Every proposal with interviews or focus groups gets the same question from a supervisor, an ethics committee or a reviewer: why that many? “Until saturation” is the usual answer, and on its own it is not enough. You need a number to plan the fieldwork and the budget, a reason for that number, and a rule for when to stop.
1. Why there is no formula
A survey sample is sized for precision: how close an estimate must be to the true value. A qualitative study estimates nothing. It looks for the range of experiences, reasons and meanings, and how they fit together. So the question changes from “how many to be precise?” to “how many until more people stop telling us anything new?” That depends on the study, but not as much as people think: it has been tested many times.
2. What the evidence says
Researchers have coded large sets of interviews one at a time and counted when new codes stopped appearing. The results are close to each other:
| Study | What they found |
|---|---|
| Guest, Bunce & Johnson, 2006 | 60 interviews with women in two West African countries: most codes appeared within the first 12, and the main themes were visible after 6. |
| Hennink, Kaiser & Marconi, 2017 | Code saturation (hearing every topic) came at 9 interviews; meaning saturation (understanding each topic fully) needed 16 to 24. |
| Guest, Namey & McKenna, 2017 | Focus groups: 2 to 3 groups found 80% of the themes, 3 to 6 groups found 90%. |
| Hennink & Kaiser, 2022 | A systematic review of these tests: saturation in 9 to 17 interviews or 4 to 8 focus groups. |
These were mostly fairly similar participants answering a focused guide. A broad aim or a very mixed group needs more. That is where information power comes in.
3. Information power: where your study sits in the range
Malterud, Siersma and Guassora (2016) proposed that the more information each interview holds for your question, the fewer you need. They named five things that raise it:
- A narrow aim. “Why mothers who attended antenatal care still deliver at home” needs fewer interviews than “maternal health in the district”.
- Participants who closely match the aim. People who have lived the experience you are asking about, not a general sample.
- An established theory or framework to guide the questions and the analysis.
- Strong dialogue. Experienced interviewers and focused conversations. Working through interpreters, or with new data collectors, usually means weaker dialogue and more interviews.
- In-depth analysis of a few cases, rather than comparing across many.
The more of these hold, the nearer the bottom of the range you can plan. If none hold, plan above it. A rough guide, built from the evidence above:
| Strengths that hold (of 5) | Interviews | Focus groups |
|---|---|---|
| 0 | 17–25 | 6–8 |
| 1 | 15–20 | 5–8 |
| 2 | 12–17 | 4–6 |
| 3 | 9–15 | 3–6 |
| 4 | 8–12 | 3–4 |
| 5 | 6–10 | 2–3 |
This is a guide for planning, not a rule. The stopping rule decides in the field.
4. One number per group, not one for the study
Saturation is reached within a group of similar people. If you interview mothers, health workers and husbands, each group has its own experiences, and each needs its own number. Twenty interviews spread across four groups is five per group, which is rarely enough for any of them. Decide the groups first, then size each one.
5. A stopping rule
“We stopped at saturation” means little unless you say how you judged it. Francis and colleagues (2010) proposed a rule that is easy to follow and to report:
- Analyse a first set of interviews, for example 10 in a group.
- Then keep interviewing, and stop when 3 more interviews in a row add no new codes.
This only works if you code as you go, not after all the fieldwork is finished. Keep a simple table of which codes first appeared in which interview: when the new ones stop coming, you can show it.
6. A worked example
A study of why women who attended antenatal care still deliver at home, with three groups:
| Group | Strengths that hold | Range | Planned |
|---|---|---|---|
| Mothers who delivered at home (in-depth interviews) | Narrow aim, participants match it (2) | 12–17 | 14 |
| Health extension workers (key informant interviews) | Narrow aim, participants match it, strong dialogue (3) | 9–15 | 10 |
| Husbands (focus groups of 8) | Narrow aim, guided by a framework (2) | 4–6 | 5 |
That is 14 + 10 interviews and 5 focus groups of 8 men: 64 participants in all. The plan sets the budget and the fieldwork days. In the field, the stopping rule decides: if the mothers’ interviews stop adding codes at 12, you can stop there; if they are still adding codes at 14, you continue.
7. What to write in the methods section
“The planned sample (14 in-depth interviews, 10 key informant interviews and 5 focus group discussions, about 64 participants in all) was guided by information power (Malterud et al., 2016) and by empirical evidence that saturation is usually reached within 9 to 17 interviews or 4 to 8 focus groups (Hennink & Kaiser, 2022). Data collection continued until no new codes emerged: after the first 10 interviews in each group, it stopped once 3 further interviews in a row added no new codes (Francis et al., 2010).”
In the results, report the number you actually reached in each group, and whether the stopping rule was met.
8. Common mistakes
- Claiming saturation without saying how it was judged. Reviewers increasingly ask. Give the rule.
- One total for different groups. Size each group.
- Analysing only after fieldwork ends. Then you cannot know when to stop, only guess afterwards.
- A round number with no reason, such as 30 because a previous study used 30. The number can be the same; the reason has to be yours.
- Using saturation with every approach. In reflexive thematic analysis, Braun and Clarke (2021) argue that meaning is made by the researcher, not found waiting in the data, so “saturation” fits poorly. There, justify the sample by information power and the depth of analysis instead.
Do this in Researchite
The study planner suggests a range for each participant group from its information power, sets your stopping rule, and writes this methods paragraph with the references. Then it drafts an interview or focus-group guide for each group.
References
- Braun V, Clarke V. To saturate or not to saturate? Questioning data saturation as a useful concept for thematic analysis and sample-size rationales. Qual Res Sport Exerc Health. 2021;13(2):201–16.
- Francis JJ, Johnston M, Robertson C, Glidewell L, Entwistle V, Eccles MP, Grimshaw JM. What is an adequate sample size? Operationalising data saturation for theory-based interview studies. Psychol Health. 2010;25(10):1229–45.
- Guest G, Bunce A, Johnson L. How many interviews are enough? An experiment with data saturation and variability. Field Methods. 2006;18(1):59–82.
- Guest G, Namey E, McKenna K. How many focus groups are enough? Building an evidence base for nonprobability sample sizes. Field Methods. 2017;29(1):3–22.
- Hennink M, Kaiser BN. Sample sizes for saturation in qualitative research: a systematic review of empirical tests. Soc Sci Med. 2022;292:114523.
- Hennink MM, Kaiser BN, Marconi VC. Code saturation versus meaning saturation: how many interviews are enough? Qual Health Res. 2017;27(4):591–608.
- Malterud K, Siersma VD, Guassora AD. Sample size in qualitative interview studies: guided by information power. Qual Health Res. 2016;26(13):1753–60.
Tesfahun Mulatu
Epidemiologist and biostatistician with more than 10 years in academia and research, including multi-country and mixed-methods studies with local and international research institutions, and monitoring, evaluation and learning across Ethiopia’s development and humanitarian sectors on projects funded by UN agencies and international donors. Tesfahun’s work spans maternal and child health, sexual and reproductive health, gender-based violence and mental health, with over 19 publications.
Epidemiologist and biostatistician, founder of Researchite · LinkedIn