Simon Porter is Digital Science’s Director of Innovation. Simon came to Digital Science from the University of Melbourne, where he has worked for the past 15 years in roles spanning the Library, Research Administration, and Information Technology. Beginning from a core strength in the understanding of how information on research is collected, Simon has forged a career transforming university practices in how data about research is used, both from administrative and eResearch perspectives. In addition to making key contributions to research information visualization and discovery within the university, Simon is well known for his advocacy of Research Profiling Systems and their capability to create new opportunities for researchers. Over the past three years, Simon has established and run the annual Australasian conference on research profiling. In 2012, Simon was the program chair of the third annual VIVO conference.

In over a decade of creating research university collaboration diagrams, the response that I’ve come to expect when presenting them is usually, “Pretty isn’t it…”, followed by a pause, and then, “…what does it mean?” – which is fair enough. Collaboration graphs without any surrounding narrative can really only be encountered as art. Identifying interesting features in a network requires some knowledge of how they were made.

That said, research collaboration graphs do play a productive role in highlighting the cultural differences between different types of disciplines and institutions. In our 2019 project, ‘What does a University look like?’, we showed that a large-scale comprehensive institution looks very different to an institution focused on science and engineering. Country-level differences in culture can also be seen. For example, two related posters in this project highlight the differences between a distributed system of institutions in New Zealand that focuses on different and complementary areas of research, and a larger university sector in Australia that produces many more comprehensive and competing institutions. Coming soon after the launch of Dimensions, the original 2019 project showcased both the quality of the research network and research classifications within Dimensions, while demonstrating how much it was possible to do with the Dimensions API – though it did take a while to download all the data! With the launch of Dimensions on Google BigQuery and the ability to gather the data for networks in

seconds rather than hours, I thought I would revisit the project again, this time with a focus on narrative. Imagine you are in a gallery, standing in front of a large framed picture of a network diagram.   What story does it tell you? It is most likely to be a story about the interplay and size of disciplines within an institution. How does physics interact with medicine? How does engineering connect with both the biological sciences and pure chemistry? Where are the intersections with education? In exploring these questions we also gain a sense of distance between disciplines. In this way, network diagrams act as a map as our eyes wander across a landscape of cooperation.

A story that tells us how disciplines are connected requires knowledge of where researchers are from. Using Dimensions we can assign a discipline to an individual by looking for the research classification that they most commonly publish in. Dimensions has many different classifications to choose from however, as the human eye can only differentiate a limited number of colours, I chose to use the Units of Assessment classification created in the UK for the Research Excellence Framework (the REF) as a high level, yet sufficiently broad scheme. Exploring interconnecting disciplines also requires a way to group researchers that are working together. To shift focus from individuals to groups within the network I used the Leiden algorithm from CWTS (also used in VOSviewer) to identify well-connected clusters of researchers. Having already used colour to indicate discipline, clusters are highlighted on the graph by hiding the links between clusters and accentuating the links within a cluster through the use of edge bundling. Clusters that are related to each other can be inferred by their proximity on the network, as well as the discipline mix of researchers within the cluster.   In an analogous method to assigning a discipline to each researcher, I assigned a colour to the edges within a cluster based on the majority discipline of the researchers within it.

Playing with edge bundling, as it can help reveal structure in graphs, but how much is too much? For me, it looks like setting a (hammer_bundle) initial param of .3 seems about right. Thoughts? (animation @matplotlib w camera) 1/3 pic.twitter.com/On13C8A64f

— Simon Porter (@sjcporter) November 28, 2020

As clusters have been derived independently from the algorithm that lays out the network, the final challenge is to differentiate between clusters that overlap across the network. Indeed, it is often interesting to identify instances when they do. To handle this, I chose to animate the graph, highlighting each cluster above a certain size in sequence, starting with Clinical Medicine (A01), and moving progressively through the Units of Assessment through science, engineering, social sciences, and the humanities. In this way, the network is presented as a tour through the disciplines providing a narrative structure for the entire network.  As Units of Assessments are quite broad, I also used the Fields of Research categorisation at the 4 digit level to provide another description for each cluster. By highlighting the four researchers within a cluster that have published the most, each cluster gains a further identity. To help identify clusters from different disciplines that overlap with one another, I created a 3-dimensional model of the network and gave each Unit of Assessment a different layer in the representation. Unit of Assessment A01 (Clinical Science) forms the base of the diagram. From there the layers move up through medicine, engineering, sciences, and humanities. The end result, created in this case for the University of Cambridge, is presented below. Pretty isn’t it? Although hopefully now the story it has to tell is embedded in the encounter.

Some Additional Technical Things of Note This project makes use of a new graph layout engine called BatchLayout, which I use in conjunction with Graphviz to reduce the number of overlapping nodes. Edge bundling is achieved using the Datashader library. Although the end result is rendered in Blender, all of the 2D development was rendered using matplotlib. You can find a Google Colab Notebook with all the steps up to 3D rendering here. The code for the whole project, including the Python script that builds the network in Blender can be found in my Gigantum repository here. Gigantum was really helpful here, as it made it easy to switch between my local computer for development, and a more powerful machine that didn’t mind being on for several days for 3D rendering. Have a go yourself, and tell us how you get along!

The post Bringing Narrative to Research Collaboration Networks in 3D appeared first on Digital Science.

Juergen Wastl is Director of Academic Relations and Consultancy at Digital Science. He previously headed up the Research Information team at the University of Cambridge’s Research Strategy Office and worked for BASF managing BMBF-funded projects internationally. Briony Fane is a Research Analyst at Digital Science. She has a higher education background, having gained a PhD from City, University of London, and has worked as both a researcher and a research manager. Bo Alroe has worked with research management and administration since 2004,  and currently as Director of Strategy with Digital Science. Bo is from Aalborg, Denmark, where he studied and currently lives with his family.

Introduction

The United Nations Sustainable Development Goals (UN SDGs) are global targets set by the UN across 17 areas that will give rise to a better and more sustainable world for all. Research relating to these SDGs can therefore be seen as socially impactful, and analysing trends in SDG-related research can indicate how a researcher, institution, funder or country is contributing to meeting these targets.

Continuing our SDG blog series, this blog focuses on insights in SDG related allocation of research funding related to the SDGs. Having broadened its categorisation of grants data to include classification by SDG codes, we can use Dimensions to gain insights into how competitive funding supports the Goals. By applying Dimensions’ SDG classifications to its grants database, we discovered over 6 million grants worth more than £1.37 trillion, from over 600 funders worldwide that can be searched and analysed. We took a dive into the data to discover which UK research councils support SDG-related research, where funding is focused across the UN SDGs, how much is allocated to sustainability research, and more.

Figure 1: A Dimensions screenshot showing how a single grant can be assigned multiple SDG classifications; SDGs 2, 7, and 12 in this instance

How much UKRI funding supports SDG-related research?

Dimensions’ classification system was developed jointly with the Dutch universities (via the Association of Universities in the Netherlands, VSNU), SpringerNature, and Digital Science. The aim was not only to design a classification system that can categorise grants that mentioned sustainability or the UN’s work but also to assign SDG classifications to research – including grants and publications – to better support the goal of sustainability. Supervised machine learning was used to classify content in Dimensions. For a publication abstract or grant to merely mention sustainability or related concepts such as ‘pollution’ would not be enough to earn an SDG classification. This means that Dimensions can identify grants that support sustainability improvements both explicitly – eg, by mentioning the UN’s sustainable development goals – and implicitly.

After extracting all UKRI grants indexed in Dimensions from eight UK research councils between 2011-2020, we applied the SDG classification to determine the proportion of UKRI funding that supports the SDGs.

Figure 2: The sum in GBP of SDG-classified UKRI grants awarded between 2011 and 2020

By applying the 17 SDG-classified grants records and publications in Dimensions, we can evaluate how funders support research towards more sustainable development. Figure 2 provides an overview of the sum in GBP of UKRI grants that have supported sustainable development research between 2011 and 2020. We have selected a public research funder as an exemplar for this analysis on the basis that a competitive public research funder, such as UKRI, is the origin of most public research funding in the UK, with arguably the most impactful strategic waypoint for research on sustainable development. Such funders have considerable influence on the type and focus of research conducted.

Figure 3: The total number of UKRI grants with and without an SDG classification awarded between 2011 and 2020

Figure 3 shows the total number of SDG-related UKRI grants versus all UKRI grants, using the same base data as for Figure 2. There is a notable increase in the number of grants awarded after 2016, the year the UN SDGs were implemented. The graph also reveals that an average of 24.9% of all UKRI grants each year aimed to support sustainable development research, with a growth rate of 218.0% over the period. This is a higher rate than for all UKRI grants in the period, which was 128% over the same period. The Global Challenges Research Fund (GCRF) UK aid strategy is administered through UKRI and aims to assist in making progress on the global effort to address the UN SDGs. It has committed £1.5bn funding to address the UN SDGs between 2016 and 2021. This has also contributed to both the number and value of grants for research with a focus on sustainability.

Unlike Figure 3, Figure 2 does not show a similar trend for awarded amounts. This could indicate an increased number of smaller sums of funding per grant awarded by UKRI.

The proportion of UKRI funding with SDG classifications by year shows an approximately linear growth, with SDG funding having almost tripled in 2020 compared to 2011. This trend is very likely to continue and we may see an even greater increase in funding as we move towards the 2030 deadline of achieving the Goals, especially as UKRI have committed to supporting the ambitions of the UK government’s aid strategy and progressing the UN SDGs¹. If and how the recent adjustments to GCRF funding by the UK government will affect the grant landscape and visibility remains to be seen.

How is sustainability research funding distributed across the 17 SDGs?

Figure 4: The value and number of UKRI grants awarded by SDG classification between 2011 and 2020

Figure 4 sheds light on the focus of UKRI’s funding in support of the UN 2030 Agenda. The graph shows the total SDG funding amounts by the total number of SDG grants awarded, as classified in Dimensions. It is clear that SDG7, ’Affordable and Clean Energy’, appears to have been prioritised as a funding objective. It has the greatest number of grants awarded and the highest total funding amount compared to other SDGs over the 10-year period. Similarly SDG13, ‘Climate Action’, is also prioritised. Given the climate crisis we face and the role that energy has to play in this, it makes sense that increased funding would be focused in these areas, as the transition towards reaching climate neutrality is now so urgent.

How is funding split across the three pillars of sustainability; societal, environmental, and economic?

We can analyse SDG-related UKRI funding in Dimensions through the lens of the three pillars of sustainability (societal, environmental and economic, also depicted as the ‘wedding cake’, as seen in a previous blog) as a means of assessing the proportion of UKRI research funding that is concentrated in these three components.

Figure 5 visualises the prioritisation of UKRI’s sustainability research funding by each pillar of sustainability. The size of each circle is directly proportional to the total amount of funding that has been awarded to support the SDGs within each pillar.  The big hitters in the Social pillar are SDG7, ‘Affordable and Green Energy’, and SDG3, ‘Good Health and Well Being’. In the Environmental Pillar funding is prioritised in SDG13, ‘Climate Action’, and for the Economic Pillar SDG8, ‘Decent Work and Economic Growth’, and SDG12, ‘Responsible Consumption and Production’, are the most highly funded research areas.

Figure 5: The value of UKRI’s sustainability research funding between 2011 and 2020, split by the three pillars of sustainability; societal, environmental and economic sustainability

Conclusion

UKRI has allocated close to £10 billion or 28% of all its awards from 2011 to 2020 in ways that would offer support to the SDGs. Grants appear to be more likely than publications to receive SDG classifications. One reason for this is that a grant abstract is more focused on what needs to be achieved and the intention behind the funding, while a publication is reporting on what has been achieved, which may not be as comprehensive in terms of its SDG focus. Our showcasing of the grants data here shows how this funding has been allocated to each of the SDGs.

This blog shows that funding research aligning with sustainable development is prominent on UKRI’s funding programme agenda and that research capacity in this area is growing. With research and innovation having such a vital role to play in helping to find sustainable solutions to address global challenges, it is reassuring to see that SDG related research by a competitive funder is so extensive. It is also particularly gratifying to see that all 17 Goals are included to some extent in UKRI funding.

References

1: UKRI announces International Development Research Programme Awards

The post How is UK Funding Allocated to Support Sustainability Research? appeared first on Digital Science.

UoA Classification for the Research Excellence Framework

We developed a REF Unit of Assessment classification system to enable institutions to assess the quality of their research.  The system has been implemented using automated allocation of UoA codes (based on REF 2021) to documents on the Dimensions platform based on category definitions defined by machine learning. Using the system institutions are able to compare their institutionally held data in Dimensions with a world view on their staff outputs in a REF context to retrieve a fuller picture of its output portfolio, including OA status and accompanying bibliometrics.

The system has the capability to give an institutional view on publications and grants and a Unit of Assessment view in a REF timeframe.  The system enables institutions to compare their output with other institutions, providing the ability to see UOA classified research output on an ongoing basis rather than seeing it as a snapshot every REF period, thereby providing further and deeper institution-wide insights.  

Find out more

The post UoA Classification for the REF appeared first on Digital Science.

In association with Springer Nature, the Digital Science Consultancy and Dimensions teams have worked together to build a classification system of research associated with the United Nations Sustainable Development Goals (SDGs) from the Dimensions publications database. This ongoing work uses machine learning and semi-automatically generated training data.

In the first phase of the project, an automated approach was used to categorise scholarly articles into five of the SDGs employing supervised machine learning whereby curated training data fed machine learning algorithms to automatically build a classification model that was then used to categorise new articles without human involvement.

The second phase saw the project widened to include all 17 SDGs. Springer Nature and the Digital Science Consultancy and Dimensions teams collaborated in this phase and Springer Nature subject matter experts assisted by manually assessing keywords for the training sets and identifying any false positives. The same automated approach was used for the research classification in Dimensions, the results of which will lead to the next iteration in Dimensions automatically assigning an SDG category to publications and grants, etc, when they are added to the Dimensions database.

Find out more

The post Sustainable Development Goals Classification appeared first on Digital Science.

March 20th is celebrated around the world as the International Day of Happiness. This day of celebration and learning was founded by the United Nations back in 2012, and this year is no exception, as the UN released their latest Happiness Report. As part of our SDG blog series on the research driving us closer to meeting the UN’s Sustainable Development Goals (SDGs) we asked Digital Science’s very own Well-Being Trainer, Danielle Feger, to tell us more about the science behind SDG 3, Good Health and Well-Being, and how to use the powers of positive psychology to keep us centred in the midst of a global pandemic. With thanks to Juergen Wastl for his assistance in delving into Dimensions to uncover the hottest trends in happiness research.

About the Authors

As a stress counsellor and wellbeing trainer, Danielle developed her own well-being concept, Equillbrium4Wellbeing, which is based on Cognitive Behavioural Therapy, Mindfulness and Yoga. She leads the well-being programme at Digital Science and runs workshops on stress management and well-being in the UK as well as in German-speaking countries. Her programmes support employees’ mental well-being, helping them to shift from ‘coping’ to ‘thriving’.

Previously, Danielle had a long career in research development at the University of Cambridge where she gained extensive experience in fundraising and managing international research projects. In addition, she had the opportunity to coach researchers and academics and has established a mentoring programme for early-career researchers.

Juergen Wastl is Director of Academic Relations and Consultancy at Digital Science. He previously headed up the Research Information team at the University of Cambridge’s Research Strategy Office and worked for BASF managing BMBF-funded projects internationally.

The History of Happiness Research

Zhuangzi, who was born about 2,300 years ago in China, probably wrote the first piece of work in history devoted to happiness; an essay called “Supreme Happiness”. Since then, a total of almost 39k research publications on happiness have been added to this corpus of knowledge. Figure 1 shows how many publications have been released each year over the last 25, and the steady increase in the rate of research outputs relating to happiness, where the word “happiness” features in the title and abstract of the publication.

Happiness is a human emotion, but does happiness research take place wherever there are people? Not quite. Figure 2 shows the global distribution of publications relating to happiness. The US has had the highest overall happiness research output with about 7,700 publications, followed by the UK with almost 2,700. China (with 1,386), Canada (with 1,375) and Germany (with 1,257) make it a close-run race for 3rd place, while other countries such as Australia, the Netherlands, and Japan are hot on their heels.

In recent years, the increase in the number of publications relating to happiness outpaced the general increase of publications overall: Between 2017 to 2020, the number of happiness-related publications increased by 27.5%, whereas the general increase was 19.6%.

Measuring the Impact of Happiness Research

The impact of happiness research can be measured in traditional citations as well as in alternative metrics (or altmetrics) that track the real-world attention received by a piece of research. When we look at publication citations, we can see there have been 350,000 citations of research from US publications, 92,000 from UK publications, and 47,000 from Canadian publications. It is interesting to note that, although China has published a comparable amount of research around happiness within Canada, citation of this work is much lower at 14,000. Is this down to a research language barrier, or something else?

When we track the volume of research publications over time, we see that research into happiness and subsequent publications start increasing significantly in the early 2000s. This interest in happiness is connected to the study of Positive Psychology.

Positive Psychology

For about 200 years, Psychology research mainly focused on the study of disease. In the 1990s, psychologists started to look at what increases well-being and what makes people happier. Since its beginning, Positive Psychology has developed significantly; millions of dollars have been invested in research, and thousands of findings have been published.

The focus on happiness is growing everywhere: Countries adopt happiness policies, companies and educational institutions introduce wellbeing and happiness programmes, and books about happiness are regularly on bestseller lists. The team at Digital Science understand the value and importance of employee happiness and have invested in boosting well-being among their employees through a range of courses and sessions led by my organisation.

The UN Happiness Report 2020 and Sustainable Development Goals

Chapter 6 of the 2020 World Happiness Report takes a close look at the link between SDGs and happiness and shows that the countries with a higher SDG index score tend to do better in terms of subjective well-being.

If we ask people what their ultimate goal is in life, most would say they want to be happy. Happiness is something we all strive to find. We believe that we will be happy once we achieve certain things like status, higher income or more possessions. However, research has shown that these things do not increase our happiness. Happiness is a state of inner fulfilment, not the gratification of inexhaustible desires for outward things.

So how do we achieve happiness?

One of the main points I teach in my programmes is to embrace experiences. Plenty of research findings show that experiences can increase our happiness. Let’s break this down from the perspective of positive psychology, to better make sense of the things that make us happy – and how we can achieve a greater sense of well being and happiness.

Danielle’s Guide to Achieving Happiness Through Experiences

1. There is no hedonic adaptation to experiences
We adapt quickly to possessions: Whenever we get what we are seeking, it quickly becomes less valuable, so we return to our original baseline of happiness and the cycle repeats like a treadmill. Experiences are temporary and there is no time to get used to them. Since the stimuli in experiences aren’t constant, we keep paying attention to them.

2. Experiences cannot be quantified or compared
Possessions foster comparisons but we tend to think of experiences more on their own terms, rather than in comparison with other things. Experiences are so individual that we have no reference points to compare them to or to quantify the relative value of any two experiences.

3. Experiences give us unforgettable memories and increase positive emotions such as joy, inspiration, pride, confidence or gratitude
According to Barbara Fredrickson, when people experience positive emotions, their minds broaden and they open up to new possibilities and ideas. At the same time, positive emotions help people build their personal wellbeing resources, ranging from physical, to intellectual, to social resources.

4. Experiences increase our engagement and enhance our sense of achievement
Activities that meet our need for engagement flood the body with positive neurotransmitters and hormones that elevate our sense of happiness. ‘Flow’ engagement stretches our intelligence, skills, and emotional capabilities. Achievement is not just about a strong sense of success or ‘winning’ – it also refers to how much we challenge ourselves in positive and progressive ways that enable us to further develop our strengths and skills, both practically and emotionally.

5. Experiences help us connect with people and improve social relationships
Relationships and social connections are a crucial part of our sense of wellbeing and happiness. Sharing experiences connects us to other people more than sharing consumption. Since experiences are very individual, jealousy and envy aren’t as much of an issue as with possessions. Therefore we are more interested in hearing about other people’s experiences than about their latest purchases.

6. Experiences help define our purpose and passions
We are the sum total of our experiences or as Oliver Wendell Holmes put it: “A mind that is stretched by new experiences can never go back to its old dimensions.” A new iPhone doesn’t define who we are, but having travelled in various countries truly enriches our lives and forms our identity.

Further Reading

• “Buy Experiences, Not Things” The Atlantic
• The UN Happiness Report
• Richard Layard, co-editor of the annual World Happiness Report (with Jeffrey Sachs and John Helliwell) is also co-founder of the movement “Action for Happiness” that brings together like-minded people from all walks of life and helps them take practical action, drawing on the latest scientific research. AfH is backed by leading experts from diverse fields including psychology, education, economics and social innovation.

The post The Pursuit of Happiness – Researching Good Health and Well-Being appeared first on Digital Science.

Juergen Wastl is Director of Academic Relations and Consultancy at Digital Science. He previously headed up the Research Information team at the University of Cambridge’s Research Strategy Office and worked for BASF managing BMBF-funded projects internationally.

Briony Fane is a Research Analyst at Digital Science. She has a higher education background, having gained a PhD from City, University of London, and has worked as both a researcher and a research manager.

Bo Alroe has worked with research management and administration since 2004,  and currently as Director of Strategy with Digital Science. Bo is from Aalborg, Denmark, where he studied and currently lives with his family.

Introduction

In this analysis, we have used Dimensions on Google BigQuery to analyse the Impact Case Studies from REF 2014, the first exercise to evaluate the impact of research outside of academia. Impact Case Studies form an important part of all university REF submissions, with each describing both the research and its impact beyond academia. The inclusion of impact as a component of the REF reflects a growing interest in demonstrating the value of academic research in society.

Of the 6,975 Impact Case Studies submitted to REF 2014, 6,737 were included in the publicly available HEFCE dataset. Each case study has a unique digital object identifier (DOIs). We have taken these DOIs and used Dimensions on Google BigQuery to connect the REF Impact Case Studies to SDG categories using Dimensions’ SDG classification scheme launched last year.

Using Google BigQuery to connect external datasets to Dimensions to gain important insights

Thanks to HEFCE’s (now Research England) freely available database of REF research submissions, we were able to merge this with Dimensions. By applying Dimensions’ SDG filter it was easy to ascertain the proportion of research output associated with sustainable development submitted to REF 2014 underpinning the Impact Case Studies. We were also able to determine other trends within this information, including ‘types of impact’ associated with the case studies using publicly available code in the Dimensions BigQuery Lab; these include technological, societal, environmental, political, legal, cultural, health and economic types of impact, as defined in the post-REF report we published with King’s College London and HEFCE (now Research England).

In this case, we were able to leverage Dimensions data by enriching it with additional information in the form of impact case study data via Google BigQuery. This levels the playing field across databases, allowing users to carry out informative analysis and gain a greater depth of understanding of REF Impact data. This methodology provided the basis for the following two analyses.

Prevalence of SDG-related research underpinning Impact Case Studies

By analysing the number of SDG-related research publications from total publications between 1993 to 2013 (the two decades before REF 2014 submission) and the number and percentage submitted to the ‘output’ (REF2) and ‘impact’ (REF3) elements of REF 2014, the data show that SDG-related research was significant before the UN SDG goals were introduced in 2015. We see that proportionately more research with a sustainable development focus was submitted to demonstrate impact in REF 2014 than for research outputs. Interestingly, there was 25% overlap of impact case study research publication references with outputs submitted to REF2 in 2014¹.

Alignment of SDGs within the Dimensions categorisation and REF-related ‘types of impact’

SDGs can be clustered into three overarching pillars: Environmental, Economic and Social. The Stockholm Resilience Institute³ has assembled the SDGs by these overarching groups in a diagram that resembles a wedding cake (with Social labelled Biosphere here). This hierarchical aggregation of the individual goals allows us to better visualise and analyse overall trends, and enables comparisons with the REF Main Panel structure (A-D) to retrieve a high level view of sustainable development goals in REF Impact.

Underpinning research categorised by the HEFCE ‘impact types’ and by the three pillars of SDGs as outlined above nicely validates the outcome of the SDG classification system in Dimensions. Table 2 reveals that 62% of underpinning research classified within SDG13, Climate Action, correlates with HEFCE’s Environmental impact type. 45% of research classified as SDG3 Good Health and Well-Being correlates with HEFCE’s Health impact type, and 47% of research associated with SDG16 Peace, Justice and Strong Institutions correlates with HEFCE’s Societal impact type.

As we have already noted, the time period for REF 2014 precedes the formalisation of the UN SDGs but, as we demonstrated in our last blog, UK research exhibits a high proportion of SDG-tagged publications. This is even more pronounced in the underpinning research accompanying the REF Impact Case Studies, showcasing the existing contribution of research in the context of sustainable development to both ‘Impact’ and ‘Excellence’ in the REF 2014 submission.

Comparing citation averages for SDG-related underpinning research versus non SDG-related underpinning research

REF Impact Case Studies were categorised into eight types by HEFCE (now Research England): Cultural, Economic, Environmental, Health, Legal, Political, Societal and Technological. Figure 2 outlines the citation averages for underpinning research publications categorised by impact type that have been SDG-tagged in Dimensions, versus those not SDG-tagged. In particular, we see that citation averages are twice as high for SDG-related publications associated with Economic and Environmental impact types than those that are not SDG-related. Conversely, citation averages for underpinning research outputs associated with Technological and Cultural impact types fall below the citation average of those outputs that are not associated with SDGs. It could be that, for both Environmental and Economic impact types, there is an existing awareness of SDG-related sustainable development thinking, particularly environmental impact; we know that climate change didn’t emerge in 2015 with the introduction of the SDGs. For research related to Technological impact, the focus may have been on more non-sustainable development related technologies.

Looking ahead, we might expect that research underpinning Impact Case Studies in REF 2021 will see a notable increase in SDG-related publications as we continue the UN’s decade of action for realising the sustainable development goals.

Weaving the four REF Main Panels into the story provides us with a slightly different insight into the contribution that the overarching REF Panels made to sustainable development research in REF 2014, as seen in Figure 3:

• Main Panel C (social sciences) contributes most and is evident in all SDGs across all HEFCE impact types. This Panel contains the greatest quantity of underpinning research publications associated with Impact Case Studies overall and confirms the contribution of this Panel to REF2 outputs
• Unsurprisingly, Main Panel A (medicine, health and life sciences) contributes predominantly through SDG3 Health and Well-Being, and across to HEFCE’s Health impact type related Impact Case Studies
• Main Panel B (physical sciences, engineering and maths) mostly contributes to SDG7 Affordable and Clean Energy, and SDG13 Climate Action through to Technology and Environmental ‘impact types’
• Finally, Main Panel D (arts and humanities) which accommodates a more diverse portfolio of underpinning research associated with SDGs, but dominated by SDG16 associated research, does not solely contribute to cultural impact type categorised case studies as might be expected

Research in the context of sustainable development (using Dimensions SDG Classification scheme) acts as a pivot between HEFCE’s ‘Impact Type’ and the four REF Main Panels, showcasing the diverse interaction between the underpinning research in the Impact Case Studies component of REF 2014, its discipline and contribution to the UN SDGs, even before these were formally implemented.

Conclusion

This blog has revealed a starting point for us looking at the scope for interpreting SDGs through the lens of Impact. Utilising the reference lists submitted with the case studies provided us with a context within which to explore research associated with sustainable development in REF 2014, ahead of the formalisation of the UN SDGs in 2015.

Using Google BigQuery allows us to draw interesting conclusions about the impact of research submitted to REF 2014. By integrating Dimensions data with a relevant external dataset we have been able to showcase the potential for future application in developing insights on and beyond sustainable development with impact case studies from the forthcoming REF 2021.

Footnotes

1: The document can be viewed here

2: Each underpinning research publication may be associated with more than one SDG classification. The total number of publications we are working with is 21,829 but, with the overlap of SDG- and UoA-associations, these 21,829 publications are represented 32,684 times

3: The SDG ‘wedding cake’, developed at the Stockholm Resilience Centre, Stockholm University by the Centre’s science director Carl Folke

The post In search of SDGs in REF Impact Case Studies appeared first on Digital Science.

Juergen Wastl is Director of Academic Relations and Consultancy at Digital Science. He previously headed up the Research Information team at the University of Cambridge’s Research Strategy Office and worked for BASF managing BMBF-funded projects internationally.

Briony Fane is a Research Analyst at Digital Science. She has a higher education background, having gained a PhD from City, University of London, and has worked as both a researcher and a research manager.

Bo Alroe has worked with research management and administration since 2004,  and currently as Director of Strategy with Digital Science. Bo is from Aalborg, Denmark, where he studied and lives with his family.

Introduction

Assessment of research in the UK helps fund and shape the development of academic disciplines in Universities. The recurring seven-year cycles of assessment exercises that Higher Education takes part in has a long history, starting in 1986 and evolving and culminating in the latest versions known as the Research Assessment Exercise (RAE) in 2001 and 2008 and the Research Excellence Framework (REF) in 2014 and 2021. The outcomes of these exercises tell us about the state of research that is taking place across the United Kingdom, from its academic strength and impact to the significance placed on research from social, economic, clinical and policy perspectives.

Reflecting on the last assessment exercise, REF 2014, outcomes would endorse that the Higher Education sector performs exceptionally well both academically, and in respect of the impact of its research in a real-world setting, as our in-depth post-REF 2014 analysis revealed.

The UN SDGs seek to tackle the biggest societal challenges in our world today, in part by asking us to fundamentally question long-held beliefs about economic growth and their compatibility with maintaining a healthy balance with the natural world. As higher education institutions have a uniquely important role to play in delivering solutions to the SDGs it would make sense that a proportion of the research submitted to the UK’s assessment exercises will have clear links to the SDGs. This focus on societal, economic and environmental challenges facing the world can be well captured by the need for the provision of impact (showcased by case studies) that was introduced in REF 2014.

We have used Digital Science’s Dimensions to shed light on the contribution of SDG-related literature to the UK’s REF 2014 submission. Dimensions offers a comprehensive collection of interlinked data in a single platform that reflects the entire research lifecycle. As Dimensions covers research inputs, outputs and data on pathways to impact, it is uniquely suited to teasing out contributions to SDG-related research. For both SDG and REF Unit of Assessment (UoA) classifications, Dimensions offers filters developed using supervised machine learning based on custom-made training sets. As a consequence, the UK’s assessment of research can be viewed in the context of sustainable development.

We start by examining REF 2014, in an attempt to ascertain whether SDG-aligned research is growing in content beyond the normal base rate growth of research as we progress towards the deadline for achieving the Goals in 2030¹.  Dimensions includes classification filters based on the REF 2021 UoA classification system comprising 34 UoAs.  These can be easily applied to the previous assessment exercises, despite research areas being defined slightly differently in previous assessments,  with 67 and 36 UoAs in REF 2014 and RAE 2008 respectively. Dimensions was used to retrieve SDG-classified research publications that were submitted in the REF 2014 period from 2008 to 2013.

Exploring REF 2014 and SDG-related research outputs in Dimensions

We start the blog series with examples of the percentage of SDG-related research submitted per UoA in REF 2014, and will address the following research questions:

1: Is research in the context of Sustainable Development submitted and deemed excellent research as defined by REF 2014?

2: What can the spread of SDG-related research submitted to REF 2014 tell us?

3: Can trends be found in SDG research that contributes not only to REF 2014 but across all years that have been part of research assessment? Is there evidence of SDG associated research before the SDGs were formally implemented in January 2016²? Will we see a steady increase in growth of SDG linked research across the research assessment timespan?

Table 1: Examples of the proportion of SDG-related research submitted to UoAs in REF 2014

Table 1 provides specific examples of the proportion of SDG-related research submitted in the UoAs featured. It highlights not only the importance of SDGs to academic research in units of assessment in clinical sciences and areas of social sciences, but also areas in the physical sciences where SDG research does not contribute significantly to REF 2014

1: SDG research is ‘excellent’ research

We can show that SDG tagged research is considered ‘excellent’ research in the UK in the context of the national research evaluation exercise.  By its very nature, we recognise that SDG-related research is, to varying degrees, ‘applied research’ because it addresses societal challenges, and ‘excellent’ research because it contributes to the national research assessment submissions. The evidence for this lies in the fact that, for research outputs submitted to the REF, we can identify what percentage are tagged as SDG research in Dimensions. If the percentage of SDG research is high, it demonstrates its level of excellence as perceived by the submitting academic and institution. The UK’s REF 2014 submission of research outputs has, in a number of Units of Assessment, a higher content of SDG-associated research than the average across the timeframe of the exercise.

Figure 1: Number of SDG-related research outputs submitted to REF 2014Figure 2: SDG-related research outputs submitted to REF 2014 as a percentage of total submissions per Unit of Assessment

2: What can SDG-related research tell us about the UK’s REF 2014 submission?

SDG tagged publications vary in number and share (see Figures 1 and 2 above) across the UoAs however trends, or hot spots, of individual SDGs per UoA are revealed. For example, SDG7, Affordable and Clean Energy (green) across Main Panel B UoAs in the physical and technological focussed UoAs, and SDG 3, Good Health and Wellbeing among Clinical Medicine and Life Sciences based UoAs (red), we see that SDG-related research coverage is considerable.

In order to have a summary view of the REF 2014 submission (and beyond) at the level of Main Panels, aggregating research in the context of SDGs helps to give us an overall picture with two main insights. Applying the SDG filter across the evaluation time frames of RAE 2008, REF 2014 and the current REF 2021, we can affiliate the SDG tagged research to the Main Panels over a period of 20 years (noting that the main panels have remained the same, but the UoA structure slightly varied across the three assessment exercises).

Figure 3: SDG-related research outputs submitted to REF 2014

Looking at figure 3, the first main insight that is immediately apparent for Main Panels A and B, which focus on STEM subjects, that there is a steady increase in SDG-related research, which for Panel A (in particular UoAs 1-3) climbs sharply as we get closer to 2020.  This can be explained by an increase in COVID-19 publications in the context of SDG3 ‘Good Health and Well-Being’.  The second insight is that when looking at Panel C, which has a focus on the social sciences, there is a notable peak in publications in 2013 which is the year for submission of publications to REF 2014, with a marked decline immediately thereafter. This potential REF 2014 effect seems very likely after we made a comparison with SDG tagged publications classified by UoAs across the same timeframe for Germany, seen in figure 4, which does not undergo the same research evaluation exercise, and reveals no marked spike and decline in publications categorised in UoAs related to social sciences between 2013 and 2014.  Panel D, which covers research in the arts and humanities and includes much more heterogeneous output types, make SDG and REF classifications less reliable as many outputs go beyond those having identifying features for classification purposes.

Figure 4: SDG-related research outputs in Germany, mapped onto the REF classification scheme in Dimensions

3: Can trends be found in SDG research that contributes not only to REF 2014 but across all years that have been part of research assessment?

When we look at year-on-year growth of SDG related research in each UoA, across both the REF 2014 submission period as well as across all RAE and REF time periods, we observe that the REF 2014 submission of research outputs has a higher content of SDG-related research than non-SDG research. Based on this, it would be our expectation to see REF 2021 contain even more SDG-related research, continuing the upward trend we see in previous assessment periods, where SDG research outpaces normal growth in research submitted to REF.

Conclusion

It has been argued that none of the SDGs can be achieved without the contribution of research carried out in higher education institutions.  Universities have the ability to generate, translate and disseminate knowledge relevant to achieving the SDGs. They also have the potential to increase societal impact through translational research.  Given this, using Dimensions to probe the vast corpus of research relating to both REF 2014 and SDGs and how the two entities interact will go some way to providing insight into the contribution of academic research to the United Nations Sustainable Development Agenda.

Footnotes

1: The SDG classification system is the most recent addition to classification systems in the Dimensions database and is not yet fully developed, not only in terms of the complexities involved in defining SDGs but also in difficulties with SDG language. It is rapidly evolving with more SDG-related research being classified on an ongoing basis

2: It is important to note that, although the SDGs were not implemented at the time of the two RAE exercises in 2001 and 2008, there was a previous global action plan, the Millenium Development Goals, which set out to reverse the poverty, hunger and disease affecting billions of people, therefore research associated with SDGs was already in evidence. It is likely that more funding was made available at this time to address SDGs

The post UK Academic Research Contributions to the SDGs Observed Through National Assessment Submissions appeared first on Digital Science.

External (left) and internal (right) collaboration patterns are presented here for Australasian Research Organisations (selected by top 20 ). Researchers are coloured by the field of research that they most commonly publish in, and sized by total number of journal articles that they have published (relative to the network). To create the networks, journal articles published between 2015 and 2018 were analysed.

If you want to find out more check out our interactive dashboard.

Download Now

The post What does New Zealand Research Look Like? appeared first on Digital Science.

Consultancy

Custom reporting and analysis

Make better decisions faster

Get in touch

Take your data visualisations to the next level

It’s not easy to get clear takeaways by looking at lines of numbers and stats. You need to have the data presented in a logical, easy-to-understand way so you can apply your learnings in an effective way. We are here to support and deliver insightful Data visualisations. Our most recent work involved the concept of “real-time” bibliometrics as a resource for analyzing COVID-19 and collaboration networks for Cambridge University.

We can support your decision making

Contact us

New research shows global divide in pharmaceutical research is significant – but closing

New research from Digital Science shows global divide in pharmaceutical research is significant – but closing

Bringing Narrative to Research Collaboration Networks in 3D

Identifying interesting features in a collaboration network requires some knowledge of how they were made.

Contact us

Supporting research classification

UoA Classification for the REF

We developed a REF Unit of Assessment classification system to enable institutions to assess the quality of their research.

Sustainable Development Goals Classification

We have built a classification system of research associated with the SDGs from the Dimensions publications database.

“We gave Jürgen Wastl and his team one hell of a challenge – to classify over 100 million scholarly publications with the UN’s x17 Sustainable Development Goals, in just seven months. And they managed it – absolutely brilliant! Innovative, hard-working, reliable. And a pleasure to work with. I cannot recommend them highly enough.”

Timon Oefelein, Springer Nature

Our research reports

Discover more

The Ascent of Open Access

While most trends are encouraging the research community is now demanding more enforcement of mandates.

Contextualizing Sustainable Development Research

Our report highlights the growth in research around the UN’s Sustainable Development Goals (SDGs).

How Covid-19 is Changing Research Culture

We investigate the research landscape trends and cultural changes in response to COVID-19.

Discover more

Our in-house data sources

Dimensions

Find out more

Altmetric

Find out more

Overleaf

Find out more

Symplectic

Find out more

Get in Touch

Dr Juergen Wastl | Director of Academic Relations and Consultancy

Contact Us

The post Consultancy appeared first on Digital Science.

Expectations around reproducible research are clear, particularly in the area of computational research. A research paper is more than an account of the research that was undertaken; it is a narrative that surrounds an orchestration of research assets from the raw data and code to the processed data and visualisations that result. A paper should invite a reader to trace the results back. How was this figure produced? What was the code that produced that particular result? The reader’s transition from narrative to exploring data or code should be as easy as turning the page.

Seen from the researcher’s perspective, the ideal computational paper arises organically from the research – the data that is created is the data that ends up in the paper. The code as it is written is the code that can be accessed in the paper. As analysis bubbles up from research into images for publication, those images keep their providence back to the data, and back to the code that produced them.

How close are we to this ideal today? Within the Digital Science family, methods for openly publishing data are ably supported by Figshare. Overleaf allows researchers to easily publish their research collaboratively using LaTeX.  As part of a poster presentation for the 2019 VIVO conference we took a broad research question that could be answered with Dimensions data, and undertook the research using workflows that knit these tools together. In doing so, our project, documented in our white paper, demonstrates an approach to undertaking reproducible computational science that operates on multiple levels. Specifically, it addresses:

• How can data assets be structured and organised throughout the life of a project inside Figshare (and not just at the end of a project)?
• What is a good approach to tying code, data, and papers together using identifiers?

In this paper we demonstrate that not only is our poster reproducible, but that the methods we have adopted are useful to others as well. We feel we learnt a lot throughout this project, and hope to continue to refine these approaches in our analysis projects moving forward.  From analysis through to publication, we would love to hear about some of the ways that you have used research productivity tools in similar ways. Get in touch!

PS – This is the first Digital Science Report to be made entirely in Overleaf…

Technical Report: https://doi.org/10.6084/m9.figshare.9741890

Poster: https://doi.org/10.6084/m9.figshare.9742055

Online Version: https://wdaull.ds-innovation-experiments.com/

The post This Poster is Reproducible appeared first on Digital Science.