Research Methods
Reproducible methods for applied economics, organized into two tracks: Python tutorials with replication code, and guides for integrating AI into rigorous research workflows.
Open-Source Methods
Python tutorials for researchers transitioning from licensed software. Difference-in-differences, spatial joins, SHAP interpretation, GTFS validation, and Census API pipelines with working code.
AI Research Workflows
Claude Code guides for academic research. CLAUDE.md context files, skills, hooks, MCP servers, and the documentation practices that make AI assistance reliable.
Screening for Medicaid Fraud with Public Data
HHS released 227 million rows of Medicaid billing data. Within hours, amateur analysts claimed billions in fraud. This 4-part series walks through what the data can actually support, from data quality to classifier performance.
Frequently Asked Questions
What programming language do the tutorials use?
All tutorials use Python with standard data science libraries: pandas, statsmodels, scikit-learn, GeoPandas, and SHAP. Every article includes complete working code and links to GitHub replication repositories.
Do I need to know Python to follow the AI guides?
No. The AI workflow guides focus on configuring Claude Code for research through CLAUDE.md context files, skills, hooks, and MCP servers. They document how to structure AI-assisted research, not how to write Python.
Can I replicate the analysis from these tutorials?
Yes. Every tutorial links to a public GitHub repository with all data and code needed to reproduce the analysis. The main replication repository contains 18 research projects.
What is a CLAUDE.md file?
CLAUDE.md is a context file that loads automatically when Claude Code starts a session. It captures project requirements, variable definitions, and methodological decisions, eliminating re-explanations. See our context file article.