Purposive Sampling Examples: 8 Types and How to Write Yours

Most purposive sampling examples you find online stop at one sentence, something like "a researcher studying chronic pain picks people with chronic pain." That's true, and it doesn't help when you have to write your methods chapter. This guide gives you the full criteria for each type, plus a paragraph you can adapt.

Purposive sampling means choosing people on purpose because of what they know. You're not pulling names from a hat. Think of it like casting a play: you need someone who has actually done the job you want to hear about.

What purposive sampling is

Purposive sampling goes by a few names. Purposeful, judgmental, and selective sampling all mean the same thing.

The goal is to find what researchers call information rich cases. These are people who can teach you a lot because they sit close to the heart of your question.

It's a non probability method. That means you can't calculate a margin of error or claim your findings apply to everyone. It also only works if you already know enough about your topic to say who the right people are.

A worked example with real criteria

Here's a study written out the way a method chapter needs it. The data is made up for this guide.

Question: Why do newly qualified nurses leave their first hospital job within two years?

Compare that to "I interviewed nurses who left." The table takes ten minutes and removes most of the questions an examiner would ask.

Notice the exclusion has a reason attached. "Excluded if they relocated" is a rule. "Excluded if they relocated, since that's a different reason for leaving" is a decision you can defend.

Purposive sampling examples for all eight types

1. Criterion sampling. Everyone who meets a set standard is eligible. Example: patients readmitted to hospital within 30 days of going home. This is the most common type in practice.

2. Maximum variation sampling. You pick cases that differ as much as possible. Example: mental health services in cities and villages, big teams and small ones. Anything that shows up everywhere is worth reporting.

3. Homogeneous sampling. The opposite. A narrow group who share what matters. Example: students who are first in their family at university, all in year one, all on one course. Good for focus groups, where shared ground helps people open up.

4. Typical case sampling. You pick a normal case to show what it usually looks like. Example: a rural school sitting in the middle for size, funding, and results.

5. Extreme case sampling. You pick the outliers. Example: the three schools with the best attendance and the three with the worst. Just remember outliers don't represent anyone.

6. Critical case sampling. One case chosen so the result carries weight. Example: test a new records system at the best staffed hospital. If people there can't make it work, "they just need more training" stops being convincing.

7. Expert sampling. People with real specialist knowledge. Define what counts. "Experts in the field" isn't a criterion. "Five years working on this, plus a publication or a named policy role" is.

8. Total population sampling. Everyone in a small, clearly bounded group. Example: all 11 people who have ever held one national role.

Which type fits your question

Start with your question, not the list.

If you're asking what something is like for people who lived through it, use criterion sampling. If you want to know what holds true across very different settings, use maximum variation. If you want one tight group explored in depth, go homogeneous.

You can also combine two. In the nurse example, criterion sampling decided who was eligible, and maximum variation decided how to spread the sample across wards. They do different jobs, so they sit together fine.

Is your sample really purposive?

This one's worth being honest about, because examiners notice.

Convenience sampling means taking whoever is available and willing. Purposive sampling means deciding in advance who counts, then going out and finding those people.

Ask yourself four things. Did you write your criteria before recruiting or after? Did you turn anyone away? Can you name someone you wanted but couldn't reach? Did you go looking, or did you post once and take the replies?

Plenty of good student projects are partly convenience based. That's fine. What hurts a chapter is calling it purposive when it wasn't.

How to justify your sample size

There's no fixed number, and writing "there's no fixed number" isn't a justification.

The usual tool is information power, from a 2016 paper by Malterud, Siersma and Guassora. The idea is simple: the more relevant information your sample holds, the fewer people you need.

Five things shrink the number you need. A narrow aim. Participants very close to the topic. Theory applied. Strong interviews. In depth case analysis. Broad aims and weak interviews push the number up.

So you write something like: "Sample size was guided by information power. The aim was narrow, participants were highly specific to the question, and interview quality was strong. On that basis, 14 interviews were judged sufficient."

That beats claiming saturation, which invites a follow up question you may not want. Our thematic analysis example covers why saturation sits awkwardly with some qualitative methods.

What to say about limitations

Be specific. A vague paragraph about bias reads as filler.

You can't generalise statistically, because your sample wasn't random. Your own judgement shaped who got picked. People who volunteer often differ from people who don't. And if you recruited through one forum, you reached people who use that forum.

What you claim instead is transferability. You describe your sample and setting clearly enough that a reader can judge whether your findings apply to theirs. That's exactly why the criteria table matters.

Where all this sits in your wider design is covered in our guide to research methodology, and the boundaries of your study belong in your scope and delimitation section.

FAQs

How many people do I need?

There's no set number. Small interview studies often run between six and 20. Justify yours with information power, not a rule of thumb.

What's the difference between purposive and quota sampling?

Quota sampling sets number targets so your sample mirrors a population. Purposive sampling picks people for what they know, with no matching required.

Is snowball sampling the same thing?

No, but they're often used together to reach people who are hard to find. If you used both, name both.

Can I use purposive sampling in a survey?

Yes, as long as you aren't reporting percentages meant to represent a wider group. For that you need a random sample.

How can Canyam help with my sampling section?

You can search across millions of papers and read summaries that show how similar studies chose their participants, then open the originals to check the details before you cite them.

Start with the table

Open a blank document and fill in the criteria table from this article for your own study. Who counts, who doesn't, how you'll find them, and why that number.

If you can't fill a row, you've just found the decision you still need to make. Working through purposive sampling examples like the ones above takes an afternoon. Discovering the gap in your viva takes a lot longer.