Education Analytics
Dropout prediction, curriculum analysis, and equity in education (2021–2026)
Education Analytics studied student dropout, curriculum design, and equity in education with predictive models and visual analytics, in collaboration with Universidad Catolica San Pablo (Peru). We compared WHO and WHEN approaches for the early identification of students at dropout risk, measured how curricular changes impact computer science programs, and built SDA-Vis and CSDA-Vis, visual systems that combine counterfactual explanations and survival analysis so that professors and mentors can explore what-if-and-when interventions for individual students. We also analyzed the equity of the Brazilian National High School Exam (ENEM) by validating its differential item functioning.