Choosing Between Probability and Non-Probability Sampling
Sampling decides who takes part in your study, and it has a direct effect on what you can claim from your results. Supervisors and examiners often ask students to justify their sampling method. In this blog, we'll compare probability and non-probability sampling, outline the main techniques, and explain how to choose.
Probability Sampling:
- Every member of the population has a known, non-zero chance of being selected.
- It requires a sampling frame, which is a complete list of the population, such as a staff register or student database.
- It supports statistical generalisation from your sample to the wider population.
Main Probability Techniques:
- Simple random sampling: every member has an equal chance, for example selecting names using a random number generator.
- Systematic sampling: selecting every nth person from a list after a random start.
- Stratified sampling: dividing the population into groups, such as departments or age bands, and sampling randomly within each group.
- Cluster sampling: randomly selecting whole groups, such as schools or branches, and then studying people within them.
Non-Probability Sampling:
- Participants are selected based on judgement, availability or referral, not random chance.
- It is common when no sampling frame exists or when the population is hard to reach.
- Results cannot be statistically generalised in the same way, so claims need more care.
Main Non-Probability Techniques:
- Convenience sampling: selecting people who are easy to reach. It is quick but carries a high risk of bias.
- Purposive sampling: selecting people who have specific knowledge or experience relevant to your question. It is common in qualitative research.
- Snowball sampling: asking participants to refer others. It is useful for hidden or hard-to-reach groups.
- Quota sampling: setting targets for groups, such as a set number of men and women, and filling them non-randomly.
How to Choose:
- If you want to generalise statistically to a defined population and you have a sampling frame, use probability sampling where possible.
- If your aim is depth and understanding rather than generalisation, purposive sampling is usually appropriate.
- If you use convenience sampling, acknowledge its limits clearly and avoid overstating how far your findings apply.
The right sampling method depends on your research aim, your population and the resources you have. Whichever you choose, describe it precisely, explain why it suits your study, and be honest about what it allows you to conclude.