Conglomerate sampling is a technique that explores the existence of groups in the
population. These groups adequately represent the total population in relation to the characteristic that we want to measure. In other words, these groups contain variability in the entire population. If this happens, you can select only a few of these conglomerates to conduct the study. Unlike the previous techniques in which the sample units coincide with the individuals to be studied, in the sample by clusters, the sample units are study groups, which can be very beneficial in relation to the sampling cost itself. In return, it is common to obtain less precision when using this technique, caused by the lack of heterogeneity within the conglomerates. For example, if we want to know the education of the residents of a neighborhood in a given Brazilian metropolis, we will divide this neighborhood into small areas on a map. After this division, a simple random sampling of these small areas is made and, in them, all their residents will be interviewed to learn about their schooling.