The challenges facing social science research are numerous and wicked – being undervalued by funders, surmounting barriers between knowledge produced and its uptake, and ensuring they get the credit they deserve in public discourse. However, as Juergen Wastl and Kathryn Weber-Boer outline, solutions could lie in demonstrating their ability to catalyse STEM research, in effect providing a ‘secret sauce’ to stimulate research and innovation in the UK.

When preparing a new meal, we might look at several recipes. One has an ingredient that is missing from the cupboard, another would take too long to make; in the end, we prepare a dish that balances the needs of the kitchen with those of the diners. The same thinking explains why we have focused on the specific elements chosen for our report on social science research. The choices were driven broadly by their coherence to the message of supporting UK research and innovation (R&I), the timeliness of artificial intelligence and regulatory frameworks, and to draw attention to the global impact of UK social sciences.

Commissioned by Sage and the Academy of Social Sciences, the report, ‘Reimagining the Recipe for Research & Innovation’, was both a challenge and an opportunity. The challenge for us was to derive insights into social science contributions to research and innovation, with an opportunity to improve our understanding of the mechanisms of research funding, outputs and influence. To do this we needed to dig our fingers into the detail of UK research data – sifting through disciplines, combining data, analysis, policy and evidence – and using this abundance of information to develop a coherent picture of the role played by the social sciences in the UK’s research and innovation ecosystem.

What follows is our ‘recipe’ for the report.

From left: Prof James Wilsdon of UCL and ESRC Chief Officer Stian Westlake at the launch of the AcSS report. Credit: Juergen Wastl.

Finding the right (data) recipe for Social Science’s secret sauce

Our contribution to the report on the Digital Science side built on the capabilities/capacities of Dimensions to reveal the Social Sciences as the (not so) secret ingredient in research and innovation. Innovative ways of looking into available data – helped by the availability of a linked database with grants, publication and policy documents, and the data being unsiloed – enabled us to probe into the depth and breadth of data, and evaluate its robustness. We also needed to examine the contribution of social sciences with and without its links to STEM subjects in a UK context.

A big part of the challenge was how to represent these complex data visually in meaningful and comprehensible ways; only a sliver of the whole picture could be shown. The report focuses generally on policy documents, publications, and grants from 2012 to 2022. To better understand those of most relevance to our study, we need to understand what Fields of Research apply to them. Dimensions data is mapped to Fields of Research (FoR), which are defined by the Australian and New Zealand Standard Research Classification (ANZSRC) and applicable to any research worldwide.

We used Machine Learning to map the data in Dimensions to these FoR – this carried out at the article level where an abstract is available, or where an abstract is not available at the journal level. Grants are also categorised using the FoR classification system. For this report, we extracted the first-level Fields of Research for each of those documents (excluding the fields of the humanities), using the FoR to assign each document (policy, publication, or grant) to the category of STEM, social sciences, or both.

This work is important to help translate the data into a visual representation. The alluvial charts in the report (Figs.3A, 3B, and 3C) look at grants, publications, and policy documents categorised by their connection to the social sciences, STEM, or both (again, excluding the fields of the humanities and here also excluding biomedical and clinical sciences). Each shows the counts of a different data type (awarded grants, publications, and policy documents respectively) connected to the mention of publications in governmental policy documents (excluding policy documents relating to the Humanities and the field of Biomedical and Clinical Sciences).

We started by looking at all research outputs mentioned in policy documents published from 2012 to 2022. We then extracted the FoR and categorised the UK-affiliated publications which had been cited in the same way. Finally, we identified all grants which had supported those publications (and extracted their fields of research and categorised them, as well). Then we used RawGraphs 2.0 to show the number of grants (Fig.3A), publications (Fig.3B), and policy documents (Fig.3C) by category (displayed as the left side of the alluvial chart) and the count of grants, publications, and policy documents per category by the governmental policy organisation (the right side of the chart). We relied on the ‘organisations’ dataset to identify these documents as governmental policy. This dataset consisted of 15,098 grants, 12,863 publications, and 7,386 policy documents.

Fig.3A: Grants by policy maker)

Fig.3B: Publications by policy maker

Fig.3C: Policy documents by policy maker

The alluvial charts below show one additional sliver: a breakdown of the policy impact of the grants awarded by the Engineering & Physical Sciences Research Council (EPSRC) and the Economic and Social Research Council (ESRC). For these diagrams, we began with publications affiliated with research institutions in the UK, with supporting grants awarded by each funding agency (2,360 publications supported by 1,569 EPSRC grants and 3,706 publications supported by 1,813 grants from the ESRC), categorising these as above, and then identified the policy documents which had cited those publications (2,360 policy documents associated with EPSRC grants and 5,092 associated with ESRC grants).

The figures below show the count of policy documents per category (STEM, Social Sciences, or joint STEM/Social Sciences) of supporting grants, cited publications, and policy documents themselves. These show the same blending of flavours found in the figures (above), but also highlight that, regardless of the source of funding, policy makers rely heavily on research produced by the social sciences and that the social science component is very clearly integral throughout the research cycle. It is worth noting that here the biomedical and clinical sciences are not excluded.

Figs.4:  Contribution of policy documents per category (STEM, Social Sciences, or joint STEM/Social Sciences) of supporting grants, cited publications, and policy documents.

Not only are we in a position to analyse the flow in a linear fashion, based on one of our most recent technical advances in Dimensions we were able to add additional spice to the analytical possibilities: with Google BigQuery we can enlarge our database with additional datasets. Previously focused only on research-led data (publications, citations) the dataset now can become enriched by data from adjacent datasets, including REF Impact case studies.

Showcasing analysis on broader data with a regional, economic, environmental (eg. UN Sustainable Development Goals) or even geopolitical stance allowed us to drill deeper into the positioning of the social sciences and its links to STEM when looking at the underpinning research of the eight different types of REF Impact.

Taking just one of these elements – geopolitical – we could also have presented the REF Impact types by national and/or international region (as per the REF Impact case study database), sorted by Social Sciences and its constituent disciplines. We can see the strength of the social sciences’ contribution to Economic-type REF Impact case studies, and the joint STEM/SocSci overlapped research in the Environment, which is the strongest SocSci/STEM overlap for REF Impact case studies.

Fig.5: Linked datasets underpinning a geopolitical analysis of social science research.

Data and data science has an increasingly significant role to play in the social sciences as a field, for institutions, funders, policy makers and for individual researchers. While for social scientists engaged in innovative research across the spectrum of disciplines these findings may seem second nature, the tools of data science are increasingly able to map out these complex relationships in various ways.

As the data has shown, the social sciences contribute to enabling world-leading and future-focused UK research, including in areas primarily seen as relating to STEM. For the full potential of the social sciences to UK R&I to be realised, the report calls for investment in inter- and multi-disciplinary research to be scaled up.

The content generated on this blog is for information purposes only. This Article gives the views and opinions of the authors and does not reflect the views and opinions of the Impact of Social Science blog (the blog), nor of the London School of Economics and Political Science. Please review our comments policy if you have any concerns on posting a comment below.

Image Credit: All images reproduced with permission of the authors.