Optimal Auxiliary Variable Assisted Two-Phase Sampling Designs.

摘要

Two-phase sampling is a procedure in which sampling and data collection is conductedin two phases, aiming at achieving increased precision in estimation at reduced cost. The rst phase typically involves sampling a large number of elements and collecting data onvariables that are easy to measure. In the second phase, a subset is sampled for whichall variables of interest are observed. Utilization of the information provided by the dataobserved in the rst phase may increase precision in estimation by optimal selection ofsampling design the second phase.This thesis deals with two-phase sampling when a random sample following some generalparametric statistical model is drawn in the rst phase, followed by subsampling withunequal probabilities in the second phase. The method of maximum pseudo-likelihoodestimation, yielding consistent estimators under general two-phase sampling procedures,is presented. The design inuence on the variance of the maximum pseudo-likelihoodestimator is studied. Optimal subsampling designs under various optimality criteria arederived analytically and numerically using auxiliary variables observed in the rst samplingphase.