The sample is a part of a set, population or universe correctly selected that is submitted to scientific observation on behalf of the set, in order to obtain results valid also for the universe of research (Sierra Bravo, 1992: 174), since it is intended to be representative; the subgroup (sample) is considered representative to be able to extrapolate the results to the whole population or universe.

Sample (n) = units to be observed

Universe or population (N) (finite) = units that could have been observed

Among the conditions required by the sample, we emphasize that it must contain part of the population, not its entirety. The amplitude must be statistically proportional to the size of the population. It has to be representative or a faithful reflection of the population.

Sampling is a widespread practice in the social sciences due to the following reasons: the universe is unlimited or unfathomable; it is not accessible; available time or resources are not enough. The sampling technique is also necessarily used in experiments (it is practically impossible to practice an experiment if it is not in small groups and representative of the population or universe) and in pre-tests, i. e. when the goal is not so much to obtain results. but to prove the usefulness of research design, methods and techniques. It is worth working with samples for speed, decreased research costs, reliability and controls. On the other hand, making a sample is a difficult task.