Doubly robust inference with missing data in survey sampling

e de Montr eal Abstract: Statistical inference with missing data requires assumptions about the population or about the response probability. Doubly robust (DR) estimators use both relationships to estimate the parameters of interest, so that they are consistent even when one of the models is misspecified. In this paper, we propose a method of computing propensity scores that leads to DR estimation. In addition, we discuss DR variance estimation so that the resulting inference is doubly robust. Some asymptotic properties are discussed. Results from two limited simulation studies are also presented.

[1]  Peter J. Diggle,et al.  Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models: Comment , 1999 .

[2]  J. Robins,et al.  Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models , 1999 .

[3]  Roderick J. A. Little,et al.  Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models: Comment , 1999 .

[4]  David Firth,et al.  Robust models in probability sampling , 1998 .

[5]  Jae Kwang Kim,et al.  Nonresponse weighting adjustment using estimated response probability , 2007 .

[6]  D. Rubin.  INFERENCE AND MISSING DATA , 1975 .

[7]  Joseph Kang,et al.  Demystifying Double Robustness: A Comparison of Alternative Strategies for Estimating a Population Mean from Incomplete Data , 2007, 0804.2958.

[8]  Ronald H. Randles,et al.  On the Asymptotic Normality of Statistics with Estimated Parameters , 1982 .

[9]  J. Deville.  Variance estimation for complex statistics and estimators: Linearization and residual techniques , 2009 .

[10]  Carl-Erik Särndal,et al.  Model Assisted Survey Sampling , 1997 .

[11]  J. Robins,et al.  Estimation of Regression Coefficients When Some Regressors are not Always Observed , 1994 .

[12]  Mark J van der Laan,et al.  Empirical Efficiency Maximization: Improved Locally Efficient Covariate Adjustment in Randomized Experiments and Survival Analysis , 2008, The international journal of biostatistics.

[13]  A. Winsor.  Sampling techniques. , 2000, Nursing times.

[14]  R. Fay.  Alternative Paradigms for the Analysis of Imputed Survey Data , 1996 .

[15]  J. Shao,et al.  Variance Estimation for Survey Data with Composite Imputation and Nonnegligible Sampling Fractions , 1999 .

[16]  Phillip S. Kott,et al.  Using Calibration Weighting to Adjust for Nonresponse Under a Plausible Model (with full appendices) , 2007 .

[17]  Zhiqiang Tan,et al.  A Distributional Approach for Causal Inference Using Propensity Scores , 2006 .

[18]  Ted Chang,et al.  Using Calibration Weighting to Adjust for Nonignorable Unit Nonresponse , 2010 .

[19]  J. Robins,et al.  Doubly Robust Estimation in Missing Data and Causal Inference Models , 2005, Biometrics.

[20]  M. Davidian,et al.  Improving efficiency and robustness of the doubly robust estimator for a population mean with incomplete data , 2009, Biometrika.

[21]  J. N. K. Rao,et al.  A unified approach to linearization variance estimation from survey data after imputation for item nonresponse , 2009 .

[22]  J. Shao,et al.  Jackknife variance estimation with survey data under hot deck imputation , 1992 .

[23]  Jae Kwang Kim,et al.  Imputation using response probability , 2006 .

[24]  Phillip S. Kott.  A Note on Handling Nonresponse in Sample Surveys , 1994 .

[25]  Wayne A. Fuller,et al.  Fractional hot deck imputation , 2004 .

[26]  A. Singh.  THE GENERALIZED EXPONENTIAL MODEL FOR SAMPLING WEIGHT CALIBRATION FOR EXTREME VALUES, NONRESPONSE, AND POSTSTRATIFICATION , 2002 .

[27]  R. Fay.  When Are Inferences from Multiple Imputation Valid ? , 2002 .

[28]  P. Kott.  Using Calibration Weighting to Adjust for Nonresponse and Coverage Errors , 2006 .

[29]  D. O. Scharfstein.  Adjusting for nonignorable dropout using semiparametric nonresponse models (with discussion) , 1999 .

[30]  Wayne A. Fuller,et al.  Hot Deck Imputation for the Response Model , 2005 .