Terms of engagement

What you are buying

• Statistical Manuscript Review: a flat-fee statistical review of the manuscript, or methods and results, you send. The deliverable is a written report.
• Manuscript Results Check (pilot): up to 80 agreed numeric claims in your Word manuscript, traced to the R or Python output you identify. The deliverable is a claim-by-claim record with findings and stated limitations.
• Reproducibility Check (pilot): a rerun of one agreed output from your R project, in one environment. The deliverable is a packet with the checklist, captured session, output comparison, and prioritized fixes.
• Reanalysis Execution Sprint (pilot): for one reviewer letter, up to three agreed analyses run on your existing data. The deliverable is reproducible R code, revised tables beside your locked original results, a change log, and a Word crosswalk from each reviewer comment to its result, with each model independently recomputed.

Nedim Yel, PhD, may use locally run AI-assisted tools for inspection and first drafts. Your materials are not sent to online AI services, and all deliverables are reviewed and finalized by Nedim.

Quote and approval

You receive a flat-price quote within 72 hours of your request. The price is set by the complexity of the analysis and the size of the material. No work begins until you approve the quote. The quote states what is in scope; anything outside it is quoted separately.

For a Reanalysis Execution Sprint, the quote comes with a written analysis plan: each reviewer request in the reviewer's words, the exact quantity to be estimated, the model, and how missing data are handled. No analysis is run until you approve the plan. A change to an approved plan, such as a different model, covariate set, or missing-data method, is a change of scope: it is agreed in writing before it is run and may be quoted separately.

Payment

Fifty percent of the quoted fee is due when you approve the quote, and the remaining fifty percent is due when the work is delivered. Work starts once the first payment is received.

Turnaround

Standard turnaround for a Statistical Manuscript Review is 5 business days, counted from the day we have your complete materials and the first payment. For the pilot services, turnaround is confirmed in your quote. For a Reanalysis Execution Sprint the clock starts when we have the reviewer letter, the analysis dataset, your approved plan, and the first payment. Rush turnaround, where offered, is agreed in the quote and may carry a rush fee. If we need clarification or a missing file, the clock pauses until we have it.

Refunds

If you cancel before work starts, you receive a full refund. Once the work is delivered, fees are not refundable, because the work is a judgment-based service that cannot be returned. If we are unable to complete work we have started, we will refund the portion you have paid for work not delivered.

If, after we receive your data, a requested analysis cannot be run as agreed (for example, the data cannot support the model, or we cannot reproduce your original results), we will tell you before doing further work on it and refund the portion you paid for that part.

Follow-up

Your fee includes one clarification call or one email thread about what we delivered. Re-reviewing a revised manuscript, a new reviewer letter, analyses beyond those in an approved plan, or a rerun after you change your code, data, or manuscript is a new engagement and is quoted separately.

What our services are, and are not

• A review is an independent check of the statistical analysis and reporting in the material you provide, against accepted practice and reporting standards.
• None of our services guarantees that a journal, reviewer, or committee will accept your work, or that they will agree with what we report.
• We rely on the accuracy and completeness of what you send. We do not audit your raw data, ethics approvals, or data collection unless that is in the agreed scope.
• Our services are not legal, regulatory, or medical advice.
• A Manuscript Results Check compares the numbers in your manuscript with the output you identify. It does not judge whether the analysis is appropriate. Where we rerun code, results we reproduced are identified separately from output you supplied.
• A Reproducibility Check reruns the agreed output in one environment on one date. A pass means that output matched then, in that environment. It is not a certification and does not cover other outputs, operating systems, or later software versions.
• A Reanalysis Execution Sprint reports what the data show. We do not choose models, exclusions, or specifications to reach a particular result, and our fee does not depend on how any result turns out.
• In a sprint, reviewer-requested analyses are labelled exploratory. Your original results are run first, locked, and reported unchanged beside the revised ones. Each model is recomputed in separate software; agreement means two programs produced the same numbers for the agreed model, not that it is the right model for your question. We will tell you if a requested analysis conflicts with your original design.
• A sprint does not include writing your response letter. You remain responsible for how you report results and answer reviewers. The pilot covers linear models; other model types are scoped separately or declined. Multiple imputation cannot yet be cross-checked to the same standard, and the quote will say so.

Pilot services

The Manuscript Results Check, Reproducibility Check, and Reanalysis Execution Sprint are pilots. Scope, format, and pricing may change, spots are limited, and we confirm the terms for your project in the quote before any work starts.

Data you share

• You confirm that you have the right to share what you send. You are responsible for the IRB approvals, data use agreements, and consent that govern your data.
• Do not send names, contact details, or other direct identifiers, and do not send data covered by an agreement that forbids sharing. For a Manuscript Results Check or Reproducibility Check, output tables, code, and synthetic or public data are enough.
• A Reanalysis Execution Sprint needs your analysis dataset. Send a de-identified file only, only if your IRB approval or data use agreement allows you to share it with an outside analyst, and only after you approve the quote. Share it through your own shared link; do not put data in the quote form or in an email.
• If a file turns out to contain identifiers or restricted data, we may pause work and ask you to send a compliant file.
• Do not include passwords, API keys, or tokens in files or code.

Your work stays yours

You keep all rights to your manuscript, code, and data. The code, tables, and documents we produce for you are yours to use, adapt, and publish with your work. We claim no authorship; whether to acknowledge us is your choice. We will not publish or describe your project, or name you, without your written permission.

Confidentiality

Materials you send are used only to complete your engagement and are deleted within 30 days of delivery, as described in the privacy policy and on the confidentiality page. A confidentiality agreement or NDA is available on request, before you send any file.

Limits of responsibility

We provide the service with professional care, but we cannot promise that a review or check will find every issue or discrepancy. To the extent the law allows, our total responsibility for any claim connected to an engagement is limited to the fees you paid for that engagement, and we are not responsible for indirect or consequential losses, such as lost publication or funding opportunities.

Governing law

These terms and any dispute about an engagement are governed by the laws of the Commonwealth of Massachusetts, and any court action will be brought in the state or federal courts located in Massachusetts. Nothing in these terms limits rights you have under law that cannot be waived.

Changes and questions

We may update these terms; the version on this page is the current one, and the terms in force when you approve a quote apply to that engagement. For questions, use the quote form and note that it is a question about these terms.