The stages in the selection of a sample for non-experimental, qualitative research are as follow:

1. Deciding whether to study an entire population, referred to as a census approach, or to study samples of the target population, based on specified sampling criteria. The following factors limit the use of census surveys:

• high costs of census surveys;
• census surveys may take too much time to plan and execute;
• it is unethical to include more participants than needed for a study;
• census surveys are logistically challenging; and
• homeless and nomadic persons cannot be represented

These limitations of census surveys result in more widespread use of sampling techniques.

2. Deciding on the sample frame, the groups of participants selected from the target population based on the study's sampling process. Examples of sample frames include:

• school lists;
• hospital records;
• mailing lists; and
• telephone lists.

Sample frames have limitations. For example, school lists may only be available for schools in the public sector; and not everyone suitable for a study may have a phone, which limits the utility of telephone lists for the study.

3. Sampling, the process by which participants in a study are selected from the sample frame. There are a number of sampling methods, such as probabilistic and non-probabilistic sampling.

Probabilistic sampling is representative of the target population. The findings of studies that use this sampling method are generalizable. Non-probabilistic sampling methods are unrepresentative of the target population. The findings of studies that use this sampling method are not generalizable.

It is important to ensure prior determination of the sampling method for a study, as it may affect sample size.

Determining the sampling method for a study in advance may prevent selection bias; and may enhance the validity and generalizability of the study.