Machine learning · education analytics

Student performance insights

Estimate a student's math score from their background and their reading and writing results, see exactly which inputs moved the estimate, and explore what 1,000 students' results say about lunch, test preparation and parental education.

Student details

Change any input. The estimate updates instantly.

Estimated math score

–/ 100

What moved the estimate

Points added or removed compared with an average student in the data.

Effects are measured with reading and writing scores held fixed. For example, test preparation mostly lifts reading and writing, so once those are known it adds little on its own; see the averages below for its overall link with results.

What the data shows

Average exam score by group across all 1,000 students. Axes start at 50; the dashed line is the overall average.

About the model

A linear regression trained on 800 students and tested on 200 it had never seen, the split used by the project's ML pipeline. Each category is measured against a baseline category, so every coefficient reads directly as points.