Variance Estimation for Sampling on Topo Occasions
Heidi Kalbfleisch · 1997
Sampling on two occasions is a vduab3e sampling scheme when observations are taken across two time points, and the correlation between these observations is high. under this framework, estimates are based on a weighted average of a two-phase sampling estimator and an estimator from an independent random sample. There has been recent work in the literature on resampling methods such as the bootstrap and the jackkkfe in developing variance I . estimators under a two-phase sampling scheme. Resampling variance estimators have oper$ ationd advantages over linearisation variance estimators. We extend these developments to sampling on two occasions. An attempt to use new linearisation variance estimators which 3 make more complete use of the data available is made. Through simulation, a study of the unconditional properties of the different estimators is performed. /-