MUESTREO ADAPTATIVO BAJO CONTACTOS ALEATORIOS

Abstract

Strategies are studied under the hyphotesis of random connections among the units. We model the case in which these conections are random. They fix networks and the sample size varies as a function of the contatcs. Hence an additional randomization stage is present. Estimators and their errors are developed. This problem is studied for simple random sampling and stratification. As a solution for computing the errors Bootstrap methods are proposed and computing algorithms are provided