From Deep Learning to Digital Insights: Twenty-Five Years of Understanding How Students Learn in Higher Education
David Gijbels
Abstract
This article marks the 25th anniversary of Active Learning in Higher Education and offers a reflective overview of how research on student learning has evolved since the journal’s inception. Drawing from my own academic journey, I first revisit the origins of deep and surface approaches to learning and the subsequent development of influential questionnaires. I then discuss how early research primarily relied on cross-sectional, correlational designs that linked students’ perceptions of the learning environment to their approaches to learning, consistently showing that positive perceptions were associated with deeper engagement. Over time, however, researchers recognized the limitations of these designs and shifted toward longitudinal studies. Although it is often assumed that higher education naturally fosters deeper approaches to learning, systematic reviews reveal that changes in learning approaches are neither linear nor universal; instead, they are influenced by individual differences, learning contexts, and disciplinary practices. In the past decade, the field has increasingly embraced multimodal and behavioral data, integrating tools such as eye tracking to gain deeper insight into students’ learning processes. This shift has opened new avenues for understanding how learners engage with texts, videos, and other instructional materials. The article concludes by outlining emerging opportunities at the intersection of artificial intelligence and multimodal learning analytics, illustrated through the EYE-TEACH project, which seeks to provide higher education teachers with actionable, ethically grounded insights to better support students’ active learning in real time.
Evidence weight
Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40
| F · citation impact | 0.50 × 0.4 = 0.20 |
| M · momentum | 0.50 × 0.15 = 0.07 |
| V · venue signal | 0.50 × 0.05 = 0.03 |
| R · text relevance † | 0.50 × 0.4 = 0.20 |
† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.