CASRAI Dictionary

Category: Guides & Explainers

Practical how-to guides, templates, checklists, and career pathways for research administrators, authors, and institutional teams.

  • AI Governance UK: What Universities Hire For

    AI governance UK hiring is real but narrow: employers are advertising standalone “AI governance” titles mainly in consultancies and tech firms, while UK universities and research funders are folding AI oversight into existing research-governance, integrity and data-protection roles rather than minting a new job category. Certifications such as IAPP’s AIGP and the ISO/IEC 42001 Lead Auditor credential map to genuinely different parts of that work — one to policy and compliance, the other to formal audit.

    AI governance is the set of policies, controls and accountability structures an organisation uses to ensure AI systems are developed, procured and used safely, lawfully and transparently across their lifecycle.

    What is driving the AI governance UK hiring wave?

    Search interest in AI governance credentials has accelerated sharply. Keyword-demand data tracked into June 2026 shows “ai governance certification” search volume in the UK up 129% year-on-year, and “ai governance job” postings now surface daily on LinkedIn, Indeed and Totaljobs for roles spanning “Director AI Governance” to “Responsible AI Specialist.”

    The trigger is regulatory, not academic. Under the government’s 2023 White Paper AI regulation: a pro-innovation approach, the UK deliberately chose not to pass a single AI statute. Instead, existing regulators — the Information Commissioner’s Office, Ofcom and the Competition and Markets Authority — enforce five cross-sector principles: safety, security and robustness; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress.

    That distributed model pushes the compliance burden into individual organisations, which is exactly the vacuum “AI governance” job titles are appearing to fill. Employers with EU exposure are also hiring against the EU AI Act, whose obligations extend to UK organisations that deploy AI systems into EU markets.

    What are UK research institutions actually hiring for?

    A survey of current academic job boards shows standalone “AI Governance Officer” titles remain rare in UK higher education. What is expanding instead is AI content grafted onto established research-governance and research-integrity posts.

    • The University of Oxford has run a postdoctoral researcher post inside its Oxford AI Governance Initiative, focused on AI and risk research rather than institutional compliance.
    • The University of Bristol advertises a Head of Research Governance role — the University’s lead officer for research regulation, ethics and integrity across human participants, tissue and data, a remit now stretching to cover AI-enabled research methods.
    • The Alan Turing Institute’s AI Ethics and Governance in Practice programme, an eight-workbook resource for project teams, functions as the de facto training reference most UK research-intensive institutions point staff toward, in place of a dedicated internal certification.
    • The Russell Group published sector-wide principles on generative AI in education in January 2025, giving member universities a shared policy baseline rather than each hiring separate AI governance specialists.

    The pattern is consistent: research institutions are governing AI through their existing research-integrity, ethics-committee and data-governance infrastructure, supplemented by sector guidance from the Turing Institute and Russell Group, rather than building a parallel AI governance function from scratch.

    Which certifications map to the job?

    Two credentials dominate current job advertisements, and they are not interchangeable. IAPP’s AIGP is a policy and compliance credential; ISO/IEC 42001 Lead Auditor is a formal management-systems audit qualification built on the international AI management system standard published in 2023.

    Certification Body Format Best fit
    IAPP AIGP International Association of Privacy Professionals 100 multiple-choice questions, 180 minutes Privacy, legal and policy staff who need to interpret AI law and risk, not audit systems
    ISO/IEC 42001 Lead Auditor Accredited training bodies (e.g. PECB, BSI) against the ISO/IEC 42001:2023 standard Multi-day course plus exam Auditors and compliance managers validating a formal AI management system (AIMS)
    Vendor foundational courses (e.g. Securiti) Commercial vendors Short on-demand modules, 2–3 hours Awareness-level onboarding, not a substitute for either credential above

    Neither certification is a licence to practise. Both function as evidence that a candidate has studied a defined body of knowledge — AIGP for law and policy, ISO/IEC 42001 Lead Auditor for management-system audit method — which is why job advertisements almost always list them as “desirable,” not mandatory.

    Genuine career pathway or rebadged compliance role?

    The honest answer is both, depending on sector. In consultancies and large tech employers, “AI governance” is emerging as a distinct, senior, well-paid track — UK job boards currently list Director-level AI governance roles paying well above general compliance-officer rates. In research institutions, it is largely a rebadged extension of research integrity, data protection and ethics-committee work that already existed.

    That does not make it hollow. It means the credential value differs by employer type: a corporate AI governance hire benefits most from IAPP’s AIGP or an ISO/IEC 42001 audit qualification, while a university research-governance officer gains more from Turing Institute and Russell Group sector guidance, since their day job already sits inside an ethics and integrity framework those resources were built for.

    Which is the best AI governance certification?

    There is no single “best” credential; fit depends on function. IAPP’s AIGP suits policy, legal and privacy specialists working across jurisdictions and the EU AI Act. ISO/IEC 42001 Lead Auditor suits professionals who must formally audit an organisation’s AI management system rather than advise on policy.

    Is AI governance certification worth it?

    It is worth it for candidates whose work already touches AI policy, compliance, risk management or privacy, where it demonstrates structured knowledge to employers. It adds little on its own without underlying domain experience, since UK job advertisements consistently list these credentials as desirable evidence rather than a mandatory gate.

    How to become an AI governance professional?

    Most current UK postholders arrive via data protection, legal, risk or research-integrity backgrounds, then add AI-specific knowledge through a credential such as AIGP or ISO/IEC 42001. Direct entry-level “AI governance” hiring remains limited; experience in an adjacent regulated function is the more common route in.

    What skills are needed for AI governance?

    Core skills include risk assessment, regulatory interpretation, bias and fairness evaluation, and stakeholder communication across legal, technical and leadership teams. Employers also expect familiarity with the AI lifecycle and enough technical literacy to question a model’s design without needing to build one.

    What this means for research institutions

    For UK research administrators and institutional leaders, the near-term implication is not to create a new “AI Governance Officer” post by default. It is to audit whether existing research-integrity, data-governance and ethics-committee functions already cover AI risk, and where they do not, to close the gap with targeted training — Turing Institute workbooks or an IAPP AIGP course — rather than an immediate new hire.

    Over the next 12–24 months, expect the corporate and research-sector paths to converge somewhat as funders begin asking institutions to document AI oversight within grant compliance and wider research administration processes. Institutions that get ahead of that by mapping certifications to real duties now, rather than hiring a title, will be better placed when funders start asking for evidence.

  • AI Legislation Tracker: Free Tools Compared for Research Offices

    An AI legislation tracker is a curated, continuously updated resource that monitors the progress of artificial intelligence bills, statutes and regulations across jurisdictions. For research offices, three free options cover the ground a paid GRC subscription would otherwise charge for: the IAPP’s US State AI Governance Legislation Tracker for state-level bills, White & Case’s AI Watch for a global regulatory sweep, and the AI Act Explorer for line-by-line navigation of the EU AI Act. Used together, they give research administrators enough coverage to flag compliance and procurement risk without a dedicated legal-intelligence budget.

    An AI legislation tracker is a legal-intelligence tool — usually maintained by a law firm, professional association, or legislature — that indexes AI-related bills and regulations by jurisdiction, status and topic so non-specialists can monitor change without reading primary legislative text. For a research office, that means catching a new state disclosure requirement or an EU AI Act compliance deadline before it lands in an audit finding.

    Table of contents

    What is an AI legislation tracker, and why does a research office need one?

    Research offices sit at the intersection of three regulatory pressures: institutional AI-use policy, funder terms and conditions, and the AI laws of every jurisdiction in which their institution operates, procures software or receives funding. No single regulator publishes a consolidated feed of all three, which is why legal-intelligence trackers — built by law firms and associations to serve their own clients — have become the de facto public monitoring layer for everyone else.

    Three gaps make this monitoring hard for a research office specifically. First, state-level fragmentation in the US: MultiState.ai reported tracking 1,561 AI-related bills across 45 states in early 2026, and a bill’s status can change between a legislative session’s opening and a grant’s renewal date. Second, phased EU obligations: the AI Act (Regulation (EU) 2024/1689) entered into force on 1 August 2024 but applies in stages — prohibited-practice provisions since 2 February 2025, general-purpose AI model obligations since 2 August 2025, and the bulk of high-risk system obligations from 2 August 2026. Third, procurement-clause drift: institutional purchasing teams increasingly need to know whether a vendor’s AI tool falls under a “high-risk” classification before a contract is signed, not after.

    Comparing the free trackers: IAPP, White & Case AI Watch and the AI Act Explorer

    Each of the three core tools covers a different layer of the regulatory stack. None requires a paid subscription for the baseline tracker view, though firms use them as client-development tools, so update cadence and depth of legal commentary vary.

    Tool Publisher Geographic scope Best use for a research office Cost
    US State AI Governance Legislation Tracker IAPP US state legislatures Flagging new state disclosure/consumer-protection bills affecting AI-assisted research tools Free
    AI Watch: Global Regulatory Tracker White & Case US, EU, UK, China and other core markets Cross-jurisdiction horizon-scanning for institutions with international partners Free
    AI Act Explorer Future of Life Institute (artificialintelligenceact.eu) European Union Locating the exact article/annex governing a specific AI use case before procurement sign-off Free
    Artificial Intelligence Legislation Database National Conference of State Legislatures (NCSL) US state legislatures Official-source cross-check against law-firm trackers, filterable by policy topic Free
    OECD.AI Policy Navigator OECD 80+ countries and international bodies Global baseline for institutions with funders or partners outside the US/EU Free

    Two law-firm trackers rarely agree exactly on bill status, since each applies its own inclusion criteria — the IAPP chart, for example, deliberately excludes government-only AI bills to focus on rules affecting private-sector organisations. A research office should treat the NCSL database as the authoritative cross-check whenever a law-firm tracker and an internal compliance log disagree, since NCSL draws directly from legislative records rather than curated commentary.

    How to monitor AI law without a paid GRC subscription

    A practical monitoring routine needs three components: a jurisdiction list, a check cadence, and an escalation trigger. Map the institution’s actual footprint — states where staff or partner sites are located, countries with active funder relationships, and any EU-based collaborators — against the five tools above, rather than trying to watch all 45+ US states with active bills at once.

    • Set a monthly review of the IAPP tracker and NCSL database for the institution’s home state plus any state with a satellite campus or major subcontractor.
    • Set a quarterly review of White & Case AI Watch for jurisdictions tied to international grant or publishing partners.
    • Check the AI Act Explorer whenever procuring or renewing an AI-enabled research tool from an EU-based or EU-selling vendor, since Article 53 transparency obligations for general-purpose AI providers already apply.
    • Escalate to institutional counsel the moment a tracked bill moves from “introduced” to “enacted” in a jurisdiction on the footprint list — status changes, not initial filings, are the actionable signal.

    This cadence substitutes staff time for the subscription cost of a commercial GRC platform. It will not catch everything a paid legal-intelligence service would, but it closes the gap between “no monitoring” and “monitoring proportionate to institutional risk,” which is the realistic target for most research offices.

    Which AI rules actually affect grant compliance and procurement

    Not every tracked bill is relevant to a research office. The ones that matter cluster into two categories: funder-facing disclosure requirements and vendor/procurement obligations. On the funder side, publishers already require disclosure of generative-AI use in manuscript preparation under guidance from bodies such as ICMJE and COPE — a policy layer that sits alongside, not inside, the legislative trackers above, and one research offices should monitor through authorship policy channels rather than a legislation tracker.

    On the procurement side, the EU AI Act’s general-purpose AI model obligations — applicable since 2 August 2025 — require providers to maintain technical documentation and, for systemic-risk models, conduct model evaluations; institutions procuring AI research tools from in-scope vendors should expect updated contract terms reflecting this. Separately, under Article 57 of Regulation (EU) 2024/1689, each EU member state must establish at least one national AI regulatory sandbox operational by 2 August 2026 — a detail the AI Act Explorer surfaces clearly but general news coverage rarely mentions, and one that matters to institutions running EU-based pilot deployments of AI research tools.

    In the US, state consumer-protection style AI bills increasingly impose obligations on “deployers” as well as developers — meaning an institution using a third-party AI tool, not just the vendor that built it, can carry compliance obligations. This is the single most consequential fact a research office should extract from the state trackers: deployer obligations mean procurement due diligence, not just vendor selection, is now a compliance function.

    Common questions research administrators ask

    Are there any regulations on AI?

    Yes. There is no comprehensive federal AI statute in the United States, but individual US states have enacted targeted laws, the European Union’s AI Act (Regulation (EU) 2024/1689) is in force with phased obligations through 2027, and dozens of other jurisdictions maintain sector-specific or principles-based AI policy frameworks tracked by the OECD.

    Does Europe have AI regulations?

    Yes. The EU AI Act is the first comprehensive AI-specific legal framework, entering into force on 1 August 2024. Prohibited-practice rules applied from February 2025, general-purpose AI model obligations from August 2025, and most high-risk system requirements apply from August 2026 onward.

    Where are the AI regulations?

    AI rules are distributed across national statutes, EU regulation, and US state legislatures rather than one source — which is precisely why trackers such as IAPP’s state chart, White & Case’s AI Watch, and the AI Act Explorer exist: each consolidates one layer of a fragmented, multi-jurisdiction landscape into a single reference point.

    The regulatory landscape a research office must monitor will keep expanding rather than consolidating: more US states are expected to move bills from “introduced” to “enacted” through 2026 and 2027, and the EU AI Act’s remaining compliance deadlines run to August 2027. A footprint-mapped, tiered-cadence monitoring routine built on these five free trackers is a realistic, sustainable substitute for a paid GRC subscription — provided it is reviewed and re-scoped as the institution’s own AI use, partnerships and procurement expand.

  • Types of Material Transfer Agreement Guide: One-Way, Two-Way and Commercial-Use MTAs

    Types of material transfer agreement fall into two overlapping categories: direction of transfer (one-way versus two-way) and nature of the parties (academic/non-profit versus commercial-use). Choosing the wrong category adds weeks of unnecessary negotiation to a simple exchange, or leaves an institution exposed on intellectual property and liability in a complex one. A material transfer agreement (MTA) is a contract that governs the transfer of tangible research materials — cell lines, plasmids, reagents, antibodies, animal models, or genetic constructs — between a provider and a recipient, setting terms for permitted use, publication, and downstream rights.

    What is a material transfer agreement, and why does type matter?

    A material transfer agreement is the legal instrument accompanying the physical movement of a research material between organisations. It records who owns the original material, what the recipient may do with it, who owns derivatives the recipient creates, and what happens if the material leads to a publication or invention.

    Institutions do not use one template for every transfer. The right MTA type depends on two independent variables: direction of the exchange, and the nature of the parties — non-profit or commercial. Mismatching the template to either variable is a common cause of avoidable negotiation delay.

    One-way vs two-way MTAs: what is the difference?

    A one-way MTA covers a single direction of transfer: one provider sends material to one recipient, who accepts the provider’s terms. Most institutional MTAs are one-way and are further split into two sub-types depending on which side of the transaction the institution sits on.

    • Incoming MTA — the institution is the recipient. The priority is understanding and accepting the provider’s restrictions: permitted research use, any prohibition on commercial use, and publication or embargo terms.
    • Outgoing MTA — the institution is the provider. The priority shifts to protecting the institution’s own intellectual property, limiting liability for the material’s performance, and controlling further distribution by the recipient.

    A two-way (or reciprocal) MTA is used when both parties send materials to each other, typically in an active collaboration where each lab holds a resource the other needs. Rather than negotiate two separate one-way agreements, the parties combine both transfers into a single reciprocal agreement with symmetric obligations. This is administratively efficient but requires both sides to specify their respective materials and restrictions with equal precision — asymmetric two-way MTAs are a frequent source of later disputes over derivative rights.

    MTA type Direction Typical parties Primary administrative focus
    Incoming (one-way) Provider → institution Academic-to-academic or vendor-to-academic Compliance with provider’s use and publication restrictions
    Outgoing (one-way) Institution → external recipient Academic-to-academic or academic-to-industry IP retention, liability limitation, distribution control
    Two-way / reciprocal Bidirectional Collaborating academic labs Symmetric terms for both transferred materials
    Academic/non-profit Either direction Non-profit-to-non-profit Non-commercial research use only; minimal negotiation
    Commercial-use Either direction At least one for-profit party IP ownership, licensing options, publication delay, indemnification

    Academic/non-profit vs commercial-use MTAs: how do the terms differ?

    The second axis of classification is the nature of the parties, and it changes negotiation complexity more than direction does. A biological material transfer agreement between two non-profit universities is usually a light-touch document; the same transfer involving a commercial partner routinely takes months longer to close.

    Academic and non-profit MTAs exist to facilitate open scientific exchange. The US National Institutes of Health (NIH) has stated that unique research resources arising from NIH-funded work should be shared on terms no more restrictive than its own model agreements, because repeated case-by-case negotiation between non-profits delays the point at which a research tool reaches the laboratory bench. These agreements typically restrict use to non-commercial research, require no royalty, and rarely need individual negotiation once a standard template is adopted.

    Commercial-use MTAs — where a for-profit company is provider, recipient, or both — carry additional, negotiated terms that a non-profit template does not anticipate:

    • Intellectual property rights over inventions made using the transferred material, including whether the provider retains an option to license.
    • Publication rights, including any pre-publication review period the commercial party can invoke to protect confidential information.
    • Scope of permitted use, distinguishing internal research from development toward a commercial product, which may trigger the need for a separate licence agreement.
    • Indemnification and liability allocation, which non-profit-to-non-profit templates typically waive or cap at a nominal level.

    A material transfer agreement policy should specify, in advance, which of these terms are non-negotiable defaults and which require case-by-case legal review — allowing routine academic MTAs to clear in days, reserving negotiation capacity for the commercial-use cases where it is genuinely needed.

    Which standard templates exist, and how do you choose the right one?

    Four template families cover the great majority of transfers, and matching a transfer to the correct family is the fastest route to a signed agreement.

    • NIH Simple Letter Agreement (SLA), published in 1995 by the NIH, is a one-page model for low-risk, non-commercial transfers of routine research materials between non-profit institutions.
    • Uniform Biological Material Transfer Agreement (UBMTA), also introduced in 1995, is the master agreement signed once by an institution; individual transfers under it use a short “Implementing Letter” rather than a fresh negotiation, and it is the standard route for a two-way material transfer agreement between two UBMTA-signatory non-profits.
    • AUTM Model MTAs extend the UBMTA framework to materials — and non-US institutions — outside the UBMTA’s original definition of biological material; a 2011 AUTM member survey found low adoption of standard templates was itself a major cause of unnecessary delay, which the toolkit’s decision tree was built to address.
    • FAO Standard Material Transfer Agreement (SMTA), adopted by the Governing Body of the International Treaty on Plant Genetic Resources for Food and Agriculture in 2006, is a distinct global instrument governing plant genetic resources held in the Treaty’s Multilateral System, with mandatory benefit-sharing terms that have no equivalent in the UBMTA or NIH templates.

    To choose: start by identifying direction (one-way or two-way) and party type (non-profit or commercial). If both parties are non-profit and the material is biological, default to the UBMTA or SLA. If a commercial party is involved, route the transfer to legal review rather than a standard template. If the material is a plant genetic resource in the Multilateral System, the FAO SMTA applies regardless of the parties’ institutional type — a distinction general MTA guidance frequently omits.

    Frequently asked questions

    What is a material transfer agreement?

    A material transfer agreement is a contract governing the transfer of tangible research materials — such as cell lines, plasmids, reagents, or animal models — between a provider and a recipient. It sets terms for permitted use, ownership of derivatives, publication rights, and liability, and is negotiated before the physical material is shipped.

    What is the difference between an MTA and an NDA?

    An NDA (non-disclosure agreement) protects confidential information exchanged between parties, while an MTA governs the physical transfer and permitted use of a tangible material. The two are often signed together — an NDA may protect data about the material, while the MTA governs the material itself — but neither substitutes for the other.

    What is the standard material transfer agreement?

    The Standard Material Transfer Agreement (SMTA) is the FAO instrument used specifically for plant genetic resources held under the International Treaty on Plant Genetic Resources for Food and Agriculture’s Multilateral System. It is distinct from the UBMTA used for general biological materials and includes mandatory benefit-sharing obligations tied to the Treaty.

    Implications and outlook for research administrators

    Institutions that map every incoming request against this taxonomy before drafting — direction first, party type second, then template — process the majority of routine biological material transfer agreement requests without individual legal review. That triage is what standard templates like the UBMTA and NIH SLA were designed to enable, and what a written material transfer agreement policy should formalise as institutional default practice.

    The remaining minority — commercial-use transfers, cross-border plant genetic material, and asymmetric two-way exchanges — is where administrative time should concentrate, since these are the categories where the wrong template creates the greatest downstream IP and compliance risk. As collaborations increasingly span academic, industry, and international-treaty jurisdictions at once, classifying a transfer correctly at intake, rather than after a dispute arises, remains the most effective control a research administration office can apply.

  • Cost Sharing in Grants: Mandatory vs Voluntary

    Cost sharing on a grant is the portion of a project’s true cost that the sponsor does not pay, covered instead by the recipient institution, a third party, or in-kind contributions. It can be mandatory (a condition of the award, set out in the funding announcement) or voluntary (offered by the applicant and not required). A growing number of funders — most notably the US National Science Foundation — have moved away from requiring or even rewarding voluntary cost sharing, on the grounds that it disadvantages under-resourced institutions and adds compliance burden without improving research quality.

    What is cost sharing in a grant budget?

    Cost sharing (also called matching) is the share of a sponsored project’s total cost that is not reimbursed by the funding agency. It is contributed instead by the recipient institution, a subrecipient, or a third-party collaborator, either as cash or as an in-kind resource such as donated staff time, waived facilities-and-administration (F&A) costs, equipment, or space.

    Under the US federal Uniform Guidance (2 CFR Part 200, §200.306), cost sharing and matching are defined as the portion of project costs “not borne by the Federal Government.” Any contribution counted this way must be verifiable from the recipient’s own records, not double-counted against another federally funded project, and necessary and reasonable for the project. This is the baseline definition US sponsored programs offices apply when reviewing a proposal’s grant budget justification.

    Mandatory vs voluntary cost sharing: what’s the difference?

    The distinction between mandatory and voluntary cost sharing determines whether a commitment is legally enforceable. Mandatory cost sharing is imposed by the sponsor and stated explicitly in the funding opportunity; without it, the proposal is ineligible. Voluntary cost sharing is offered by the applicant even though the sponsor did not require it — and once quantified in a funded federal proposal, it becomes just as binding and auditable as a mandatory commitment.

    Type Who requires it Reporting obligation once awarded
    Mandatory cost sharing Sponsor, stated in the solicitation Documented, tracked and reported to the sponsor for the life of the award
    Voluntary committed cost sharing Applicant, quantified in the proposal budget or narrative Treated as binding and auditable once the award is made, on federal awards
    Voluntary uncommitted cost sharing Applicant, contributed after award but never quantified in the proposal Not tracked or reported to the sponsor

    The trap is the second row. A PI who writes “the PI will devote 20% effort at no cost to the sponsor” creates a quantified, reportable commitment — even though the sponsor never asked for one. This is why sponsored programs offices train investigators to use non-quantified language (“will provide expert consultation, as needed”) whenever cost sharing is not actually required.

    Why are funders moving away from mandatory cost sharing?

    The clearest example is the National Science Foundation. Following its own Cost Sharing Task Force review, NSF’s Proposal & Award Policies and Procedures Guide (PAPPG) states that cost sharing is not required except where a specific program solicitation invokes a statutory requirement, and that reviewers may not factor voluntary committed cost sharing into merit review. NSF’s rationale was that cost sharing had become a competitive filter favouring wealthier institutions rather than an indicator of project quality.

    Three arguments recur across funder policy statements and research-administration literature on this reform:

    • Equity between institutions. A fixed percentage match is far harder for a community college or small non-profit to absorb than for a well-endowed research university — skewing award patterns by wealth rather than merit.
    • Administrative burden. Cost sharing must be certified through effort reporting and reconciled at closeout; auditors treat under-delivered cost share as a disallowed cost, risking clawback.
    • Review integrity. A visible voluntary contribution can bias scoring toward applicants who over-promise resources they may struggle to deliver.

    Cost sharing has not disappeared. It remains common — and often mandatory — on infrastructure and construction grants, public-private partnership schemes, and Department of Justice (DOJ) Office of Justice Programs awards, where the required match varies by programme and is set out in each solicitation’s guide sheet.

    How do UK and EU funders structure cost sharing?

    US-centric discussions of cost sharing rarely mention that the UK and EU systems build an equivalent principle directly into their core funding formulas, rather than treating it as a discretionary add-on.

    UK Research and Innovation (UKRI) funds most Research Council grants at up to 80% of a project’s Full Economic Cost (fEC), calculated via the sector’s Transparent Approach to Costing (TRAC) methodology. The host university funds the remaining 20% itself — a structural, near-universal form of mandatory cost sharing built into the grant terms, not a clause institutions can negotiate away project by project.

    Under Horizon Europe, reimbursement rates differ by action type rather than a flat match: Research and Innovation Actions (RIA) are typically funded at 100% of eligible direct costs, while Innovation Actions (IA) are reimbursed at 100% for non-profit entities but only 70% for profit-making organisations — meaning commercial participants effectively cost-share 30% of their own costs as a condition of taking part.

    This is a genuinely different model from the US project-by-project mandatory/voluntary framework. A US-style “voluntary cost sharing is discouraged” mindset does not transfer cleanly to a UKRI fEC or Horizon Europe budget, where the shortfall is baked into the reimbursement rate itself, not offered or declined proposal by proposal.

    Common questions about cost sharing

    What is cost share on a grant?

    Cost share on a grant is the share of a sponsored project’s total cost that the funding agency does not pay, covered instead by the recipient institution, a subrecipient, or a third party. It can be cash (salary, direct funding) or in-kind (donated time, waived facilities-and-administration costs, equipment) and must be verifiable, allowable, and incurred within the project period.

    What are the three types of cost sharing?

    The three recognised categories are mandatory, voluntary committed, and voluntary uncommitted cost sharing. Mandatory is required by the sponsor as a condition of funding; voluntary committed is offered by the applicant and becomes binding once awarded; voluntary uncommitted is contributed after the award but never quantified in the proposal, so it carries no reporting obligation.

    What is a cost sharing requirement?

    A cost sharing requirement is a condition, stated explicitly in a funding announcement, that obliges applicants to contribute a defined percentage or dollar amount of project costs from non-sponsor sources. Requirements vary widely by programme — from a flat percentage match to a formula tied to Modified Total Direct Costs — and must be documented and reported to the sponsor if the proposal is funded.

    How does cost sharing work?

    Cost sharing works by allocating a defined portion of a project’s budget to the recipient rather than the sponsor, expressed either as a percentage of total cost or as a match ratio (for example, 1:1). Once quantified in a funded proposal’s grant budget justification, the commitment must be tracked through effort reporting or financial records and reconciled at the project’s grant closeout report.

    Implications for institutional budget commitments

    For sponsored programs offices, the decline of mandatory cost sharing at agencies like NSF does not reduce the compliance workload — it relocates it. Institutions must train investigators to recognise when descriptive language in a proposal narrative inadvertently creates a quantified, auditable commitment, distinct from genuinely required match on programmes (DOJ, construction grants, many state and foundation awards) where cost sharing is still mandatory and enforced at closeout.

    Under-delivered cost sharing is treated by auditors as a disallowed cost, triggering a proportional reduction in drawable funds regardless of whether the shortfall was mandatory or voluntary. A “decline all voluntary cost share” policy calibrated to NSF norms misfires against a UKRI fEC award, where the 20% institutional contribution is structural, not optional. A no-cost extension can buy time to complete an outstanding commitment, but it does not waive the obligation — the shortfall must still be resolved before the award can close.

    The direction of travel across US federal science funders is towards evaluating proposals on merit rather than an applicant’s ability to co-invest. Institutions that update proposal-review checklists and budget-justification templates accordingly — while keeping separate, funder-specific guidance for programmes where cost sharing remains mandatory or structural — will reduce both audit exposure and the administrative overhead cost sharing has historically imposed.

  • Office of Grants Management vs Program Offices

    The Office of Grants Management is the part of a federal department — at the Department of Health and Human Services (HHS), the Office of Grants (OG), under the Assistant Secretary for Financial Resources (ASFR) — that sets department-wide policy, issues the Notice of Award, and enforces financial and compliance rules across every award. Individual program offices, by contrast, judge scientific and programmatic merit within their own subject area. Grantee institutions deal with both, for different reasons, throughout the life of an award.

    In one sentence: the Office of Grants Management is the administrative and financial authority that governs how federal grant funds are awarded, monitored, and closed out, while program offices decide what gets funded and why. HHS is the largest federal grant-making agency in the United States, and the distinction between its central grants office and its dozens of program offices is one of the most consistently misunderstood parts of the federal award lifecycle for institutional research administrators.

    What Does the Office of Grants Oversee?

    The HHS Office of Grants formulates department-wide grants policy and oversees its implementation across every HHS operating division. It does not decide which research or service proposals get funded; it decides how the resulting awards are administered, financed, and audited.

    According to a December 2023 U.S. Government Accountability Office review (GAO-24-106008), the Office of Grants “provides department-wide leadership on grants” and serves several government-wide roles beyond HHS itself. In January 2021, the Office of Management and Budget designated HHS to house the government-wide Grants Quality Services Management Office (Grants QSMO), which supports other federal agencies in adopting shared, standardised grants-management systems.

    • Developing and issuing department-wide grants policy, including the HHS Grants Policy Statement (GPS), last revised October 2024
    • Applying the Uniform Administrative Requirements, Cost Principles, and Audit Requirements codified at 45 CFR Part 75
    • Issuing the official Notice of Award (NoA) that legally obligates federal funds
    • Overseeing financial reporting, audit resolution, and closeout across all HHS awards
    • Running the Grants QSMO Marketplace, launched September 2022, which offers other agencies shared grants-management and payment platforms

    The scale is substantial: GAO reports the federal government distributed approximately $1.2 trillion in grants in fiscal year 2022 — roughly 19 percent of total federal spending, and over $400 billion more than FY 2019. HHS accounts for the largest share of any single federal grant-making agency.

    How Does the Office of Grants Differ From Program Offices?

    The core distinction is “how” versus “what.” The Office of Grants governs the administrative, financial, and regulatory mechanics of an award — eligibility of costs, reporting deadlines, audit requirements, closeout. Program offices — the National Institutes of Health institutes, the Health Resources and Services Administration bureaus, the Administration for Children and Families divisions, and similar bodies — set programmatic priorities, write the Funding Opportunity Announcement’s scientific or service requirements, and judge whether a grantee is meeting technical objectives.

    Function Office of Grants (Grants Management) Program Office
    Primary question answered Is this cost allowable and compliant? Is this science/service meeting its goals?
    Issues Notice of Award Yes No
    Sets scientific/programmatic scope No Yes
    Reviews financial/progress reports Financial reports, audit findings Technical/programmatic progress reports
    Governs closeout mechanics Yes Provides final technical sign-off
    Typical grantee contact Grants Management Specialist Project Officer / Program Officer

    Grantee institutions need two working relationships per award: a technical relationship with the program office’s project officer, and an administrative relationship with the grants management specialist. Sending a budget modification to a project officer instead of the specialist is a routine, avoidable source of delay.

    Where Does OASH’s Own Grants Function Fit In?

    A frequent source of confusion is the phrase “OASH Office of Grants Management.” The Office of the Assistant Secretary for Health (OASH) operates its own grants and cooperative agreements function, published at health.gov/grants, covering programmes such as Title X family planning and adolescent health initiatives that OASH itself administers.

    This is not a separate, competing authority to the department-wide Office of Grants under ASFR. OASH’s grants activity operates within the HHS-wide policy framework — the same Grants Policy Statement and 45 CFR Part 75 requirements apply — but OASH runs its own competitions, issues its own Funding Opportunity Announcements, and assigns its own grants management staff for the awards it makes. A grantee dealing with OASH therefore interacts with an OASH-specific contact who still answers to department-wide policy. This layered structure — one policy authority, multiple operating-division grants functions beneath it — is largely absent from generic explainer pages, which describe either the federal picture or a single state office, not HHS’s two-tier structure.

    Every accredited research institution maintains an institutional counterpart to the federal grants office: the sponsored programs office (sometimes called Office of Research Administration or Grants and Contracts). Its function mirrors the Office of Grants Management’s role, but from the recipient side.

    The sponsored programs office is the institution’s authorised signatory for award acceptance, its central point for compliance with 45 CFR Part 75 and OMB Uniform Guidance (2 CFR Part 200), and its liaison to the HHS grants management specialist rather than the program office’s project officer. Bodies such as the National Council of University Research Administrators (NCURA) and INORMS document this division of labour consistently: principal investigators own the science; the sponsored programs office owns the compliance interface. For a broader view of this interface within institutional research administration practice, see CASRAI’s research administration resources.

    What Happens at Closeout and With Cost Sharing?

    Two compliance touchpoints sit squarely with the Office of Grants Management rather than the program office: closeout and cost sharing.

    A grant closeout report is the set of final documents — the Federal Financial Report, the final progress report, and any property disposition report — that a recipient must submit once the period of performance ends. Under the Uniform Guidance framework that 45 CFR Part 75 incorporates for HHS awards, these reports are due within a fixed post-performance window, after which unspent funds are deobligated and the award is formally closed by the grants management office, not the program office.

    Cost sharing (sometimes called matching) is the portion of total project cost that the recipient institution — not the federal award — commits to fund, whether required by statute or offered voluntarily in the proposal. The Office of Grants Management verifies documented cost-sharing commitments were actually met before an award can close; a shortfall found at closeout is a grants-management finding, even when the project was scientifically successful.

    Frequently Asked Questions

    What does a grants manager do?

    A grants manager at a federal Office of Grants administers the financial and compliance lifecycle of an award: reviewing budgets, issuing the Notice of Award, monitoring reporting compliance, and processing closeout. This role is distinct from a project officer, who judges technical or scientific performance.

    What is the grant management function?

    The grant management function is the administrative infrastructure — policy, systems, and staff — that a funding agency uses to award, monitor, and close federal financial assistance. At HHS this sits with the Office of Grants under ASFR, applying the Grants Policy Statement and 45 CFR Part 75 across every operating division.

    What are common mistakes in grant management?

    The most common mistakes are routing compliance questions to a project officer instead of the grants management specialist, missing the fixed closeout deadline, and failing to document cost-sharing commitments contemporaneously rather than reconstructing them at award end.

    What are grant management services?

    Grant management services cover pre-award risk assessment, Notice of Award issuance, ongoing compliance monitoring, and closeout processing. HHS centralises much of this through its Recipient Data Insights tool, which automates pre-award risk scoring department-wide.

    Implications and Outlook

    For institutions holding HHS awards, the practical takeaway is structural, not procedural: two distinct offices govern every award, and each has authority the other cannot override. A program office cannot waive a 45 CFR Part 75 cost-allowability rule, and the Office of Grants Management cannot override a program office’s technical judgement on scientific merit.

    HHS’s modernisation record shows this split hardening rather than dissolving. The ReInvent Grants Management initiative (2017–2020) and the September 2022 Grants QSMO Marketplace launch both centralised administrative infrastructure further, while leaving programmatic decisions with the operating divisions. Institutions that route compliance questions to their sponsored programs office, and technical questions to the program office, will keep seeing faster processing than those that conflate the two.

  • Leiden Manifesto Checklist for Research Offices

    The Leiden Manifesto for Research Metrics sets out ten principles, published as a comment in Nature in 2015, for the responsible use of quantitative indicators in research evaluation. Research offices can convert each principle into a direct audit question, testing whether KPI dashboards, promotion criteria and grant-review rubrics rely on a single metric, ignore field norms, or substitute for qualitative judgement.

    The Leiden Manifesto for Research Metrics is a ten-principle framework for the responsible use of bibliometric and other quantitative indicators in evaluating research, published by Diana Hicks, Paul Wouters, Ludo Waltman, Sarah de Rijcke and Ismael Rafols in Nature on 22 April 2015. It was formulated at the 19th International Conference on Science and Technology Indicators, held in Leiden, the Netherlands, in September 2014, and has since been cited more than 4,000 times, according to Google Scholar’s tracking of the original paper.

    What is the Leiden Manifesto for Research Metrics?

    The Leiden Manifesto is a response to what its authors called “impact-factor obsession” — the tendency of universities, funders and promotion committees to substitute a single number for expert judgement. It does not ban metrics. It requires that quantitative indicators support, rather than replace, informed peer assessment of research quality.

    The manifesto’s home institution is the Centre for Science and Technology Studies (CWTS) at Leiden University, where co-author Paul Wouters served as director. CWTS also produces the CWTS Leiden Ranking, a separate bibliometrics-based university ranking — a distinction research offices should not conflate when citing the source.

    What are the ten principles of the Leiden Manifesto?

    Each principle addresses a specific failure mode observed in metric-driven research assessment. The table below states each principle exactly as published, alongside the practical audit question a research office should ask of its own KPI or promotion framework.

    # Principle (Hicks et al., 2015) Audit question for your office
    1 Quantitative evaluation should support qualitative, expert assessment Does any committee decision rest on a metric alone, with no narrative peer input?
    2 Measure performance against the research missions of the institution, group or researcher Are KPIs generic, or tailored to the unit’s stated mission (teaching-intensive, applied, translational)?
    3 Protect excellence in locally relevant research Does the framework penalise work published in non-English or regionally focused outlets?
    4 Keep data collection and analytical processes open, transparent and simple Can an academic reproduce their own score from publicly documented methodology?
    5 Allow those evaluated to verify data and analysis Is there a formal, timely route to challenge or correct metric data before a decision is made?
    6 Account for variation by field in publication and citation practices Are raw citation counts compared across disciplines without field normalisation?
    7 Base assessment of individual researchers on a qualitative judgement of their portfolio Does promotion criteria require a portfolio narrative, or just an h-index threshold?
    8 Avoid misplaced concreteness and false precision Are decimal-point differences in impact factor or citation rate treated as meaningful?
    9 Recognise the systemic effects of assessment and indicators Has the office assessed whether its KPIs create incentives to game submission counts or venues?
    10 Scrutinise indicators regularly and update them Is there a scheduled review cycle for the KPI framework itself, not just for scores against it?

    How can a research office audit its KPI and promotion framework against it?

    Running the manifesto as a live audit tool means working through each principle against real artefacts: the appraisal form, the promotion rubric, and the departmental dashboard.

    1. Mark every clause in the promotion/tenure criteria naming a specific metric (impact factor, h-index, citation count).
    2. Check each marked clause has a qualitative narrative requirement alongside it (Principles 1 and 7).
    3. Confirm KPI targets are set per unit mission, not copied institution-wide (Principle 2).
    4. Check non-English-language or applied outputs score on the same scale as high-impact-journal outputs (Principle 3).
    5. Verify each dashboard metric’s data source and calculation method is documented and accessible (Principles 4 and 5).
    6. Confirm citation indicators are field-normalised, not raw counts compared across disciplines (Principle 6).
    7. Look for false precision — ranking staff by two-decimal citation averages (Principle 8).
    8. Ask whether the KPI framework has driven any unintended behaviour, such as salami-slicing publications or discouraging risky research (Principle 9).
    9. Set a fixed review date for the framework itself, independent of individual appraisal cycles (Principle 10).

    A framework that fails more than two or three of these checks is not aligned with the manifesto, regardless of how sophisticated its dashboard software looks. The most common failure in practice is Principle 6: comparing raw citation counts across a mathematics department and a cell biology department, where top-ranked mathematics journals carry impact factors around 3 while top-ranked cell biology journals carry impact factors around 30 — a field-scale gap the manifesto’s authors cite directly as evidence that uncorrected cross-field comparison is meaningless.

    How does the Leiden Manifesto compare with DORA and CoARA?

    The Leiden Manifesto did not appear in isolation. The 2013 San Francisco Declaration on Research Assessment (DORA) preceded it, while the Coalition for Advancing Research Assessment (CoARA) has since built a sector-wide agreement on reforming assessment practice. Research offices are frequently asked which one to adopt.

    Framework Published Format Primary focus
    Leiden Manifesto 22 April 2015 (Nature comment) 10 principles Correct use of quantitative indicators across disciplines and settings
    DORA 2013 (San Francisco Declaration) General recommendations + signatory pledge Eliminating journal impact factor as a proxy for article or researcher quality
    CoARA 2022 (Agreement on Reforming Research Assessment) Institutional commitment agreement Sector-wide reform of hiring, promotion and funding assessment criteria

    DORA has been signed by more than 27,000 individuals and organisations, according to DORA’s own published tally as of March 2026, making it the higher-profile pledge. But when Loughborough University’s LIS-Bibliometrics committee chose a framework for its own policy in 2018, policy manager Elizabeth Gadd selected the Leiden Manifesto because it takes a “broader approach to the responsible use of all bibliometrics across a range of disciplines and settings” — not only journal-level metrics. Elsevier separately announced on 14 July 2020 that it would use the manifesto’s principles to guide its CiteScore methodology.

    In the UK, the independently commissioned Metric Tide review (2015), led by James Wilsdon for the then Higher Education Funding Council for England, reached compatible conclusions and recommended metrics support, not replace, peer review within the research administration processes underpinning the Research Excellence Framework. A research office building a REF-adjacent KPI policy should treat the two as aligned, not competing, references.

    Common questions and what comes next for research offices

    Who wrote the Leiden Manifesto for Research Metrics?

    The manifesto was written by Diana Hicks, professor of public policy at Georgia Institute of Technology, and Paul Wouters, then director of CWTS at Leiden University, together with co-authors Ludo Waltman, Sarah de Rijcke and Ismael Rafols. It was published as a comment in Nature, volume 520, on 22 April 2015.

    Does the Leiden Manifesto ban the use of bibliometrics tools?

    No. The manifesto does not prohibit bibliometrics tools such as Web of Science, Scopus or Dimensions. It requires that any output from these tools — citation counts, h-indices, journal metrics — be interpreted alongside qualitative expert review and adjusted for field-specific citation norms before it informs a decision.

    Why does the importance of bibliometrics remain contested?

    Bibliometrics matter because they scale evaluation across thousands of researchers where individual peer review is impractical. The contested part is misuse: treating a single indicator as an objective proxy for quality, rather than one input alongside portfolio review, mission fit and field context, as the manifesto’s ten principles specify.

    How often should a research office review its KPI framework under the manifesto?

    Principle 10 requires indicators to be “scrutinised regularly and updated,” but sets no fixed interval. Good institutional practice, reflected in library and research-office guidance built on the manifesto, is an annual technical review of data sources plus a full policy review on the same three-to-five-year cycle as promotion-criteria revisions.

    The Leiden Manifesto’s ten principles were written as durable evaluation ethics, not a one-time compliance exercise. As institutions layer AI-assisted analytics, altmetrics and funder-mandated open-data reporting onto existing KPI frameworks, the manifesto’s core requirement — that quantitative evaluation support, not replace, expert judgement — becomes harder to satisfy by default and more important to audit deliberately. Research offices that build the checklist above into their annual promotion-criteria review cycle, rather than treating the manifesto as background reading, are the ones actually applying it.

  • OpenAlex API: Building a Metrics Dashboard

    The OpenAlex API is a free, fully open REST interface to a catalogue of hundreds of millions of scholarly works, authors, institutions and funders, and it is the most practical data source for building an in-house institutional research metrics dashboard without a subscription. Query the /works endpoint with an institution filter, aggregate with group_by, and you have publication counts, open-access share and citation-percentile data in a single JSON response.

    OpenAlex is an open, CC0-licensed catalogue of the global research system — works, authors, institutions, sources, funders and topics — built and maintained by the non-profit OurResearch as a successor to the discontinued Microsoft Academic Graph. Because every record and the API itself are free to query, research offices can build metrics dashboards without licensing a commercial bibliometrics platform, provided they understand the filter syntax, pagination limits and the metric gaps this guide covers.

    What is the OpenAlex API and what does it cover?

    The OpenAlex API exposes entity endpoints — Works, Authors, Institutions, Sources, Topics, Funders and Awards — each accessed at https://api.openalex.org/{entity}. Every entity supports four operations: list, get (by ID), filter, and group_by (server-side aggregation), which together are the building blocks of a dashboard.

    Each entity carries a persistent OpenAlex ID and, for institutions, a cross-walked ROR identifier — the Research Organization Registry ID also used by ORCID, Crossref and DataCite. Filtering on an institution’s ROR-linked OpenAlex ID, rather than a free-text name match, is what keeps a dashboard’s institutional attribution stable as an organisation’s name or subsidiary structure changes.

    Entity endpoint Dashboard use case Example filter
    /works Publication counts, open-access share, citation percentiles authorships.institutions.id
    /authors Researcher productivity, h-index-style summary stats affiliations.institution.id
    /institutions Peer benchmarking, collaboration networks ror
    /topics Subject-area concentration and trend detection works_count

    How do you query the Works endpoint for institutional metrics?

    Every institution-level query starts with the authorships.institutions.id filter set to the institution’s OpenAlex ID, which you resolve once via /institutions?filter=ror:https://ror.org/{your-ror-id}. From there, combine filters with commas (AND logic) and pipes (OR logic), and add group_by to turn a list query into an aggregation query in one request — no client-side loop required.

    • Publication trend: /works?filter=authorships.institutions.id:I123...,publication_year:2020-2026&group_by=publication_year
    • Open-access share: add &group_by=oa_status to the same filter to split output into gold, green, hybrid, bronze and closed counts.
    • Field distribution: &group_by=primary_topic.field.id reveals subject concentration across an institution’s output.
    • Collaboration mapping: &group_by=authorships.institutions.id returns co-publishing partner institutions ranked by shared-work count.

    Use the select parameter to strip unused fields from large responses, and switch from offset-based page/per_page pagination to cursor pagination once a query’s meta.count exceeds roughly 10,000 results — offset pagination is capped and will silently stop returning new pages beyond that depth.

    How do you approximate field-weighted citation impact with OpenAlex data?

    Field-weighted citation impact (FWCI) is a proprietary metric popularised by Elsevier’s SciVal and Scopus products, calculated by comparing a work’s citations to the average for same-year, same-subject, same-document-type publications; OpenAlex does not expose a field literally called “FWCI”, and no open API replicates the Scopus subject-classification baseline it is normalised against.

    OpenAlex’s nearest open equivalent is the cited_by_percentile_year object returned on every work record, which gives a min/max percentile rank of that work’s citation count against all works of the same publication year and type. Aggregating this field across an institution’s output — for example, the share of works in the top decile (percentile ≥ 90) per year — produces a transparent, reproducible citation-impact proxy that a dashboard can compute without a commercial licence, though it is not interchangeable with SciVal’s FWCI for benchmarking against institutions that report the Scopus figure.

    For most dashboards the honest approach is to present both: raw citation counts (context-dependent, not comparable across fields) and the percentile-year proxy (comparable within OpenAlex’s corpus), clearly labelled as distinct from any vendor-reported FWCI value cited in external reports.

    What are the authentication, rate-limit and pricing rules?

    OpenAlex’s underlying dataset, website and API are free and the data is CC0-licensed, so no purchase is required to query or redistribute results. Every request should still include a contact identifier — either a mailto query parameter with your email address or a registered api_key — to enter the “polite pool”, which OurResearch prioritises over anonymous traffic for faster, more consistent response times.

    Requests without a mailto parameter or API key are routed to a slower, lower-priority pool and are more likely to be throttled during peak load; this single parameter is the most common fix for intermittent 429 or timeout errors reported by developers building batch-harvesting scripts. Dashboard builders scheduling nightly refresh jobs should always set mailto or an API key rather than relying on the anonymous pool.

    Common developer questions

    Is the OpenAlex API free?

    Yes. OpenAlex is free to query, and the underlying data is licensed under CC0, meaning it can be reused and redistributed without royalties. Registering an email via the mailto parameter or an API key gives access to the faster “polite pool” but does not change the underlying no-cost model.

    Does OpenAlex have an API for institutional data?

    Yes. The Institutions endpoint returns disambiguated organisation records cross-walked to ROR identifiers, and the Works endpoint accepts an authorships.institutions.id filter, which is the standard way to scope any query to a single institution’s publication output for a dashboard.

    What is OpenAlex used for in research administration?

    Research offices use OpenAlex to track publication trends, open-access compliance, collaboration networks and topic concentration without paying for a commercial bibliometrics subscription. Its open licence also makes it suitable for public-facing institutional reporting, since results can be republished without redistribution restrictions.

    Implications for institutional research offices

    A dashboard built directly on the OpenAlex API gives research administration teams a free, auditable alternative to proprietary bibliometrics tools for routine reporting — publication counts, open-access compliance tracking and collaboration mapping — while reserving paid platforms for tasks that genuinely require vendor-normalised metrics such as reported FWCI. The trade-off is that teams take on the engineering work themselves: handling pagination beyond 10,000 results, keeping institution ID mappings current as ROR records change, and documenting clearly that a percentile-based proxy is not the same figure a funder or ranking body may expect from Scopus.

    As OpenAlex’s topic classification and percentile fields mature, the gap between what a free, transparent API can deliver and what a paid platform delivers continues to narrow for most day-to-day institutional reporting needs, making a well-built in-house dashboard an increasingly credible default rather than a stopgap.

  • What Is Bibliometrics? A Research Office Primer

    Bibliometrics is the quantitative analysis of scholarly publications and the citations between them, used to measure research output, impact and collaboration patterns. For a research office, the practical challenge is rarely gathering these numbers — library systems, funders and university dashboards supply them constantly — but recognising which of the three main types of bibliometrics a given report represents, and what it can and cannot responsibly tell you.

    In its simplest form, bibliometrics is the statistical analysis of books, articles and other publications, most often using citation counts to describe patterns in scholarly communication. That one-line definition, drawn from the OECD’s usage and echoed by university library guides, is the starting point for everything that follows.

    What is bibliometrics?

    Bibliometrics applies statistical methods to bibliographic data — publication counts, citation counts, co-authorship networks and, increasingly, download and mention data — to describe and evaluate scholarly activity. It sits alongside scientometrics, a closely related field that extends the same statistical logic to science and technology output more broadly; in practice research offices treat the two terms as near-synonyms.

    Eugene Garfield, founder of the Institute for Scientific Information and creator of the Science Citation Index in 1964, is widely credited as a founding figure of modern bibliometrics. His citation-indexing work established the infrastructure — later commercialised as Web of Science — that most present-day bibliometric reporting still depends on.

    A metrics report a research office receives is rarely a single “bibliometric score.” It is usually a blend of three distinct analytical modes, and conflating them is the single most common source of misread reports.

    What are the three types of bibliometrics?

    Library and information science distinguishes descriptive, evaluative and relational bibliometrics. Each answers a different question, and each carries a different risk of misinterpretation when applied outside its proper scope.

    Type Core question it answers Typical output Main risk if misread
    Descriptive How much has been published, by whom, where? Publication counts, output by year, discipline or department Treated as a quality signal when it only measures volume
    Evaluative How much impact or influence has that output had? Citation counts, h-index, Journal Impact Factor, Field-Weighted Citation Impact Used to rank individuals directly, ignoring field and career-stage differences
    Relational How are researchers, topics or institutions connected? Co-authorship networks, co-citation maps, research-front clustering Read as a measure of quality rather than of structure or collaboration

    Descriptive bibliometrics is the safest category for research offices to report externally, because it counts rather than judges. Evaluative bibliometrics is the category most prone to misuse — a single h-index or Journal Impact Factor figure says nothing about an individual paper’s quality. Relational bibliometrics is the least familiar to non-specialists but the most useful for identifying emerging collaboration opportunities or research strengths across a department.

    What bibliometric indicators will appear in a metrics report?

    Most institutional metrics reports combine a handful of recurring indicators. Knowing which category each one belongs to prevents a descriptive count being read as an evaluative judgement.

    • Citation count — the raw number of times a work has been cited; evaluative, but highly field- and age-dependent.
    • h-index — an author-level figure meaning a researcher has h publications each cited at least h times; evaluative, and known to disadvantage early-career researchers and those in low-citation-rate fields.
    • Journal Impact Factor (JIF) — the average citations per article in a journal over the preceding two years; a journal-level, not an article-level, indicator.
    • Field-Weighted Citation Impact (FWCI) — a normalised indicator comparing a publication’s citations against the global average for its subject, document type and publication year; a value above 1 indicates above-average performance for that field.
    • Altmetrics — non-citation signals such as policy-document mentions, news coverage, social media activity and downloads, which supplement rather than replace citation-based evaluation.

    These indicators are drawn from different underlying databases, and coverage varies. Web of Science and Scopus apply curated, subscription-based indexing; Google Scholar offers broad, free coverage with less curation; Dimensions links publications to grants and clinical trials on a freemium basis. A report’s headline number can shift depending on which source supplied it.

    How should research offices use bibliometrics responsibly?

    Bibliometrics should inform, not replace, expert judgement. Three widely referenced frameworks set out how research offices can operationalise that principle rather than treat it as an aspiration.

    The San Francisco Declaration on Research Assessment (DORA), launched in 2012, commits signatory institutions to avoid using journal-based metrics such as the Journal Impact Factor in hiring, promotion or funding decisions. Imperial College London, for example, states it has applied this commitment since becoming a DORA signatory in 2017.

    The UK’s Metric Tide review, commissioned by the then Higher Education Funding Council for England (now part of UK Research and Innovation) and published in 2015, set out five principles for responsible metrics: robustness, humility, transparency, diversity and reflexivity. Those five principles remain the reference point most UK research offices cite when drafting internal metrics policies.

    INORMS’ Research Evaluation Working Group publishes the SCOPE framework — Start, Context, Options, Probe, Evaluate — a five-step method research administrators can apply before commissioning or interpreting any metrics report, rather than defaulting to whichever indicator a database happens to surface first.

    • Start by clarifying the purpose of the evaluation before selecting any indicator.
    • Establish the context: discipline, career stage, output type and comparator group.
    • Identify the options available, including qualitative alternatives such as peer review or narrative CVs.
    • Probe the suitability and limitations of each proposed indicator.
    • Evaluate the process itself once the assessment is complete, and refine it for next time.

    Momentum toward narrative-based assessment has also grown outside the UK: the 2022 Coalition for Advancing Research Assessment (CoARA), joined by many Horizon Europe-affiliated funders and institutions, commits signatories to reduce reliance on journal-based and output-count metrics in funding and hiring decisions.

    It is worth distinguishing bibliometrics from contributor-level attribution. Bibliometrics counts citations and outputs; it does not record who did what on a given paper. CASRAI originated the CRediT contributor role taxonomy in 2014 for that separate purpose, and the standard is now stewarded by NISO as ANSI/NISO Z39.104-2022. A research office reconciling a bibliometrics report with authorship disputes should reach for CRediT roles, not citation counts.

    Common questions about bibliometrics

    What is the meaning of bibliometric?

    “Bibliometric” describes any measure derived from the statistical analysis of published scholarly output — most commonly publication counts and citation counts. The term covers the underlying data point (a bibliometric) and the wider field that studies it (bibliometrics), and applies equally to authors, journals, institutions and individual articles.

    What is an example of bibliometrics?

    The h-index is the most commonly cited example: an author with an h-index of 20 has 20 publications that have each received at least 20 citations. Other everyday examples include a journal’s Impact Factor, a department’s annual publication count, and a co-authorship map showing collaboration between institutions.

    What is bibliometrics in simple terms?

    In simple terms, bibliometrics counts and analyses publications and citations to show how much research is being produced and how much attention it receives. It turns scattered publication records into structured evidence — useful for funding reports and CVs, but never a full substitute for reading the work itself.

    Who is the father of bibliometrics?

    Eugene Garfield (1925–2017) is widely regarded as the founding figure of bibliometrics and scientometrics. As founder of the Institute for Scientific Information, he created the Science Citation Index in 1964, establishing the citation-indexing infrastructure that underpins most bibliometric analysis conducted today.

    What this means for research offices

    A metrics report that blends descriptive, evaluative and relational bibliometrics without labelling which is which will inevitably be misread by whoever receives it next — a promotion panel, a funder, or a departmental head. Labelling each figure by type, and pairing evaluative indicators with the field-normalisation context they need, is a low-cost fix most research offices can apply immediately.

    As narrative-assessment frameworks such as CoARA and DORA gain signatories, research offices should expect bibliometric reports to sit alongside, not instead of, qualitative evidence in funding and promotion decisions. Building that dual capability now — clear metrics literacy plus a credible narrative-CV process — will matter more with each assessment cycle, not less.

  • Narrative CV (R4RI): Format, Sections and a Worked Example

    A narrative CV, called the Résumé for Research and Innovation (R4RI) at UK Research and Innovation, replaces bullet-pointed achievement lists with a structured, evidence-based prose account of a researcher’s contributions across four fixed modules, each mapped to a specific funding-opportunity word count. This guide walks through the R4RI format section by section, with a worked example for each module.

    The Declaration on Research Assessment (DORA) defines a narrative CV as a CV format that provides a structured written description of a person’s contributions and achievements, reflecting a broader range of relevant skills and experience than a traditional academic CV typically shows — a definition UKRI’s own guidance cites directly.

    What is a narrative CV (R4RI)?

    The Résumé for Research and Innovation (R4RI) is UKRI’s flexible narrative CV template. UKRI describes it as an evolved version of the Royal Society’s Résumé for Researchers, redesigned to be inclusive of the full range of sectors and roles that make up the global research and innovation community.

    Unlike a conventional academic CV built around publication lists and grant totals, the R4RI asks applicants to explain, in prose, how they achieved impact — through ideas, people, community leadership and societal reach. UKRI’s narrative CV explainer page (last updated 13 November 2024) frames this as aligning with DORA and Coalition for Advancing Research Assessment (CoARA) principles on responsible research assessment, alongside equality, diversity and inclusion goals and support for non-linear careers.

    R4RI is not the only narrative CV format in use — funders including NIHR and Cancer Research UK run comparable schemes — but it is the template most UK researchers will meet first, since UKRI’s seven research councils, Research England and Innovate UK all draw on it.

    What are the four required R4RI modules?

    UKRI’s R4RI-specific guidance, last updated 30 April 2026, is explicit: “You need to structure your answer into the four module headings. The word count is specified in the funding opportunity.” There is no fifth mandatory module — only an optional “Additional information” section for context, not extra achievements.

    Module Title What it evidences
    Module 1 Generation of new ideas, tools, methodologies or knowledge How you communicated ideas and results; key outputs such as datasets, software, research and policy publications; skills acquired from past projects
    Module 2 Development of others and effective working relationships Expertise critical to a team’s success; teaching, workshops and mentoring; leadership shaping a team’s direction
    Module 3 Contributions to the wider research and innovation community Positions of responsibility; reviewing, editing and committee work; strategic leadership influencing a research agenda
    Module 4 Contributions to broader users, audiences and societal benefit Engagement with the public sector, clients and the public; research contributing to public understanding or policy

    The optional “Additional information” section covers career breaks or voluntary work that provide context — for example, an interruption that affected output volume. UKRI states this section “will be seen by the panel and reviewers even if it references a sensitive issue,” and warns applicants not to use it to smuggle in extra skills or outputs, since that content will not be assessed.

    How do you write an R4RI? A worked example

    Each module should read as a short, evidenced narrative rather than a list. The University of Oxford’s Narrative CV Guide advises that, absent a specific funding call, applicants should aim for the whole CV to sit under 1,000 words and should never attach a separate publications list unless the funder asks for one.

    A simplified, illustrative Module 1 entry might read:

    “My work on [research area] identified a gap in existing measurement methods, which I addressed by developing a new protocol now used by three collaborating laboratories. I communicated this through a peer-reviewed publication and two conference presentations, and the underlying dataset is archived for reuse. Contributing to this project built my skills in [named technique], which I will apply directly to the proposed work.”

    Note the pattern: a specific problem, a concrete action, a verifiable output, and a link forward to the funding bid — not a bare list of publications. The same discipline applies to Modules 2 to 4: name the relationship, activity or audience, describe the action taken, and state the outcome.

    Practical drafting tips that recur across UKRI, Oxford and Cancer Research UK guidance:

    1. Write in the first person, in continuous prose — no bullet points within a module.
    2. Support every claim with a specific, checkable example rather than a general assertion.
    3. Select five to ten of your most relevant contributions rather than attempting an exhaustive record.
    4. Tailor the narrative to the specific funding opportunity’s assessment criteria before each submission.
    5. Use active verbs — “led,” “developed,” “convened” — to make your role unambiguous.

    Applicants describing intellectual contributions in Module 1 can also borrow precision from established contributor-role vocabulary. CASRAI originated the CRediT contributor role taxonomy in 2014; the standard is now stewarded by NISO as ANSI/NISO Z39.104-2022, and its named roles (such as methodology, data curation or writing — original draft) offer a ready vocabulary for stating precisely what a contribution involved.

    Individual vs team R4RI: what changes?

    A team submits one R4RI for the whole group, even where members sit in different organisations — UKRI’s guidance is explicit that individual achievements can be highlighted, but the four modules should demonstrate the appropriateness of the team overall, with different members showing complementary skills.

    Assessors also use the R4RI differently from a scored checklist. UKRI states that assessors “will not view it in isolation” and “will never be asked to score individual modules” — the R4RI informs the overall application assessment rather than generating its own numeric mark. For some funding opportunities it is instead used to establish team eligibility, so applicants should always check the specific opportunity’s guidance on how the R4RI will be used before drafting.

    Completed R4RIs are shared with reviewers and panel members without anonymisation, and treated the same way as a traditional CV within UKRI’s evaluation process.

    Frequently asked questions

    What is the difference between a narrative CV and a traditional CV?

    A narrative CV replaces bulleted lists and metrics with structured prose describing contributions and achievements across fixed modules, while a traditional academic CV lists publications, grants and posts chronologically. UKRI’s own definition frames the narrative format as capturing a broader range of skills than a conventional CV typically shows.

    Is a narrative CV written in first person or third person?

    Most guidance, including the Mental Health Research Incubator’s explainer, recommends writing in whichever person you find clearest to write, but notes that first person often suits questions phrased as “how have you…”. Check whether the specific funding call’s wording implies a preferred voice before drafting.

    How long is a narrative CV?

    Length is normally set by the individual funding opportunity’s word count. The University of Oxford’s guide recommends that, without a specific call in mind, applicants keep the whole document to under 1,000 words and avoid attaching a separate publications list unless requested.

    How does a narrative CV look compared with a covering letter?

    The University of Edinburgh describes a narrative CV as “a blend of a traditional academic CV and covering letter,” since it explains the context — the “how” — behind achievements, typically across four modules covering contributions beyond immediate research outputs.

    What this means for applicants and institutions

    Narrative CV adoption is spreading beyond UKRI: NIHR, Cancer Research UK and a growing number of UK universities now request R4RI-like formats for recruitment as well as funding, and UKRI signposts the Peer Exchange Platform on Narrative CVs for applicants who want structured support drafting one. Institutions are responding by building local guidance and review services rather than leaving researchers to interpret module headings unaided.

    For research administrators supporting applicants, the practical task is less about policing word counts and more about coaching researchers to replace generic claims with specific, evidenced sentences — the single change UKRI, Oxford and Cancer Research UK guidance converge on most consistently. As more funders formalise R4RI-style requirements, institutions that build this coaching capacity now will reduce rework at the application stage rather than after rejection.

  • Horizon Europe Financial Reporting Field Guide

    Horizon Europe financial reporting is the process by which every beneficiary in a funded consortium declares its incurred eligible costs to the European Commission at the end of each reporting period, via a financial statement lodged on the Funding & Tenders Portal. A periodic financial report is the beneficiary-level cost declaration — covering personnel, subcontracting, purchase and indirect costs — that accompanies the technical report, triggers EU reimbursement, and (above a EUR 430,000 cumulative threshold) requires independent audit sign-off.

    For grant and research administrators, the financial statement is where compliance is won or lost. Cost-category misallocation, missing evidence and audit-threshold miscalculation are the recurring causes of rejected or queried claims. This field guide sets out the reporting cycle, the cost-category breakdown, the Certificate on Financial Statements (CFS) threshold, and the specific errors that trigger Commission pushback.

    What is the Horizon Europe periodic reporting cycle?

    A Horizon Europe periodic report combines a technical/scientific report with a financial report made up of each beneficiary’s financial statement. Accounting periods in Horizon Europe projects usually cover 18 months, and the coordinator must submit the complete periodic report within 60 days of the period’s end, per the Model Grant Agreement’s standard reporting terms.

    Each beneficiary prepares its own financial statement, has it electronically signed by a designated financial signatory, and submits it to the coordinator, who consolidates all statements before submission to the Commission via the Funding & Tenders Portal. From the second reporting period onward, a beneficiary can adjust a prior period’s declared costs — for example, correcting under- or over-declared amounts — which then flows through to the current period’s requested contribution.

    Missing the 60-day deadline does not usually void the claim outright, but it can defer the beneficiary’s costs into the following reporting period, delaying reimbursement. Deadline extensions are generally only available for the final reporting period.

    What cost categories must appear in the financial statement?

    The Horizon Europe financial statement, set out in Annex 4 of the Model Grant Agreement, requires beneficiaries to declare actually incurred, eligible costs against a fixed set of categories. Indirect costs are not itemised: they are calculated automatically as a flat rate of 25% of eligible direct costs (excluding subcontracting, financial support to third parties, and unit-cost volunteer time), under Article 6 of the Model Grant Agreement.

    Category What it covers Key reporting rule
    A. Personnel costs Employees, natural persons under direct contract, seconded staff, SME-owner/natural-person beneficiary time Daily-rate method; capped at 215 declarable day-equivalents per person per calendar year across all EU/Euratom grants
    B. Subcontracting Tasks outsourced to a third party under the beneficiary’s usual purchasing practices Must follow best-value-for-money procurement; ideally listed in Annex 1
    C. Purchase costs Travel and subsistence; equipment (usually depreciation-only); other goods, works and services Itemised explanation required once purchase costs exceed 15% of declared personnel costs
    D. Other direct costs Financial support to third parties (if permitted by the call); internally invoiced goods and services Same 15%-of-personnel-costs itemisation trigger as Category C
    Indirect costs Overheads not directly attributable to the action Flat rate of 25% of eligible direct costs, calculated automatically — no separate evidence required

    The “Use of Resources” itemisation rule is a common trip point: once Category C or D spend exceeds 15% of a beneficiary’s declared personnel costs for the period, the portal requires a detailed breakdown of major cost items down to that threshold, starting with the highest-value items first.

    When does a Certificate on Financial Statements (CFS) apply?

    A Certificate on Financial Statements is an independent auditor’s report — or, for public bodies, a report from a competent independent public officer — that verifies the eligibility and accuracy of a beneficiary’s declared costs. Under Article 24.2 of the Horizon Europe Annotated Model Grant Agreement, and as set out in Data Sheet Point 4.3 of the Grant Agreement, a CFS is mandatory once a beneficiary’s cumulative EU contribution requested for costs reaches EUR 430,000 over the life of the grant.

    • The threshold is assessed per beneficiary, not per consortium or per project — each partner is evaluated individually.
    • It applies to the EU contribution actually requested for costs, not the total project budget.
    • A CFS is typically required at final payment, though a Grant Agreement’s Data Sheet can also require one at an interim reporting period.
    • Costs already audited by the granting authority do not need to be re-included in the CFS and do not count toward the threshold.

    Some older guidance in circulation still cites a EUR 325,000 threshold, which was the figure attached to the original wave of Horizon Europe Model Grant Agreements. Administrators relying on legacy templates or older third-party summaries should check their own Grant Agreement’s Data Sheet Point 4.3 directly, since the currently applicable AMGA text sets the CFS trigger at EUR 430,000 per beneficiary.

    Why do financial statements get queried or rejected?

    Cost rejection at review or audit stage is rarely arbitrary — it follows a predictable pattern of documentation and classification failures. The most frequent causes are:

    • Misclassified costs — for example, subcontracted work declared as personnel costs, or purchase costs booked under the wrong sub-category.
    • Insufficient supporting documentation — missing invoices, timesheets, employment contracts, or delivery confirmations that an auditor or the Commission cannot trace back to the accounting system.
    • Personnel cost calculation errors — daily-rate miscalculations, timesheet-to-payroll mismatches, or exceeding the 215 day-equivalent annual cap across combined EU and Euratom projects.
    • Ineligible costs — expenditure incurred outside the eligible period, disallowed hospitality costs, or items not connected to the action described in Annex 1.
    • Non-compliant subcontracting — subcontracts awarded without following the beneficiary’s usual best-value-for-money purchasing practices.
    • Technical/financial inconsistency — a financial statement that does not match the effort or progress described in the accompanying technical report, such as unexplained over- or under-spending against person-months.

    Most of these are preventable with real-time cost tracking rather than end-of-period reconstruction. Reconciling declared costs against the general ledger before submission, and keeping personnel time records current throughout the period, removes the majority of documentation-based rejections before they reach the Commission.

    Common questions on Horizon Europe financial reporting

    What is the reporting period for Horizon Europe?

    Horizon Europe accounting periods typically run for 18 months, and the consortium coordinator must submit the full periodic report — technical and financial — within 60 days of the period’s end. The final reporting period may be shorter, and deadline extensions are generally only available for that last period.

    What is the audit threshold for Horizon Europe?

    A Certificate on Financial Statements becomes mandatory once a beneficiary’s cumulative EU contribution requested for costs reaches EUR 430,000 over the grant’s duration. The threshold is assessed per beneficiary under Article 24.2 of the Annotated Model Grant Agreement, not against the consortium’s combined budget.

    What this means for research administrators

    Horizon Europe financial reporting rewards continuous cost tracking over end-of-period reconstruction. Institutions running multiple Horizon Europe grants should monitor each beneficiary’s cumulative requested EU contribution against the EUR 430,000 CFS threshold from the outset, since crossing it late in a project leaves little time to engage a qualified independent auditor. Equally, personnel time recording against the 215 day-equivalent annual ceiling needs to be checked across all EU and Euratom grants a researcher is involved in, not project by project in isolation.

    For institutional research administration teams, the practical takeaway is procedural: reconcile the financial statement against the general ledger before locking it for review, confirm the CFS requirement against the specific Grant Agreement’s Data Sheet Point 4.3 rather than generic guidance, and treat the technical report and financial statement as a single, cross-checked submission rather than two separate exercises. As reporting cycles repeat every 18 months for the life of a project, the administrative overhead compounds — building the tracking discipline into normal project operations, rather than into a pre-deadline scramble, is what keeps periodic reports moving through the Commission without dispute.