Research Power: Why Publication Volume Alone Does Not Define Research Strength

Research activity can be counted. Research strength has to be interpreted.

For universities, publication volume is useful information. It shows that scholarly work is being produced, provides a view of activity across departments and can reveal changes in research capacity over time. The problem begins when output is asked to carry a much larger claim: that more papers automatically mean stronger research.

Responsible research assessment has been moving away from that assumption for good reason. The San Francisco Declaration on Research Assessment (DORA) urges institutions to judge the scientific content and value of work rather than rely on publication metrics or journal identity, and to recognise outputs beyond papers, including data and software. The Leiden Manifesto makes a related case for contextual, field-aware use of quantitative indicators.

Research power, then, is not the rejection of measurement. It is the ability to measure without confusing activity with achievement.

Output Is a Signal, Not a Verdict

A publication count answers a narrow question: how many recorded scholarly outputs were produced during a defined period? That can be strategically useful. It may reveal whether research activity is expanding, whether new centres are becoming productive, or whether particular disciplines require additional support.

But the count does not tell leadership whether the work altered understanding in a field, was methodologically robust, influenced professional practice, created reusable research infrastructure, informed policy, supported innovation, or opened a genuinely new line of enquiry.

It also cannot distinguish between a large quantity of incremental work and a smaller body of research that becomes foundational. Both have a place in a healthy research system, but they are not analytically interchangeable.

Disciplinary Context Changes the Meaning of the Number

Publication cultures differ substantially across disciplines. Experimental sciences may involve large teams and frequent journal articles. Some engineering and computing fields place considerable value on conference proceedings. Humanities scholarship may culminate in books, critical editions or long-form work developed over several years. Practice-oriented disciplines may create datasets, designs, protocols, software, exhibitions or other outputs that conventional paper counts capture poorly.

Comparing these fields through one undifferentiated volume benchmark creates a false precision. It can reward the publishing rhythm of one discipline while making another appear less productive even when both are performing strongly by the standards of their fields.

A credible research assessment system therefore starts with disciplinary context before it starts with comparison.

Research Strength Includes the Environment That Produces the Work

A university's research position is partly visible in its outputs, but the conditions behind those outputs matter just as much. Are doctoral researchers well supervised? Are early-career academics given time to develop ambitious work? Is infrastructure reliable? Can teams access data, laboratories, libraries, technical expertise and research management support? Are integrity, reproducibility and ethical review treated as capabilities rather than administrative hurdles?

These questions matter because research strength is not simply what an institution produced last year. It is also whether the institution is building the capacity to produce credible, consequential work five and ten years from now.

This is one reason research-environment measures have become important in formal assessment systems. The UK's Research Excellence Framework, for example, assesses not only outputs and impact but also the environment that supports research and enables impact.

Quality, Influence and Durability Need Different Evidence

No single metric can establish research quality. Peer review remains important because expert judgement can examine novelty, rigour and disciplinary contribution in ways a database cannot. Citation-based indicators can add information about scholarly uptake when they are normalised for field and publication year, but a citation is not automatically praise: work can be cited because it is influential, contested, replicated or criticised.

Wider influence requires still other evidence. Adoption in professional standards, use by industry, incorporation into public policy, clinical or technical application, licensing, open-source reuse and documented social or cultural effects all represent different pathways from research to consequence.

Durability adds a further horizon. Some work attracts rapid attention and fades; other work accumulates significance slowly. Research strategy should be capable of recognising both without forcing every project into an annual-performance timetable.

A Better Research-Power Dashboard

For leadership, the practical answer is not to replace one headline metric with another. It is to build a balanced dashboard.

That dashboard might combine field-normalised scholarly influence, research quality review, doctoral completion and researcher development, external research income interpreted by discipline, collaboration patterns, open-research practices, policy or professional uptake, translational outcomes and carefully evidenced impact narratives.

The mix should reflect institutional mission. A specialist design university, a medical research institution and a comprehensive public university should not be expected to demonstrate research strength in identical ways.

What matters is that the institution can explain what it measures, why those measures fit its research mission and how quantitative evidence is combined with expert judgement.

HERA Perspective: Research Power Is Capacity With Consequence

The strongest research cultures are not defined by the longest publication lists. They are defined by the ability to produce rigorous work, attract capable people, sustain serious enquiry, build useful research infrastructure and create knowledge that other researchers or communities can use.

Publication volume belongs in that picture. It simply should not be allowed to become the picture.

For universities seeking a more credible account of research strength, the question should move from 'How much did we publish?' to a more demanding set of questions: What did we contribute? Who used it? What changed because of it? And are we creating the conditions for the next generation of important work?