Principles
Scientific work is trusted because readers can assume that the data are real, the methods are described honestly, the authors are who they appear to be, and the conclusions follow from the evidence. A service that supports researchers has a particular duty to protect these assumptions, because it works close to the point where they can be weakened. Our work is organized around a short list of principles that are drawn from widely used research integrity guidance, including the Singapore Statement on Research Integrity and the European Code of Conduct for Research Integrity.
- Honesty. We report what was done and found, including results that are inconvenient, and we do not shape analyses to reach a conclusion.
- Responsibility. The researchers who conceive and direct a study are responsible for it. Our support does not move that responsibility.
- Transparency. Methods, deviations from plan and limitations are recorded and reported, and help received is acknowledged.
- Confidentiality. Unpublished work is protected.
- Care. Statements, numbers and citations are checked before they are relied on.
Authorship
Authorship is not a matter of who typed the words. The recommendations of the International Committee of Medical Journal Editors (ICMJE), which many journals outside medicine have also adopted, give four criteria that a person should meet to be an author. They require a substantial contribution to the conception or design of the work, or to the acquisition, analysis or interpretation of data. They require that the person drafted the work or revised it critically for important intellectual content. They require final approval of the version to be published. And they require agreement to be accountable for all aspects of the work, including its accuracy and integrity.
The fourth criterion is the one that matters most here. An author can be questioned by an editor, a reader or an institution about any part of the paper, and must be able to answer. The researchers who direct a study, interpret its findings and approve its conclusions are in that position. Support with statistics, search methods, writing or language does not make the provider the author and does not relieve the authors of that accountability.
In practice this means three things. First, we do not draft or submit work in a way that conceals who is responsible for it. Second, the authors are expected to take part in the substantive decisions, such as the question, the criteria, the interpretation and the conclusions, and we ask for their review and approval at each stage. Third, if someone on our side contributes at a level that meets the authorship criteria, for instance by designing the analysis and taking part in its interpretation, the matter is raised with the research team at the start so that authorship is settled openly and in line with the target journal's policy. Journals set their own rules, and the rules of the target journal prevail.
Acknowledgment and disclosure of help
People who contribute but do not meet the authorship criteria should be acknowledged. The ICMJE recommendations specifically list help with writing, technical editing, language editing and proofreading among the contributions that merit acknowledgment, and they recommend that the source of funding for such help be disclosed where relevant. Many journals now ask authors to state this in an acknowledgments section or a declaration.
We encourage every client to disclose the support they received in the way the journal requires, and we will provide a description of what our contribution was on request, for example "statistical analysis and manuscript editing". Where a journal uses contributor roles, the CRediT taxonomy gives a standard vocabulary for them, covering items such as formal analysis, data curation, methodology, writing of the original draft and writing, review and editing. Describing support accurately is an ordinary part of responsible publication and not an admission of weakness.
The decision on how support is acknowledged belongs to the authors and is governed by the policy of the target journal. We will not agree to an arrangement that hides a contribution which the journal requires to be declared.
Data and analytical integrity
The most serious breaches of research integrity are fabrication and falsification: inventing data or results, and changing data or results so that they misrepresent what was found. We do not do either, and we will not carry out work that requires them. This applies to the fabrication of studies in a systematic review, to the removal of data points or entire studies without a stated, pre-specified reason, and to the alteration of extracted values.
A subtler problem is selective reporting, in which analyses are repeated with different choices until one gives a desired result, and only that one is reported. In evidence synthesis, it takes forms such as changing the eligibility criteria after seeing the studies, switching the effect measure or the model because the first choice was not significant, or reporting only the outcomes that reached significance. The safeguards are the ones used throughout this service. A protocol and an analysis plan are written before results are seen. Deviations from the plan are recorded with their reason. Sensitivity analyses show how much the conclusions depend on analytic choices. The reporting follows PRISMA 2020 and, where relevant, MOOSE, so that readers can see what was done.
When a result is unexpected, or a study turns out to have data that cannot be reconciled, we report it to the research team and discuss how it should be handled and described. We do not make that decision silently.
Citations, quotations and originality
A citation is a claim that a source supports a statement. We check each reference against the source it points to, including the authors, the journal, the year and the details of the publication, and we do not include references that we cannot verify. Fabricated or inaccurate citations are a form of misrepresentation and, since automated writing tools can invent plausible-looking references, they are also a practical risk that this check is meant to remove.
Text is written for the project and is not copied from other sources. Quotations are marked and attributed. Material is not reused from other clients' work, and the same work is not offered to more than one party. Authors are responsible for confirming that the manuscript has not been submitted to more than one journal at once and that earlier publication of any part of it is disclosed.
Conflicts of interest and funding
A conflict of interest is a circumstance that could reasonably be thought to influence the way a study is designed, analyzed or reported, such as a financial relationship with a company whose product is being evaluated. Conflicts are not always avoidable, but they must be declared. Journals require authors to disclose funding and relationships, and the information should be complete.
When a project is funded or sponsored by an organization with an interest in the result, please tell us at the start. The role of the funder in the design, analysis and reporting should be stated in the manuscript, and the analysis plan carries more weight as a safeguard in that situation. We will not accept instructions that make the conclusions conditional on the funder's preference.
Confidentiality
Unpublished manuscripts, data, protocols and ideas that clients share with us are confidential. They are used only for the project they were provided for, access is limited to the people working on it, and they are not disclosed to third parties without the client's permission. When data come from human participants, they should be de-identified before they are shared with us, and the client is responsible for confirming that sharing them is permitted by the consent obtained and by the rules of the ethics committee or institution concerned. How personal information in enquiries is handled is described in the privacy policy.
Journal selection and publication ethics
Choosing a journal is part of responsible publication. We consider whether the journal's scope fits the paper, whether it accepts the type of review, whether it is indexed in the databases that matter to the field, and whether its policies and peer review process are transparent. We try to steer clients away from predatory journals, which charge fees without providing genuine editorial and peer review services. There is no complete definition of the term, which is why the assessment looks at several features rather than a single list.
We do not make promises about acceptance, speed of review or impact, and we do not offer to arrange publication, to influence reviewers or editors, or to supply fake peer reviews or reviewer contact details. Guidance from the Committee on Publication Ethics (COPE) on author and reviewer conduct is the reference for these matters.
Automated writing and analysis tools
Journals and publication ethics bodies have set out expectations about the use of generative artificial intelligence in research and writing. In summary, such tools cannot be authors because they cannot take responsibility for a work, and their use should be disclosed as the journal requires. Fabricated data and fabricated references produced by such tools are the same misconduct as when a person produces them. Our own position on the use of automated tools is set out in a separate page, the AI-use policy.
When we decline work
We decline or stop work when a request requires us to fabricate or alter data, to hide a contribution that should be declared, to write a paper for submission in a way that misrepresents who did the work, to reach a conclusion that was set before the analysis, or to disregard a journal's rules. We also decline projects outside our competence, and we say so early. Clients who are unsure whether a plan is acceptable are welcome to ask before they commit to it, and we will explain our view.
Raising a concern
If you have a concern about the integrity of work carried out with our support, or about our conduct, please write to us and describe it. [OWNER VERIFICATION REQUIRED: concern-reporting email address and how reports are handled.] Concerns about a published article can also be reported to the journal's editor, and COPE provides a framework that journals follow for such cases.
Frequently asked questions
Is it acceptable to get help with writing or analysis for a journal paper?
Yes, when it is done openly. Help with statistics, writing and language editing is common, and the ICMJE recommendations say that such contributions should be acknowledged. The authors remain responsible for the work and its conclusions.
Does the person who helps with the paper become an author?
Only if they meet the authorship criteria, which include a substantial contribution to the study, drafting or critical revision, approval of the final version and accountability for the work. This is discussed with the research team at the start and is governed by the target journal's policy.
Will you write references that I can cite without checking?
No. Every reference is checked against its source, and authors should also check the references in their own manuscript. Unverifiable references are not included.
What if the analysis does not give the result I expected?
The result is reported as it is. The analysis plan, sensitivity analyses and honest reporting are there so that unexpected results are understood and not hidden. Unexpected findings are often informative.
Do you accept work from sponsored studies?
Funding does not prevent us from supporting a project, but it must be disclosed in the manuscript, and the funder cannot dictate conclusions. Please tell us about any sponsor at the start.
Where can I read the guidance this statement is based on?
The main sources are the ICMJE recommendations, the COPE core practices, the Singapore Statement on Research Integrity and the European Code of Conduct for Research Integrity. They are listed under References.
References
- International Committee of Medical Journal Editors. Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals. Section on defining the role of authors and contributors. Available at icmje.org (check the current version).
- Committee on Publication Ethics. Core Practices. Available at publicationethics.org.
- Committee on Publication Ethics. Authorship and AI tools: COPE position statement. Available at publicationethics.org.
- Brand A, Allen L, Altman M, Hlava M, Scott J. Beyond authorship: attribution, contribution, collaboration, and credit. Learn Publ. 2015;28(2):151-155.
- Second World Conference on Research Integrity. Singapore Statement on Research Integrity. 2010.
- ALLEA. The European Code of Conduct for Research Integrity. Revised edition. Berlin: ALLEA; 2023.
- Grudniewicz A, Moher D, Cobey KD, et al. Predatory journals: no definition, no defence. Nature. 2019;576:210-212.
- Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. doi:10.1136/bmj.n71