In this post, Neil Jacobs (Jisc’s head of open science and research lifecycle) outlines a small project that is just started, to take further steps in addressing the difficult challenge of deciding what research data should be kept.
Jisc is working to develop shared infrastructure, influence policy and provide guidance to support institutions with the growing need for robust research data management. There is a wide-range of needs and existing provision for creation, collection, storage and preservation, and reuse of data for UK HE.
What research data should be kept?
Researchers, data curators and policy makers all need to answer the question, what research data should be kept? We can’t keep it all, because that would be too expensive and time-consuming. However, we have to keep data that is irreplaceable and unique in its value for future research; to enable it to be reused and validated; to enable peer review to be informed; and to enable there to be trust in research findings. Types of data needing to be retained vary and may include related materials such as software and documentation. But how much and what is enough? Obviously, there is no single answer to that; it depends on many factors, but what are those factors, and how should we weight them? These remain difficult and open questions, but this year Jisc is working with the Charles Beagrie consultancy to take a step toward answering them.
How can we identify what to keep?
We are setting out to explore, what actually is the optimal data to keep from research projects conducted at UK institutions? Over the course of the rest of 2018, our project will work with a small number of research areas to find out. What conditions, such as openness or timescales, might be ideal? We will consult the views of researchers (as data creators and data users), research funders, ethics professionals, archivists, research data managers, peer reviewers, other research users, and others on these questions. We will dig into the reasons for their views, and into whether research data is currently kept in line with those views, or not.
Why are we carrying out this investigation now?
This work comes at a critical time in the evolution of research data management and sharing. At the policy level, the recommendations from the UK Open Research Data Taskforce are expected shortly. These may take into account both the recommendations to Government of the 2017 report by Dame Wendy Hall and Jérôme Pesenti into the future of the UK artificial intelligence industry and the recent Government announcements around this, where research data can be a key input into AI tools. The availability of research data is also a matter of concern to those interested in research integrity and reproducibility, though some are worried that such concerns are being used for political purposes. Relevant infrastructure investments include both the Jisc research data shared service and the increasing activity around the European Open Science Cloud.
Both policy and infrastructure investments need better information about the extent and nature of the research data that needs to be kept, under what conditions, and for how long. Our 2018 project will not provide all this information, but it will explore current practices and take the next step.