L'indagine ISTAT sulle forze di lavoro: l'effetto del disegno

Abstract

This paper reports results of a wide empirical study carried out in order to evaluate the design effect components of the Italian labour force survey sampling design. After a brief description of the sampling scheme, we introduce the post-stratified separate ratio estimator of a total and establish an interesting property of self weighting samples. Then we describe the empirical study developed using data from the 1981 demographic census in Umbria. The results concern the contribution of stratification, clustering, systematic selection, and post-stratification to the final design effect in estimating various labour force characteristics. At the end, some comments on the obtained results are given.