Review the methods behind your conclusions.
RigorMD examines whether your study design and analysis support the conclusions you draw. It recalculates statistics where the reported numbers allow and identifies concerns with evidence from your manuscript. Use the report to plan revisions before submission or reassess a revised draft. The review is automated, using two independent AI engines and statistical checks.
| 01 | Draft ready | Pre-submission report |
| 02 | Major revision | Revise against the findings |
| 03 | Revised draft | Finding-by-finding re-review |
§01 — Author paths
Review the draft before you submit.
Examine whether the design supports your claims, the analysis fits the study, and the reported numbers agree. See the concerns and suggested revisions in a sample report.
Plan your revision around the findings.
Each finding explains the concern and points to a supporting manuscript passage, calculation, or reference record. Use that evidence as you revise the manuscript and prepare your response to reviewers.
Compare the revised draft with your first report.
Re-review shows which findings are no longer flagged, which remain, and which are new. The comparison is a reading aid; it does not establish that an issue was fixed.
Plan your revisions with the evidence in hand.
The report explains each concern, its severity, and the suggested change. A methodological finding may call for a different analysis, a narrower conclusion, or a clearer account of the study's limitations. A recomputation finding shows the discrepancy and the recalculated value. The review also considers how your contribution relates to the prior work retrieved for the report.
Methods and interpretation
Does an observational study support the causal language in its conclusion? Does the analysis account for patients grouped within hospitals or for repeated measurements? Are missing data and multiple comparisons handled in a way that fits the claim?
Statistical recomputation
Applicable checks use the counts, estimates, confidence intervals, and test statistics your paper reports. When a value does not reconcile, the report shows the calculation and the discrepancy. The review uses your manuscript and supporting files; it does not rerun the analysis on your underlying dataset.
See a worked calculation →