Absolutely, Angus - and I'm not supporting researchers handing over their data to publishers at all  .. But rather the (inter)national infrastructure development that indeed JISC, DCC and others are and should continue to support. 

And I've never said CRIS are the answer to everything ... But just now for data I'm haven't yet been convinced we should invest in anything else at the institutional level .. But that may well change in the future

All the best

Anna

Anna Clements
Head of Research Data and Information Services
University of St Andrews
St Andrews, Fife,KY16 9AL
@annakclements

> On 20 Jun 2014, at 11:53, "Angus Whyte" <[log in to unmask]> wrote:
> 
> Anna,
> 
> I just wanted to add a few comments on your earlier post...
> 
> -----
> On 19/06/2014 19:45, Anna Clements wrote:
> Angus, all .. I think collecting info on which systems are being used for what is very useful ..and extending the UCISA list to cover research support systems or all kinds would be very useful and as I mentioned I think the landscape here in terms of new and enhanced systems will change pretty rapidly.
> 
> On 'long-tail' research ... I suppose my nirvana is that institutions shouldn't have to worry about this as the subject or national data centre infrastructure should be developed to support this ... but I recognise this will take time and leadership.  But surely this is more cost-effective for the sector than everyone doing it themselves at institutional level?
> 
> With research data also, given how it is so intrinsically part of the research process ... and how it differs from discipline to discipline in terms of how it is manifested and created/analysed/used then surely disciplines are better placed to provide infrastructure and services that the researchers themselves will find relevant
> 
> ... I just keep seeing that headline response from our academic-led review on research data
> 
> 'We need to avoid the idea of a data repository as "a place where data goes to die"'
> 
> -----
> A great quote and I doubt anyone would disagree with it, but I see a risk of exactly that in institutions that don't act like they really mean that "data is an asset". There are good reasons to deposit in disciplinary data centres, not least of which is that they really do treat it as an asset.
> 
> I agree with you that more discipline-based support from more data centres and learned societies would be helpful. But many of the same researchers who will claim only disciplines have the know-how will be the very same who turn to Google or Dropbox- neither exactly disciplinary solutions, but generic ones that deal intelligently with specific requirements.  So it also makes sense for institutions to collaborate on RDM solutions, and for DCC and JISC to develop cross-sector ones (e.g. the Research Data Registry and Discovery Service),  and make it easier to deploy commercial ones, as Janet framework agreements do.
> 
> Institutions have a lot to deal with just to drag the majority of researchers into handing over their data for the sake of compliance with funder mandates. Some of those researchers may be happy to see their data stuck on an online shelf with a few basic descriptors that allow others to cite it. Others will want to make their data dance, sing, and shake it all about. Why should they have to hand it over to publishers to give them nice analytic and visualisation tools, recognise their vocabularies, or integrate their workflows with library systems?
> 
> Why shouldn’t institutions themselves benefit from having good data co-located with good researchers, as they currently do just from having good researchers co-located with each other (and good facilities)? Ideally and maybe eventually, institutional support services should be able to help their researchers exploit the opportunities in mining and reusing data that institutional colleagues have produced, to make new connections with them and draw on their tacit knowledge of how, where and why.
> 
> I’m sure that grant information does impart some ‘context’ but a small fraction of what researchers typically need if they really to reuse data. And just because much of it should be open to public access doesn’t mean local know-how will count for nothing. Truly open reproducible research is a great goal, but it is someway off being a reality.
> 
> It’s still early days for RDM and as you say the landscape is changing fast. Whatever new variations on the data repository theme emerge I’m sure requirements will become more elaborate, not less, and that CRIS systems will be supplying some of the glue to stick solutions together. I'm sure you'd agree they're not the whole solution.
> 
> Angus
> 
> 
> 
> -- 
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.