Know where a manuscript needs closer review.

RigorMD Editorial Evaluation examines study design, analysis, and conclusions against the criteria your journal supplies. It considers concerns such as confounding, missing data, multiple comparisons, and unsupported interpretation. Separate numerical checks recalculate statistics when the reported inputs allow. Your editors receive evidence, specific author questions, and areas for reviewer or statistical-editor attention.

The screen is automated. Two AI appraisers read independently, and the evaluation records where they agree or differ.

Focus your reviewers — identify methodological questions and areas for closer review.
Review manuscripts consistently — provide a consistent methodological review before reviewer assignment.
Support clinical and statistical review — give clinical reviewers and statisticians a shared account of concerns, with statistical reasoning where applicable.
Show the editor what a manuscript needs before review is assigned — whether it can use the review it is about to get, or needs work first. The report informs that call; it never makes it.

Structured editorial decision support. Not an editorial decision. Not a replacement for peer review. This public page does not collect manuscripts.

Reviewer attentionPortfolio consistencyConfidential pilot first

Request an editorial pilot See a real evaluation A paper its journal corrected Read a historical example How the screen works Reviewer and reader modes — early access

§01 Where editorial review helps

Editorial offices use the report to decide where human attention adds the most value, before reviewer and statistical-editor time is spent.

Examine methods and reported numbers

The evaluation examines concerns such as imprecise comparisons, unaddressed confounding, and p-values that do not recompute from the reported inputs.

Focus reviewer attention

Focus reviewer attention on the methods and conclusions that need closer examination.

Sharpen the questions you send

Every concern is anchored to something your editors can check — the sentence in the manuscript where a reading of the text is at issue, the recomputed number where the arithmetic is — so the queries you send are specific, not a vague “please have a statistician look at this.”

Identify questions before reviewer assignment

The report identifies methodological questions before reviewer assignment, so your office can triage with evidence in hand. The accept, reject, or revise call stays with the journal.

Check the reported evidence

The evaluation links concerns to manuscript passages and applicable recalculations, giving editors evidence to examine and questions to send to authors.

§02 How an editorial pilot works

STEP 01

Journal context

Your editorial office sends the manuscript and any supplementary files, plus the instructions or criteria you want it read against.

STEP 02

Rigor screen

RigorMD flags methods, statistics, and reporting risks — and notes where the conclusions outrun the data.

STEP 03

Reviewer focus

The report frames author queries and reviewer focus points. Decisions stay with your editors.

§03 Pilot boundaries

Four things are settled before any pilot report is produced or delivered.

Only manuscripts your journal has already decided, de-identified, or otherwise authorized

Confidentiality terms agreed before we read anything

No model training on your manuscripts

Reports are for internal editorial-office use only

What this is not. Editorial Review does not make the journal’s decision, judge novelty, replace peer review, or stand in for an independent statistical editor. Pilot manuscripts are never used for model training, and manuscripts and their reports are kept only for the retention window agreed in the confidentiality terms.

Request an editorial pilot Security and confidentiality