The sampling technique from the general population under investigation, which meets the following two conditions: 1) each unit of the population has a positive and known probability of entering the sample; 2) for each set of population units, it is possible to determine the probability that it will be in the sample. The first condition refers to the situation of selecting individual units to be examined (people, enterprises, things), the second condition applies to the situation when whole sets of units are drawn (eg student groups, census enumeration areas). It is worth emphasizing that the first part of the definition does not require at all that the probability of getting into the sample is the same for each unit of the population. However, since this probability is to be known, a researcher using a given random sampling technique must be able to calculate the probability of being in the sample for each individual.
GUS – Departament Programowania, Koordynacji Badań i Rejestrów