This analysis is going into a government report, so it has to hold up under scrutiny. For a research institute in Gothenburg you build an inferential analysis that separates an air quality regulation's effect on particulate levels from seasonal, weather, and long-term trends across seven years of monitoring data.
The discipline skills: trend analysis with confounders using statsmodels, assumption checking for inferential methods, effect size computation with scipy, and communicating uncertainty honestly for a government report.
The AI-direction lesson: every previous project handed you a CLAUDE.md file that made AI productive from the first prompt. This time, you write that file yourself. You run the analysis once without project memory and see what AI does by default. Then you encode your analytical conventions -- temporal splitting, effect sizes, assumption checks -- into persistent infrastructure and run it again. The difference is the lesson. Specific, testable constraints produce specific compliance. Vague entries produce vague compliance. The skill is learning to direct AI through infrastructure, not just through prompts.
Your Role
You deliver a methodologically defensible analysis of whether the regulation changed PM2.5 levels. The analytical terrain is familiar. What is new is building the AI infrastructure yourself -- encoding what you have learned into files that shape AI's behavior from the first prompt of every session.
What's New
Last time, you determined the question type from an ambiguous brief, designed a validation strategy, and used meta-prompting to verify unfamiliar territory. You owned the methodology.
This time, you still own it — but you also build the infrastructure underneath. You will run the analysis once without any project memory and see what AI does by default. Then you build the memory file, run the analysis again, and compare. The difference is the lesson.
The hard part is not the analysis. It is writing infrastructure specific enough that AI follows your conventions from the first prompt — and recognizing when vague infrastructure produces vague compliance.
Tools
- Python 3.11+ via your conda "ds" environment
- Jupyter Notebook for the analysis
- pandas for data handling
- statsmodels for hypothesis tests and assumption checks
- scipy for statistical tests and effect size calculations
- matplotlib / seaborn for visualization
- Claude Code as the AI you direct
- Git / GitHub for version control
Materials
You receive:
- Seven years of air quality monitoring data (daily averages from 40 stations across three cities)
- Weather data matched to each station (temperature, wind speed, precipitation)
- Station metadata (locations, types, cities)
- A data dictionary describing all three datasets
- A methodology memo template
- An archived project memory file from a previous NERI analysis (reviewed in Unit 3)
- No CLAUDE.md — you create this yourself