Research Authorship: Discussing Credit, Contributions, and Order Before Submission
This guide is for anyone involved in planning, leading, or contributing to a research study: principal investigators, postdocs, graduate students, and research staff. The direct answer is straightforward. Authorship conversations should start before data collection begins. You must define each person’s role using a standard taxonomy, document the agreement, and review the author order at predefined project milestones. A final check before submission prevents last minute disputes. The Committee on Publication Ethics provides core principles that should underpin every discussion: authorship requires a substantial intellectual contribution, accountability for the work, and willingness to defend the findings. The NIH Data Management and Sharing Policy indirectly reinforces the need for early clarity because data management plans often involve multiple contributors whose roles must be distinguished for sharing and attribution.
At a Glance
| Key Action | When | Why | Tool / Resource |
|---|---|---|---|
| Define contributor roles | Before data collection | Prevents ambiguity about expected work | Contributor Roles Taxonomy (CRediT) |
| Assign authorship order | At project launch | Sets expectations for credit and career advancement | Shared project document |
| Document agreement | After initial discussion | Creates a written record that can be revisited | Signed or emailed confirmation |
| Revisit at milestones | At data analysis, manuscript drafting | Accounts for changing contributions | Regular lab meeting agenda |
| Finalize before submission | One week before journal submission | Resolves any remaining disagreements | Checklist with co-author approval |
Decision Criteria for Authorship
Authorship is not a reward for being the lab head, a funding provider, or a technician who only followed instructions. The core standard is a substantial intellectual contribution. The ORCID registry allows researchers to link specific contribution types to their persistent identifier, which supports transparency. Use the Contributor Roles Taxonomy (CRediT) to define roles such as conceptualization, methodology, investigation, formal analysis, writing the original draft, and supervision. Each person who performs at least two of these roles and participates in revision or approval typically qualifies for authorship.
A study on the U shaped association between dietary flavonoid intake and gallstone prevalence U shaped association between dietary flavonoid and gallstone prevalence with sex-specific effects illustrates why clear criteria matter. The authors needed to distinguish between those who designed the cross sectional analysis and those who handled data entry. Without a prediscussed framework, a data manager might have been overlooked or incorrectly added as a gift author. Similarly, a bibliometric analysis of colonic diverticular disease Colonic diverticular disease: a bibliometric and visual analysis shows that publication counts affect career progression. Including someone solely to boost their list of papers violates ethical guidelines from Committee on Publication Ethics.
When deciding authorship order, consider the norms of your field. In many biomedical disciplines, the first author is the primary contributor, the last author is the senior researcher, and corresponding authors handle submission. But these norms are not universal. A systematic review on colon cancer classification using convolutional neural networks A systematic review on colon cancer classification by convolutional neural networks listed twenty two authors across multiple institutions. The group used a predetermined rule based on hours contributed and analytic leadership. You should write down the agreed ordering rule at the start and specify how changes will be negotiated.
A Practical Workflow for Transparent Authorship
Use this sequence to move from vague promises to a clear, documented plan.
Step 1: Hold a kickoff authorship meeting. Invite everyone who will interact with the study, including trainees and data managers who may contribute intellectually. Present a draft list of anticipated roles using CRediT categories. Discuss who will draft sections, perform analyses, and interpret results. Record the meeting minutes. Reference the Topics of the WHO Patient Safety Curriculum Guide Proposals as an example of how training programs embed authorship ethics into curricula, you can adopt a similar teaching moment for your group.
Step 2: Draft a written authorship agreement. Write a one page document listing each person, their tentative roles, and the proposed author order. Include a clause that contributions will be reassessed after data collection and after the final analysis. Store the document in a shared repository (like a lab wiki) that everyone can access. The ORCID record can eventually link to this agreement or its CReDiT assignments.
Step 3: Assess contributions at the halfway point. When you finish data collection or a major analysis stage, revisit the agreement. Someone who initially planned to lead the statistics may have moved to a different role. Adopt the approach used in the development of the cascade of care framework for anal cancer screening guidelines Application of the Cascade of Care Framework to Guide Evidence Informed Implementation: map each person’s tasks against a stepwise process and adjust credit accordingly.
Step 4: Finalize before writing begins. Decide authorship order and confirm that everyone accepts their role. This is the last time to negotiate without the stress of an approaching submission deadline. In fields where senior authors traditionally handle correspondence, confirm that the designated corresponding author is willing and available.
Step 5: Confirm and submit. One week before submission, send the final author list to all co authors. Ask each to reply with explicit approval. Do not proceed until you have written responses from everyone. This mirrors the consensus building process used in developing speech and language therapy guidelines in Finland Defining the Level of Need and Total Intervention Time in Children's Speech and Language Therapy in Finland, where multiple stakeholders contributed to a single agreed document.
Common Mistakes in Authorship Discussions
Waiting until the manuscript is written. This leads to pressure to add people who did little or to omit someone who contributed heavily. The Committee on Publication Ethics has many case examples of disputes that could have been avoided with early conversation.
Using a general criterion like “made a significant contribution” without defining what that means. A postdoc who runs a hundred assays may feel they contributed more than the senior author who provided the idea. Discuss specific roles, not vague feelings.
Gift authorship. Including a department chair or a colleague who provided no intellectual input is dishonest and violates journal policies. It also inflates the author list, making it harder for readers to know who did what.
Ghost authorship. Not naming a professional medical writer or a junior researcher who wrote large portions is equally unethical. The NIH Data Management and Sharing Policy emphasizes transparency, which applies to people as much as data.
Assuming order by seniority or alphabetical listing without discussion. While some fields default to alphabetical order, most expect a transparent explanation. Document the rationale in the manuscript’s author contributions statement.
Limits and Uncertainty
No single system fits every research context. Authorship norms vary between disciplines (e.g., physics and economics differ greatly from biomedicine). In large consortium projects with dozens of members, the CRediT taxonomy may become unwieldy. You may need to adopt a group authorship model where all members are listed with a named writing group.
Even with early discussion, contributions can shift unexpectedly. A student may leave the program, or a collaborator may fall seriously ill. Build flexibility into your agreement: include a clause that says if someone cannot complete their agreed role, they may be moved to an acknowledgment after a group vote. However, note that the Committee on Publication Ethics advises against removing someone without their consent unless they refuse to provide necessary data or critically review the manuscript.
Uncertainty also surrounds the role of research support staff a data manager who sets up the database but does not interpret the results. In many institutions, such roles fall below the authorship threshold. Yet they are essential. Acknowledge them explicitly and consider offering co authorship if they contribute to the study design or manuscript revision. The ORCID record can capture these contributions even if the person is not an author, by using the “acknowledged contributor” flag.
Finally, remember that authorship is a form of credit that persists in the literature forever. If you make a mistake, it cannot easily be corrected. Err toward inclusion when in doubt, but only if the person meets the intellectual contribution standard.
Frequently Asked Questions
Q: Do I have to list everyone who contributed to data entry as an author? No. Data entry typically does not involve intellectual input unless the person also helps interpret, design the analysis, or write the methods. Those who only perform data entry should be thanked in the acknowledgments.
Q: What if a co author disagrees with the order during the final check? Return to the documented criteria you agreed on at the start. If the criteria are met, the order stands. If the person’s contribution changed significantly, adjust accordingly. An in person meeting (via video call) often resolves the dispute faster than email.
Q: Is it okay to add a new author after the manuscript is drafted? Only if that person made a substantial intellectual contribution during the interpretation or writing phase that was not originally anticipated. Adding someone to improve the submission chance is not acceptable. Document the reason in the cover letter.
Q: Should I use a signed contract for authorship agreements? A signed or acknowledged email is sufficient. Formal legal contracts are unusual in academic research and may create unnecessary barriers. A clear email thread or a signed note in a shared document is enough. The ORCID system can later reference the contribution as part of the researcher’s record.
References and Further Reading
- Committee on Publication Ethics , Provides case studies and guidelines on authorship and publication ethics.
- NIH Data Management and Sharing Policy , Official policy that encourages transparent planning of roles in data intensive research.
- ORCID , Platform for persistent researcher identifiers that can record contributor roles.
- U shaped association between dietary flavonoid and gallstone prevalence with sex-specific effects , Example of a study requiring clear role delineation.
- Topics of the WHO Patient Safety Curriculum Guide Proposals , Illustrates how authorship ethics can be taught in medical curricula.
- Colonic diverticular disease: a bibliometric and visual analysis , Demonstrates how authorship counts affect career metrics.
- A systematic review on colon cancer classification by convolutional neural networks , Example of collaborative authorship in machine learning research.
- Application of the Cascade of Care Framework to Guide Evidence Informed Implementation , Shows stepwise assignment of contributor roles.
- Defining the Level of Need and Total Intervention Time in Children's Speech and Language Therapy in Finland , A consensus based guideline that models inclusive contribution agreements.
Related Articles
- Responding to Peer Review: A Structured Revision and Response Process
- Research Collaboration Agreements: Aligning Roles, Data, and Authorship Early
- Scientific Communication for Mixed Audiences
- Research Project Management: A Lightweight System for Complex Studies
- Research Ethics in Data-Intensive Biology: Questions to Ask Before Sharing