Submission: One .html file rendered from a Jupyter Notebook. The report should read like a concise mini research paper with narrative + code + tables + figures.
.html exported from your notebook.| Section | What to include | Tiny example |
|---|---|---|
| 1. Title & Introduction | State your research question and why it matters. | “What is the effect of class size on test scores? We study U.S. 4th grade districts (2015–2019) to inform policy on optimal class sizes.” |
| 2. Data Description | Source, unit, key variables, and descriptive stats/visuals. | Outcome score, main regressor class_size, controls income, urban. Show a summary table and a histogram of score. |
| 3. Empirical Model | Baseline equation & brief justification. Note multicollinearity & OVB risks. | y = β₀ + β₁·class_size + γ′controls + u. OVB risk: teacher quality unobserved. Multicollinearity risk: income vs urban. |
| 4. Results | One main table/figure, interpret β₁ in plain English. | “A 1-student reduction per class is associated with +0.15 points (SE=…).” |
| 5. Robustness & Discussion | One simple alternative spec or subsample; limitations. | Add spending as a control; β₁ similar. Limitation: measurement error in class size. |
| 6. Conclusion | 2–3 bullets + one implication. | Smaller classes improve scores; effect largest in low-income districts; target crowded schools. |
.html submissionTip: Keep sections concise; prioritize clarity over volume.