Because good research needs good data

Organisational perspectives at #idcc11

Summary of presentations in the organisational perspectives session at #idcc11 Data Matters: A view from a research funder David Lynn, Head of Strategic Planning and Policy, Wellcome Trust   Taking care of the university's research data Jeff Haywood, Vice Principal Knowledge Management,...

Kirsty Pitkin | 07 December 2011

In this session we heard from David Lynn from the Wellcome Trust, Jeff Haywood from the University of Edinburgh and Andrew Charlesworth from the University of Bristol, who provided an overview of the issues faced by funders, institutions and legal departments when considering the issue of open data.

Data Matters: A view from a research funder
David Lynn, Head of Strategic Planning and Policy, Wellcome Trust
Lynn provided us with a brief overview of the Wellcome Trust's role in making data available, and their role as a leading advocate in the open access publication of data.
As a funder, the Wellcome Trust established a data management and sharing policy in 2007, which required all of their funded researchers to think about how they would maximise access to their research data with as few restrictions as possible. Researchers were required to provide a data management and sharing plan, which was peer reviewed internationally, and the Trust committed to meet the costs associated with implementing that plan. This policy was renewed in 2010 and they have since provided new guidance, drawing on resources such as the DCC's data management toolkit.
Lynn identified some of the main challenges to greater data sharing from a funder's perspective, including infrastructure issues, cultural issues (such as incentives and recognition for researchers who share their data), a group of technical challenges, a group of professional challenges and a number of ethical considerations. However, he observed that they are keen to promote research access to public data, as there is evidence to suggest sharing data does really make a difference. 
As part of their advocacy in this area, the Wellcome Trust is part of a consortium of 17 health funders, which issued a statement in January 2011 to promote research data sharing to improve public health. They have also been engaging with government departments in the UK to look at how best to open up data and promote e-health. Lynn highlighted the Prime Minister David Cameron's statement yesterday about how the government would open up some of that data, a move which has resulted from some of these discussions. The Wellcome Trust working with government departments to look at data protection legislation to make sure it is proportional and does not unduly repress research.
Lynn's talk provided us with a flavour from one funder of why data matters and the efforts that are underway to influence major decision makers in this area..
Taking care of the university's research data
Jeff Haywood, Vice Principal Knowledge Management, Chief Information Officer and Librarian, University of Edinburgh
Haywood opened by emphasising that universities have a duty of care to the future:
“We are the moral responsibility to manage research data for the present (reuse) and for the future for scholars, students and the wider public.”
He stressed that there need to be incentives at every level, outlining the “wins” that need to exist along the chain from the PhD student, the researcher, the research team, universities and supra-universities. He focussed very much on the wins, as he argued that “carrots are so much better than sticks.”
Haywood provided a structural diagram demonstrating how they are taking care of the University of Edinburgh's research data, noting that the initial pressures to do this came from the individual researchers within the university, who were concerned about open data and practical issues such as accessing their materials when on the move. He traced the history of their efforts through to the university court approved policy, which has formalised what their researchers were asking for.
He stressed that underlying their strategy is the aim to collect the smallest number of datasets possible, and get the funders to take up the others so there is a joined up record with minimal duplication. They are now working through how to put systems into place so they can offer a start-to-finish service for researchers, broken down from pre-award to post-archiving, across the University.
One of the specific “itches” he identified that researchers want scratched is research data storage, rather than the management. He stressed that we already have a mental model of what this should look like, which includes many Dropbox-like features.
Haywood concluded by outlining the challenges ahead, which include finding, finding a way of integrating with current practices in subject domains, finding enough staff with the right skills to support the service, and encouraging key people to bring others on board. He emphasised the need to stay agile, so that they don't move so far that we are out of sync with national developments.
Data Transparency: Managing the legislative risk
Andrew Charlesworth, Director of the Centre for IT and Law (CITL), University of Bristol
Charlesworth observed that legislation surrounding data and the sharing of data is now getting really complicated for the average researcher to handle. As the general public become more interested in the backstory about how data is being created, it is obvious that many researchers are not really ready for this, or ready for people to use our data in ways that we have never considered. He noted that this goes for the researchers and the institutions, who are concerned with the legal implications of letting people loose with their data and the risks that may be associated with this.  
Charlesworth feels that we need to discuss and debate our data collection and methodologies, and be clear up front about the legal and ethical risks. He stressed that we have to ask who is ultimately responsible for storing the data in the long term, and what they need to do it properly. In the past it was down to the individual researcher to keep it for a certain amount of time, but they are not always good at hanging on to data in a secure way. This becomes a significant issue if the institution needs to get to that data to answer a Freedom of Information request in the very short timescale stipulated by the legislation.
He observed that a lot of the legal and ethical issues that arise usually come about because no one is sure who is responsible for the data. It may be stored somewhere, but no one is designated as being responsible for that data. Policies are often outdated and not updated often enough, and often there is no sanction for screwing up.  
Charlesworth emphasised that we need to be reviewing our policy constantly, identifying who the should be the “data guardian” and who has oversight over them. We also need to be looking at how to make the metadata stick to the data, and educate our researchers about the legal issues.  
He observed that there is a caveat to all of this: we need to careful about undue legalisation. He noted that is not clear whether the ethics committee is the appropriate body within the university to be oversee these issues. We are also increasingly moving towards applying medical ethics to social science and it is not clear that this should be the case.
Charlesworth did acknowledge that FoI has caused a number of problems for institutions, citing the recent example of the University of Stirling's response to a request for research data into the effects of smoking. This case highlights how the university needs to know about the relevant exemptions that apply to their data and how to justify their decisions to withhold data where appropriate.
During the questions, Charlesworth was asked whether this is an issue that has an institutional solution or do we need to be looking for another solution to provide the appropriate expertise when interpreting such legal issues. He noted that there is an institutional solution, which is to train researchers from a much earlier stage in their career to understand these issues.