Guides · For students and researchers

How many interviews do you need for a dissertation?

It's one of the first questions every qualitative student asks. Here's what the research actually says, how to pick a sensible number for your project, and how to explain it in your methods section.

About a 7-minute read · For interview-based dissertations

The short answer

There's no magic number, and anyone who gives you one without asking about your project is guessing. For most undergraduate and master's dissertations using interviews and thematic analysis, somewhere between 6 and 15 interviews is common. What matters more is that you can explain why your number was enough for your question.

Always check your module handbook and ask your supervisor first. Some departments set a minimum or a range, and that beats any general advice.

What the research says

1

The "12 interviews" study

The most quoted study is Guest, Bunce and Johnson (2006). They coded 60 interviews in batches of six and watched how many new codes appeared. Most of the codes turned up early: the main ideas were already there in the first six interviews, and by about twelve, new codes had almost stopped appearing.

The catch: their participants were similar to each other and the interview questions were focused. With a more varied group or a broader question, you'd expect to need more.

2

Hearing an idea versus understanding it

Hennink, Kaiser and Marconi (2017) separated two things. "Code saturation", where you've heard the range of issues, came at around nine interviews in their study. "Meaning saturation", where you understand each issue in depth, took roughly 16 to 24. In other words, spotting a theme is quicker than being able to write about it with confidence.

3

Bigger, more varied studies need more

Hagaman and Wutich (2017) looked at interviews across several different sites and cultures, and found that themes shared across all the sites needed far more interviews, roughly 20 to 40. If you're comparing groups, plan for enough people in each group, not just enough overall.

4

"Information power"

Malterud, Siersma and Guassora (2016) suggest thinking about how much useful information each interview gives you, rather than counting. You need fewer participants when your aim is narrow, your participants know a lot about the topic, you're using an established theory, the conversations are rich, and you're analysing a few cases in depth. You need more when the opposite is true.

A note on "saturation": Braun and Clarke, who developed reflexive thematic analysis, argue that saturation doesn't fit their approach well, because themes are built by the researcher rather than simply found. If you're using their method, it's often better to justify your sample with information power and the practical limits of a dissertation than to claim you "reached saturation".

How to decide your number

  1. Start with your department's guidance. If the handbook says 8 to 12, aim for that.
  2. Look at your question. A narrow question about a specific group needs fewer people than a broad one about a varied group.
  3. Count your groups. Comparing nurses and doctors? You need enough of each to say something about both.
  4. Be realistic about time. Each hour of interview can take several hours to transcribe and check, then more to code. Ten good interviews you analyse well beat twenty you rush.
  5. Analyse as you go. Code your first few interviews before you finish recruiting. If new interviews keep bringing up important new ideas, consider doing a few more.

How to justify it in your methods section

Examiners look for a clear, honest reason, not a big number. A short paragraph covering four things usually does it: how many people you interviewed, who they were, why that was enough for your aim, and any practical limits. For example:

"I interviewed ten primary school teachers in South Wales. Because the study had a narrow aim and the participants had direct, recent experience of the topic, each interview gave rich, relevant information (Malterud et al., 2016). Coding continued alongside recruitment, and the final two interviews raised no important new ideas. The sample size also reflected the time available for a master's dissertation."

Change the details so it's true for your project. Don't claim saturation unless you actually checked for it.

Coding your interviews as you go?

Coding alongside recruitment is the best way to know when you have enough, and it's easier when your codes live next to your transcripts. Themely is a Google Docs add-on that counts how many interviews each code appears in, so you can see when new interviews stop adding new ideas.

See how Themely works

Free for 3 interviews.

Questions people ask

Is 5 interviews enough for a dissertation?

It can be for a small undergraduate project with a narrow question and similar participants, especially if the interviews are long and rich, but many departments expect more. Check your handbook and ask your supervisor, and explain your reasons clearly in your methods section.

Is 12 interviews enough for thematic analysis?

Twelve is often enough for a focused question with similar participants, which is where the well-known Guest, Bunce and Johnson (2006) figure comes from. With a broader question or a varied group, you may need more.

What is data saturation?

Data saturation is the point where new interviews stop adding new information. It's widely used to justify sample sizes, but it can only really be judged while you analyse, and some researchers argue it doesn't suit every method.

Should I count focus groups as one interview?

A focus group gives you one discussion with several voices, so describe it as such: say how many groups you ran and how many people were in each.

Further reading

Braun, V. and Clarke, V. (2021) To saturate or not to saturate? Questioning data saturation as a useful concept for thematic analysis and sample-size rationales. Qualitative Research in Sport, Exercise and Health, 13(2), 201–216.

Guest, G., Bunce, A. and Johnson, L. (2006) How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1), 59–82.

Hagaman, A.K. and Wutich, A. (2017) How many interviews are enough to identify metathemes in multisited and cross-cultural research? Field Methods, 29(1), 23–41.

Hennink, M.M., Kaiser, B.N. and Marconi, V.C. (2017) Code saturation versus meaning saturation: how many interviews are enough? Qualitative Health Research, 27(4), 591–608.

Malterud, K., Siersma, V.D. and Guassora, A.D. (2016) Sample size in qualitative interview studies: guided by information power. Qualitative Health Research, 26(13), 1753–1760.

More guides