Many people in academic circles have been amazed by how rapidly and completely data-driven approaches have started to transform scientific research.
Researchers can analyse greater volumes of data faster and more cheaply than ever before, and collaborate easily with colleagues around the globe. It is all making the outputs more robust and visible, but the speed at which the tide of data is rising brings its own set of challenges.
Research Councils UK (RCUK) now expects project bids to include explicit plans for how research data collected and generated by a project will be managed and (wherever possible) made available openly to others in the long term. That’s not something that researchers can do alone, and nor should they; institutions have to play their part.
Looking after data properly makes sure that research can be robustly evidenced, verified and reproduced while making it easier for researchers to be open and collaborative with data. At the same time it helps a research institution to build and maintain its own reputation for research excellence.
Roadmap for research data
Recently, I chaired a working group for the League of European Research Universities (LERU) and in January 2014 we produced the LERU roadmap for research data. The first document in the world to explore the issues that research data management raises explicitly for research universities, it is a timely intervention that describes in detail the many demands that data-driven research places on research universities and offers a blueprint for how they could respond.
It reveals how these challenges have an impact throughout the research functions within universities, summarising them under six main headings:
- Policy and leadership, and the importance of developing institution-wide research data management policies
- Advocacy, promoting data sharing and its benefits, and overcoming researchers’ reluctance to share their data
- Legal issues and standards, to iron out copyright issues and standardise protocols for description, discovery and citation
- Infrastructure, including tools and staffing, and how to organise these to provide optimal support for research efforts
- Costs – how to assess and meet them
- Skills, roles and responsibilities at all levels
The roadmap also makes a series of 44 recommendations for robust institutional research data management. The time for every research-intensive university to make a start is now, because the data deluge can only get heavier. Over the past few years we have seen researchers in science subjects becoming increasingly comfortable with the data-driven approach and now the arts, humanities and social sciences are setting out along the same path.
The benefits of the data-driven approach
An example of the benefits of this type of approach can be seen in the ProQuest's Early European Books collections 1-4, which has been licensed by Jisc on behalf of the UK’s research community. The initiative puts digitised versions of 26,000 volumes published across Europe between 1450 and 1700 at researchers’ fingertips.
It is an important bid to open up the possibility of data-driven research to arts, humanities and social science subjects and that is a genuinely revolutionary prospect: for the first time humanities researchers can review the full text of all these works without having to spend months or years visiting archives and libraries across the European continent.
To the historical researcher examining - for example - the epic battle to nullify Henry VIII’s marriage to Catherine of Aragon, the primary texts created at the time are the data that he or she needs to examine to identify nuances in the debate that may be hidden from us in the 21st century but which were apparent to both writer and propagandists in the 16th.
The insights that researchers can gain from examining the primary texts could be huge and that is only the start. The digitised collections will support the application of text- and data-mining tools and techniques, raising that prospect that researchers might uncover connections that could otherwise remain forever hidden.
At a symposium in London in the middle of October the research community will have an opportunity to find out more about Early European Books and how people are putting the collections to work in their research projects. I hope that what delegates discover will spark fresh ideas for ways to use the collections.
For my part, I have prepared a paper that will be presented there by my UCL colleague Ben Meunier, looking in detail at research data management challenges and the principles of the LERU roadmap and applying them to Early European Books. It is free to attend but booking is required.
In the meantime you can browse through the full text of the LERU roadmap to find out more.