Wednesday 24 August 2011

CRIS + Repositories at UK Universities

On the back of this recent post I've been invited to speak at the RSP Autumn School about whether the developing CRIS / repositories landscape at UK Universities might present an opportunity to re-focus on Open Access.

Since the RAE2008, the uptake of integrated electronic research administration systems (CRIS, ERA, RMAS...choose your favourite acronym / abbreviation) has been dramatic with the primary driver being to oil the administrative wheels of the REF in 2013.

As the ever-growing UKCoRR membership attests (currently 254 members) Institutional Repositories are now well established across the HE sector and there are several approaches that institutions are taking to utilise this existing repository infrastructure for research administration and/or embed their repository as a component of a broader research administration infrastructure with many either implementing additional commercial software or developing a bespoke solution in-house.

A significant initiative in this area is the JISC funded
RePosit project (final report due in October 2011) which aims to "increase uptake of a web-based repository deposit tool embedded in a researcher-facing publications management system" and comprises a consortium of 5 institutions (University of Leeds , Keele University, Queen Mary University of London, University of Exeter and University of Plymouth) in partnership with Symplectic Ltd as a commercial partner. The project runs a Google group at http://groups.google.com/group/reposit where there has been a very active thread recently discussing this developing environment; I have used the thread to collate a list of CRIS + repositories at UK institutions and set up a public Google doc if anyone would like to add their institution (I won't post the link here but it's already been shared via the RePosit and UKCoRR mailing lists.)

As can be seen from the list so far, the most common solutions are commercial software implementations of Atira Pure (11 instances) and Symplectic Elements (16 instances) [+4 instances of Avedas Converis and 1 bespoke]. Though this hastily compiled document almost certainly reflects the membership of the RePosit Google group* with only 32 institutions so far represented out of 142 Institutional Repositories in the UK listed on OpenDoar I think it's still a big enough sample to be significant, especially as Symplectic, Atira and Avedas are arguably the only real options currently in the market-place (notwithstanding the ongoing development of CRIS-like functionality into EPrints itself which is by far the most popular repository platform in the UK (run by 71 UK institutions listed on OpenDoar - exactly half of the total.)

* Similar information is also being captured on the new RSP wiki -http://www.rsp.ac.uk/pmwiki/index.php?n=Institutions.HomePage

Both Symplectic and Pure are designed to enable research staff to manage their research profile both manually and by pulling data from online databases via their APIs (Web of Science, SCOPUS, Mendeley etc) but one observation that is worth commenting upon is the different core functionality of the two systems with Symplectic providing a modular solution designed to integrate with an existing repository whereas Pure is arguably more fully featured software, capable of managing full-text and with full version control, functionality to manage embargo, visibility of items based on business rules, and fully indexable by search engines. Indeed, the comprehensive nature of Atira Pure raises the intriguing possibility that it could effectively replace a repository altogether; at this stage, however, I believe the majority of institutions running the software have chosen to integrate with an existing repository in the Symplectic model – a good example is the University of St Andrews who are running Pure – research portal here - alongside their DSpace repository - http://research-repository.st-andrews.ac.uk/ - which is just used for full-text.

There are some very interesting perspectives on the RePosit thread on why we should or should not maintain two systems (CRIS + repository) and it is clear that the decision will depend to a large extent on the particular systems at a given institution and their specific configuration. Janet Aucock of St Andrews suggests that the integration of the best features of both CRIS and IR will evolve over time going on to say that an important consideration should be "not to lose flexibility and options. Teams across research offices and libraries can be well coordinated, communicate well and have regular contact and debate"..."But even then...there can be differences in emphasis in what the CRIS is about and what service it offers. The Library tends to emphasise open access and discovery. Research office is undoubtedly more focused on research assessment. (Also see James Toon’s comment on this blog).

Arguments in favour of retaining repositories include the issue of creative arts research outputs and Jackie Wickham of the RSP (and now UKCoRR secretary) points out that “there has been considerable investment in developing repositories (focused on EPrints, KULTUR plug in) to enable them to showcase this type of research e.g. UAL - http://ualresearchonline.arts.ac.uk/. The visual impact of the repository is critical to arts researchers and many institutions are using the KULTUR plug in – and not just the specialist arts ones”. I think this is a persuasive argument, also for those using their repositories for "non-REF" output / grey literature / Open Educational Resources (though I have posted elsewhere that I’d actually like to investigate the CRIS model for managing OER as there seems no fundamental reason why such a system could not be used to support the workflow for both OA research and OER.

As to whether all of this does indeed present an opportunity to re-focus on Open Access I think is still a moot point. There is perhaps a danger that the administrative burden of the REF will overshadow the objective of providing Open Access to research but there is also the opportunity to integrate the various infrastructural components in such a way to facilitate what are ultimately complementary objectives; to increase visibility of institutional research, improve awareness and advocacy initiatives around OA (and OER) and to more effectively link institutional research administration with access to the actual research outputs.



4 comments:

  1. Hi Nick,

    Good post. My view is that it doesn't matter which of the many buckets you put the data in as long as you get out the info you need. There are many ways to skin the cat. It is the quality (ok quantity too ;-)) of the data that matters. Look forward to hearing how things develop.

    ReplyDelete
  2. Hi Nick,

    Pure can integrate with repositories, as you write. That has been our suggested way forward for years, Pure's integrations with ePrints and DSpace has long been standardised, for example. But institutions increasingly inquire about Pure replacing the IR. Some seem driven by savings, others aim to simplify their IT infrastructure and internal business processes. Only one institution has made the decision so far though, that is University of Dundee. The relevant contact person at University of Dundee is Hannah Whaley, Assistant Director (Research and Systems), Library and Learning Centre (h.whaley@dundee.ac.uk, details posted here with permission). To our knowledge, 2-3 other UK institutions are considering the same.

    Our position is, that we are happy to set up integrations with IRs, but that we also find projects like Dundee's interesting. We have developed features in Pure to support such projects for years; the recent addition of improved Open Access support is one example:

    http://www.atira.dk/en/pure/functionality/#open.access.awareness.and.support

    Sufficient functionality in the CRIS to act as an IR is only one of several problem areas when an institution is making the decision though, at least in our experience.

    There is also the question of what will happen to IR managers' jobs when a CRIS takes over populating the IR or if a CRIS takes over the IR function entirely. Our observation is that People sometimes fear greater change than actually take place. Our experience is that the same IR staff will work in Pure with similar tasks:

    1. Enforcing the rules of individual journals; checking licenses, setting embargo dates, making sure fulltext is the right version

    2. Managing the pool of Journals and Publishers in Pure

    3. Validating bibliographic information entered by academics and/or clerical staff

    4. Handling duplicates - not just Publications, but also other content types such as External Organisations and External Persons. They are important CRIS content types, but vulnerable by nature to data quality problems and therefore can be maintenance heavy.

    5. Provision of reports and statistics to management and other stakeholders; sometimes bibliometric reports but also in other areas

    Another observation of ours is, that some IRs may have been populated centrally by repository managers rather than by academics, although that may not have been the vision initially. And that can change if Pure takes over the IR function because it is good at constituting a distributed model for population (if that is desired, but it often is). Largely, Pure succeeds in getting academics on board: I would guess because of the richer offering to academics (e.g. richer online profiles and CVs including PGR supervisor relations, personal grant info, projects, esteem, external collaborations, etc.) and because top-level managers often are involved in Pure projects personally. Rector herself took up Pure at one university, for example.

    When academics take up Pure as "their" tool, it entails a request for user support that might be different from what was requested before, when users interfaced directly with a repository system. One consequence for “former” repository mangers can therefore be less time spent populating the CRIS and more time spent supporting and guiding a broader user community. Academics and other users want to be supported when using the system but also when classifying data, choosing submission categories, suggesting REF outputs, and in many other situations, where qualified support can be the difference between achieving high quality data or not. The point here may be that if anything, repository managers are in more demand when introducing a CRIS, not in less demand.

    Anyway, great article and your description of Pure’s current technical abilities with regards to IRs is absolutely correct.

    Atira

    ReplyDelete
  3. Posting this comment on behalf of Atira:

    Pure can integrate with repositories, as you write. That has been our suggested way forward for years, Pure's integrations with ePrints and DSpace has long been standardised, for example. But institutions increasingly inquire about Pure replacing the IR. Some seem driven by savings, others aim to simplify their IT infrastructure and internal business processes. Only one institution made the decision so far though, that is University of Dundee. The relevant contact person at University of Dundee is Hannah Whaley, Assistant Director (Research and Systems), Library and Learning Centre (h.whaley@dundee.ac.uk). Details posted here with her permission. To our knowledge, 2-3 other UK institutions also consider having Pure replace the repository.

    Our position is, that we are happy to set up integrations with IRs, but that we also find projects like Dundee's interesting. We have developed features in Pure to support such projects for years; the recent addition of improved Open Access support is one example (http://www.atira.dk/en/pure/functionality/#open.access.awareness.and.support).

    Sufficient functionality in the CRIS to act as an IR is only one of several problem areas when an institution is making the decision though, at least in our experience.

    There is also the question of what will happen to IR managers' jobs when a CRIS takes over populating the IR or if a CRIS takes over the IR function entirely. Our experience is that the same IR staff will work in Pure with the same and new tasks:

    1. Enforcing the rules of individual journals; checking licenses, setting embargo dates, making sure fulltext is the right version

    2. Managing the pool of Journals and Publishers in Pure

    3. Validating bibliographic information entered by academics and/or clerical staff

    4. Handling duplicates - not just Publications, but also other content types such as External Organisations and External Persons. They are important CRIS content types, but vulnerable by nature to data quality problems and therefore can be maintenance heavy.

    5. Provision of reports and statistics to management and other stakeholders; sometimes
    bibliometric reports but also in other areas

    Another observation of ours is, that some IRs may have been populated centrally by repository managers rather than by academics, although that may not have been the vision initially. And that can change if Pure takes over the IR function because it is good at constituting a distributed model for population (if that is desired, but it often is). Largely, Pure succeeds in getting academics on board: I would guess because of the richer offering to academics (e.g. richer online profiles and CVs including PGR supervisor relations, personal grant info, projects, esteem, external collaborations, etc.) and because top-level managers often are involved in Pure projects personally. Rector herself took up Pure at one university, for example.

    When academics take up Pure as "their" tool, it entails a request for user support that might be different from what was requested before, when users interfaced directly with a repository system. One consequence for “former” repository mangers can therefore be

    less time spent populating the CRIS and more time spent supporting and guiding a broader user community. Academics and other users want to be supported when using the system but also when classifying data, choosing submission categories, suggesting REF outputs, and in many other situations, where qualified support can be the difference between achieving high quality data or not. The point here may be that if anything, repository managers are in more demand when introducing a CRIS, not in less demand.

    Anyway, great article and your description of Pure’s current technical abilities with regards to IRs is absolutely correct.

    ReplyDelete
  4. Should say the comment is from Bo Alroe Product director at Atira - thanks Bo and sorry for problems commenting!

    ReplyDelete