Weighting Regressions by Propensity Scores
- 1 August 2008
- journal article
- Published by SAGE Publications in Evaluation Review
- Vol. 32 (4), 392-409
- https://doi.org/10.1177/0193841x08317586
Abstract
Regressions can be weighted by propensity scores in order to reduce bias. However, weighting is likely to increase random error in the estimates, and to bias the estimated standard errors downward, even when selection mechanisms are well understood. Moreover, in some cases, weighting will increase the bias in estimated causal parameters. If investigators have a good causal model, it seems better just to fit the model without weights. If the causal model is improperly specified, there can be significant problems in retrieving the situation by weighting, although weighting may help under some circumstances.Keywords
This publication has 25 references indexed in Scilit:
- Examination of the association between syringe exchange program (SEP) dispensation policy and SEP client‐level syringe coverage among injection drug usersAddiction, 2007
- Intergenerational mobility and sample selection in short panelsJournal of Applied Econometrics, 2006
- On The So-Called “Huber Sandwich Estimator” and “Robust Standard Errors”The American Statistician, 2006
- Hospice Enrollment and Evaluation of Its Causal Effect on Hospitalization of Dying Nursing Home PatientsHealth Services Research, 2006
- Comparing Experimental and Matching Methods Using a Large-Scale Voter Mobilization ExperimentPolitical Analysis, 2006
- Doubly Robust Estimation in Missing Data and Causal Inference ModelsBiometrics, 2005
- Statistical Models: Theory and PracticePublished by Cambridge University Press (CUP) ,2005
- Nonexperimental Versus Experimental Estimates of Earnings ImpactsThe Annals of the American Academy of Political and Social Science, 2003
- Inverse problems as statisticsInverse Problems, 2002
- The Price of Bias Reduction when there is no Unbiased EstimateThe Annals of Statistics, 1989