How Many Participants Does Qualitative Research Need? Saturation, Information Power, and the Myth of a Magic Number

Is 10, 15, 20, or 30 Participants Enough for a Qualitative Study?

Few questions in qualitative research attract as many suspiciously precise answers as this one.

You may hear that phenomenology needs a certain number of participants, that interviews should continue until saturation, or that 20 participants are “enough” because another published study used 20. These numbers can feel reassuring when a proposal requires a recruitment target.

The difficulty is that qualitative sample adequacy cannot generally be reduced to one participant count that applies across research questions, methodologies, populations, and analytical strategies.

A narrowly focused interview study involving highly specific participants and rich dialogue may require fewer participants than a broad study seeking substantial variation across several groups. A grounded theory study using theoretical sampling follows a different logic from a qualitative descriptive study. Even the familiar concept of saturation has multiple meanings.

The better question is therefore not simply “How many participants do qualitative studies need?” but “How will I know when this particular qualitative sample provides enough information for the analytical work the study is designed to do?”

Qualitative Sample Size Is About Informational Adequacy, Not a Universal Number

How to Think About Sample Adequacy in Qualitative Research

Why There Is No Universal Qualitative Sample Size

Qualitative studies differ too much for one participant count to serve as a general rule.

Consider an interview study focused narrowly on the experiences of a highly specific group, an ethnography involving prolonged observation across several settings, a grounded theory study that samples iteratively as categories develop, and a multi-site qualitative study comparing several stakeholder groups.

All are qualitative, but “How many participants?” does not mean quite the same thing across them.

Qualitative sample size therefore needs to be considered in relation to the methodology and the kind of information the study requires. This is also why sampling participants in qualitative research should be planned before deciding that a particular number is adequate.

What Does Saturation Mean?

Saturation is commonly used to describe a point at which additional data collection is no longer contributing sufficiently new information of the type relevant to the analysis.

That broad description hides substantial variation.

Researchers use terms such as data saturation, thematic saturation, code saturation, meaning saturation, and theoretical saturation, sometimes with different definitions. Malterud and colleagues explicitly note that saturation is closely tied to methodology and has been applied inconsistently across qualitative research.

This matters because simply writing “interviews continued until saturation” does not tell readers what actually happened. What was expected to saturate? Codes? Themes? Categories? Conceptual relationships? How was the judgment made?

Code Saturation and Meaning Saturation Are Not the Same

A useful illustration comes from research distinguishing code saturation from meaning saturation.

Code saturation concerns whether additional interviews continue to identify new issues or codes. Meaning saturation asks a deeper question: whether the researcher has developed a sufficiently rich understanding of the dimensions, nuances, and meanings of those issues.

A researcher may therefore reach the point where few new topics appear while still needing more data to understand existing topics adequately.

This distinction helps explain why asking only when “nothing new” appears can underestimate what a qualitative analysis actually needs.

Theoretical Saturation Has a More Specific Meaning

In grounded theory, theoretical saturation has a methodological role connected to theoretical sampling and the development of categories and their relationships.

It should not be used casually as a generic synonym for “we interviewed enough people.”

If your study does not use grounded theory, borrowing the phrase theoretical saturation may imply an analytical process you did not actually conduct. Qualitative terminology, like statistical terminology, becomes less useful when it is recruited mainly for decorative purposes.

What Is Information Power?

Malterud, Siersma, and Guassora proposed information power as an alternative way to think about adequate sample size in qualitative interview studies. Their central proposition is straightforward: the more information relevant to the study that a sample holds, the fewer participants may be needed.

They identify five dimensions relevant to information power:

The framework does not provide a formula that converts these dimensions into an exact participant number. It provides a structured argument for judging whether the sample is likely to contain sufficient relevant information.

A Narrow Study May Need Fewer Participants Than a Broad One

Suppose one study asks how experienced neonatal nurses describe one highly specific decision they repeatedly make in a defined clinical context. Participants are carefully selected for direct experience, interviews are detailed, and the analysis is tightly focused.

Another study asks how healthcare professionals generally experience digital transformation. It includes physicians, nurses, administrators, allied-health professionals, several hospital types, multiple technologies, and many forms of organizational change.

The second study demands considerably more variation and analytical breadth. Expecting both projects to need the same number of interviews because both are “qualitative” would make little methodological sense.

Sample Heterogeneity Usually Increases the Information You Need

If meaningful differences across participant groups are central to the analysis, the sample must contain enough evidence to investigate those differences.

A study comparing novice and experienced teachers, for example, needs sufficient information from both groups. Add public and private schools, several disciplines, multiple geographic regions, and different AI-use profiles, and the analytical burden grows further.

This does not mean every subgroup needs an equal quota. It means the intended comparisons and variation should be reflected in the sampling and adequacy logic.

The more dimensions of variation you promise to analyze, the harder it becomes to defend a very small sample simply by saying that “qualitative research uses fewer participants.”

Data Richness Matters, Not Just Participant Count

Twenty one-hour interviews producing detailed, reflective accounts are not informationally equivalent to twenty brief interviews in which participants provide only a few relevant sentences.

Information power explicitly includes quality of dialogue for this reason.

The same principle extends beyond interviews. Qualitative evidence may include observations, documents, diaries, field notes, visual material, online interactions, or repeated encounters. Participant count alone may poorly represent the actual volume and richness of the evidence.

This is one reason a sample-size statement should be connected to the data-collection strategy rather than reported as an isolated number.

Do Published Numerical Recommendations Help?

Empirical methodological studies can provide useful reference points by examining when particular forms of saturation occurred in specific datasets. Such evidence may help researchers anticipate recruitment and evaluate whether an initial target is plausible.

But observed saturation in one set of interview studies does not create a universal threshold for all qualitative research. Differences in study aim, sample heterogeneity, interview quality, coding granularity, methodology, and analytical ambition can change what counts as adequate.

Numerical recommendations are therefore better treated as planning evidence than as commandments.

You May Need an Initial Recruitment Target Before Saturation Can Be Assessed

Ethics applications, budgets, recruitment plans, and research proposals often require researchers to estimate participant numbers before data collection begins. “We will know when we get there” may be methodologically defensible in spirit but administratively unhelpful.

You can provide an initial anticipated range or recruitment target based on the methodology, study aim, sample specificity, expected variation, prior methodological literature, and practical experience, while explaining that final adequacy will be evaluated using the prespecified qualitative criterion.

This is different from pretending the initial target is an exact calculation.

If you anticipate approximately 20 to 30 interviews, for example, explain why that range is plausible for this design and how you will decide whether recruitment should stop, continue, or become more targeted as analysis proceeds.

Sampling and Analysis May Need to Proceed Together

Some qualitative approaches require analysis during data collection because emerging findings inform what information is still missing.

If interviews are all completed before meaningful analysis begins, claims that saturation guided recruitment can become difficult to substantiate. Researchers cannot readily use an analytical stopping criterion that they did not assess until after recruitment had already stopped.

Iterative analysis is particularly important when sampling is refined in response to emerging categories, contrasts, or gaps.

More Participants Are Not Automatically Better

Quantitative intuitions about sample size do not transfer neatly to qualitative research.

Adding participants can broaden variation and strengthen an analysis when important perspectives remain underdeveloped. But more data also create analytical demands. A very large interview dataset analyzed superficially may provide less insight than a smaller, appropriately sampled dataset examined with substantial depth.

The objective is therefore not to maximize participant count. It is to obtain enough relevant, rich, and appropriately varied information to support the intended qualitative analysis.

This broader principle is related to why a bigger sample does not automatically make a study better.

Sample Adequacy Does Not Establish Transferability by Itself

Reaching saturation or judging that the sample has sufficient information power does not mean the findings statistically represent a broader population.

Those concepts concern adequacy for the qualitative analysis. Questions about whether insights may be informative in other contexts require attention to study setting, participant characteristics, contextual similarity, theoretical reasoning, and the form of inference appropriate to the methodology.

The distinction becomes important when considering generalizability, external validity, and transferability. A saturated sample is not automatically a representative sample, and those concepts should not be treated as synonyms.

Planning Qualitative Sample Size Without Pretending You Know the Exact Number

The final participant count is still reported, of course. What makes the number defensible is the methodological reasoning around it, not the number's resemblance to a familiar rule of thumb.

Common Mistakes When Justifying Qualitative Sample Size

Plan a Defensible Range, Then Evaluate the Information You Actually Obtain

A useful justification should answer three questions: Why was the initial target plausible? How was sample adequacy evaluated during the study? Why was recruitment stopped when it was?

That explanation is considerably more informative than “the sample consisted of 20 participants, which is sufficient for qualitative research.”

Before You Finalize a Qualitative Sample-Size Plan

Questions About Qualitative Sample Size

Stop Looking for the One Number Every Qualitative Study Supposedly Needs

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.