Research Saturation
A context-dependent judgment that further qualitative data collection is adding insufficient new information or meaning for a study’s analytic purpose.
Download Markdown (.md)Soroush’s Product Wiki · https://wiki.soroush.uk/topic/research-saturation/
01What it means
Research saturation is used in some qualitative approaches to assess whether additional data is still contributing new codes, properties, themes, or understanding. The relevant kind of saturation depends on the research question, population, method, analytic depth, and whether the goal is code coverage, meaning, or theory.[1][2]
- 01Define analytic goalCode coverage, meaning, theory, or another contribution
- 02Collect + analyzeDo not postpone analysis until all sessions end
- 03Track new informationCodes, properties, exceptions, variation
- 04Challenge repetitionCheck sample and guide breadth
- 05Document stopping judgmentEvidence, limits, unresolved populations
Apparent repetition may reflect adequate coverage—or a narrow sample and interview guide. State the kind of saturation and unresolved gaps.
02Why it matters
Saturation can support a reasoned decision to continue or stop qualitative collection instead of relying only on a conventional participant count. Used transparently, it focuses attention on data adequacy and variation.
03When to use it
Consider it in iterative qualitative studies with concurrent analysis and a reasonably focused population and question. Do not use it as a universal rule for every qualitative method, a fixed sample-size formula, or a substitute for including important segments.
04How it works
- Define what kind of new information would matter for the study’s analytic goal.
- Collect and analyse data iteratively rather than waiting until all sessions are complete.
- Track new codes, meanings, exceptions, and subgroup variation across successive cases.
- Test whether apparent repetition reflects adequate coverage or a narrow sample and interview guide.
- Document the stopping judgment, evidence, unresolved gaps, and populations to which it does not apply.
05A practical example
A team interviewing one homogeneous group about a narrow billing task sees no new process steps across several successive sessions but continues with screen-reader users because that relevant context remains underrepresented.
06Where it breaks
There is no universal saturation number, and researchers disagree about the concept’s fit across qualitative traditions. Poor questions or homogeneous recruitment can create false repetition, while rare but important cases may never become common themes.
Common mistakes
- Declaring saturation after a preset number of interviews without concurrent analysis.
- Using no new codes as proof that all meanings and segments are understood.
- Stopping when familiar themes repeat while relevant outliers remain unsampled.
07Sources
- Sample sizes for saturation in qualitative research: A systematic review of empirical testsSocial Science & Medicine · Systematic reviewSample sizes for saturation in qualitative research: A systematic review of empirical tests
- A simple method to assess and report thematic saturation in qualitative researchPLOS ONE · Peer-reviewed researchA simple method to assess and report thematic saturation in qualitative research