What is metascience?

Geanina Beres, Jude Fransman and James Wilsdon from the Research on Research Institute (RoRI) reflect on the findings from a bibliometric analysis that accompanies a recent report by the UK Metascience Unit: ‘A Year in Metascience’.

The recent explosion in research on research through funding calls, new centres and organisations, strategic programmes, policy and training initiatives has created something of a metascientific moment. Here, we discuss findings from a bibliometric analysis to reflect on the evolution of metascience and its meaning today.

What is metascience?

Despite the recent proliferation of initiatives, the concept of metascience (and related terms such as ‘metaresearch’, ‘science of science’ and ‘research on research’) has a deep and multifarious history, with each strand evolving through its own genealogy, context and networks. Yet across these strands runs a shared commitment to the application of robust methods to the analysis of research systems, cultures and decision-making.

To better understand how the metascience landscape has evolved, the Research on Research Institute (RoRI) conducted a bibliometric analysis using the Dimensions database. Dimensions provides access to a vast, multidisciplinary collection of publications with rich metadata, enabling the kind of large-scale, topic-based analysis needed to map a cross-cutting field like metascience. Rather than anchoring the study in a single definition of metascience, the aim was to create a comprehensive delineation of the main research topics, communities, and disciplines that now constitute metascience.

The evolving metascience landscape

The rich theoretical and empirical heritage of metascience dates back decades and spans disciplines and subdisciplines including the philosophy and history of science; science and technology studies (STS); innovation studies; economics; scientometrics; public policy studies; higher education studies; library, data and information sciences. It also encompasses the changing relationship between policy, science and innovation. Understanding this heritage is essential to the credibility of metascience. Ignoring it risks erasing valuable contributions or, worse, rebranding established work in other fields as if it were a novel discovery (sometimes referred to as ‘Columbusing’).

As early as 1939, J.D. Bernal’s The Social Function of Science laid the groundwork for what became known as a science of science. Twenty five years later, the Science of Science Foundation (later the Science Policy Foundation) was established by a group of British scientists with a mission “to promote the application of scientific methods to the understanding of science itself, especially in its relations with society.”

From the 1960s onward, science and innovation studies gained momentum at the universities of Sussex, Edinburgh and Manchester, with parallel developments in countries including the US, Norway, Germany, China, the Netherlands, South Korea, and Japan and later in South African and Brazil. By the mid-1990s, science, technology and innovation (STI) policy analysis began shifting from a focus on internal science dynamics and towards to broader innovation systems and economic impact, influenced by OECD’s “knowledge economy” frameworks.

In the 2000s, movements for science reform surged, catalyzed by several high profile interventions. White House science adviser John Marburger addressed the American Association for the Advancement of Science (AAAS) in 2005 on the need for a “science of science policy” and in the same year John Ioannidis published his pivotal article on research unreliability. This period also saw a surge in open science and open access advocacy, and all these different streams flowed into what we now consider part of metascience.

Five vectors of change

Our paper identifies metascience as evolving through five vectors of change. First, a shift from specialized disciplines and toward distributed engagement, with researchers across fields increasingly addressing metascientific questions as a “side-hustle” - a part-time or occasional focus, pursued on the margins of their primary research role. Instead of reducing metascience to a singular, disciplinary frame, we therefore argue it is better understood as an orientation or way of engaging with questions that many researchers encounter in their networks, disciplines and institutions.

Second, the field is marked by rapid uptake and lower barriers to entry for the use of innovative methods, including AI, deep learning, and advanced bibliometrics, which are reshaping how science is evaluated.

Third, targeted policies and investments from governments and funders globally—such as the UK Metascience Unit and initiatives in the EU, China, and Canada—signal growing institutional support.

Fourth, expanding coalitions such as DORA (San Francisco Declaration on Research Assessment) and COARA (Coalition for Advancing Research Assessment) are driving reform in research assessment, ethics, reproducibility and open science, reflecting broad systemic concerns and overlapping stakeholder interests, as well a commitment to strategic collaboration.

Finally, new institutions are emerging that embody metascience in their design—such as Arc Institute and Convergent Research—alongside novel philanthropic models, infrastructures, and alliances like the upcoming Metascience Alliance. Collectively, these trends position metascience as a diverse, decentralised movement – and a source of collective intelligence for governments, funders and scientific leaders, rather than a traditional academic discipline.

The past, present and future of UK Metascience

Our scientometric analysis offers a window into the field’s deep roots, diverse communities, and accelerating momentum. Metascience has grown into a rich and dynamic community - multidisciplinary in its roots, global in its reach, and increasingly transdisciplinary and applied in its orientation.

Looking ahead, two trends stand out: metascience is emerging both as a discourse coalition that brings together diverse actors around shared system challenges, and as a form of collective intelligence, where distributed, self-organised and highly contextualised expertise, enhanced by computational, AI and other methods, is directed towards deeper understanding of the inner workings of the scientific landscape.

We hope that RoRI’s partnership-based approach can help to consolidate shared priorities, foster stronger connections, and chart the next phase in the evolution of metascience as a force for making research fairer, more efficient and impactful to the benefit of us all.

RoRI was founded in 2019 and aims to be a hub for new ways of doing collaborative, applied metascience. Thier consortium of funders work in close collaboration with researchers to develop impactful meta-research and open up their own funding systems for comparative analysis. We structure our research around three priority areas: experiments, systems and platforms, and all of our projects involve funder organisations from multiple countries, ensuring the relevance of our work beyond the context of a single funder or national system. A Year in Metascience is new paper for the UK Metascience Unit, launched by Lord Vallance, UK science minister, at Metascience 2025.