Information Power

Qualitative research has a guideline that sounds like common sense and almost never is.

You interview people. You stop interviewing them when you are not learning anything new. This is called saturation, and for decades it was treated as the quality standard for how many participants your study needed. Keep going until the patterns repeat. Stop when they do. Beautifully simple, and in practice it lets researchers get away with sample sizes that are too small for the questions they are actually asking. A Norwegian pair of researchers, Malterud and colleagues, pointed out that saturation alone is not enough. Whether your sample is adequate depends on how ambitious your question is, how specific your participants are, how strong your theoretical framework is, how good the quality of your dialogue with each participant is, and how focused your analysis is.

They called this information power. A narrow question with a sharply defined group and a strong guiding theory and excellent interviews and tight analysis can produce real insight from a handful of participants. A broad question with a loose population and no theoretical anchor and hurried interviews might require dozens and still fall short. This reframes the whole conversation about sample size, which in qualitative research has always been defensive, always anxious about being dismissed as not enough participants. It is not about enough. It is about enough for what you are actually trying to learn, in the conditions you are actually learning it in. Have you ever read a study that felt convincing with a small sample, and another that felt hollow with a much larger one?

Last week we said themes do not just sit in data waiting to be found. This week we look at the question of how many people you actually need to interview, and at how that question keeps being asked the wrong way.