A team evaluates a classroom learning app and reports a clear gain in test scores. Two years later it emerges that the lead investigator held shares in the company behind the app. The data may be perfectly honest, yet the paper has lost its value to readers. Not because a conflict existed, but because it was never declared.
That is the core point. Almost every researcher has a conflict of interest at some stage. Full disclosure lets readers weigh the work for themselves; concealment turns an ordinary situation into an integrity question.
What counts as a conflict in practice
A conflict of interest exists when a relationship or interest could reasonably be seen as affecting your judgement. The word “could” matters: you do not need to have actually been influenced for disclosure to be required. A simple test: if readers learned about this after publication, would they feel they should have been told?
Financial and non-financial
| Category | Examples | Often forgotten? |
|---|---|---|
| Direct financial | Salary, consulting fees, equity, royalties on related patents | Rarely |
| Research funding | Grants from a company that benefits from the findings | Rarely |
| In-kind support | Donated equipment, free software licences, conference travel | Often |
| Personal relationships | A partner or relative employed by an interested organisation | Often |
| Intellectual | Evaluating a method you developed; direct rivalry with another group | Very often |
| Institutional roles | Sitting on the editorial board of the journal you submit to | Often |
Non-financial conflicts are harder to see because they leave no invoice. A lecturer studying the effectiveness of a programme her own department launched, or a researcher reviewing work by a long-standing rival, should both think about disclosure.
How far back should you go?
Many journals use a window of around three years before submission, and the ICMJE disclosure form also asks about a fixed period. Follow the form your target journal uses; if it says nothing, declare whatever is still relevant to the topic. Over-disclosure almost never hurts. Under-disclosure does.
Model disclosure wording
A good statement says who, what relationship, with which organisation, and how it relates. Avoid vague lines such as “some authors have links to industry”.
- None: “The authors declare no competing interests relevant to this work.”
- Funding: “This study was funded by [organisation]. The funder had no role in study design, data collection and analysis, the decision to publish, or preparation of the manuscript.”
- Consulting: “J.K. received consulting fees from [company] in 2024–2025 for work unrelated to the subject of this article.”
- Intellectual: “A.N. developed the instrument evaluated in this study.”
- Editorial: “L.M. is a member of the journal's editorial board and was not involved in the peer review of this manuscript.”
Only write “the funder had no role” if that is true. If the funder commented on a draft, say which parts.
Reviewers and editors have conflicts too
When you receive a review invitation, decline or tell the editor if:
- You recognise an author as a close colleague, former supervisor, former student or recent co-author.
- You are writing a directly competing paper on the same question.
- You have a financial interest in the product being evaluated.
- You have already reviewed the same manuscript for another journal. This does not always require declining, but the editor should know.
Anonymous review is not invisible review. In a small field authors often guess who reviewed them, and a biased report can follow you for years.
Managing a conflict, not just declaring it
For serious conflicts, disclosure alone may not be enough. Common safeguards include:
- Having a team member without the conflict run the primary analysis, or an independent re-analysis.
- Preregistering the study so the analysis plan demonstrably predates the data.
- Sharing data and code so others can verify the results.
- Stepping back from decision-making roles, such as final sign-off on conclusions.
If you realise you under-declared
Contact the journal as soon as possible and state the missing information. Before publication, the editor will simply update the statement. After publication, a correction is the usual outcome. COPE guidance treats early self-reporting very differently from discovery by a third party. Staying silent only raises the odds of something worse, including retraction.
A pre-submission checklist
- In recent years, have I received money, goods or funding from anyone who benefits from these findings?
- Do I or a close family member hold equity, patents or a role in a related organisation?
- Am I evaluating a tool, programme or instrument I created?
- Do I have any role at the journal receiving the paper?
- Has every co-author answered these four questions?
Question five is the one most often skipped, because the corresponding author signs for everyone. Your next step: send these five questions to each co-author and file their answers with the manuscript records.
Câu hỏi thường gặp
Do I need a statement if I have no conflicts?
Yes. Most journals require an explicit statement that there are no competing interests. Leaving the section blank can delay processing.
Is a free software licence a conflict of interest?
It can be, if the software is related to the research or is what you are evaluating. In-kind benefits are financial interests even without cash changing hands.
What is a non-financial conflict of interest?
It is a personal, intellectual or institutional interest that could be seen as influencing judgement, such as evaluating your own programme or reviewing a rival's work. Declare it clearly and consider independent analysis.
Should I review a paper written by someone I know?
Tell the editor and let them decide. For close colleagues, recent co-authors or former students, the norm is to decline.
What happens if I forgot to declare a conflict of interest?
Contact the journal promptly to correct the record. Self-reporting usually leads to a correction, while later discovery by others can trigger an investigation.
