Student Performance in Math and Portuguese with 32 Demographic and Behavioral Features

Two datasets from the UCI Machine Learning Repository, authored by P. Cortez and A. M. G. Silva, containing student records for secondary school Math and Portuguese language courses. Each record includes 32 attributes covering demographics, family background, school life, and alcohol consumption, with final grades (G1, G2, G3) as target variables. The temporal coverage and exact number of rows are not specified in the provided metadata.

Use Cases

• Predict final grades (G3) based on demographic, family, and behavioral features like studytime and failures.
• Analyze correlations between alcohol consumption (Dalc, Walc) and academic performance or absences.
• Investigate the impact of family support (famsup), parental education (Medu, Fedu), and internet access on student outcomes.
• Compare performance predictors across different subjects (Math vs. Portuguese) using the two separate datasets.

Strengths

• Includes two distinct datasets (Math and Portuguese) for comparative analysis.
• Contains 32 detailed features per student, covering demographics, family, school, and lifestyle factors.
• Grades are recorded across three periods (G1, G2, G3), allowing for longitudinal analysis within a school year.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Last update date is unknown; freshness unverified.
• Geographic and temporal coverage are not specified, limiting generalizability conclusions.

Provenance

Source

UCI Machine Learning Repository, authors P. Cortez and A. M. G. Silva.

Collection Method

Likely collected via school surveys, as suggested by the demographic and self-reported lifestyle attributes.

Time Range

null

Freshness

null

Geography

null