Sampling explained

Probability and non-probability sampling explained simply, with guidance on sample size and how to justify your choice.

7 min read · Reviewed 2026

Sampling is where ambition meets reality. Your sample determines what you can claim, and claiming more than your sample supports is one of the fastest ways to lose marks.

There is no universal correct sample size. There is only a sample that is adequately justified for your design.

Probability sampling

Random, systematic, stratified and cluster sampling all aim at statistical representativeness so that findings can be generalised to a defined population. They require a sampling frame — a list of the population — which most student projects do not have.

Non-probability sampling

Purposive, convenience, snowball and quota sampling select participants for relevance or accessibility rather than representativeness. This is entirely legitimate for qualitative work, provided you claim transferability rather than generalisability.

How many participants?

For qualitative work, justify size by the depth of data and the point at which new interviews stop producing new codes. For quantitative work, justify it by the analysis you intend to run and the precision you need. In both cases, state the reasoning rather than a number alone.

Common mistakes

Calling a convenience sample random.
Generalising to a population from a self-selected online sample.
Choosing a sample size with no stated rationale.
Not reporting who declined or dropped out.

Try this: Write your sampling paragraph

Name your sampling strategy precisely.
State your inclusion and exclusion criteria.
State your target size and the reason for it.
Write one sentence on what this sample cannot support.

Do it in SuperDiss

Track recruitment, consent and dropout as you go, so the write-up is accurate.

Open the Fieldwork tracker