We propose a jackknife variance estimator for the population average from two, two-phase samples after imputation. The jackknife method has long been used to estimate and reduce bias, but has now become a valuable tool for variance estimation. We apply two different sampling methods, (simple random sampling and stratified random sampling) to derive jackknife variance estimators for the twosampl...

A new Retail Commodity Survey was launched by Statistics Canada in January 1997, to obtain more information on commodity by sales in Canada. The survey is based on a two-phase sample design, where the first phase is the Monthly Retail Trade Survey, in place since 1988. Information from the first phase is used in all steps of the commodity survey in order to maximize the efficiency of the design...

Two-phase outcome dependent sampling (ODS) is widely used in many fields, especially when certain covariates are expensive and/or difficult to measure. For two-phase ODS, the conditional maximum likelihood (CML) method very attractive because it can handle zero Phase 2 selection probabilities and avoids modeling covariate distribution. However, most existing CML-based methods use only sample th...

We establish a mathematical framework that formally validates the two-phase " super-population viewpoint " proposed by Hartley and Sielken [Biometrics 31 (1975) 411–422] by defining a product probability space which includes both the design space and the model space. The methodology we develop combines finite population sampling theory and the classical theory of infinite population sampling to...