Random sampling techniques are required when seeking to establish scientific evidence. Random sampling helps ensure that research findings are representative of the study population. In quantitative research, the use of random or convenience sampling techniques will directly relate to the interpretation of the study findings. When non-probability sampling techniques are used, the results of the study cannot be generalizable. Generalizability describes how well the findings can be attributed to the population as a whole.

Some random sampling approaches include simple random sampling, stratified sampling, systematic sampling, cluster sampling, and multi-stage sampling. Simple random sampling refers to the technique of using random number selection to choose study participants. When more is known about the study population, stratified sampling can be used to create like subgroups within the study. Cluster sampling is a random sampling technique that divides the entire study population into clusters and randomly selects a cluster rather than an individual participant. Systematic clustering chooses participants from each cluster based on specific criteria. Multi-stage sampling techniques select participants several of the previously mentioned sampling methods.

If non-probability sampling techniques, such as convenience sampling, are used, the study findings cannot be generalizable. Non-probability sampling techniques are typically used in qualitative research because qualitative studies are not designed to be representative of the entire population. Qualitative research can, however, be transferable. Transferability describes the degree to which the findings can be applied to other settings.

Non-probability sampling techniques include convenience sampling, purposive sampling, quota sampling, and snowball sampling. Convenience sampling is a technique that is used when the researcher accepts any eligible participant that expresses interest. Purposive sampling techniques describe the selection of participants based on specific predefined criteria. Quota sampling techniques create a sample with characteristics that are similar to the population at large. Snowball sampling techniques ask the first participant to recommend future participants for the study. In each of these non-probability sampling techniques, it is impossible to obtain a representative sample of the population. Especially in qualitative research, where the sample size is small.