You are here

Propensity Score Analysis
Share
Share

Propensity Score Analysis
Statistical Methods and Applications

Second Edition


August 2014 | 448 pages | SAGE Publications, Inc

Fully updated to reflect the most recent changes in the field, the Second Edition of Propensity Score Analysis provides an accessible, systematic review of the origins, history, and statistical foundations of propensity score analysis, illustrating how it can be used for solving evaluation and causal-inference problems. With a strong focus on practical applications, the authors explore various strategies for employing PSA, discuss the use of PSA with alternative types of data, and delineate the limitations of PSA under a variety of constraints. Unlike existing textbooks on program evaluation and causal inference, this book delves into statistical concepts, formulas, and models within the context of a robust and engaging focus on application.

 
List of Tables
 
List of Figures
 
Preface
 
About the Authors
 
Chapter 1: Introduction
Observational Studies

 
History and Development

 
Randomized Experiments

 
Why and When a Propensity Score Analysis Is Needed

 
Computing Software Packages

 
Plan of the Book

 
 
Chapter 2: Counterfactual Framework and Assumptions
Causality, Internal Validity, and Threats

 
Counterfactuals and the Neyman-Rubin Counterfactual Framework

 
The Ignorable Treatment Assignment Assumption

 
The Stable Unit Treatment Value Assumption

 
Methods for Estimating Treatment Effects

 
The Underlying Logic of Statistical Inference

 
Types of Treatment Effects

 
Treatment Effect Heterogeneity

 
Heckman’s Econometric Model of Causality

 
Conclusion

 
 
Chapter 3: Conventional Methods for Data Balancing
Why Is Data Balancing Necessary? A Heuristic Example

 
Three Methods for Data Balancing

 
Design of the Data Simulation

 
Results of the Data Simulation

 
Implications of the Data Simulation

 
Key Issues Regarding the Application of OLS Regression

 
Conclusion

 
 
Chapter 4: Sample Selection and Related Models
The Sample Selection Model

 
Treatment Effect Model

 
Overview of the Stata Programs and Main Features of treatreg

 
Examples

 
Conclusion

 
 
Chapter 5: Propensity Score Matching and Related Models
Overview

 
The Problem of Dimensionality and the Properties of Propensity Scores

 
Estimating Propensity Scores

 
Matching

 
Postmatching Analysis

 
Propensity Score Matching With Multilevel Data

 
Overview of the Stata and R Programs

 
Examples

 
Conclusion

 
 
Chapter 6: Propensity Score Subclassification
Overview

 
The Overlap Assumption and Methods to Address Its Violation

 
Structural Equation Modeling With Propensity Score Subclassification

 
The Stratification-Multilevel Method

 
Examples

 
Conclusion

 
 
Chapter 7: Propensity Score Weighting
Overview

 
Weighting Estimators

 
Examples

 
Conclusion

 
 
Chapter 8: Matching Estimators
Overview

 
Methods of Matching Estimators

 
Overview of the Stata Program nnmatch

 
Examples

 
Conclusion

 
 
Chapter 9: Propensity Score Analysis With Nonparametric Regression
Overview

 
Methods of Propensity Score Analysis With Nonparametric Regression

 
Overview of the Stata Programs psmatch2 and bootstrap

 
Examples

 
Conclusion

 
 
Chapter 10: Propensity Score Analysis of Categorical or Continuous Treatments
Overview

 
Modeling Doses With a Single Scalar Balancing Score Estimated by an Ordered Logistic Regression

 
Modeling Doses With Multiple Balancing Scores Estimated by a Multinomial Logit Model

 
The Generalized Propensity Score Estimator

 
Overview of the Stata gpscore Program

 
Examples

 
Conclusion

 
 
Chapter 11: Selection Bias and Sensitivity Analysis
Selection Bias: An Overview

 
A Monte Carlo Study Comparing Corrective Models

 
Rosenbaum’s Sensitivity Analysis

 
Overview of the Stata Program rbounds

 
Examples

 
Conclusion

 
 
Chapter 12: Concluding Remarks
Common Pitfalls in Observational Studies: A Checklist for Critical Review

 
Approximating Experiments With Propensity Score Approaches

 
Other Advances in Modeling Causality

 
Directions for Future Development

 
 
References
 
Index

Supplements

Companion Website
The site contains programming syntax for all the examples found in the book, by chapter and section.

    Over the past 35 years, methods of program evaluation have undergone a significant change, and the researchers have recognized the need to develop more efficient approaches for assessing treatment effects from studies based on observational data and for evaluations based on quasi-experimental designs.

    Written by experts, this volume is updated and fully reflects the current changes to the field. It offers a systematic review of the history, origins, and statistical foundations of propensity score analysis,  and more.

NeoPopRealism Journal
NeoPopRealism Journal
Key features

NEW TO THIS EDITION:

  • Propensity score sub-classification and propensity score weighting are treated as separate models to give thorough attention to each.
  • Newly expanded coverage of analyzing treatment dosage in the context of propensity score modeling broadens the scope of application for readers.
  • New coverage of modeling heterogeneous treatment effects includes two nonparametric tests and a discussion of modeling issues to ensure students are on the cutting edge.
  • Expanded content on propensity score analysis with multilevel data includes new discussions of four multilevel models for estimating propensity scores and two strategies for controlling clustering effects in outcome analysis.
  • The principles and issues related to running propensity score models with sub-classification and weighting are covered in depth.
  • The authors demonstrate new software and include clear illustrations for analyzing treatment dosage with GPS.

 

KEY FEATURES:

 

  • The authors present key information on model derivations and summarize complex statistical arguments—omitting their proofs to challenge readers to apply their learning.
  • Each method, and its empirical examples, is linked to specific Stata programs for seamless integration of learning and application. 
  • Two conceptual frameworks—the Neyman-Rubin counterfactual framework and the Heckman econometric model of causality—provide a foundation for understanding key topics.
  • Examples in every chapter demonstrate real challenges found in social and health sciences research.
  • Data simulation is used to illustrate key points. 
  • New statistical approaches necessary for understanding the seven evaluation methods are included.

The most significant change of the second edition is discussion of propensity score subclassification, propensity score weighting, and dosage analysis from Chapter 5 to separate chapters. These methods are closely related to the Rosenbaum and Rubin’s (1983) seminal study of the development of propensity scores—it is for this reason that Chapter 5 of the first edition pooled these methods together. Because subclassification and weighting methods have been widely applied in recent research and have become recommended models for addressing challenging data issues (Imbens & Wooldridge, 2009), we decided to give each topic a separate treatment. There is an increasing need in social behavioral and health research to model treatment dosage and to extend the propensity score approach from the binary treatment conditions context to categorical and/or continuous treatment conditions contexts. Given these considerations, we treated dosage analysis in the second edition as a separate chapter. As a result, Chapter 5 now focuses on propensity score matching methods alone, including greedy matching and optimal matching.

Sample Materials & Chapters

Chapter 1

Chapter 2


For instructors

Select a Purchasing Option