Statistical Methods and Applications

, Volume 18, Issue 2, pp 257–273 | Cite as

Estimating and using propensity score in presence of missing background data: an application to assess the impact of childbearing on wellbeing

  • Alessandra MatteiEmail author
Original Article


Propensity score methods are an increasingly popular technique for causal inference. To estimate propensity scores, we must model the distribution of the treatment indicator given a vector of covariates. Much work has been done in the case where the covariates are fully observed. Unfortunately, many large scale and complex surveys, such as longitudinal surveys, suffer from missing covariate values. In this paper, we compare three different approaches and their underlying assumptions of handling missing background data in the estimation and use of propensity scores: a complete-case analysis, a pattern-mixture model based approach developed by Rosenbaum and Rubin (J Am Stat Assoc79:516–524, 1984), and a multiple imputation approach. We apply these methods to assess the impact of childbearing events on individuals’ wellbeing in Indonesia, using a sample of women from the Indonesia Family Life Survey.


Ignorability Propensity score Missing data Childbearing Wellbeing 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Becker S, Ichino A (2002) Estimation of average treatment effects based on propensity scores. STATA J 2(4): 358–377Google Scholar
  2. Frangakis CE, Rubin DB (1999) Addressing complications of intention-to-treat analysis in the combined presence of all-or-none treatment-noncompliance and subsequent missing outcomes. Biometrika 86: 365–380zbMATHCrossRefMathSciNetGoogle Scholar
  3. Gu XS, Rosenbaum PR (1993) Comparison of multivariable matching methods; structures, distances, and algorithms. J Comp Graph Stat 2: 405–420CrossRefGoogle Scholar
  4. Heckman JJ, Ichomura H, Todd PE (1998) Matching as an econometric evaluation estimator. Rev Econ Stud 65: 261–294zbMATHCrossRefGoogle Scholar
  5. Hill J (2004) Reducing bias in treatment effect estimation in observational studies suffering from missing data. Working paper series, School of International and Public Affairs, Columbia University, New YorkGoogle Scholar
  6. Klepinger D, Lundberg S, Plotnick R (1995) Instrumental selection: the case of teenage childbearing and women’s educational attainment. DP 1077-95, 1995 29 ppGoogle Scholar
  7. Little RJA, Rubin DB (1987) Statistical analysis with missing data (2nd edition in 2002). Wiley, New YorkGoogle Scholar
  8. Royston P (2004) Multiple imputation of missing values. Stata J 4: 227–241Google Scholar
  9. Rosenbaum PT, Rubin DB (1983) The central role of propensity score in observational studies for causal effects. Biometrika 70(1): 41–55zbMATHCrossRefMathSciNetGoogle Scholar
  10. Rosenbaum PT, Rubin DB (1984) Reducing bias in observational studies using subclassification on the propensity score. J Am Stat Assoc 79: 516–524CrossRefGoogle Scholar
  11. Rubin DB (1976) Inference and missing data. Biometrika 63: 581–592zbMATHCrossRefMathSciNetGoogle Scholar
  12. Rubin DB (1978) Multiple imputation in sample survey: a phenomenological Bayesian approach to nonresponse. In: The Proceedings of survey research methods section of the American Statistical Association, pp 20–34Google Scholar
  13. Rubin DB (1987) Multiple imputation for nonresponse in surveys. Wiley, New YorkCrossRefGoogle Scholar
  14. Rubin DB (1996) Multiple imputation after 18+ years (with discussion). J Am Stat Assoc 91: 473–489zbMATHCrossRefGoogle Scholar
  15. Rubin DB, Thomas N (1992a) Affinely invariant matching methods with ellipsoidal distribution. Ann Stat 20: 1079–1093zbMATHCrossRefMathSciNetGoogle Scholar
  16. Rubin DB, Thomas N (1992b) Characterizing the effect of matching using linear propensity score methods with normal distributions. Biomatrika 79: 797–809zbMATHCrossRefMathSciNetGoogle Scholar
  17. Rubin DB, Thomas N (1996) Matching using estimated propensity score; relating theory to practice. Biomatrics 52: 249–264zbMATHCrossRefGoogle Scholar
  18. van Buuren S, Boshuizen HC, Knook DL (1999) Multiple imputation of missing blood pressure covariates in survival analysis. Stat Med 18: 681–694CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2007

Authors and Affiliations

  1. 1.Dipartimento di Statistica “G. Parenti”Università di FirenzeFirenzeItaly

Personalised recommendations