Being Bored Isn't Saturation: How to Actually Justify Your Qualitative Sample Size

A doctoral student once told me she'd stopped at interview seventeen because she "wasn't hearing anything new anymore." When I asked how she knew that, she said she just felt it, sitting across from her eighteenth potential participant, and decided to cancel the session.

That feeling might have been correct. It might also have been fatigue, pattern blindness, or simple boredom with a repetitive interview protocol wearing on her attention. The trouble is that a feeling, on its own, can't tell the difference, and neither can a committee reading a methodology chapter that offers no other justification.

Saturation was never supposed to be a feeling

Theoretical saturation, as originally defined in grounded theory methodology, is a specific, demonstrable claim: that continued data collection is no longer producing new codes, categories, or conceptual properties relevant to the emerging theory. It's a statement about the data itself, verified through systematic analysis, not a statement about the researcher's subjective sense of repetition.

This distinction matters because the two experiences feel identical from the inside. A researcher who has genuinely reached saturation and a researcher who is simply exhausted after seventeen long interviews will both report the same sensation: nothing new seems to be coming up anymore. Only one of them has actually verified that claim against the data. The other has mistaken personal fatigue for an empirical finding, and a tired mind is a notoriously unreliable instrument for detecting subtle new patterns it has stopped looking closely enough to see.

What verified saturation actually requires

Proving saturation, rather than simply feeling it, requires the same disciplined process that turns raw interview transcripts into usable qualitative data in the first place: systematic coding paired with constant comparison. Each new interview gets coded using the same categorical framework applied to every prior one, and the researcher explicitly checks whether that specific interview introduced any new code, or whether it simply reinforced categories already established.

This is detailed, comparative work, typically supported by qualitative software like NVivo or ATLAS.ti, which let a researcher track exactly when a new code was last introduced and how the frequency of new categories has changed across the sequence of interviews. A saturation table showing that no new codes emerged across the final three or four interviews is a defensible, auditable claim. "I felt like I'd heard it all before" is not, however true the feeling may have turned out to be.

Why this distinction matters in a defense

A committee member who understands qualitative methodology will ask, directly, how saturation was determined, and "I sensed it" is one of the fastest ways to lose credibility in that specific moment. Rigorous qualitative coding and saturation analysis produces a specific, citable answer instead: a documented point in the coding process where new category emergence demonstrably stopped, backed by the coding record itself rather than the researcher's memory of how the interviews felt to sit through.

This isn't just a matter of covering yourself in a viva. The distinction protects the actual integrity of the finding. A sample size justified by systematic evidence of saturation is a genuinely defensible methodological claim. A sample size justified by researcher fatigue is a coincidence that happened to produce a number, and there's no way to know, without the coding record, whether stopping one interview earlier would have missed a meaningful category, or whether five more interviews were entirely unnecessary.

The fix is procedural, not instinctual

None of this requires interviewing indefinitely out of anxiety that saturation was never really proven. It requires building the coding and comparison process into the data collection itself, from the first few interviews onward, so that the decision to stop is made by tracking an actual, visible decline in new code emergence rather than waiting for a subjective sense of repetition to arrive.

In practice, this means coding interviews as they happen rather than waiting until all of them are collected, so the researcher can see, in real time, whether interview twelve introduced anything new that interview eleven hadn't already covered. The stopping point becomes a specific, visible moment in that ongoing record, not a private feeling that happens to align with a suspiciously round number of completed sessions.

Prove it, don't just feel it

The difference between a researcher who has reached genuine saturation and one who has simply run out of patience is invisible from the inside, and that's exactly why it can't be left to instinct. Only a systematic coding record can actually distinguish the two, and only that record will hold up when a committee member asks the question every qualitative researcher should expect to face.