Sankhya: The Indian Journal of Statistics
2002, Volume 64, Series A, Pt. 2, 364--378
ESTIMATING EQUATIONS FOR THE ANALYSIS OF SURVEY DATA USING POSTSTRATIFICATION INFORMATION
J.N.K. RAO, Carleton University, Ottawa, Canada W. YUNG and M.A. HIDIROGLOU, Statistics Canada, Ottawa, Canada
SUMMARY. Finite population parameters may be expressed as solutions to suitable ``census'' estimating equations (EE). Parameter estimates are obtained by solving sample EE which involve design weights as well as adjustment factors (also called g-weights) based on poststratification information. Taylor linearization variance estimators that take account of poststratification adjustments are given and practical implications are noted. For stratified simple random sampling and stratified multistage sampling, jackknife linearization variance estimators are developed. Wald tests and analogues of C.R. Rao's score tests that take account of the survey design features and poststratification are also given. Finally, the implementation of the proposed methodology into a computer system is outlined.
AMS (1991) subject classification}. 62D05, 62G09, secondary 62P20.
Key words and phrases: Generalized regression estimator, jackknife, stratified multi-stage sampling.
Full paper (PDF)