Propensity Score Analysis

Shenyang Guo is the author of numerous research articles in child welfare, child mental health services, welfare, and health care. He has expertise in applying advanced statistical models to solving social welfare problems and has taught graduate courses that address survival analysis, hierarchical linear modeling, structural equation modeling, propensity score analysis, and program evaluation. In addition, Guo serves as the editor of SAGE Publication’s Advanced Quantitative Techniques in the Social Sciences Series and as a frequent reviewer for journals seeking a critique of advanced methodological analyses.Guo is a fellow of American Academy of Social Work and Social Welfare. In addition, Guo serves on behalf of Washington University as the assistant vice chancellor for International Affairs—Greater China and as the McDonnell International Academy ambassador to China working with Fudan University. He was appointed by China’s Ministry of Education in 2016 as the Yangtze-River Chaired Professor at Xi’an Jiaotong University, China.Mark W. Fraser, PhD, holds the Tate Distinguished Professorship at the School of Social Work, University of North Carolina where he serves as associate dean for research. He has won numerous awards for research and teaching, including the Aaron Rosen Award and the Distinguished Achievement Award from the Society for Social Work and Research. His work focuses on risk and resilience, child behavior, child and family services, and research methods. Dr. Fraser has published widely, and, in addition to Social Policy for Children and Families, is the co-author or editor of eight books. These include Families in Crisis, a study of intensive family-centered services, and Evaluating Family-Based Services, a text on methods for family research. In Risk and Resilience in Childhood, he and his colleagues describe resilience-based perspectives for child maltreatment, substance abuse, and other social problems. In Making Choices, Dr. Fraser and his co-authors outline a program to help children build sustaining social relationships. In The Context of Youth Violence, he explores violence from the perspective of resilience, risk, and protection, and in Intervention with Children and Adolescents, Dr. Fraser and his colleagues review advances in intervention knowledge for social and health problems. Intervention Research: Developing Social Programs describes the design and development of social programs. His most recent book is Propensity Score Analysis: Statistical Methods and Applications. Dr. Fraser serves as editor of the Journal of the Society for Social Work and Research. He is a fellow of the National Academies of Practice and the American Academy of Social Work and Social Welfare.

• List of TablesList of FiguresPrefaceAbout the AuthorsChapter 1: IntroductionObservational StudiesHistory and DevelopmentRandomized ExperimentsWhy and When a Propensity Score Analysis Is NeededComputing Software PackagesPlan of the BookChapter 2: Counterfactual Framework and AssumptionsCausality, Internal Validity, and ThreatsCounterfactuals and the Neyman-Rubin Counterfactual FrameworkThe Ignorable Treatment Assignment AssumptionThe Stable Unit Treatment Value AssumptionMethods for Estimating Treatment EffectsThe Underlying Logic of Statistical InferenceTypes of Treatment EffectsTreatment Effect HeterogeneityHeckman’s Econometric Model of CausalityConclusionChapter 3: Conventional Methods for Data BalancingWhy Is Data Balancing Necessary? A Heuristic ExampleThree Methods for Data BalancingDesign of the Data SimulationResults of the Data SimulationImplications of the Data SimulationKey Issues Regarding the Application of OLS RegressionConclusionChapter 4: Sample Selection and Related ModelsThe Sample Selection ModelTreatment Effect ModelOverview of the Stata Programs and Main Features of treatregExamplesConclusionChapter 5: Propensity Score Matching and Related ModelsOverviewThe Problem of Dimensionality and the Properties of Propensity ScoresEstimating Propensity ScoresMatchingPostmatching AnalysisPropensity Score Matching With Multilevel DataOverview of the Stata and R ProgramsExamplesConclusionChapter 6: Propensity Score SubclassificationOverviewThe Overlap Assumption and Methods to Address Its ViolationStructural Equation Modeling With Propensity Score SubclassificationThe Stratification-Multilevel MethodExamplesConclusionChapter 7: Propensity Score WeightingOverviewWeighting EstimatorsExamplesConclusionChapter 8: Matching EstimatorsOverviewMethods of Matching EstimatorsOverview of the Stata Program nnmatchExamplesConclusionChapter 9: Propensity Score Analysis With Nonparametric RegressionOverviewMethods of Propensity Score Analysis With Nonparametric RegressionOverview of the Stata Programs psmatch2 and bootstrapExamplesConclusionChapter 10: Propensity Score Analysis of Categorical or Continuous TreatmentsOverviewModeling Doses With a Single Scalar Balancing Score Estimated by an Ordered Logistic RegressionModeling Doses With Multiple Balancing Scores Estimated by a Multinomial Logit ModelThe Generalized Propensity Score EstimatorOverview of the Stata gpscore ProgramExamplesConclusionChapter 11: Selection Bias and Sensitivity AnalysisSelection Bias: An OverviewA Monte Carlo Study Comparing Corrective ModelsRosenbaum’s Sensitivity AnalysisOverview of the Stata Program rboundsExamplesConclusionChapter 12: Concluding RemarksCommon Pitfalls in Observational Studies: A Checklist for Critical ReviewApproximating Experiments With Propensity Score ApproachesOther Advances in Modeling CausalityDirections for Future DevelopmentReferencesIndex

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.