CRediT contribution statements: 14 roles and copy-ready examples.

A CRediT author contribution statement assigns every author one or more of the 14 roles in the Contributor Roles Taxonomy, written as a plain list: "S. Al-Rashid: Conceptualization, Formal analysis, Writing – original draft. M. Duarte: Supervision, Funding acquisition, Writing – review & editing." PLOS ONE states the floor in one sentence: "Provide at minimum one contribution for each author in the submission system using the CRediT taxonomy." The sections below translate the 14 roles into plain language and give worked statements for two, four, and six author papers.

What is CRediT and why do journals require it?

CRediT (Contributor Roles Taxonomy) is a controlled vocabulary of 14 contributor roles, formalized as the US national standard ANSI/NISO Z39.104-2022 in February 2022. Journals adopted it because free-text contribution lines had stopped carrying information.

"All authors contributed to the manuscript and approved the final version" tells an editor nothing. It cannot distinguish the student who ran every experiment from the department head who read one draft, and it cannot be checked, indexed, or compared across papers. A fixed vocabulary can. When each author is tagged with named roles, the record shows who built the software, who did the statistics, and who wrote the first draft, in terms that mean the same thing at every journal using the standard.

The practical consequence for you is a form. At submission, the system asks you to tick roles per author, and at PLOS ONE that form will not accept an author with zero roles. The thinking has to happen before you reach that screen.

What do the 14 CRediT roles mean in plain language?

The standard defines 14 roles, and none of them refers to seniority or byline position. Here is each one, glossed for a working researcher:

• Conceptualization: forming the research questions and the overall aims of the project.
• Data curation: cleaning, annotating, and maintaining the data so someone else could reuse it.
• Formal analysis: applying statistical, mathematical, or computational techniques to the data.
• Funding acquisition: winning the money that paid for the work.
• Investigation: running the experiments, fieldwork, or interviews and collecting the evidence.
• Methodology: designing the methods or building the models the study rests on.
• Project administration: coordinating the plan, the people, and the timeline.
• Resources: providing materials, instruments, samples, patients, or computing capacity.
• Software: writing and testing the code the study depends on.
• Supervision: carrying oversight and mentorship responsibility for the research activity.
• Validation: verifying that the results and outputs replicate or reproduce.
• Visualization: preparing the figures and other presentations of the data.
• Writing – original draft: producing the first full text of the manuscript.
• Writing – review & editing: revising, commenting on, or reworking drafts at any stage.

Two features of the taxonomy do most of the work. An author can hold several roles, and a role can belong to several authors. Nothing requires all 14 to appear; a theory paper with no dataset simply never uses Data curation.

Worked examples by team size

The format is the author's name or initials, a colon, then their roles, matching the byline exactly. Adapt the mappings below to what actually happened, never the reverse.

A two-author paper, in the common student-and-supervisor shape:

S. Al-Rashid: Conceptualization, Methodology, Investigation, Formal analysis,
Data curation, Writing – original draft.
M. Duarte: Conceptualization, Supervision, Funding acquisition,
Writing – review & editing.

A four-author paper where the technical work is split:

L. Chen: Conceptualization, Methodology, Investigation, Writing – original draft.
P. Novak: Software, Data curation, Visualization.
T. Okafor: Formal analysis, Validation, Writing – review & editing.
H. Karimi: Supervision, Project administration, Funding acquisition,
Writing – review & editing.

A six-author paper across two labs:

A. Farouk: Conceptualization, Investigation, Writing – original draft.
J. Mbeki: Investigation, Data curation.
R. Silva: Methodology, Software.
K. Tanaka: Formal analysis, Visualization.
D. Petrova: Resources, Validation, Writing – review & editing.
N. Haddad: Supervision, Project administration, Funding acquisition,
Writing – review & editing.

Notice what the six-author version admits. Two people did nothing but collect and manage data or build tools, and the statement says so without embarrassment. That is the point of the taxonomy. A contribution statement where all six authors hold all the same roles reads as either an extraordinary team or an unexamined form, and editors have seen enough of both to guess which.

Check your target journal's preferred name format before copying any of these. Some want initials, some want full names, and the names must match the byline you entered in the system.

How do you handle shared roles and settle disputes early?

Any number of authors can share a role. Where several hold the same one, the standard optionally lets you mark each person's degree of contribution as lead, equal, or supporting.

The harder problem is human, and it has a cheap solution: circulate a draft role mapping while the manuscript is still being written, not the week of submission. Credit disagreements surface eventually. Surfaced early, they are a short conversation about who did what; surfaced at the submission form, they hold the paper hostage while the argument runs. The mapping exercise also exposes the awkward case directly, because an author for whom you cannot honestly tick a single role is not a statement problem. That is an authorship problem, and renaming a courtesy inclusion as Supervision does not solve it, it just writes it down.

Shared first authorship and byline order are separate machinery from CRediT roles. The taxonomy records what each person did; the byline order carries whatever conventions your field attaches to it. Keep the two negotiations apart and each gets easier.

Where does the contribution statement go in your submission?

At PLOS ONE, it goes into the submission system as per-author metadata rather than into the manuscript file. Other journals want an Author Contributions section in the manuscript itself, and their guide for authors names the exact position.

So this is a check-the-guide item, like everything else on the first-submission checklist. The system will typically collect the roles alongside each author's affiliation and ORCID iD, which is one more reason every coauthor should have an ORCID before submission week. Your cover letter does not need to repeat the roles; the editor reads them in the system.

The statement is also one of the declarations a journal's technical check screens before review, together with ethics, funding, and data availability. A pre-submission review that checks your declarations against the target journal's own rule set flags the author with no assigned role before the submission form does, when fixing it costs a conversation instead of a delay.

The vocabulary is fixed at 14 roles, and the floor is one role per author. Map honestly, settle disputes early, and the statement is a formality by the time you reach the submission form.