In details

Probabilistic sample


In order to make inferences about the population, it is necessary to work with probabilistic sampling. It is the method that guarantees safety when investigating any hypothesis.

Usually the investigated individuals are just as likely to be selected in the sample.

Simple Random

It is the most widely used sampling process. Practical and effective, it gives precision to the sampling process. Usually a table of random numbers is used and the individuals are named, drawing one by one until completing the calculated sample. The random draw available in spreadsheets such as Excel is commonly used.®.

A variation of this type of sampling is systematic. In a large number of examples, the researcher comes across the ordered population. In this sense, there are individuals arranged in sequence, which makes the exact application of this technique difficult.

When working with house block draw for example, there is a growing rule for house numbers. In cases like this, the population is divided by the sample and a coefficient (Ħ) is obtained. The first box will be number x, the second will be number x + Ħ; the third will be number x + 3. Ħ.

Assuming this coefficient is 6. The first element will be 3. The second will be 3 + 6. The third will be 3 + 2.6. The fourth will be 3 + 3.6, and so on.

Stratified Random

When you want to keep a proportionality in the heterogeneous population. Each subpopulation is stratified by criteria such as social class, income, age, gender, among others.

Conglomerate

In ordinary situations, it is difficult to collect characteristics of the population. In this sampling mode, one set is drawn and the whole set is studied. It is an example of cluster sampling, families, organizations and blocks.

Next: Sample Scaling