Combining Multiple Imputation and Inverse‐Probability Weighting

Abstract

Summary Two approaches commonly used to deal with missing data are multiple imputation (MI) and inverse‐probability weighting (IPW). IPW is also used to adjust for unequal sampling fractions. MI is generally more efficient than IPW but more complex. Whereas IPW requires only a model for the probability that an individual has complete data (a univariate outcome), MI needs a model for the joint distribution of the missing data (a multivariate outcome) given the observed data. Inadequacies in either model may lead to important bias if large amounts of data are missing. A third approach combines MI and IPW to give a doubly robust estimator. A fourth approach (IPW/MI) combines MI and IPW but, unlike doubly robust methods, imputes only isolated missing values and uses weights to account for remaining larger blocks of unimputed missing data, such as would arise, e.g., in a cohort study subject to sample attrition, and/or unequal sampling fractions. In this article, we examine the performance, in terms of bias and efficiency, of IPW/MI relative to MI and IPW alone and investigate whether the Rubin’s rules variance estimator is valid for IPW/MI. We prove that the Rubin’s rules variance estimator is valid for IPW/MI for linear regression with an imputed outcome, we present simulations supporting the use of this variance estimator in more general settings, and we demonstrate that IPW/MI can have advantages over alternatives. IPW/MI is applied to data from the National Child Development Study.

Shown for reading. Its publisher allows quoting it, not republishing it.

References 27

1. 1.

   Atherton K, Fuller E, Shepherd P, Strachan DP, Power C. Loss and representativeness in a biomedical survey at age 45 years: 1958 British Birth Cohort. Journal of Epidemiology and Community Health. 2008;62:216-223.

   Not in our library10.1136/jech.2006.058966

2. 2.

   Caldwell TM, Rodgers B, Clark C, Jefferis BJMH, Stansfeld SA, Power C. Lifecourse socioeconomic predictors of midlife drinking patterns, problems and abstention: Findings from the 1958 British Birth Cohort study. Drug and Alcohol Dependence. 2008;95:269-278.

   Not in our library10.1016/j.drugalcdep.2008.01.014

3. 3.

   Gelman A, Carlin JB, Stern HS, Rubin DB. Bayesian Data Analysis. 2004.

   Not in our library

4. 4.

   Goldstein H. Handling attrition and non-response in longitudinal data. Longitudinal and Life Course Studies. 2009;1:63-72.

   Not in our library

5. 5.

   Höfler M, Pfister H, Lieb R, Wittchen H. The use of weights to account for non-response and drop-out. Social Psychiatry and Psychiatric Epidemiology. 2005;40:291-299.

   Not in our library10.1007/s00127-005-0882-5

6. 6.

   Jones MP. Indicator and stratification methods for missing explanatory variables in multiple linear regression. Journal of the American Statistical Association. 1996;91:222-230.

   Not in our library

7. 7.

   Kim JK, Brick JM, Fuller WA, Kalton G. On the bias of the multiple-imputation variance estimator in survey sampling. Journal of the Royal Statistical Society: Series B (Statistical Methodology). 2006;68:509-521.

   Not in our library

8. 8.

   Little RJA, Rubin DB. Statistical Analysis with Missing Data. 2002.

   Not in our library

9. 9.

   Meng X-L. Multiple-imputation inferences with uncongenial sources of input. Statistical Science. 1994;9:538-573.

   Not in our library

10. 10.

    Nielsen SF. Proper and improper multiple imputation. International Statistical Review. 2003;71:593-627.

    Not in our library

11. 11.

    Power C, Elliott J. Cohort profile: 1958 British Birth Cohort (National Child Development Study). International Journal of Epidemiology. 2006;35:34-41.

    Not in our library10.1093/ije/dyi183

12. 12.

    Priebe S, Fakhoury W, White I, Watts J, Bebbington P, Billings J, Burns T, Johnson S, Muijen M, Ryrie I, Wright C. Characteristics of teams, staff and patients: Associations with outcomes of patients in assertive outreach. British Journal of Psychiatry. 2004;185:306-311.

    Not in our library10.1192/bjp.185.4.306

13. 13.

    Robins JM, Gill RD. Non-response models for the analysis of non-monotone ignorable missing data. Statistics in Medicine. 1997;16:39-56.

    Not in our library10.1002/(sici)1097-0258(19970115)16:1<39::aid-sim535>3.0.co;2-d

14. 14.

    Robins J, Wang N. Inference for imputation estimators. Biometrika. 2000;87:113-24.

    Not in our library

15. 15.

    Robins J, Rotnitzky A, Zhao L. Estimation of regression coefficients when some regressors are not always observed. Journal of the American Statistical Association. 1994;89:846-866.

    Not in our library

16. 16.

    Royston J. Multiple imputation of missing values: Update of ice. Stata Journal. 2005;5:527-536.

    Not in our library

17. 17.

    Rubin DB. Multiple Imputation for Nonresponse in Surveys. 1987.

    Not in our library

18. 18.

    Schafer JL. Multiple imputation in multivariate problems when the imputation and analysis models differ. Statistica Neerlandica. 2003;57:19-35.

    Not in our library

19. 19.

    Schenker N, Welsh AH. Asymptotic results for multiple imputation. Annals of Statistics. 1988;16:1550-1566.

    Not in our library

20. 20.

    Stansfeld SA, Clark C, Caldwell TM, Rodgers B, Power C. Psychosocial work characteristics and anxiety and depressive disorders in midlife: The effects of prior psychological distress. Occupational and Environmental Medicine. 2008a;65:634-642.

    Not in our library10.1136/oem.2007.036640

21. 21.

    Stansfeld SA, Clark C, Rodgers B, Caldwell TM, Power C. Childhood and adulthood socio-economic position and midlife depressive and anxiety disorders. Drug and Alcohol Dependence. 2008b;95:269-278.

    Not in our library10.1016/j.drugalcdep.2008.01.014

22. 22.

    Thomas C, Hypponen E, Power C. Prenatal exposures and glucose metabolism in adulthood. Diabetes Care. 2007;30:918-924.

    Not in our library10.2337/dc06-1881

23. 23.

    Van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Statistical Methods in Medical Research. 2007;16:219-242.

    Not in our library10.1177/0962280206074463

24. 24.

    Vansteelandt S, Carpenter J, Kenward MG. Analysis of incomplete data using inverse probability weighting and doubly robust estimators. Methodology. 2010;6:37-48.

    Not in our library

25. 25.

    Wang N, Robins JM. Large-sample theory for parametric multiple imputation procedures. Biometrika. 1998;85:935-948.

    Not in our library

26. 26.

    White IR, Carlin JB. Bias and efficiency of multiple imputation compared with complete-case analysis for missing covariate values. Statistics in Medicine. 2010;29:2920-2931.

    Not in our library10.1002/sim.3944

27. 27.

    White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Statistics in Medicine. 2010;30:377-399.

    Not in our library10.1002/sim.4067

Cited by papers we hold 3

• Strategies for assessing the impact of loss to follow-up on estimates of neurodevelopmental impairment in a very preterm cohort at 2 years of age · 2021
  Not checked
• Skipping breakfast among 8-9 year old children is associated with teacher-reported but not objectively measured academic performance two years later · 2017
  Not checked
• Initial Outcomes From a 4-Week Follow-Up Study of the Text4baby Program in the Military Women’s Population: Randomized Controlled Trial · 2014
  Not checked

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.