Quantitative Methods of Data Collection | Samples

Probabilistic and Non-probalistic Samples

Basically, two types of samples can be distinguished: probabilistic and non-probabilistic. In probabilistic sampling, the probability of being included in the sample is the same for each member of the population (or at least known). The best-known variant is probably random sampling: the desired sample is drawn at random from a list of all population members, with each population member having exactly the same "chance" of being included in the sample.

On the one hand, a random sample is the gold standard when it comes to making statements about the population on the basis of the sample results, but on the other hand it is difficult to achieve in practice, if only because the complete list of population elements necessary for probabilistic samples is rarely available in practice. Most samples in research practice are therefore non-probabilistic samples (unknown selection probability), more precisely so-called convenience samples: one takes the participants one can get.

The results on the basis of such a convenience sample are thus biased in comparison to a random sample and cannot be transferred to the population in the same way. How serious this bias is must be assessed in each individual case and depends on the population, the sample and also very much on the research question under investigation.

To summarise briefly: True random sampling (or probabilistic sampling) is the highest quality sampling, but difficult to implement in practice. Thus, it is probably not expected for your Bachelor's or Master's thesis, but in return you should also not misuse the term "random sample": if your sample consists of those persons who were "randomly" available and willing to participate, then it is a convenience sample.

Literature

Bortz, Jürgen & Döring, Nicola (2015). Forschungsmethoden und Evaluation in den Sozial- und Humanwissenschaften. Wiesbaden: Springer

NOTE: Find answers to Frequently Asked Questions here: FAQ