Predict students' dropout and academic success
Realinho, Machado, Baptista, Martins (Polytechnic Institute of Portalegre)
Enrollment and early performance of thousands of students, labeled dropout, enrolled, or graduate.
A dataset from a Portuguese higher education institution about students in different undergraduate degrees. It includes information known at enrollment (academic path, demographics, socio-economic factors) and first- and second-semester performance, to classify students as dropout, enrolled, or graduate. It is a popular teaching data set.
Access
- Cost
- FreeFree download from Zenodo; no account needed.
- License
- License not statedNot stated in the record
- Sign-up
- No sign-up
- How you get it
- Bulk downloadCSV
What's in it
- One row is
- One row per student
- Covers
- Portugal (one institution)
- Years
- Students enrolled around 2008 to 2019
- Updated
- One-time release
- Size
- About 4,400 students
Quality
Cleaned and well described, but covers a single institution.
- Single institution
- Class imbalance and country-specific context
Connects to
- UCI Student Performance Dataset: Other education datasets for comparison
- Open University Learning Analytics Dataset: Online learning engagement data
- Eurostat tertiary education statistics: Country-level context
Questions it could answer
- Which early indicators best predict dropout?
- How do scholarships relate to graduation?