. For this purpose, a quasi-experimental design was implemented with 780 first-semester veterinary medicine students. From this cohort, 500 students were randomly assigned to two groups: a control group (N = 250) that used traditional teaching methods, and an experimental group (N = 250) that used an adaptive AI system. The variety of data sources enabled a mixed-methods methodological design that integrated quantitative analysis (t-tests, ANOVA, linear regression, and time series) with qualitative analysis (sentiment analysis using natural language processing). It was found that the experimental group scored 8% higher than the control group on the final evaluation, demonstrated more dynamic learning, and had an overall more positive experience. The research also highlights critical tensions, such as the protection of sensitive data and the lack of pedagogical grounding in AI systems. It is concluded that AI in education presents significant opportunities, which must be developed through teacher training and a robust ethical framework for its use.