This study adopts a multi-method approach to explore the role of artificial intelligence (AI) in higher education, focusing on its impact on student learning. A bibliometric analysis was conducted using Scopus-indexed publications from 2000 to 2024 to examine research trends, thematic developments, and influential contributions in the field. Text mining techniques were applied to extract keywords from titles and abstracts, followed by TF-IDF weighting. K-means clustering and Latent Dirichlet Allocation (LDA) were used to identify key research themes, while citation networks were analyzed using the PageRank algorithm to highlight major publications. Complementing the bibliometric work, an experimental study was carried out to evaluate ChatGPT as a formative assessment tool. Students submitted written responses, which were processed by ChatGPT to generate automated feedback and grades. These outputs were compared with human-generated assessments to evaluate accuracy and usefulness. The findings suggest that students who received AI-supported feedback performed better overall, with particularly notable gains among lower-performing students. The feedback generated by ChatGPT combined corrective guidance, elaborative explanations, and motivational elements, contributing to improved understanding and engagement. Although the grades given by ChatGPT were mostly consistent with human assessments, some small differences were noticed in areas that involved judgments about writing style and clarity. However, further empirical research is necessary to explore how these tools can be effectively implemented in ways that align with instructional goals and the practical realities of higher education contexts.
Zhang et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: