Using Propensity Scores in Quasi-Experimental Designs

William Holmes is a faculty member at the University of Massachusetts, Boston, in the College of Public and Community Services. He has evaluated criminal justice and community programs serving families, children, individuals who have suffered abuse, and those with substance abuse problems. He coauthored with Kay Kitson Portrait of Divorce, which won the William Goode Award from the Family Section of the American Sociological Association, and coauthored Family Abuse: Consequences, Theories, and Responses with Calvin Larsen and Sylvia Mignon. Dr. Holmes has conducted research funded by the U.S. Bureau of Justice Statistics, the National Institute of Justice, the National Institute of Mental Health, the National Center on Child Abuse and Neglect, the U.S. Children's Bureau, United Way, foundations, and many community agencies. He received a merit award from the Office of Justice Programs for evaluation of criminal justice programs, as well as the G. Paul Sylvester Award for contributions to criminal justice statistics.

Preface Acknowledgments About the Author Chapter 1. Quasi-Experiments and Nonequivalent Groups Chapter 2. Causal Inference Using Control Variables Chapter 3. Causal Inference Using Counterfactual Designs Chapter 4. Propensity Approaches for Quasi-Experiments Chapter 5. Propensity Matching Chapter 6. Propensity Score Optimized Matching Chapter 7. Propensities and Weighted Least Squares Regression Chapter 8. Propensities and Covariate Controls Chapter 9. Use With Generalized Linear Models Chapter 10. Propensity With Correlated Samples Chapter 11. Handling Missing Data Chapter 12. Repairing Broken Experiments Appendix A. Stata Commands for Propensity Use Appendix B. R Commands for Propensity Use Appendix C. SPSS Commands for Propensity Use Appendix D. SAS Commands for Propensity Use References Author Index Subject Index

"I find the accessibility of propensity scores to be the most appealing contribution of this text. As the authors pointed out, many articles on propensity scores use statistical equations and programs that many users are unfamiliar with. Most students that take workshops from me want how-to instructions for computing and using propensity scores. I like that this book would present them from a methodological and applied approach, rather than the more-common theoretical approach." -- M. H. Clark "The worked up examples in different software programs are a definite strength." -- Tina Savla "The discussion of alternatives in order to control sources of influence is very good." -- Michael A. Milburn "I was most intrigued by some of the material covered near the end of the outline, in particular the chapters on missing data and repairing broken experiments. It is one thing to cover the statistical theory, but in my experience students really need guidance in how to handle messy research design and data situations. In the same vein, I liked seeing how many of the chapters appear to end with sections on assessing the adequacy and sufficiency of the techniques covered in those chapters." -- Douglas Luke