Week 3: Directed Report with AI Use Log

Assignment

Using a provided dataset, create a directed report demonstrating data engineering and formal statistical analysis.

Components 1-2: Data documentation and analysis notebook

Submit as a single Jupyter notebook (.ipynb) rendered as .html that runs top to bottom without errors.

1. README and data dictionary - as markdown cells at the top of the notebook: document the dataset’s source, unit of analysis, every variable (name, type, description, missingness pattern, issues, cleaning applied), transformations, and exclusions. Verify AI-drafted entries against your code output.
2. Reproducible analysis - import, cleaning with assertions after each step, formal statistical test, at least two visualizations, and a cross-check verifying your key finding.

Component 3: Directed report (3-4 pages)

Submit as a PDF. Five sections: question, methods (cleaning decisions, sample size, test choice and assumptions), results (effect sizes, confidence intervals, visualizations), interpretation (what it means in context), and limitations.

Component 4: AI use log

See the AI use log guide.

Rubric

Criterion	Excellent (5)	Adequate (3)	Needs revision (1)
Data dictionary	All variables documented, missingness analyzed, cleaning justified	Most documented but some vague	Incomplete or unverified
Code quality	Runs end-to-end, assertion after every step, decisions explained	Runs but assertions sparse	Doesn’t run or major gaps
Statistical analysis	Appropriate test, assumptions checked, effect size and CI reported	Test present but assumptions unchecked or effect size missing	No formal test
Report structure	All 5 sections with specific evidence	Most sections but methods or limitations thin	Missing sections
Visualizations	2+ informative plots tied to analysis	Plots present but disconnected	No plots
AI use log	Detailed discovery, verification, and judgment	Present but vague	Missing

Total: 30 points