WEIGHTING ADJUSTMENTS FOR UNIT NONRESPONSE WITH MULTIPLE OUTCOME VARIABLES

Weighting is a common form of unit nonresponse adjustment in sample surveys where entire questionnaires are missing due to noncontact or refusal to participate. Weights are inversely proportional to the probability of selection and response. A common approach computes the response weight adjustment cells based on covariate information. When the number of cells thus created is too large, a coarsening method such as response propensity stratification can be applied to reduce the number of adjustment cells. Simulations in Vartivarian and Little (2002) indicate improved efficiency and robustness of weighting adjustments based on the joint classification of the sample by two key potential stratifiers: the response propensity and the predictive mean, both defined in Section 2. Predictive mean stratification has the disadvantage that it leads to a different set of weights for each key outcome. However, potential gains in efficiency and robustness make it desirable to use a joint classification. Here, we consider the efficiency and robustness of weights that jointly classify on the response propensity and predictive mean, but the base the predictive mean dimension on a single canonical outcome variable. WEIGHTING ADJUSTMENTS FOR UNIT NONRESPONSE WITH MULTIPLE OUTCOME VARIABLES Sonya L.Vartivarian, Department of Statistics, University of Michigan. E-mail: slvartiv@umich.edu Roderick J. Little, Department of Biostatistics, University of Michigan. E-mail: rlittle@umich.edu

[1]  James M. Robins,et al.  On Profile Likelihood: Comment , 2000 .

[2]  R. Little.  Survey Nonresponse Adjustments for Estimates of Means , 1986 .

[3]  Richard A. Johnson,et al.  Applied Multivariate Statistical Analysis , 1983 .

[4]  Jack R. Anderson.  Design and estimation for the National Health Interview Survey, 1995-2004. , 2000, Vital and health statistics. Series 2, Data evaluation and methods research.

[5]  A. W. van der Vaart,et al.  On Profile Likelihood , 2000 .

[6]  D. Rubin,et al.  Statistical Analysis with Missing Data , 1988 .

[7]  R. Little,et al.  On the Formation of Weighting Adjustment Cells for Unit Nonresponse , 2003 .

[8]  W. G. Cochran.  The effectiveness of adjustment by subclassification in removing bias in observational studies. , 1968, Biometrics.

[9]  Roderick J. A. Little,et al.  Statistical Analysis with Missing Data , 1988 .

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

[11]  David Judkins,et al.  NONRESPONSE ADJUSTMENTS FOR A TELEPHONE FOLLOW-UP TO A NATIONAL IN-PERSON SURVEY by H(Lseyin , 2002 .

[12]  R J Little,et al.  Causal effects in clinical and epidemiological studies via potential outcomes: concepts and analytical approaches. , 2000, Annual review of public health.

[13]  R. Little,et al.  On weighting the rates in non‐response weights , 2003, Statistics in medicine.

[14]  D. Rubin,et al.  The central role of the propensity score in observational studies for causal effects , 1983 .

[15]  R. Little.  SURVEY NONRESPONSE ADJUSTMENTS , 2002 .