Lab Authorship Policy: A Template Built on ICMJE and CRediT

A lab authorship policy is a short written rule, agreed at the start of a project, that says who qualifies as an author, who is acknowledged instead, how author order is decided and what happens when people disagree. The most widely used test is the four-part ICMJE definition: a substantial contribution, drafting or critical revision, final approval, and accountability for the work. CRediT, a list of 14 contributor roles, records what each person did.

Most authorship disputes are not about the test. They are about expectations nobody voiced until the paper was nearly done. This guide gives you the standard definitions from their sources, a way to use them in a lab, and a template. A caution first: the ICMJE recommendations come from medical journal editors, and fields differ in their conventions on order and credit. Check the target journal's policy, and your institution's, before you finalize.

The four criteria

The International Committee of Medical Journal Editors (ICMJE) recommends that authorship rest on all four of these:

A substantial contribution to the conception or design of the work, or to the acquisition, analysis or interpretation of data.
Drafting the work or revising it critically for important intellectual content.
Final approval of the version to be published.
Agreement to be accountable for all aspects of the work, including investigating and resolving questions about its accuracy or integrity.

Everyone who meets all four should be an author. Anyone who does not meet all four should not be listed as an author but should be acknowledged, individually or as a group, with their specific contributions stated. ICMJE adds that editors should expect the corresponding author to get written permission from each person acknowledged, because a name in the acknowledgments can imply that they endorse the data and conclusions.

What does not earn authorship alone

ICMJE is direct about this, and it is where many lab arguments start. These activities, on their own, do not qualify someone as an author:

securing funding;
general supervision of a research group, or general administrative support;
writing assistance, technical editing, language editing or proofreading.

That means a PI who obtained the grant and supervised the group is not automatically an author. They are, if they also meet the four criteria, which in practice means shaping the design or interpretation, revising the manuscript critically and being willing to stand behind it. The same rule protects students: a student who ran the experiments but was never asked to draft or approve may fall short of criteria two and three, and the fix is to invite them in, not to leave them out.

CRediT: recording who did what

CRediT, the Contributor Role Taxonomy, is a community-owned taxonomy of 14 roles, approved as ANSI/NISO Z39.104-2022. It describes contributions; it does not decide authorship or order. The 14 roles are:

Conceptualization
Data curation
Formal analysis
Funding acquisition
Investigation
Methodology
Project administration
Resources
Software
Supervision
Validation
Visualization
Writing, original draft
Writing, review and editing

Used well, CRediT is the cheapest dispute prevention tool a lab has. At the start of a project you write down, role by role, who expects to do what. At submission you update it. If two people have claimed "Formal analysis" you find out in week two, not at the cover letter. A role can be held by one person, several people, or shared, and one person can hold several.

Deciding the order

ICMJE says the author group decides the order collectively, by whatever criteria it chooses, and that editors do not set it. So your lab has to have criteria, and the PI cannot leave them to the editor. The common questions to settle in advance:

First author. Usually the person who did the most work and drafted the paper. Decide what happens when two people split the work equally (co-first authors, with the contribution statement saying so).
Last author. In many life-science fields the senior author. In others, order differs, so state your field's convention rather than assuming.
Corresponding author. The person who will handle the journal and answer questions for years, which is not always the first author.
The middle. By contribution, alphabetical or something else. Say which.

Write the criteria in the policy and then apply them. A rule applied to the first paper and abandoned for the second causes more resentment than a rule applied consistently that someone dislikes.

Disputes, changes and people who leave

ICMJE's position on disputes is useful because it shows where authority sits. Editors do not decide who qualifies. If the authors cannot agree, the institution or institutions where the work was done should be asked to investigate. If someone asks to add or remove an author after submission or publication, editors should seek an explanation and a signed statement of agreement from all listed authors and from the person being added or removed.

So your policy should name the path before there is a problem: talk to the PI first, then the department's research integrity or graduate office. And it should cover the case that causes real trouble, which is the student or postdoc who leaves with work unpublished. Say in advance who finishes it, how the leaver's contribution is credited, and that a leaver who met the four criteria keeps their place.

A template

Adapt it and ask your research office to check it against institutional policy and your target journals' rules.

AUTHORSHIP POLICY: [Lab name], effective [date]

1. Standard
   We follow the ICMJE criteria. An author must: (1) make a substantial
   contribution to the design, data or analysis; (2) draft or critically
   revise the work; (3) approve the final version; and (4) agree to be
   accountable for it. Anyone who does not meet all four is thanked in
   the acknowledgments, with their contribution stated, with their
   permission.
   Funding the work, supervising the lab, administrative help and
   editing alone do not qualify.

2. Record of contributions
   At the start of each project we list who expects to do what, using
   the CRediT roles. We update it at the draft stage and at submission.
   The final CRediT statement appears in the paper where the journal
   allows.

3. Author order
   - First author: [the person who did most of the work and drafted
     the paper]. Co-first authorship is used when [equal contribution],
     and stated in the paper.
   - Last author: [the senior author / supervisor, per our field].
   - Corresponding author: [the person who will answer for the paper
     over the long term].
   - Middle authors: in order of [contribution / alphabetical].
   We discuss order when the project starts and again when the draft
   exists. Changes need everyone's agreement.

4. Disagreements
   Raise it with [the PI] first, in writing if you prefer. If it is not
   resolved, you may go to [graduate director / research integrity
   office] at [institution]. Nobody will be penalised for raising a
   concern about authorship.

5. When someone leaves
   Before you leave, we agree in writing who will complete unpublished
   work, how your contribution is credited, and how you will be
   contacted. If you met the criteria, you remain an author, and you
   approve the final version.

6. Data, code and acknowledgments
   Everyone who contributed data or code is credited as the journal
   allows. We ask permission before naming anyone in acknowledgments.

Making it stick

A policy in a PDF nobody reads solves nothing. Three habits help. Start every project with a ten-minute CRediT table in a shared doc. Put "authorship check" on the agenda at the draft stage of every paper, as a standing item in the lab meeting or the supervision meeting. And include the policy in the lab expectations document, where new members meet it on day one.

LetRelay's paper pipeline tracks each paper from idea to camera-ready, with its authors, deadlines (counted down in AoE where the venue uses it) and a DOI or arXiv import. That gives you a natural place for the "authorship check" step to live as a task with an owner and a date. It does not decide authorship, and it does not generate a CRediT statement; the policy is a human agreement, and the software is a reminder that it is due.

FAQ

What are the four ICMJE criteria for authorship?

A substantial contribution to the conception or design of the work, or to acquiring, analysing or interpreting data; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; and agreement to be accountable for all aspects of the work. All four are needed.

Does being the PI automatically make someone an author?

No. ICMJE says that securing funding and general supervision of a research group do not qualify someone alone. A PI who also contributes to design or interpretation, revises the manuscript and approves it, and accepts accountability, meets the criteria in the usual way.

What is CRediT?

The Contributor Role Taxonomy: 14 roles that describe what each person did on a research output, such as Conceptualization, Investigation, Software and Supervision. It is a NISO standard, ANSI/NISO Z39.104-2022. It records contributions and does not decide who is an author or in what order.

Who decides author order?

The authors, collectively, using criteria they choose. ICMJE says journal editors do not decide it. The lab should have its own written criteria so the decision is not made under deadline pressure.

What should we do if authors cannot agree?

ICMJE advises that the institution where the work was done should investigate when authors cannot agree who qualifies. In practice that means your department's research integrity or graduate office. Put the route in your policy before anyone needs it.

Can we add or remove an author after submission?

Editors will normally ask for an explanation and a signed statement of agreement from every listed author and from the person being added or removed. It is much easier to settle order before submission.

Does this apply to every field?

The ICMJE criteria come from biomedical journal editors, and other fields have their own conventions about order and credit. Use ICMJE and CRediT as a starting structure, and check the target journal's policy and your institution's before you finalize.