Data saturation in qualitative research

By Dr. Rafiq Muhammad, MD, PhD · Updated June 2026

Qualitative research can’t use a power calculation to set its sample size — so it uses saturation: the point where new data stop telling you anything new. You keep hearing the same themes, and the next participant adds nothing. It’s how you justify “why this many?” — and reviewers want to see how you judged it, not just the word.

What it actually means

Saturation is reached when additional data produce no new codes or insights. Crucially, it’s judged from the data as you go — which means analysing alongside collection, not deciding a number up front and stopping there.

The types worth naming

• Code saturation — no new codes appear.
• Meaning saturation — no new understanding of existing codes develops (usually takes longer than code saturation).
• Theoretical saturation — in grounded theory: a developing category is fully accounted for, with no new properties emerging.

So how many participants?

It depends on method and topic. Some research finds themes often saturate within roughly 9–17 interviews for fairly homogeneous samples; IPA studies use far fewer; diverse or complex topics need more. There’s no fixed number — see the broader “how many is enough?” logic: stop when you stop learning.

The honest caveat

Saturation has been critiqued as vague and sometimes used to retrofit a justification for a convenient sample size. The credible move is to define what saturation meant in your study, set it before you start where you can, and show the evidence — part of overall trustworthiness.