Universities and their IRs Can Help Monitor Compliance With Funder Mandates

SUMMARY: There just might be some hope that UK's Research Funding Councils -- all seven of which now mandate Green OA self-archiving, as recommended by the UK Parliamentary Select Committee on Science and Technology in 2004 -- could go on and take the initiative to stipulate that each fundee's Institutional Repository (IR) is to be the default locus-of-deposit (with DEPOT as the interim back-up). If adopted by the UK Funding Councils, this small change in implementational detail has a good chance of motivating all UK universities and research institutions to adopt Green OA self-archiving mandates too, for the rest of their research output. This UK model will then undoubtedly propagate globally, to bring the planet universal OA at long last!

---

Gerry Lawson [GL] (NERC Research Information Systems, RCUK Secretariat) wrote (in JISC-REPOSITORIES):

GL: Stevan, a very useful series of postings - thanks. UK Research Councils have a variety of OA mandates - including two which mandate deposition in CRs (MRC- UK PubMed and ESRC - Society Today).  WIth the exception of EPSRC (and this may well change) the others do mandate deposition, but are unspecific about where.  NERC, for example, says:

"From 1 October 2006 NERC requires that, for new funding awards, an electronic copy of any published peer-reviewed paper, supported in whole or in part by NERC-funding, is deposited at the earliest opportunity in an e-print repository.  NERC also encourages award-holders to deposit published peer-reviewed papers arising from awards made before October 2006. "

BUT its very difficult to check compliance to these mandates!  Councils have reduced their final reporting requirements on the expectation that it will be possible to collect outputs information (not just publications) electronically from grantholders.  RCUK is assessing options for doing this - either pushing/pulling from Institutional Repostories or from HEI CRIS systems, or both.  Whatever is decided its certain that that we'd be assisted by inclusion in IRs of metadata fields for a) "Funder" (perhaps using a dropdown list of funders URIs); and b) "GrantReference".  

Gerry, you are absolutely right. IRs need to have a metadata field that specifies the funder, for a variety of reasons, including verification of grant fulfillment conditions.

(As you note below, the EPrints IR software has already implemented this metadata tag.)

This is also yet another strong reason why funders should not require direct deposit in a CR, nor even simply require open-ended deposit in any repository at all (as NERC does), but should instead specify the author's own institutional IR as the designated locus of deposit (and DEPOT for those fundees whose institution has not yet set up its own IR).

Universities are already eager to do everything they can to help in ensuring compliance with funders' grant conditions. They can accordingly be invaluable aids to each funding council in verifying compliance with its deposit mandate. See: "How To Integrate University and Funder Open Access Mandates".

GL: The disadvantage of using IRs rather than Central Repositories is the absence of minimum standards and formats in the former.  Both the above fields exist in CRs (e.g. UK PubMed and Society Today)

But the standards and formats can all be implemented in IRs. EPrints is continuously upgrading its functionality to keep pace with the emerging needs of Open Access (including Open Access mandates by funders and institutions). 

Don't forget that two free IR softwares -- EPrints and DSpace -- are used to create the majority of IRs. IR software standards can be made widespread or even universal (as OAI-PMH, for example, was made) in the distributed worldwide IR community with a resultant power, scope and functionality that can not only match but exceed what can be done with CRs -- and without any of the disadvantages of CRs that Professor Rentier, author of the U. Liège mandate, and I have both described.

GL: So, three questions re IRs

1. Funder and GrantRef fields exist in EPrints (as free text) from version 3.0  - do they exist in DSpace and Fedora - and in what form?

I don't know. But EPrints -- which is the first of the IR softwares and invariably the leader in keeping upgrades lock-step with the emerging needs of OA -- will contact DSpace and Fedora developers, as it has in the past (most notably with the all-important  "request a copy" Button) to urge them to implement the GrantRef field too. (Meanwhile, institutions should just adopt EPrints!)

GL: 2. Can a standard be introduced where they allow multiple funders - like multiple authors? (its unlikely we'd want to be as sophisticated as adding a %DueToGrant field!)

I can't see any reason why not. I am branching this to Les Carr, who will be able to reply. (Perhaps it has been implemented already.)

GL: 3. If Councils were to add to their mandates a sentence like:  'By [date] such records should be tagged with Funder and Grant Reference information, and made available for harvesting', what would be an appropriate [date].  I guess this is depends on the harvisting tool.  I'm told that standard OAI-PMH doesnt handle these fields and that SWAP is not widely used?  What is the best approach?

For the technical answer, I defer to Les Carr and the EPrints development team.

But for timing, the question is slightly more complicated: The Councils should specify that the deposit must take place immediately upon the date of acceptance for publication. This date will vary from paper to paper, of course, so it cannot be specified in advance, but it is the most natural, reliable and universal reference point for authors and funders to use to time their deposit. See: Optimizing OA Self-Archiving Mandates: What? Where? When? Why? How?

With IRs (as long as we ensure that they provide the requisite functionality), harvesting need not be restricted to only metadata OAI fields. Again, I defer to Les, but the EPrints and DSpace metadata fields should surely be uniformly detectable and automatically harvestable regardless of whether they are part of the OAI protocol. (Les?)

GL: Additionally, some Councils mandate deposition only 'where a suitable repository exists'.  Should we change this to something like 'where a suitable Institutional Repository does not exist it is expected that the JISC-supported repository of last resort, 'The DEPOT' , will be used.'?

Yes, definitely! That will at last breathe some life into DEPOT so that it can at last begin to be used for its intended purpose, which was precisely that!

I am ever so grateful for your reply, Gerry, because it shows not only that the Funding Councils are listening, but it confirms how important and fruitful convergent mandates can and will be. Much gratitude also to Professor Rentier, Rector of University of Liège, whose timely and perspicacious essay on the relation between IRs, CRs, and between institutional and funder deposit mandates has triggered all this constructive discussion and coordination.

Stevan Harnad
American Scientist Open Access Forum