Complex Sample Design for Rating Estimation

Abstract

A complex sample design is developed to estimate the radio audience rating through a Panel Survey applied five times a year. The purpose of this design is the inclusion of segmentation variables such as sex, age, and family category, which gives rise to a two-stage cluster sampling and different stratification variables. In the first stage, clusters of equal size of 150 dwellings each will be selected via stratified sampling; in the second stage, two dwellings will be selected for each cluster through simple random sampling.