An open-source textbook covering program evaluation regression tools commonly used for the estimation of program or policy impact in the public and nonprofit sectors.
CITE:
Counterfactual Analysis with Regression Models
An Open Source Textbook for Program Evaluation
Authors: Jesse Lecy and Federica Fusi
The book was designed around several core principles:
- Chapters emphasize intuition over mathematical precision.
- Instead of highlighting traditional econometrics concepts of endogeneity, efficiency, and parametric assumptions the chapters focus on the counterfactual embedded in each estimator and whether the comparison adequately captures unbiased program effects.
- Concepts are explained with examples, data and code so that students can replicate analysis in the chapters.
- All analysis is done in R so that students can use the software after graduating without the burden of expensive licenses.
TABLE OF CONTENTS:
- Interrupted Time Series
- Difference-in-Difference Models
- Instrumental Variables
- Fixed Effect Models
- Regression Discontinuity Design
- Logistic Regression
- Matching
DATA used in replication examples in the chapters.
LABS:
- Interrupted Time Series Lab
- Diff-in-Diff Lab
- Fixed Effects Lab
- Instrumental Variables Lab
- Regression Discontinuity Lab
- Logistic Regression Lab
- Matching Lab
DATA used in labs.
Textbook chapters are located in the docs folder.
https://ds4ps.org/pe4ps-textbook/docs/index.html
The R Markdown (RMD) files used to create each chapter are in the root project folder:
https://github.com/DS4PS/pe4ps-textbook
Labs are in the labs folder.
The chapters are rendered with Bookdown in R. The main YAML file with global book parameters is at:
https://github.com/DS4PS/pe4ps-textbook/blob/master/_site.yml