There is no reliable universal number of interviews that makes a qualitative study “saturated.” Saturation is used in several different ways and does not fit every method. It is more useful to treat it as a documented judgement about whether additional material is changing the analysis in ways relevant to the study’s purpose.
Ask what is becoming stable
Are new cases no longer adding relevant code meanings? Are the boundaries of key categories still changing? Are important contrasts sufficiently explored? These questions are more informative than a numerical target detached from the design.
Do not confuse repetition with adequacy
Repeated statements may signal a stable pattern, but they can also reflect a narrow recruitment pathway. Conversely, new material may be analytically important even when it produces no new code label because it changes the interpretation of an existing one.
Report process, not magic
Describe the sample, analytic iteration, and the basis for deciding that further collection was or was not necessary. In reflexive approaches, avoid presenting saturation as an objective endpoint if it conflicts with the method’s assumptions.
Questions researchers often ask
Can software prove saturation?
No. Software can help document coding development and coverage, but the methodological judgement remains the researcher’s.