Learning analytics identify three distinct student profiles in higher education

A study published in Discover Sustainability has identified three distinct learning profiles among higher education students using institutional learning analytics data. The research, led by Paulina Aguirre and colleagues, analyzed 51,492 student-term observations to explore how academic performance, learning engagement, and competency-related indicators vary across different student groups. The primary objective was to determine whether unsupervised clustering methods could reveal heterogeneous learning patterns associated with systems thinking and sustainability competence.

The study utilized data from the Data Hub of the Institute for the Future of Education, providing a large-scale dataset suitable for exploring complex educational dynamics. The researchers applied K-means clustering, an unsupervised machine learning technique, to group students based on their observed behaviors and outcomes.

The analysis revealed three primary learning profiles that differed significantly in their academic and behavioral characteristics. One profile was characterized by lower academic performance, yet these students maintained levels of participation comparable to those in the other groups. The other two profiles both exhibited similarly high academic performance, but they diverged in their levels of engagement and the number of competency-related indicators they displayed.

To evaluate the factors associated with these learning profiles, the researchers employed Chi-square tests and Cramér’s V, a statistical measure of association. These methods allowed the team to assess the strength of the relationship between profile membership and various demographic and academic variables. The results indicated that program major exhibited the strongest association with learning profile membership. In contrast, nationality and gender showed statistically significant but weak associations with profile membership, indicating that these demographic factors have limited practical influence on the differentiation of learning profiles.

The study emphasizes that the identified profiles are associated with systems thinking and sustainability competence rather than directly measuring these competencies. The authors clarify that the proposed approach provides a data-driven framework for understanding learning patterns that can support curriculum design and personalized educational interventions. By identifying heterogeneous learning profiles, the framework offers a basis for continuous improvement initiatives in competency-based higher education.

The implications of these findings extend to the strategic development of systems thinking and sustainability competence in higher education. As these competencies become a priority, educational approaches must be capable of identifying and addressing diverse learning patterns among students. The study demonstrates that learning analytics, when combined with unsupervised clustering, offers a viable opportunity to explore large-scale educational data. By identifying heterogeneous student learning patterns, institutions can better tailor their educational strategies to meet the specific needs of different student groups, thereby enhancing the overall effectiveness of competency-based education.

Limitations of the study include the reliance on institutional learning analytics data, which may not capture all aspects of student learning. The associations identified are statistical and do not imply causation; for example, while program major is strongly associated with learning profile membership, it is not necessarily the cause of the specific learning behaviors observed. Furthermore, the study’s focus on specific competency-related indicators means that the profiles may not generalize to all educational contexts or competency frameworks.

By identifying three distinct learning profiles and analyzing their associations with demographic and academic variables, the research provides a nuanced understanding of how students engage with higher education. The findings underscore the importance of using data-driven approaches to support curriculum design and personalized interventions. As higher education institutions continue to prioritize systems thinking and sustainability competence, the framework proposed in this study offers a practical tool for identifying and addressing the diverse learning patterns that characterize modern student populations.

Subject of Research: Identification of learning profiles associated with systems thinking and sustainability competence in higher education using learning analytics.

Article Title: Profiles associated with systemic and sustainability competence in higher education identified through learning analytics

Article References: Aguirre, P., Carrión, P., Castro, J. R., Elizalde-Solano, R., González-Porras, R. A., & Ramírez-Montoya, M.-S. (2026). Profiles associated with systemic and sustainability competence in higher education identified through learning analytics. Discover Sustainability. https://doi.org/10.1007/s43621-026-05214-1

Image Credits: AI Generated

DOI: 10.1007/s43621-026-05214-1

Keywords: learning analytics, higher education, sustainability competence, systems thinking, K-means clustering, unsupervised learning, student engagement, academic performance, educational data mining, curriculum design, competency-based education, Principal Component Analysis, Chi-square tests, Cramér’s V, student profiles, Institute for the Future of Education, data-driven education, heterogeneous learning, personalized interventions, Discover Sustainability