The Profile of Students Retained at the State University of Mato Grosso: A Cluster-Based Analysis
retention in higher education; clustering; educational data
The present work presents an approach based on clustering of educational data as an alternative proposal to study the phenomenon of retention at the Universidade do Estado de Mato Grosso (UNEMAT). The retention problem affects all Brazilian higher education institutions and, at UNEMAT, it also presents itself as a latent issue. Mitigating retention is a way to reduce dropout rates and, consequently, increase graduation rates. To this end, the development of institutional policies and more specific pedagogical interventions depend on the knowledge of the profile of the retained student. Based on the data from the 2020 higher education census, which were later enriched with information from the institution's academic system, it was possible to build a database with 2455 records of students retained from UNEMAT and then we clustered these records which revealed 3 distinct groups of student profiles which we present in this research.