Nov 152011

After five days of workshops, break out sessions, seminars and groups tasks at RDMF7 and DCC Roadshow Cambridge, this JISC project manager is now well versed in the language and lore of Research Data Management (RDM). As is often the case with these kind of events, you come away with more questions than answers, and having embarked with a fair idea of what I was doing, the RDM ‘horizon’ receded at times into a haze of unresolved issues. To bring back some perspective and to report back to my colleagues at University of Hertfordshire I did some reflection. This is a cut down version for public consumption.

On the RDM landscape:

A) There is nothing really new in RDM, it is just an extension of professional and rigorous research practice. This ought to be motivation in itself.

There is however, a new impetus which makes RDM a hot topic across HE. The drivers are diverse and many:

– funders policy (for example, RCUK Common Principles on Data Policy, EPSRC’s near term expectation of compliance with their data policy )

– increasing number of Freedom of Information requests

– instances of reputation damage as a result data loss or security failure (cf. UEA ‘climategate’)

– the data deluge, the ever increasing volume and shape of data creation

B) There is a big gap between funder’s and regulator’s policy and institutional practice. This is true right across the sector and in the large majority of institutions.

C) RCUK is clear: RDM in an institutional responsibility. How we go about distributing that responsibility is a key issue to resolve.

D) RDM plans will be a mandated element of funding applications in the near future.

E) There is a lack of policy at the institutional level, and RDM skills at the investigator level. Right now, the capacity to deliver policy is limited which ever way you look.

F) There are some well served areas, where long term curation is built into the research lifecycle, generally driven by the existence of a discipline specific data repository.

G) Strategists think Researcher’s motivation (and early training), rather than technology, is the primary key to bridging the gap. Counterpoint this with the small group practitioner’s view: technology remains an issue because it is often too complex or in short supply.

H) I think RDM is a two level activity, and will remain so: robust, practical measures during the conduct of research (by Researchers); and specialist, conditioning measures for long term preservation and access (by Curators and information professionals).

I) The unaddressed elephant in the room is cost. Cost of bytes, cost of effort; cost of doing and cost of not doing (probably in order of increasing significance).

On the #rdtk_herts project:

J) participation in JISC MRD programme is a great opportunity. The DCC and MRD phase one projects are at the vanguard, but these represent a small wedge in a big space. There is with plenty of room to learn and contribute.

K) University of Hertfordshire has a specific and public data policy and senior management buy-in to pursue good RDM. We seem therefore, to have something to build ‘the bridge’ to.

L) The DCC is the conduit to a lot of great resources and practice. There is already a wealth of great guidance and RDM planning tools. The shortest route to the guidance element of our Research Data Toolkit will be to make existing generic advice specific to UH, rather than the seeking within our own walls.

M) work package WP1 should focus on deploying, with direct DCC assistance, two of their tools as widely as possible. Cardio to assess institutional capacity for RDM. Digital Asset Framework to audit data and needs. We can learn from these processes so that we have our own best practice audit capacity to use later on in this work, and more generally.

N) the use of external cloud services in work packages WP2, WP3 and WP8 chimes a clear chord in the MRD programme. JISC UMF and MRD phase one projects have built some cloud services, for example, ViDaas, which are being offered to HE for trials. The long term hosting arrangements for these are unclear (to me at least), so our proposed work with HRC3 to shed light on the cost and benefit of this may be an important contribution, and offer an alternative or complimentary solution to Eduserve.

O) In work package WP8, the proposed use of Shibboleth authentication layered on DSpace, also looks like a useful piece of work. The focus around access issues in the MRD programme seems to be on discovery and long term value, enabled by metadata, in the entirely ‘open’ arena. Our work will address instances where commercial or ethical sensitivities require a level of federated protection.