Survey Design

Distribution according to examination characteristics

The stratified sample

In a stratified selection, the population is divided into individual subgroups (strata). Independent random samples are drawn from each (largely homogenous) layer. The stratification characteristics should be related to the object under investigation.

We use layered selection procedures in a series of samples (not to be confused with quota samples). By default, for example, landline samples and residents' registration records are stratified according to regional (federal state, district, etc.) and regional structural characteristics (municipality size classes, region structure types, etc.). To calculate the result parameters, the shift results must be weighted according to their respective ratios.

An important advantage of layered samples is the reduction of the sampling error. This will make the results more accurate. In addition, statements within the individual layers are possible. The stratification may be proportional to the distribution of units in the population or disproportionate to a distribution other than the population.