Free checks for reported statistics and references
Four small tools that help you notice things worth a second look before you submit or review a manuscript. Each flag is a prompt for human review, not a verdict.
Privacy: the tools run entirely in your browser and work offline when you open the files. Nothing you type is uploaded. The one exception is the DOI checker, which asks the public Crossref API about the DOIs you paste, from your browser, and says so on its page.
- CI and p consistencyDoes the confidence interval agree with the p value? Operator-aware, with an approximate implied p.
- GRIM testCan a reported mean arise from whole-number data with this N? Shows the nearest attainable means.
- APA p-value formatFinds p values in text and fixes .000, leading zeros, spacing and scientific notation.
- DOI and reference checkDoes each DOI exist in Crossref, and is it registered under the title you cite?
Methodology in brief
- CI-p: alpha = 1 − CI level. The p operator decides what can be concluded; a contradiction is flagged only when both the CI and p give definite, opposite answers. The implied p uses the Altman and Bland normal approximation and is labelled approximate.
- GRIM: exact integer arithmetic over all attainable response totals (Brown and Heathers' GRIM idea).
- p format: APA 7 conventions for reporting p values.
- DOI check: Crossref registered title versus the title you cite (normalised token overlap); retraction signals are shown conservatively.
Disclaimer
These are screening aids. A flag can have an innocent explanation (rounding, missing data, a different model or test) and the absence of a flag does not show that a result is correct. Decisions about a manuscript belong to people who can see the data.