The story
A course research project investigating the relationship between sleep patterns and exam performance among undergraduate students.
I cleaned the survey dataset in Python, then moved to R for the modelling — linear regression plus a hierarchical layer to capture per-cohort variance. The goal was never a clickbait correlation; it was understanding where the data genuinely supported a claim and where it did not.
The deliverable was a written analysis with reproducible code and plots honest enough to argue against.
Marks in the margin
- Cleaned and reshaped a student-survey dataset for analysis.
- Fit linear and hierarchical models to estimate effect size.
- Produced diagnostic plots readable enough to critique the model fit.
- Wrote the results up for course review.