Sampling methods explained
Sampling methods split into two families. In probability sampling, every member of the population has a known, non-zero chance of being chosen — which is what lets you generalize your results to that population with statistics. In non-probability sampling, selection isn’t random; you can’t generalize statistically, but it’s often exactly right for qualitative, exploratory, or hard-to-reach studies.
Probability sampling (for generalizing)
| Method | How it works | Note |
|---|---|---|
| Simple random | Every member has an equal chance | Gold standard; needs a full sampling frame |
| Stratified | Random sampling within subgroups (strata) | Ensures key subgroups are represented |
| Cluster | Randomly select whole groups, sample within | Cheaper for dispersed populations; less precise |
| Systematic | Every k-th member from an ordered list | Simple; beware hidden periodic patterns |
Non-probability sampling (for depth / access)
| Method | How it works | Common in |
|---|---|---|
| Convenience | Whoever is easiest to reach | Pilots; quick studies (highest bias risk) |
| Purposive | Deliberately pick information-rich cases | Qualitative research |
| Quota | Fill preset quotas per subgroup (non-randomly) | Market/opinion research |
| Snowball | Participants refer others | Hard-to-reach / hidden populations |
How to choose
- Need to generalize to a population with statistics? Use a probability method and a calculated sample size.
- Exploring meaning, or studying a specific/hidden group? Purposive or snowball sampling usually fits.
- Be honest about bias. Convenience sampling is common in student projects — acknowledge its limits rather than over-claiming representativeness.
Get the free Research Design toolkit
Sampling decision charts, bias checklists, and worked examples from Research Design Simplified. We’ll email you the download link.
Frequently asked questions
Probability vs non-probability sampling?
Probability = known chance of selection, supports generalizing; non-probability = non-random, limits statistical generalization but fits qualitative/exploratory work.
What are the main probability methods?
Simple random, stratified, cluster, and systematic.
What is purposive sampling?
Deliberately selecting information-rich cases that fit the study’s purpose — common in qualitative research.
Which should I use?
Probability + a calculated sample size to generalize; purposive/snowball to explore or reach specific groups.