|
Getting your Trinity Audio player ready...
|
Key Takeaways:
- Screening now goes well beyond plagiarism. Journals expect declarations on authorship, conflicts of interest, funding, data availability, ethical approval, and AI use, all at the point of submission.
- Small, frequent lapses do the most damage. Most harm to the scholarly record comes from common minor misbehaviors, not rare fraud, so routine disclosure habits protect your work.
- AI cannot be an author. Any substantive use of generative tools in writing or analysis should be disclosed, and fabricated text or images are a fast-growing threat.
- A clean package protects you twice. Complete, honest declarations guard your paper against desk rejection today and retraction later, and they protect public trust in science.
Contents
- Glossary of Key Terms
- Why do ethical declarations matter at submission?
- Where can you learn about publishing ethics?
- Which declarations must accompany your manuscript?
- How should you handle plagiarism and text recycling?
- What must you disclose about AI and generative tools?
- Image integrity and data manipulation
- What happens when ethical declarations fail?
- Honest error versus misconduct
- How can you build a culture of integrity?
- Should you worry about predatory journals?
- What are your ethical duties during peer review?
- Your submission-ready ethics checklist
- Frequently asked questions
Glossary of Key Terms
Use this quick reference while you assemble your submission. Each term appears again, in context, later in the article.
| Term | What it means |
| Conflict of interest (COI) | Any financial, personal, or professional interest that could bias the work or its assessment. |
| Competing interest | A broader label many journals use for the same idea, covering non-financial interests too. |
| ICMJE criteria | The 4 conditions the International Committee of Medical Journal Editors sets for authorship. |
| CRediT taxonomy | A 14-role system for stating exactly what each contributor did on a paper. |
| Ghost, guest, and gift authorship | Omitting a real contributor, or naming someone who did not qualify, both breaches of authorship ethics. |
| Institutional Review Board (IRB) | An ethics committee that approves research involving human participants before it begins. |
| Informed consent | A participant’s documented agreement to take part after understanding the risks and purpose. |
| Consent to publish | A separate agreement allowing identifiable details or images to appear in print. |
| Data availability statement | A declaration of whether, where, and how the underlying data can be accessed. |
| FAIR principles | A standard that data should be Findable, Accessible, Interoperable, and Reusable. |
| Preprint | A manuscript posted publicly before peer review, which most journals now permit. |
| Text recycling (self-plagiarism) | Reusing your own previously published text without disclosure or citation. |
| Salami slicing | Splitting 1 study into several thin papers to inflate output. |
| Paper mill | A commercial operation that fabricates and sells manuscripts or authorship slots. |
| Similarity check | Software screening, such as iThenticate, that flags overlapping text. |
| Retraction | Formal withdrawal of a published paper from the scholarly record. |
| Corresponding author | The author responsible for communication and for collecting all declarations. |
Why do ethical declarations matter at submission?
They matter because journals treat integrity as a gatekeeping test: even novel, high-impact research can be rejected or retracted when required disclosures are missing, incomplete, or false. Trust, integrity, and honesty are the pillars that hold up the scholarly record, and editors screen for them early.
The computer scientist Bill Joy once warned that scientists cannot simply do their science and ignore the ethical questions around it. That warning has aged well. As research volume grows and tools evolve, the number of ways a submission can go wrong has grown with it, so a disciplined approach to declarations is now part of basic authorship craft.
Editors see these issues constantly. Reflecting on the problems that most concern journal editors, Prof. Ana Marušić points to the cluster of “article similarity, authorship, conflict of interest” as the declarations where transparency matters most, because editors must build policies and processes to assure the quality of what authors disclose.
There is also a practical payoff. A submission that arrives with every declaration in order signals care and competence, and it speeds the whole process. Missing statements, by contrast, generate back-and-forth queries that delay review and can sour an editor’s first impression. Treating declarations as part of the science, not as paperwork bolted on at the end, is the mindset that keeps good work moving toward publication.
Where can you learn about publishing ethics?
You can learn from a small set of authoritative, mostly free resources maintained by editors, publishers, and integrity bodies. Start with the ones below, then check your target journal’s own instructions, which always take priority.
| Resource | What it offers | Best used for |
| COPE (Committee on Publication Ethics) | Flowcharts, cases, and guidelines for handling misconduct. | Deciding what to do when something goes wrong. |
| ICMJE Recommendations | The standard for authorship, disclosures, and trial registration. | Getting authorship and COI right. |
| WAME | Guidance and policy from medical journal editors worldwide. | Editorial norms in medicine and health. |
| CSE white paper | A broad overview of ethical roles across the publishing chain. | A single reference document for authors and editors. |
| EQUATOR Network | Reporting guidelines such as CONSORT and PRISMA. | Reporting your study type completely. |
| OASPA and DOAJ | Principles for open access and for spotting predatory journals. | Vetting where you submit. |
| Retraction Watch | A database and blog tracking retractions and their causes. | Learning from real cases. |
| STM Integrity Hub | A cross-publisher effort against paper mills and image fraud. | Understanding modern detection at submission. |
Which declarations must accompany your manuscript?
At minimum, most journals require declarations covering conflicts of interest, funding, authorship, originality, ethical approval, and data availability. The sections below explain each one, plus the checks journals now run against them.
Conflict of interest and financial disclosure
A conflict of interest arises whenever a financial, personal, or professional relationship could bias your work, or could influence an editor’s or reviewer’s judgment. The safest approach is full disclosure, every time, even when you believe the interest is trivial.
- Declare funding, sponsorship, gifts, honoraria, patents, and paid advisory or consulting roles.
- Confirm that no sponsor agreement lets a funder suppress or bias results.
- State clearly when there is nothing to declare; a positive “none” is still a declaration.
- Use the ICMJE disclosure form, or your journal’s specific form, and keep author statements consistent.
- Remember that an undeclared competing interest can trigger rejection now, or retraction if found later.
Authorship and contributorship
Authorship disputes are among the most common integrity problems, and among the most avoidable. The ICMJE sets 4 criteria: each author should contribute to the work’s design or data, help draft or revise it, approve the final version, and agree to be accountable. This last point is why authorship cannot be bought or sold. As Prof. Zabta Khan Shinwari puts it, in some circles “authorship is for sale”, a practice that corrodes trust and, increasingly, ends in retraction.
- Ghost authorship: a person who did the work, often a writer or junior researcher, is left off the byline.
- Guest or gift authorship: a senior or well-known name is added without a real contribution.
- Contributorship statements: many journals now require the CRediT taxonomy, which names who did what across 14 defined roles.
- Corresponding-author duty: 1 author must collect every declaration and confirm each co-author approves the submission.
Original work, preprints, and duplicate submission
Most journals ask you to confirm that the manuscript, in whole or in part, is not published elsewhere and is not under review at another journal. Honesty here prevents the twin problems of duplicate publication and simultaneous submission.
- Disclose prior appearances of the content, including conference proceedings, theses, and journal communications.
- Treat preprints as usually acceptable, but declare them; most journals allow posting before review.
- Avoid salami slicing: do not split 1 coherent study into several thin papers.
- Obtain written permission to reuse any figure, table, or long passage from earlier published work.
Ethical approvals and informed consent
Research involving humans or animals needs approval from a properly constituted ethics committee or Institutional Review Board, obtained before the project begins. You must also secure and document informed consent from participants, and from parents, guardians, or legal representatives when minors are involved.
- Name the approving body and the approval or protocol number in your methods.
- Distinguish consent to participate from consent to publish; identifiable images or case details need the second.
- For animal studies, state the welfare guidelines and oversight body that applied.
- Keep the signed documents; editors may ask to see them during review.
Data availability and reporting standards
A data availability statement, now mandatory at many journals, tells readers whether, where, and how your data can be accessed. Aligning with the FAIR principles, that data be Findable, Accessible, Interoperable, and Reusable, strengthens both transparency and reuse.
- State plainly if data are in a repository, available on request, or restricted, and why.
- Follow the relevant EQUATOR reporting guideline for your study design.
- Deposit code and materials where you can, and cite them with persistent identifiers.
Clinical trial registration
For clinical trials, prospective registration in a recognized public registry is a precondition of publication at most journals. This norm exists in large part because editors pushed for it. Prof. Marušić, whose work helped mandate trial registration, treats it as central to the transparency of the whole publication process.
How should you handle plagiarism and text recycling?
Cite every source, quote sparingly, and disclose any reuse of your own earlier text; journals run similarity software such as iThenticate on every submission, so overlaps surface quickly. Plagiarism is not only copying others; reusing your own words without citation is text recycling, a form of self-plagiarism.
- Paraphrase and attribute rather than copy, even for a sentence or 2.
- Cite your own prior work when you build on it; do not present it as new.
- Run a similarity check before you submit so nothing surprises you or the editor.
- When in doubt about permissible overlap, ask the editor rather than guess.
Journals do distinguish accidental overlap from deliberate copying, but both slow your submission, so it pays to resolve overlaps before an editor sees them. Standard method sentences and common phrases rarely cause trouble; large verbatim blocks, reused introductions, and undisclosed prior text do. A high similarity score is not proof of misconduct, yet it invites scrutiny you can easily avoid.
What must you disclose about AI and generative tools?
Disclose any substantive use of generative AI in writing or analysis, and never list an AI tool as an author, because it cannot take responsibility for the work. Journal policies vary, so read your target journal’s AI clause before you submit. The risk is not hypothetical: the STM Integrity Hub warns that tools like ChatGPT “will make the creation of fake papers much easier”, and that AI can also fabricate images and data.
- Name the tool and describe how you used it, usually in the methods or acknowledgements.
- Keep AI to language polishing or ideation unless your journal permits more, and verify every output.
- Do not present AI-generated text, citations, or images as original human work.
- Take special care with AI-generated figures, which can introduce fabricated detail that screening will flag.
Policies are evolving quickly, so a statement that satisfies 1 journal may fall short at another. When in doubt, disclose more rather than less, and keep a short record of which tools you used and how. That record protects you if questions arise during review or after publication, and it costs almost nothing to maintain.
Image integrity and data manipulation
Inappropriate splicing, beautifying, or duplicating images is a leading cause of retraction, and detection is getting sharper. Publishers now pool expertise through initiatives like the STM Integrity Hub, which reports rising submissions carrying “manipulated images, plagiarism, or fabricated data” and builds tools to catch them at submission.
- Keep unedited originals of every blot, gel, and micrograph, with a clear processing log.
- Apply adjustments to the whole image, never to a selected region, and disclose any processing.
- Never duplicate, mirror, or reuse a panel to represent a different result.
One reassurance for honest authors: in these systems, van Rossum stresses that “humans will always make the final editorial decision”; the software only flags cases for an editor to investigate.
What happens when ethical declarations fail?
The consequences escalate quickly: desk rejection at submission, retraction after publication, lasting reputational damage, and eroded public trust. Yet the largest cumulative harm is quieter. Prof. Lex Bouter observes that “minor research misbehaviors” carry an “embarrassingly high prevalence”, so their aggregate impact outweighs the rare, headline fraud case.
- At submission: missing or inconsistent declarations trigger queries or immediate rejection.
- After publication: undisclosed conflicts, image problems, or duplication can lead to correction or retraction.
- For your career: retractions are permanent, public, and cited against the work for years.
- For science: each preventable lapse chips away at the trust the whole system runs on.
Honest error versus misconduct
Not every mistake is misconduct. An honest error, disclosed and corrected, is part of normal science; misconduct involves fabrication, falsification, or plagiarism, done knowingly. The dividing line is intent and transparency, so the moment you spot a genuine mistake, tell the editor and offer a correction.
Culture decides how often errors surface. Prof. Bouter argues for “creating a climate in which researchers feel free to discuss” their errors and dilemmas, rather than hiding them. He also notes that poor mentoring and supervision are consistently reported as important drivers of research misbehavior.
How can you build a culture of integrity?
You build it through training, honest supervision, and incentives that reward good conduct, not just output; declarations then become habits rather than hurdles. Integrity is a shared responsibility across researchers, mentors, institutions, funders, and journals.
- Invest in education. Prof. Shinwari calls for training and even an oath of integrity, honesty, responsible conduct, and service to society for every graduate.
- Mentor deliberately. Prof. Bouter highlights supervision as a key lever, since weak mentoring drives many lapses.
- Fix the incentives. The STM Integrity Hub traces much misconduct to “excessive pressure to publish” and overreliance on single metrics.
- Keep learning. Prof. Marušić advises authors to “never stop learning” and to build methodological rigor, which prevents many ethical problems at the source.
Should you worry about predatory journals?
Yes, because submitting to a predatory journal can waste your work, your fee, and your reputation; these outlets take payment but skip genuine peer review and ethical oversight. Vet every target journal before you submit, using recognized whitelists and a few plain checks.
- Confirm indexing or membership in trusted bodies such as DOAJ, COPE, or OASPA.
- Check that the editorial board is real, named, and reachable.
- Be wary of aggressive solicitation, guaranteed acceptance, or unclear fees.
- Match the journal’s stated scope and metrics against independent sources.
- Use the “think, check, submit” habit: pause and verify before you commit.
What are your ethical duties during peer review?
If you review for a journal, you take on real obligations: keep the manuscript confidential, judge it fairly on its merits, declare any conflict, and never use privileged information for your own gain. Ethics is not only an author’s concern; it runs through the whole review process.
- Decline or disclose when you have a competing interest with the authors or the work.
- Treat the manuscript as confidential; do not share, cite, or reuse its unpublished data.
- Give constructive, evidence-based feedback rather than vague or hostile comments.
- Flag suspected misconduct to the editor instead of acting on it yourself.
Your submission-ready ethics checklist
Run through this list before you hit submit. Confirm each item, attach the supporting documents, and make sure every co-author has seen the final version.
- Conflict of interest and funding declared for every author, or a clear “none.”
- Authorship meets the ICMJE criteria; contributions stated via CRediT; no ghost, guest, or gift authors.
- Originality confirmed; preprints, prior versions, and permissions disclosed.
- Ethics approval and informed consent obtained, named, and on file.
- Consent to publish secured for any identifiable person or image.
- Data availability statement included; reporting guideline followed.
- Clinical trials registered prospectively in a recognized public registry.
- Similarity check run; any text reuse cited or disclosed.
- AI use disclosed, with tools named and outputs verified.
- Images unaltered beyond disclosed, whole-image adjustments; originals retained.
- Corresponding author has collected all declarations and every co-author has approved.
Frequently asked questions
What ethical declarations are required for journal submission?
Most journals require declarations covering conflicts of interest, funding, authorship and contributions, originality, ethics approval and informed consent, data availability, and, where relevant, clinical trial registration and AI use. Always check your target journal’s instructions, which may add more.
Do I need to declare AI or ChatGPT use in my manuscript?
Yes, if the use was substantive. Disclose generative AI used in writing or analysis, usually in the methods or acknowledgements, and name the tool. You cannot list AI as an author, because it cannot take responsibility, and you must verify every AI output yourself.
What is the difference between a conflict of interest and a competing interest?
They describe the same underlying issue. “Conflict of interest” is the traditional term; “competing interest” is a broader label many journals prefer because it more clearly includes non-financial interests, such as personal or professional relationships. Follow whichever term your journal uses.
Who qualifies as an author on a research paper?
Under the ICMJE criteria, an author must contribute substantially to the work, help draft or revise it, approve the final version, and agree to be accountable for it. Someone who only secured funding or supervised generally does not qualify, and should be acknowledged instead.
Is a data availability statement mandatory?
At a growing number of journals, yes. Even where it is optional, including one is good practice. It states whether data are in a repository, available on request, or restricted, and why. Aligning with the FAIR principles makes your data easier to find and reuse.
Can I submit my manuscript to two journals at the same time?
No. Simultaneous submission to more than 1 journal is an ethics breach and grounds for rejection. Submit to 1 journal at a time, wait for a decision, and only then move on if needed. Posting a preprint is different and is usually permitted.
What is a paper mill and how do journals detect one?
A paper mill is a commercial operation that fabricates manuscripts or sells authorship slots. Journals detect them through cross-publisher tools that flag suspicious language, references, image reuse, and duplicate submissions, after which an editor investigates. Human editors, not software, make the final call.
What happens if I fail to declare a conflict of interest?
An undeclared conflict can lead to rejection at submission, or to a correction or retraction if discovered after publication. It can also damage your credibility with editors and readers. Because disclosure is simple and low-cost, full transparency is always the safer choice.
This article was originally published on December 5, 2023, and updated on July 21, 2026.
Comments (1)
注册获取100 USDTsays:
June 5, 2026 at 5:13 amYour article helped me a lot, is there any more related content? Thanks!