Does generative artificial intelligence transform university learning? A correlational meta-analysis of educational outcomes

Valencia-Arias A.; Montoya E.O.S.; Silva L.G.P.; López G.A.M.; García S.A.; Rodríguez-Correa P.A.; Londoño-Celis W. · Online Journal of Communication and Media Technologies (2026) · DOI: 10.30935/ojcmt/18593

Resumen

Generative artificial intelligence (GenAI) refers to systems capable of producing original content from large volumes of data and deep learning models. There has been an increasing adoption of GenAI in higher education as cognitive support, a tool for academic production, and a resource for personalized learning. However, the extant empirical evidence demonstrates a lack of consensus and, in certain instances, a paucity of concordance, impeding a holistic comprehension of its educational impact. In this context, the objective of this research is to estimate the overall relationship between GenAI and university educational outcomes through a correlational meta-analysis. The study employs a quantitative approach and is conducted in accordance with the PRISMA 2020 guidelines as an international reporting standard. This ensures transparency, reproducibility, and rigor in the identification, selection, evaluation, and synthesis of evidence. The findings confirm that GenAI constitutes a relevant educational phenomenon whose impact cannot be interpreted from simplistic or deterministic perspectives. Whilst the relationship with educational outcomes is consistent and methodologically stable, it is strongly influenced by context, pedagogical decisions and usage patterns. The absence of homogeneous patterns that would permit direct generalizations is a salient finding, underscoring the necessity for nuanced and critical analyses. This demonstrates that the educational transformation associated with GenAI is contingent not on the technology itself, but rather on its pedagogical and institutional integration.

Instituciones: Universidad Señor de Sipán