Un problema di campionamento stratificato in una popolazione variabile nel tempo

In this paper we present the empirical results of a sampling study of construction firms in the city of Rome over a period of eighteen months. A univariate stratification procedure was applied to the variable “number of hired workers” estimated for each firm. For each stratum, weights were calculated as means of monthly observations. The results of this approach were compared with those obtained from a dynamic model of the firms’ population. The relative efficiency of this procedure was estimated with respect to the traditional sampling procedure based on a static specification.