Check where a qualitative claim reaches too far.
Paste findings or discussion text. The check points to wording a reviewer may question.
It sees words, not your transcripts. A flag means “look here,” not “this finding is wrong.” Nothing is uploaded and no research decision is made for you.
No account · no upload · nothing leaves your browser.
What this checks
- Theme stated without a quote — a finding or theme sentence with no participant quote or excerpt in it.
- Quantifying language — "most", "the majority", "often", "%", "n=". A purposive sample can't show how common something is.
- Generalizing beyond the sample — "all nurses", "everyone", "proves", "generalizable". Purposive samples aren't representative.
- Causal claims — "caused by", "leads to", "results in", "because of". Most qualitative designs explain how and why, not cause and effect.
- Loaded or evaluative wording — "obviously", "clearly", "simply", "failed to", "shockingly". Wording that can carry your judgement rather than participants' meaning.
- Rigor-marker scan — whether the text mentions member checking, triangulation, reflexivity, saturation, a second coder, an audit trail, thick description, or COREQ/SRQR.
What it can't do. Heuristics see wording, not data. They can't open your transcripts, can't tell
whether a theme is grounded, and a clean result is not a clearance. Whether your interpretation holds is a judgement
for you, your co-authors, and your reviewers — not this tool.