On some calibration estimators of subpopulation total for longitudinal data

The problem of modeling longitudinal profiles is considered assuming that the population and elements affiliation to subpopulation may change in time. The considerations are based on a model with auxiliary variables for longitudinal data with subject specific (in this case - element and subpopulation specific) random components (compare Verbeke, Molenberghs, 2000; Hedeker, Gibbons, 2006) which is a special case of the General Linear Mixed Model. In the paper calibration estimators of subpopulation total for data from one period are presented and some modifications for the case of longitudinal data are proposed. Design-based mean squared errors and its estimators are also presented. In the simulation study accuracy of the estimators is compared with Horvitz-Thomson estimator and the best empirical linear unbiased predictor derived for the considered model.