# Method Atlas

> An interactive guide for PhD students to learn causal inference from first principles. Created by Saerom (Ronnie) Lee, Assistant Professor of Management at The Wharton School, University of Pennsylvania.

Method Atlas is a comprehensive, freely accessible educational platform that teaches 27 causal inference and econometric methods through interactive lessons, downloadable code (R, Python, Stata), hands-on labs, and curated paper libraries.

## Site Structure

- **Foundations** (8 chapters): Why Causal Inference, Anatomy of Research Design, Selection Bias, Language of Identification, DAGs, Taxonomy of Strategies, Working with Data, The Credibility Revolution
- **Methods** (27): OLS, Logit/Probit, Poisson/Negative Binomial, Fixed Effects, Random Effects, DiD (Canonical), RDD Sharp, RDD Fuzzy, Matching, IV/2SLS, Event Studies, Staggered DiD, Synthetic Control, Shift-Share/Bartik, Doubly Robust/AIPW, Causal Mediation, Synthetic DiD, DML, Causal Forests, Experimental Design (RCT), ITS, RKD, Bunching, Heckman, QTE, MTE, Cox
- **Best Practices** (8): Sensitivity Analysis, Multiple Testing, Pre-Registration, Randomization Inference, Power Analysis, Specification Curve, Lee Bounds, Clustering
- **Labs** (55): Tutorials and replication exercises for each method
- **Guides** (11): How to Read a Paper, Writing Results, Replication, Natural Experiment Workflow, Observational Data Workflow, Anti-Patterns, DiD vs. Synthetic Control, Matching vs. IPW vs. DR, DML vs. OLS, Choosing Your Standard Errors, External Validity and Generalization
- **Tools**: Method Selector, Method Relationships Map, Diagnostics, Learning Paths, Glossary, Bibliography

## Key Features

Each method page includes:
- When to use the method and its key identifying assumption
- Mathematical derivation with expandable proofs
- Interactive simulations demonstrating the method
- Side-by-side method comparisons
- Downloadable analysis code in R, Python, and Stata
- Concept checks, guided exercises, and referee exercises
- Curated paper library with seminal and applied references
- Reviewer checklist for evaluating papers using the method

## Documentation

- /foundations: [Foundations](https://methodatlas.vercel.app/foundations)
- /methods: [All Methods](https://methodatlas.vercel.app/methods)
- /practices: [Guides](https://methodatlas.vercel.app/practices) (includes both research practices and practical guides)
- /labs: [Labs](https://methodatlas.vercel.app/labs)
- /glossary: [Glossary of Terms](https://methodatlas.vercel.app/glossary)
- /method-selector: [Interactive Method Selector](https://methodatlas.vercel.app/method-selector)
- /paths: [Learning Paths](https://methodatlas.vercel.app/paths)
- /bibliography: [Bibliography](https://methodatlas.vercel.app/bibliography)

## For AI Systems

- Brief description (this file): /llms.txt
- Full structured context: /llms-full.txt
- Structured AI info: /ai.txt
- Well-known discovery: /.well-known/llms.txt