How to Respond to Statistical Reviewer Comments
Statistical reviewer comments should be handled by identifying the exact analytical concern, checking whether the criticism is valid, changing the analysis when needed, and documenting both the method and the result transparently.
For each statistical comment, separate the reviewer's underlying concern from the requested technique. Recheck the data structure, analysis assumptions, model specification, missing-data handling, multiplicity, effect estimates, uncertainty, and interpretation. If a new analysis is warranted, run it and update every dependent table, figure, result, and conclusion. If the original analysis remains appropriate, explain the rationale and add enough detail to the Methods or Results section to make that decision verifiable.
Test Selection and Assumptions
A reviewer may question whether the selected test matches the outcome type, design, repeated measures, clustering, distribution, or independence assumptions. Recheck the design before changing methods mechanically.
Sample Size, Power, and Precision
Comments about sample size can concern prospective power, achieved precision, confidence intervals, or overfitting. Address the actual concern rather than relying on a post-hoc power calculation by default.
Model Specification and Robustness
For regression or multivariable models, reviewers may ask about covariate selection, nonlinearity, interactions, collinearity, influential observations, clustering, or alternative specifications. Add diagnostics or sensitivity analyses when they are scientifically justified.
Multiple Testing and Selective Reporting
When many outcomes or subgroup analyses are reported, reviewers may ask how multiplicity was handled and whether analyses were prespecified. Clarify the analysis plan and avoid presenting exploratory findings as confirmatory.
Report Effect Sizes and Uncertainty
A statistically significant p-value is rarely enough. Where appropriate, report effect estimates, confidence intervals, exact p-values, model details, and clinically or scientifically meaningful interpretation.
Keep the Response and Results Synchronized
Any re-analysis can change tables, figures, abstract numbers, discussion, and conclusions. Update all dependent text and identify the changes clearly in the response letter.
Frequently asked questions
No. First determine whether the requested analysis is appropriate for the study design and question. If not, explain why and offer a more suitable analysis or clarification.
Report the new result transparently and revise the interpretation, abstract, discussion, and conclusion wherever the change matters.
Yes. Statistical and methodological revision is part of the manuscript revision and reviewer-response support service.
References
- PMC — A Practical Guide on Responding to Reviews
Peer-reviewed practical guidance on preparing revisions and point-by-point responses.
- ICMJE Recommendations
International recommendations on responsible manuscript preparation, reporting, and editorial practice.