Group of Eight: CRediT as a tenure-and-promotion signal

Profile

Composite institution profile

The challenge

What problem were they trying to solve?

The university had signed DORA (the San Francisco Declaration on Research Assessment) in 2019 and CoARA in 2023, but its promotion committees had continued to lean heavily on publication counts, h-index, and journal-impact-factor proxies. A 2024 internal review found that mid-career staff in non-bench disciplines (computational biology, methods-heavy social sciences, large-team physics) were being systematically disadvantaged because their contributions did not show up in first-author or last-author counts. The deputy vice-chancellor (research) commissioned a redesign of the promotion process aligned with DORA principles and using CRediT as the structured evidence layer. The constraint was that the new process had to be defensible at appeal — that is, a researcher whose promotion was declined had to be able to see exactly what contribution evidence had been used, and the committee's reasoning had to be reproducible from the published CRediT data and the candidate's narrative statement.

The approach

How they implemented it

The redesign was anchored on three principles agreed at the outset. First, CRediT roles would never be counted or ranked — the committee would read contribution patterns qualitatively, not score them. Second, the candidate would always have the final word: every dossier would include a candidate-written narrative statement (modelled on UKRI R4RI module 1, contextualised for Australia) in which the researcher described their own contribution patterns in their own words. CRediT data would be the structured evidence backing that narrative, not a substitute for it. Third, the committee training would be explicit about what CRediT could not tell them — the distinction between someone who Conceptualised a paper as a senior PI and someone who Conceptualised it as the lead first-author of their own programme of work is not visible in the CRediT data alone. The technical work, done by the central research-systems team in collaboration with the library, surfaced CRediT data three ways: a per-candidate dashboard accessible to the candidate and committee, a contribution-pattern visualisation grouped by the four NISO functional categories, and an exportable PDF of the candidate's contribution history for inclusion in the dossier. The committee training was redesigned as a two-day workshop run twice a year for all new committee members, using the CASRAI /for-institutions/training kit as the starting point and contextualising it for ERA reporting and Australian Research Council grant-application norms.

Timeline

Rollout phases

1. Months 1–3

   DORA/CoARA-aligned principles agreed

   DVC(R) commissioned the redesign. Three principles agreed by academic board: no counting, candidate has the final word, committee training is explicit about CRediT's limits.

2. Months 4–6

   CRediT capture verified across Symplectic

   Central research-systems team audited CRediT coverage across the previous five years. ~74% of outputs had structured CRediT; remainder marked as "narrative-only" or "not captured", which the committee was trained to handle.

3. Months 7–9

   Dashboard + visualisation build

   Per-candidate dashboard, contribution-pattern visualisation, PDF export. Pilot tested with twelve volunteer candidates from the previous year's promotion round.

4. Months 10–12

   Committee training + pilot promotion round

   Two-day workshop for promotion-committee members. Pilot promotion round ran in parallel with the legacy process; committee compared outcomes on the same twelve volunteer cases.

5. Year 2

   Institution-wide rollout

   New process became the standard for all promotion levels. Legacy process retired. First full year of CRediT-aware promotion data collected.

Outcomes

Illustrative outcomes

Every metric below is illustrative — synthesised from observed patterns across multiple adoption journeys, not attributed to a single real institution.

Lessons learned

What they would tell the next institution

• 01The "no counting" principle was the hardest to maintain and the most important. Committees naturally want to score things; the training had to be explicit and repeated.
• 02Candidate narrative + CRediT structured evidence is stronger than either alone. The narrative supplies context the structured data cannot; the structured data supplies verification the narrative cannot. They are complements, not substitutes.
• 03The 74% coverage figure was the realistic starting point. We did not wait for 100% coverage; we trained the committee to handle the gaps and audited progress quarterly.
• 04Committee training is the largest single cost. We underestimated it in year 1 and over-corrected in year 2; somewhere in the middle is the right answer.
• 05DORA/CoARA framing gave the reform political cover. Without it, the proposal would have been read as anti-research-impact rather than pro-research-assessment-reform.

What's next

Planned next steps

Year 3 work focuses on extending the CRediT-aware framework to research-grant assessment (where role data could inform peer-review weighting) and to early-career fellowship competitions. The university is also contributing case-study data to the CoARA practitioner network so that the approach can be replicated by other Go8 institutions without each one rediscovering the same lessons.

Q&A with the composite project lead

Composite project-lead Q&A

The questions and answers below are composite — synthesised from interview patterns across multiple real project leads. They are not attributed to a specific real person.

How do you stop the committee from informally counting roles even if the policy says not to?

Training and dashboard design. The dashboard deliberately does not show role counts — it shows contribution patterns visualised as the four NISO functional categories over time, with no numeric totals anywhere on the page. The training reinforces that the visualisation is a starting point for committee discussion, not a verdict. Committee chairs are responsible for interrupting any discussion that drifts into ranking.

What about candidates whose CRediT data is incomplete?

They are not penalised. The narrative statement is the candidate's opportunity to describe contributions that did not get captured in the structured data — and committees are trained to weight the narrative more heavily when CRediT coverage is partial. About one in six candidates relies on this provision; the committee outcomes for that group are statistically indistinguishable from the rest.

How did the academic union react?

Cautiously supportive at the principle stage, then sharper when the pilot data showed the 31% lift for non-bench disciplines — at which point the union asked (reasonably) whether the legacy process had been actively discriminating against those staff. The university's answer was "we think so, and that's why we changed it." The union now sits on the year-2 review committee.

What was the biggest single mistake?

Year 1 communication. We led the launch with the CRediT mechanics rather than the DORA/CoARA principles, and a chunk of the academic community read it as "the university is going to score us by contributor role." We spent six months walking that back. Year 2 launches led with the principles and barely mentioned CRediT in the first email.

Cited CASRAI resources

Internal CASRAI resources referenced

• ›Dictionary: DORA
• ›Dictionary: CoARA
• ›Dictionary: Narrative CV
• ›CRediT role: Conceptualization
• ›CRediT role: Supervision
• ›CRediT role: Funding acquisition
• ›CRediT for institutions
• ›Narrative CV — guidance for authors
• ›NHMRC CRediT statement guidance
• ›ARC CRediT statement guidance
• ›Workshop materials (CC-BY 4.0)
• ›Implementation checklist: responsible assessment