
Usman spent eight months collecting survey data for a study of school attendance in Punjab — 400 interviews, in 20 schools, in summer heat. When the paper was published, his name did not appear on it. The lead author had listed only the three people who wrote the manuscript, and the data collector was thanked in one line of the acknowledgements. Usman had done a third of the work and received none of the credit. An author contributions statement exists to stop exactly this from happening.
Most journals now ask for an author contributions statement: a short section that says who did what on the paper. It usually appears near the end of the manuscript, just before the acknowledgements or the references. The standard way to write it is the CRediT taxonomy, a list of fourteen contributor roles covering everything from designing the study to securing the funding. This guide explains what CRediT is, what each role means, how to assign roles fairly, and how to write a statement your journal will accept.
CRediT stands for Contributor Roles Taxonomy, a standard list of fourteen roles that describe the work involved in producing a scholarly paper. It was developed by publishers, funders, and researchers to give every contributor a consistent vocabulary, and it is now used by most major journals. Most major publishers now ask for CRediT roles during submission, and many display the roles on the published article page, next to the author list, so a reader can see at a glance who designed the study, who collected the data, and who wrote the paper.
The fourteen roles fall into four groups that are easier to remember. The ideas group: Conceptualization (forming the research idea and goals) and Methodology (designing the methods). The doing group: Software, Validation, Formal analysis, Investigation, Resources, and Data curation. The writing group: Writing the original draft, Writing — review and editing, and Visualization. The running group: Supervision, Project administration, and Funding acquisition. One person can hold several roles, and several people can share one role.
A typical written statement looks like this, using initials as most journals prefer: "Conceptualization: A.B. and C.D.; Methodology: A.B.; Investigation: E.F.; Formal analysis: C.D.; Writing — original draft: A.B.; Writing — review and editing: all authors; Supervision: G.H.; Funding acquisition: G.H." Keep it factual and specific, and use the exact CRediT role names so the statement can be read consistently by anyone.
Now the same statement with real names. Ayesha, a lecturer in Lahore, designed a study of reading habits in primary schools. Bilal, her colleague, collected the data in 12 schools. Sana, a research assistant, cleaned the dataset and ran the analysis. Professor Tariq supervised the project and secured the grant. Their statement reads: "Conceptualization: Ayesha Khan; Methodology: Ayesha Khan; Investigation: Bilal Ahmed; Data curation: Sana Raza; Formal analysis: Sana Raza; Writing — original draft: Ayesha Khan; Writing — review and editing: Ayesha Khan, Bilal Ahmed, Sana Raza; Supervision: Tariq Mehmood; Funding acquisition: Tariq Mehmood." Every author appears, every major task is covered, and a reader can see at a glance who did what. Note that Bilal, the data collector, is an author here — because collecting data for a study he helped to design is a substantial contribution, not a routine task.
Large collaborations need more care than small teams. For a six-author paper the statement might read: "Conceptualization: A.B. and F.G.; Methodology: A.B., C.D.; Software: C.D.; Investigation: C.D., D.E.; Formal analysis: A.B.; Data curation: D.E.; Writing — original draft: A.B.; Writing — review and editing: A.B., C.D., F.G.; Supervision: F.G.; Funding acquisition: F.G." Notice that every author appears, that no one claims a role they did not do, and that the two lead authors carry the roles a reader would expect them to carry.
With ten or more authors, a single statement can run to a full page, and some journals ask for the roles to be entered in the submission system instead of written as prose. In that case, the system generates the statement automatically from the ticked roles, so the job is to make sure the ticked roles are right for every author. Ask the corresponding author to circulate the generated statement to the whole team before submission, so no one is surprised by what appears in print.
The best time to agree contributions is at the start of the project, not at submission. Sit down with your collaborators and decide who will do what, using the CRediT roles as a checklist. Write the agreement down and share it, so everyone knows what is expected and what credit they will receive.
Put the agreement in writing, even for a small team. A short email after the first project meeting is enough — for example: "Dear all, Following today's meeting, our roles for the reading-habits study are: Ayesha — Conceptualization, Methodology, Writing (original draft); Bilal — Investigation; Sana — Data curation, Formal analysis; Professor Tariq — Supervision, Funding acquisition. Please reply to confirm or correct this. Best, Ayesha." Everyone can reply to confirm or correct it, and the email becomes a record you can return to when memories differ. Revisit the agreement when something significant changes — when a new collaborator joins, when a method fails and someone redesigns it, or when the analysis takes a new direction.
Authorship and contributorship are related but different. Authorship is the list of names on the paper, and it is governed by criteria such as those of the ICMJE, the International Committee of Medical Journal Editors: an author must have made a substantial contribution to the work, helped draft or revise it, approved the final version, and agreed to be accountable for it. All four conditions must be met. Contributorship is the record of what each person actually did, whatever their authorship status.
Some contributors do not qualify as authors — a person who provided a dataset or ran a routine analysis may be acknowledged rather than listed as an author — but their contribution should still be recorded. The contributions statement is where that record lives, and it should name everyone who contributed, not only the authors. Consider two cases. A technician who ran 200 blood samples following a standard protocol, and did nothing else, is a contributor to be acknowledged, not an author. A research assistant who designed the sampling plan, collected the data, and helped interpret the results has made a substantial contribution and should usually be an author. The difference is not the hours worked but the nature of the contribution. Our guide on who deserves authorship explains the criteria in full.
When two authors contributed equally, journals usually allow them to be marked as joint first authors, often with a footnote stating that they contributed equally. The contributions statement should support that claim: if two people are joint first authors, their roles in the statement should show comparable, substantial work. A footnote that says "A.B. and C.D. contributed equally" is not convincing if the statement shows that A.B. did the design, the analysis, and the writing while C.D. only collected data.
The same applies to senior authorship. The last author is usually the person who supervised the work or secured the funding, and the statement should show that through the Supervision or Funding acquisition roles. If the author order does not match the contributions statement, reviewers and readers will notice, so make sure the two agree before you submit. Check the author order against the statement as a final step: read the statement, then read the author list, and ask whether the order tells the same story.
Contributions often change during a project, and three situations come up most often. First, someone leaves: a co-author moves to another university halfway through. Record what they did up to their departure, keep their name on the paper if their contribution justifies authorship, and confirm the final draft with them by email. Second, someone new joins: a statistician is brought in to redo the analysis. Add their roles to the statement, discuss authorship with the whole team, and update the written agreement. Third, the work takes a direction nobody planned: the method fails and one author redesigns it. That author has earned the Methodology role.
Any change to the author list after submission usually needs the agreement of all authors and the editor's approval, so settle authorship questions before you submit if you can. If a change is unavoidable after submission, explain it to the editor honestly and provide the written agreement of everyone involved. A short, factual email to the editor — who is being added or removed, why, and that all authors agree — is usually enough.
Most authorship disputes come from the same cause: contributions were never agreed or never written down. A contributions statement made at the start of the project, and updated as the work changed, is the strongest protection against disputes, because it turns a matter of memory into a matter of record. If a dispute does arise, the contributions statement is the first document an editor or an institution will ask for.
Keep dated records of who did what — meeting notes, drafts, data files with authorship information — so the statement can be checked against evidence. Journals follow COPE guidance on authorship disputes, and a clear record makes their job, and yours, much easier. COPE, the Committee on Publication Ethics, advises editors to look for a written record of agreed contributions and to ask all authors to confirm any change. If your team kept the first-meeting email and the milestone updates, you can answer that request in minutes. If nothing was written down, the same dispute can take months.
The first mistake is writing the statement at the last minute from memory, which is how a data collector like Usman gets left off the paper. The second is using invented role names instead of the CRediT taxonomy — "helped with the study" tells nobody anything, while "Investigation" is a standard term every reader understands. The third is listing a supervisor as an author for supervision alone, since supervision alone does not meet the ICMJE test.
The fourth mistake is gift authorship — adding a name for a contribution that does not justify it, such as a department head who only approved the project. The fifth is ghost authorship — leaving off someone who did substantial work, such as the person who wrote the analysis code. Both damage trust, and both can lead to a formal complaint. Avoid vague entries such as "all authors contributed equally" unless that is genuinely true, and avoid listing roles that nobody performed. A statement that claims everyone did everything tells the reader nothing.
| Role | What it covers | Typical contributor |
|---|---|---|
| Conceptualization | Forming the research idea and goals | Lead or senior author |
| Methodology | Designing the methods and models | Lead author or methodologist |
| Software | Writing code and software | Analyst or programmer |
| Validation | Checking results can be reproduced | Co-author or reviewer |
| Formal analysis | Applying statistics or analysis | Analyst or statistician |
| Investigation | Running experiments, collecting data | Researcher or student |
| Resources | Providing materials or participants | Collaborator or lab head |
| Data curation | Managing, cleaning, and storing data | Researcher or data manager |
| Writing — original draft | Writing the first version | Lead author |
| Writing — review and editing | Revising and improving the draft | All authors |
| Visualization | Preparing figures and graphics | Author or designer |
| Supervision | Overseeing the project | Senior author |
| Project administration | Managing the work and timeline | Lead or senior author |
| Funding acquisition | Securing the money for the work | Principal investigator |
These authoritative external resources explain the standards and tools mentioned in this guide:
What is CRediT?
CRediT is the Contributor Roles Taxonomy, a standard list of fourteen roles that describe the work involved in producing a scholarly paper, from conceptualization and methodology to writing, supervision, and funding acquisition. Most major journals now use it, so a reader can see at a glance who designed the study, who collected the data, and who wrote the paper.
Is a contributions statement required?
Most journals now require one, either as a section in the manuscript or as roles selected in the submission system. Check the journal's author guidelines for its exact format. Some systems ask you to tick the CRediT roles for each author before the manuscript can be submitted, and the system then generates the statement automatically.
Can one person have several CRediT roles?
Yes. One person can hold several roles, and several people can share one role. The statement should record the real work, not force a one-to-one match between people and roles. A lead author, for example, often holds conceptualization, methodology, formal analysis, and writing the original draft on the same paper.
What is the difference between authorship and contributorship?
Authorship is the list of names on the paper, governed by criteria such as the ICMJE's. Contributorship is the record of what each person actually did, including people who contributed but do not qualify as authors. A person who provided a dataset may be acknowledged rather than listed as an author, but their contribution should still be recorded there.
When should contributions be agreed?
At the start of the project, using the CRediT roles as a checklist, and revisited whenever the work changes significantly. Agreeing late, from memory, is the most common cause of authorship disputes. Put the agreement in a short email after the first project meeting, so everyone has seen it and can confirm or correct it.
What are joint first authors?
Joint first authors are two or more authors who contributed equally and share the first-author position, usually marked with a footnote. The contributions statement should show comparable, substantial work for each of them. If the statement shows one author doing far more than the other, the claim of equal contribution will not convince readers or reviewers.
Submit your manuscript ← Back to GAP Insights
Useful publishers, databases, funders, tools and institutions for researchers: