Abstract (endast på engelska)

The estimation of the variance for GREG (general regression) estimators by weighted residuals is widely accepted as a method which yields estimators with good properties.

Since the optimal (regression) estimator shares vital properties with GREG estimators, we may incorporate the weighting procedure also for estimating the variance of the optimal estimator.

We will discuss the resulting variance estimator from a conditional point of view and also have a look at an alternative way of utilizing the weights.

Examples that stress conditional behaviour of estimators will be given for some sampling designs, such as Poisson sampling.

The talk will begin with some introductory remarks concerning notation and terminology in survey sampling theory.