Friday, 23 April 2010

Report on the 'Institutional Policy and Guidance for Research Data' Workshop

'How to share expertise? Where to get advice'? Just two of the questions institutions need to address in their research data management policies according to Paul Taylor of Melbourne University. On the 29th March 2010, the place for advice and sharing expertise was the EIDCSR Institutional Policy Workshop in Oxford.

A significant part of the ‘Embedding Institutional Data Curation Services in Research’ Project has been to start developing an Institutional research data management policy for the University of Oxford, so this workshop offered us a chance both to say how things were going and find out the lessons learnt from others farther down the road.

The University of Melbourne has been grappling with the issues for some time already, and we were lucky enough to be joined by several of their representatives via videoconference. Indeed, given how close we were to not being joined by their representatives due to the videoconferencing equipment, ‘lucky’ is the operative word. Paul Taylor stressed that any effective policy needs to be implementable. This involves getting the researchers themselves involved in the development process and offering somewhere where people can go for information and advice. Compliance becomes easier the more central services exist, leaving researchers to do the research.

Another university which has already done a lot of work on data curation is Southampton, and Kenji Takeda introduced their long-term ambitions. The unfortunate incident a few years back when Southhampton’s Mountbatton building burnt down led to claims against the lost research data, so this has perhaps focussed minds more there than in other institutions. A cost-benefits analysis is now being undertaken which should help institutions better appreciate the value of their data outputs. Furthermore, they are looking to make data management courses compulsory. Herding academics into classrooms sounds ambitious, but there was a general sense from the workshop that without training there was little chance of persuading researchers to adopt best practices.

Jeff Hayward, from the University of Edinburgh emphasised that when it comes to data curation it is better to identify the opportunities than enumerate the problems, but then failed to ignore the various ‘inhibitors’ to good data management. “Researchers want data management, but don’t want to do it.” Quite. Nevertheless, Edinburgh are bravely forging ahead, setting up an experimental ‘DataShare’ service and adapting the Digital Curation Centre’s 101 training, with the intention of making it compulsory for doctoral students.

Finally, David McAllister of the BBSRC explained data management policies from a Research Council’s point of few – clearly a key driver for institutional policies.

Perhaps the last word should go to Jeff Hayward who concluded the panel questions session by indicating that the world would actually be a happier place if there were fewer data repositories. Individual universities should really act as repositories of last resort, but the onus must be on them to guarantee that research data is not lost or rendered inaccessible.

For a more complete report on the workshop, plus the various sets of slides used by the presenters, go to the EIDCSR Project website: