CASRAI Dictionary

Tag: credit contributor roles taxonomy

  • CRediT Roles for Data Curation and Software

    CRediT roles such as Data Curation and Software give data managers and research software engineers a standardised, citable way to be formally recognised for research contributions that byline authorship rules routinely exclude. Two of the taxonomy’s fourteen roles map directly onto their work, and institutions can use those role tags — including the optional “degree of contribution” qualifier — as documentary evidence in annual review, promotion and tenure cases.

    The Contributor Role Taxonomy (CRediT) is a controlled vocabulary of fourteen contributor roles used to describe the specific contributions individuals make to a research output, independent of authorship order. CASRAI originated the CRediT contributor role taxonomy in 2014; the standard is now stewarded by NISO as ANSI/NISO Z39.104-2022, approved in 2022.

    What Do the Data Curation and Software CRediT Roles Cover?

    Data Curation and Software are two of the fourteen roles in the CRediT contributor role taxonomy, and together they cover most of the technical infrastructure work that keeps a research output usable and reproducible. Under NISO’s definitions, Data Curation covers management activity that annotates, cleans and maintains research data — including software code where it is needed to interpret that data — for both initial use and later reuse. Software covers programming and software development: designing programs, implementing code and algorithms, and testing existing code components.

    These definitions were written broadly enough to capture roles that rarely appear on a title page:

    • Research data managers and data stewards who build metadata schemas and manage repository deposits
    • Research data librarians who oversee data management plans and long-term preservation
    • Research software engineers (RSEs) who build, maintain and test the pipelines and analysis code a study depends on
    • Bioinformaticians and computational scientists whose code is central to the result but who did not draft the manuscript

    Several adjacent roles frequently apply to the same people: Data Curation and Software combine most often with Methodology, Validation, Resources (for computing infrastructure) and Visualization. A single individual can — and typically does — hold more than one CRediT role on the same output; NISO’s implementation guidance is explicit that a role can also be assigned to multiple contributors on the same paper.

    Why Are Data Managers and Software Engineers Excluded From the Byline?

    Byline authorship is governed by criteria such as those from the International Committee of Medical Journal Editors (ICMJE), which require substantial contribution to conception, drafting or critical revision, plus final approval and accountability. Building a data pipeline or curating a dataset frequently fails one of those tests — usually drafting or revision — even when the contribution was essential to the result.

    CRediT was designed to sit alongside, not replace, those authorship criteria. NISO’s own guidance states plainly that CRediT “is not designed to determine authorship” but instead documents the specific contributions of authors and other contributors, including people who would otherwise only appear in an acknowledgements line. That separation is precisely what makes it useful for technical staff: a data manager or RSE can receive a formal, structured, machine-readable Data Curation or Software credit on an output without needing to clear the higher bar of full authorship.

    How Can Institutions Use CRediT Roles in Performance Review?

    Research institutions can treat a contributor’s accumulated CRediT role tags as structured evidence of impact, distinct from and complementary to publication counts or authorship position. Because the roles are standardised and, where publishers tag them in JATS XML, machine-readable, they can be aggregated across a person’s ORCID record rather than re-argued from scratch at every review cycle.

    Traditional CV evidence CRediT-based evidence Use in review
    Author position (2nd, 5th, etc.) Named role (Data Curation, Software) Documents what was actually done, not just list order
    General “contributed to analysis” statement Structured, standardised role plus optional degree of contribution Comparable across papers, journals and disciplines
    Acknowledgements-only mention Formal CRediT role on the same footing as co-authors Countable, citable line in annual review or promotion dossier
    Manual claim, hard to verify Machine-readable tag exportable via ORCID / Crossref metadata Independently verifiable by a review committee

    Concrete steps an institution can take:

    • Require staff to list their specific CRediT roles — not just “co-author” — against each output in annual review and promotion documentation
    • Instruct principal investigators to assign Data Curation and Software roles at the manuscript-submission stage, before contributor memory fades
    • Ask review committees to weight a sustained pattern of Data Curation or Software roles as evidence of infrastructure contribution, on comparable footing with first authorship in relevant career tracks
    • Encourage contributors to keep their ORCID record current so CRediT roles from publishers using CRediT-tagged JATS XML populate automatically

    What Does CRediT Leave Undefined? The Degree-of-Contribution Gap

    CRediT includes an optional qualifier — “lead”, “equal” or “supporting” — that can be attached where several people share the same role. This degree-of-contribution tag is exactly what a performance-review committee needs to distinguish a data manager who led curation for a multi-year cohort study from one who supported it for a single deposit. It is the single most under-used lever institutions have available inside a taxonomy most already partially adopt.

    Two constraints matter for institutions relying on it. First, NISO’s implementation guidance is explicit that degree-of-contribution is not currently part of the CRediT standard itself — individual publishers decide whether to request it, so its presence in a contributor statement is inconsistent across journals. Second, adoption of CRediT overall is uneven: PLOS has made CRediT its sole standard across all journals, while others offer it optionally through Editorial Manager (integrated since 2016) or Clarivate’s ScholarOne (integrated since 2018). Crossref has stated it will add CRediT to its publisher metadata schema in 2026, which should make role and degree-of-contribution data far easier to aggregate at the institutional level once adopted — but until then, institutions auditing technical contributions cannot assume degree-of-contribution data exists for every paper in a portfolio, and should ask contributors to supply it directly where the publisher record is silent.

    Frequently Asked Questions About CRediT Roles

    What is a credit role?

    A CRediT role is one of fourteen standardised labels — such as Data Curation, Software, Methodology or Investigation — used to describe a specific type of contribution an individual made to a research output. Roles are assigned independently of authorship order and can be combined, so one contributor may hold several roles on the same paper.

    What are the 14 credit contributor roles?

    The fourteen roles are Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, and Writing – review & editing, as defined under ANSI/NISO Z39.104-2022.

    What does role taxonomy mean?

    A role taxonomy is a fixed, controlled vocabulary that classifies types of contribution rather than free-text description. CRediT’s role taxonomy standardises how contributions are labelled across journals and disciplines, so a “Software” credit means the same thing whether the paper is in genomics or economics.

    Implications for Research Assessment

    As Crossref’s planned 2026 metadata integration and continued publisher adoption make CRediT data more machine-readable, the practical barrier to using it in institutional review shifts from availability to policy: whether promotion committees are willing to treat a Data Curation or Software role, especially one tagged “lead”, as comparable evidence to first authorship in the relevant career track. Institutions that update review criteria now — ahead of that data becoming routine — will be positioned to credit data managers and research software engineers on the strength of a standard that, unlike free-text acknowledgements, is built to be counted.

  • MDPI Author Contributions: Compliance Guide

    MDPI requires every submitted manuscript to carry an author contributions statement built on the CRediT taxonomy — a mandatory list of the 14 CRediT roles mapped to author initials, followed by a fixed sign-off sentence. This is stricter than most publishers, many of which still treat CRediT as optional or recommend it only for research articles. Authors who submit across journal families need to know exactly what MDPI checks for, because incomplete or missing statements are a common cause of pre-submission delay.

    The Contributor Roles Taxonomy (CRediT) is a fixed, 14-term vocabulary — Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, and Writing – review & editing — used to describe what each named contributor actually did on a research output. CASRAI originated the CRediT contributor role taxonomy in 2014. The standard is now stewarded by NISO as ANSI/NISO Z39.104-2022.

    What exactly does MDPI require in the author contributions statement?

    MDPI’s Research and Publication Ethics policy states that “for complete transparency, all submitted manuscripts should include an author contributorship statement that specifies the contribution of every author.” For research articles with more than one author, this is not a suggestion — it is a submission requirement checked during manuscript preparation, alongside the standard ICMJE authorship criteria (substantial contribution, drafting or critical revision, final approval, and accountability).

    The statement must be built from the CRediT role list rather than free text. MDPI’s own instructions for authors reproduce the taxonomy directly and ask authors to attach initials to each role that applies. Review articles are treated slightly differently: because CRediT’s experiment-oriented roles (Investigation, Resources, Validation) often do not map cleanly onto a literature synthesis, MDPI instead asks review authors to clarify who conceived the review, conducted the literature search or analysis, and drafted or revised the text.

    What is the required format and wording?

    MDPI publishes a template sentence structure: each CRediT role name is followed by a comma and the initials of the contributing author(s), with roles separated by semicolons. A representative example from MDPI’s own manuscript templates reads:

    “Conceptualization, X.X. and Y.Y.; methodology, X.X.; software, X.X.; validation, X.X., Y.Y. and Z.Z.; formal analysis, X.X.; investigation, X.X.; resources, X.X.; data curation, X.X.; writing—original draft preparation, X.X.; writing—review and editing, X.X.; visualization, X.X.; supervision, X.X.; project administration, X.X.; funding acquisition, Y.Y.”

    The statement must close with a fixed sentence: “All authors have read and agreed to the published version of the manuscript.” Omitting this closing line, or listing contributions in narrative prose instead of the role-and-initials format, is one of the most frequent reasons a manuscript is returned for correction before it proceeds to peer review.

    MDPI author contributions statement — required elements
    Element Requirement
    Vocabulary CRediT’s 14 fixed role terms (no free-text substitutes)
    Attribution unit Author initials, not full names
    Multiple contributors per role List all initials, separated by commas, “and” before the last
    Single-author manuscripts Statement may be omitted; sole authorship implies all roles
    Closing sentence Mandatory: “All authors have read and agreed to the published version of the manuscript.”
    Review articles Narrative statement of conception, search/analysis, and drafting responsibility instead of full role list

    Where is CRediT optional rather than mandatory?

    MDPI’s blanket, all-journal mandate is not universal practice. Publisher policy on CRediT sits on a spectrum, and authors moving between journal families need to check each venue separately rather than reusing one house style:

    • Mandatory, standardised wording — MDPI requires the role-and-initials format described above for every multi-author research article, across all of its journals.
    • Mandatory, house-style variation — publishers such as PLOS and Springer Nature journals require an author contributions statement but permit some variation in how roles are phrased alongside CRediT terms.
    • Recommended, not enforced — some society and smaller specialist journals encourage CRediT statements per ICMJE guidance but do not reject manuscripts that omit them.
    • Journal-editor discretion — a number of journals leave the decision to use CRediT versus a free-text contributions paragraph to the handling editor or field convention.

    This inconsistency is the practical reason a compliance walkthrough matters: an author contributions statement that satisfies one journal family may need reformatting — not rewriting, just reformatting into the fixed CRediT syntax — before it satisfies MDPI.

    What are the common compliance errors authors make?

    Four errors recur across MDPI submission checks, based on the patterns visible in MDPI’s own instructions, templates, and authorship-change forms:

    • Using full names instead of initials. The template format calls for initials only, matched consistently to the author list and the acknowledgements/affiliations sections.
    • Dropping the closing sign-off sentence. The “All authors have read and agreed…” line is treated as part of the statement, not a separate formality.
    • Inventing role labels. Only the 14 defined CRediT terms are accepted; ad hoc labels like “senior author” or “corresponding” are not CRediT roles and do not belong in this statement.
    • Applying the full 14-role template to a review article. Review manuscripts need the narrative conception/search/drafting statement, not the full experimental role list.

    Answer-first Q&A

    What are author contributions for MDPI?

    MDPI defines author contributions as a mandatory statement, built from the CRediT taxonomy, specifying which named author performed which of the 14 defined roles. It sits alongside MDPI’s authorship criteria, which mirror ICMJE‘s four conditions: substantial contribution, drafting or critical revision, final approval, and accountability for the work.

    What are examples of author contributions?

    Examples include Conceptualization (formulating research goals), Methodology (designing the study), Software (writing code), Formal analysis (running statistical tests), Data curation (managing datasets), and Writing – original draft. MDPI requires initials against each applicable role, not a general description.

    What this means for multi-journal authors

    Research groups publishing across MDPI, society journals, and mixed-model publishers gain the most by drafting one internal CRediT-mapped contributions record per manuscript at submission time, then reformatting the output to match each target journal’s house style — role-and-initials for MDPI, narrative or hybrid formats elsewhere. Because CRediT is a fixed vocabulary rather than a publisher-owned format, the underlying role assignments do not change between venues; only the presentation does. Consulting the CRediT contributor roles reference before submission, and cross-checking definitions against the research administration dictionary, reduces the back-and-forth that a mismatched contributions statement otherwise creates at the editorial-office stage.

    As more funders and institutions request structured contributorship data for assessment exercises, publisher-level enforcement patterns like MDPI’s are likely to become the norm rather than the exception, making early, consistent CRediT-mapping practice a durable habit rather than a one-off compliance task.

  • Author Contribution Statement: Elsevier vs Wiley vs Springer

    Elsevier makes a CRediT author contribution statement mandatory for journals on its Editorial Manager system, Wiley widely requires a structured “Author Contribution” section built on the same 14-role taxonomy, and Springer Nature — including Scientific Reports — asks for a statement of responsibility that is often free text rather than the full CRediT structure. The result is that the same collaboration can be documented three different ways depending purely on where it is submitted.

    An author contribution statement elsevier-style disclosure is a short, structured or narrative section, published alongside a journal article, that specifies exactly what each named author did — using either the Contributor Roles Taxonomy (CRediT) or free-text prose. CASRAI originated the CRediT contributor role taxonomy in 2014; the standard is now stewarded by NISO as ANSI/NISO Z39.104-2022, and it is this NISO-stewarded version, not any single publisher’s implementation, that defines the 14 canonical roles.

    Is a CRediT statement mandatory at Elsevier?

    Elsevier requires a CRediT author statement for every submission to journals running on its Editorial Manager system. The corresponding author selects the applicable roles for each named author from Elsevier’s published list of 14 categories — Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data Curation, Writing (Original Draft and Review & Editing), Visualization, Supervision, Project administration, and Funding acquisition.

    The statement is entered during submission, can be revised through peer review, and is published above the acknowledgments section once the article is accepted. Elsevier’s own guidance is explicit that CRediT does not alter a journal’s authorship criteria — it documents contribution, not eligibility for authorship.

    How does Wiley handle author contribution statements?

    Wiley has widely adopted CRediT across its portfolio, requiring many journals to publish a dedicated Author Contribution section built on the same taxonomy. As with Elsevier, the submitting author enters roles on behalf of all co-authors, who are expected to have reviewed and agreed their listed roles before submission.

    Wiley’s author guidance also surfaces the taxonomy’s optional degree-of-contribution qualifiers — lead, equal, or supporting — allowing authors to signal relative weighting within a shared role rather than a flat checkbox. Exact formatting still varies by individual journal, since Wiley operates a federated set of editorial policies rather than one house style.

    What does Springer Nature, including Scientific Reports, require?

    Springer Nature’s editorial policy requires authors to include “a statement of responsibility… that specifies the contribution of every author,” but it does not universally mandate the 14-role CRediT structure the way Elsevier’s Editorial Manager workflow does. Many journals accept a narrative contribution statement — full sentences describing who conceived the study, collected data, or drafted the manuscript — rather than a checklist of standardised roles.

    Scientific Reports, a Springer Nature journal, is more prescriptive within that flexible framework: its submission guidelines state authors “must supply an Author Contribution Statement,” referencing the wider Nature Portfolio authorship policy rather than a bespoke CRediT interface. This makes Springer Nature the most heterogeneous of the three publishers — requirement is consistent, but format is journal-dependent.

    Elsevier vs Wiley vs Springer: a side-by-side comparison

    The table below summarises where each publisher sits on mandate, structure, and degree-of-contribution support.

    Publisher CRediT status Format Degree-of-contribution tags
    Elsevier Mandatory on Editorial Manager journals Structured — select from 14 CRediT roles Not part of the standard submission form
    Wiley Widely required across journals Dedicated “Author Contribution” section, CRediT-based Yes — lead / equal / supporting
    Springer Nature (incl. Scientific Reports) Statement required; full CRediT taxonomy not universal Structured or narrative free text, by journal Inconsistent — varies by title

    The common thread is that all three publishers trace their statements back to the same source taxonomy, standardised as ANSI/NISO Z39.104-2022, but each has built a different submission workflow on top of it.

    Common questions about author contribution statements

    What is a contribution statement example?

    A typical example reads: “A. Smith: Conceptualization, Methodology, Writing – Original Draft. B. Jones: Data Curation, Formal Analysis, Visualization.” This CRediT-style format names each author once per role, drawing directly from the 14 standardised categories rather than free-form prose.

    What are the criteria for author contribution?

    Publishers generally look to ICMJE’s authorship criteria: substantial contribution to conception or design, or data acquisition/analysis/interpretation; drafting or critically revising the work; final approval of the version published; and agreement to be accountable for accuracy and integrity.

    What are author contributions?

    Author contributions are the specific, individually attributed tasks — such as study design, data analysis, or manuscript drafting — that each named author performed on a published work. They are distinct from authorship eligibility itself, which is governed separately by each journal’s authorship policy.

    What should substantial contributions include to be credited as an author?

    Per the Council of Science Editors, a substantial contribution covers conception or design of the work, OR acquisition/analysis/interpretation of data, OR drafting/substantive revision, combined with final approval of the submitted (and any revised) version and accountability for its accuracy.

    What the inconsistency means for multi-journal authors

    Researchers who submit to more than one of these three publishers in a given year — common in fields like biomedicine or materials science — must track three separate conventions rather than one. That has practical consequences:

    • Templates are not interchangeable: an Elsevier CRediT selection list cannot simply be pasted into a Springer Nature narrative-format submission without rewriting.
    • Degree-of-contribution nuance (lead/equal/supporting) may be lost or gained depending on which publisher’s form an author uses, even for an identical collaboration.
    • Institutional research offices compiling contribution evidence for tenure, promotion, or funder reporting face inconsistent source data across a researcher’s output.

    For research administrators, the practical fix is to standardise internally on the full CRediT taxonomy — matching the CRediT contributor roles as stewarded under ANSI/NISO Z39.104-2022 — regardless of which publisher’s native format a given article ultimately uses, then map down to each journal’s submission requirements at the point of manuscript preparation.

    Where contribution statement standardisation is headed

    The gap between Elsevier’s mandatory structured workflow and Springer Nature’s narrative flexibility is unlikely to close through publisher policy alone. NISO’s stewardship of ANSI/NISO Z39.104-2022 provides the shared reference point, but enforcement remains a per-journal editorial decision. Authors and institutions publishing across multiple houses gain the most by treating CRediT as their internal default and adapting output format — structured selection versus narrative sentence — to each publisher’s submission system, rather than the reverse.

  • ICMJE Authorship Criteria vs CRediT Roles: What the Four-Point Test Still Leaves Out

    The ICMJE authorship criteria are four cumulative conditions — substantial contribution, drafting or critical revision, final approval, and accountability — that a journal-listed author must meet in full. They decide who qualifies for the byline, but they say nothing about what each named author actually did, which is why a growing number of journals now pair the ICMJE test with a granular CRediT contributor-role declaration.

    The ICMJE authorship criteria are the International Committee of Medical Journal Editors’ four-part definition of authorship, first published in the ICMJE Recommendations and now the de facto global standard referenced by COPE, most biomedical journals, and many university research-integrity offices.

    What are the four ICMJE authorship criteria?

    The ICMJE recommends that authorship rest on four criteria, all of which must be met — not a majority. An individual must have made substantial contributions to conception, design, or data work; drafted or critically revised the manuscript; given final approval of the published version; and agreed to be accountable for its accuracy and integrity.

    • Criterion 1 — Substantial contribution: conception or design of the work, or acquisition, analysis, or interpretation of data.
    • Criterion 2 — Drafting or critical revision: writing the manuscript or reviewing it critically for important intellectual content.
    • Criterion 3 — Final approval: sign-off on the exact version submitted for publication.
    • Criterion 4 — Accountability: agreement to answer for the accuracy and integrity of any part of the work, including parts done by co-authors.

    The ICMJE is explicit that these criteria are not a filter for excluding deserving colleagues: anyone who meets criterion 1 must get the opportunity to participate in drafting, review, and approval, so they can also satisfy criteria 2–4. Funding acquisition, general supervision, and technical or language editing — on their own — do not qualify a contributor for authorship; those belong in the acknowledgements, not the byline.

    A newer addition addresses generative AI directly: under the current ICMJE Recommendations, journals must require disclosure of AI-assisted technology use, and chatbots such as ChatGPT cannot be listed as authors, because they cannot be held accountable for accuracy and integrity under criterion 4.

    Why does meeting the criteria still produce authorship disputes?

    The four-point test is qualitative, self-reported, and adjudicated by the author group itself — the ICMJE states explicitly that it is “the collective responsibility of the authors, not the journal” to determine who qualifies, and that editors should not arbitrate authorship conflicts. That design leaves real gaps in practice.

    • The biostatistician who never drafts. A statistician runs the primary analysis (clearly criterion 1) but is not invited to write or revise the manuscript, so criterion 2 is never offered to them — despite the ICMJE’s own instruction that anyone meeting criterion 1 should get that opportunity. This is one of the most common authorship grievances reported to COPE.
    • Guest and honorary authorship. A senior figure who supervised the lab, but did not contribute intellectually to conception, analysis, drafting, or revision, is added to the byline for prestige or funding-renewal reasons. COPE’s authorship guidance identifies two minimum requirements across authorship definitions — a substantial contribution and accountability — and honorary authors typically fail both.
    • Ghost authorship. A medical writer or industry statistician does the drafting and analysis but is left off the byline entirely, often in industry-funded clinical trials, obscuring who is actually accountable for the reported results.
    • Large multi-author consortia. When hundreds of contributors work on a single dataset or trial, the ICMJE recommends the group decide authorship before the work starts — but retrospectively verifying that every named individual met all four criteria, including final approval, becomes practically unenforceable at scale.

    In each case, the pass/fail structure of the ICMJE test cannot show a reader, an editor, or a research-integrity investigator which specific task a disputed author did or didn’t do. That is the exact gap CRediT was built to close.

    How do CRediT contributor roles add the missing granularity?

    CASRAI originated the CRediT contributor role taxonomy in 2014 as a structured vocabulary of contribution types. The standard is now stewarded by NISO as ANSI/NISO Z39.104-2022, and it defines 14 discrete roles — Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, and Writing – review & editing.

    Where the ICMJE test asks a single binary question — author or not — CRediT asks a descriptive one: which of these 14 tasks did this named contributor actually perform, and can more than one person share a role. Journals across Elsevier, Cell Press, PLOS, and Frontiers now request a CRediT statement alongside (not instead of) an ICMJE-compliant author list, and several also publish CRediT contributions for non-author acknowledged contributors.

    Dimension ICMJE authorship criteria CRediT contributor roles
    Function Threshold test: qualifies for the byline or not Descriptive vocabulary: records specific tasks performed
    Structure 4 cumulative, all-or-nothing criteria 14 non-exclusive, combinable roles
    Who it covers Named authors only Authors and non-author contributors alike
    Steward International Committee of Medical Journal Editors NISO (ANSI/NISO Z39.104-2022), originated by CASRAI
    Resolves guest/ghost authorship? In theory, no — self-adjudicated and unverifiable at the criteria level Makes the mismatch visible: a “Writing” credit with no “Investigation” or “Formal analysis” role is a red flag

    The complementary use matters most in the disputed scenarios above. A CRediT statement that lists a senior author under Supervision only — with no Conceptualization, Investigation, Formal analysis, or Writing role — gives an editor or institutional investigator concrete evidence to test against the ICMJE’s four criteria, something a bare byline never could.

    Answer-first: common authorship questions

    What are the criteria for authorship in the ICMJE?

    The ICMJE requires all four criteria to be met: substantial contribution to conception, design, or data work; drafting or critical revision of the manuscript; final approval of the published version; and accountability for the work’s accuracy and integrity. Meeting only some criteria means acknowledgement, not authorship.

    What are the five criteria for authorship?

    Some sources describe “five criteria” by splitting the ICMJE’s fourth criterion — accountability — into two parts: taking responsibility for the work and confirming its integrity. The ICMJE’s own text remains four official criteria; the five-part version is a restatement, not a competing standard.

    What are the minimum requirements for authorship?

    COPE identifies two minimum requirements common to authorship definitions across disciplines: making a substantial contribution to the work, and being accountable for the work and its published form. These map directly onto ICMJE criteria 1 and 4.

    What are the guidelines for authorship?

    Authorship guidelines set who can be named on a publication and what they must do to earn that status. The dominant biomedical framework is the ICMJE’s four-criteria test, supplemented in practice by CRediT contributor-role statements and journal-specific policies aligned with COPE guidance.

    What this means for journals, institutions, and researchers

    For editors, ICMJE and CRediT serve different stages of one workflow: ICMJE decides the byline, CRediT documents the record. Requiring both at submission gives research-integrity offices a verifiable trail when a dispute later reaches them, since the ICMJE explicitly directs unresolved conflicts to the researchers’ institution, not the journal.

    For research administrators, a documented CRediT statement is often the fastest way to evidence individual contribution for funder and promotion-committee requirements, independent of authorship order.

    For early-career researchers and biostatisticians, raising criterion-2 access early — asking to review and comment on a draft — is the practical way to convert a CRediT-documented “Formal analysis” role into full ICMJE-qualifying authorship before submission, not after a dispute arises.

    Where authorship attribution is heading

    Neither framework is static. The ICMJE continues to revise its recommendations — most recently to address AI-assisted technology disclosure — and CRediT’s stewardship under NISO opens a formal maintenance path for role definitions as research practice evolves. The direction of travel is layering, not replacement: a qualitative gate for who is accountable, and a structured record of who did what.

    Journals, funders, and institutions that adopt both the CRediT taxonomy and ICMJE-aligned authorship policies give readers, editors, and integrity investigators the clearest possible picture of a paper’s provenance — something the four-point test was never designed to provide on its own. For definitions of individual roles, see the CRediT roles reference and the broader research-administration dictionary.