Hi Stevan,
I'd like to add my opinions on your 6 points:
Point 3 perhaps doesn't need to be as narrow as 'impact incentives' but
just incentives to use and rely on the database whether that be for
creating your own list of publications or for doing a more general bib
search. Although impact incentives are an excellent example of how this
might work, they are just one example and not necessarily essential from
the start - there simply needs to be some incentive to rely on the
system for something and not just be a black hole where authors send
their documents never to be seen by them again.
Point 4 (or a new point) I would also add needs to be someone or some
people who can also look out for sources (eg authors web sites, other
repositories, even hard copy archives of older papers) that contain
papers that are relevant to your institution and look at ways to add
those either manually or automatically.
For example, we download a lot of our full text direct from arXiv..an
easy way to add content and avoids having to get authors to deposit in
both repositories. We are also about to start downloading full text
papers direct from APS who have given permission to put their published
versions of our own papers in our own archive.
I think without this kind of activity we'll always be a long way from
reaching 100%. And as we know, the more content you have the more people
are likely to add their own so it makes sense to spend some time on this
to help build your coverage. Whether for new repositories it should be a
job done right from the start or one that needs to be started once you
have squeezed all your individual uploaders I can't say.
Point 5 is currently ignored here at CERN as collection of preprints
happened before any publishers had given any thought to defining what
they allowed, but this might be something we have to pick up on in
future. However, I wholeheartedly agree that with so many authors having
concerns about this it makes sense to start out like this from the
beginning.
All these points seem to me to be excellent guidelines for starting up.
Regards,
Joanne
********************
Joanne Yeomans
Office 3/1-012, DSU/SI Service
http://library.cern.ch/
Mail address:
Mailbox C27810
CERN CH 1211 Geneva 23
Switzerland
Tel: 70548 (externally dial +41 22 76 70548)
> -----Original Message-----
> From: Repositories discussion list
> [mailto:JISC-REPOSITORIES_at_JISCMAIL.AC.UK] On Behalf Of Stevan Harnad
> Sent: Thursday, January 19, 2006 4:21 PM
> To: JISC-REPOSITORIES_at_JISCMAIL.AC.UK
> Subject: Learning from the successful OA IRs
>
> Paula Callan of Queensland University of Technology (QUT)
> http://www.eprints.org/openaccess/policysignup/
> has summarised some extremely useful data on the actual
> efforts and costs that were involved in setting the QUT OA
> self-archiving policy onto it's now successful and
> unstoppable course. (I hope Paula will post it to this Forum.)
>
> The data, especially on QUT's false-starts and how they were
> eventually remedied are especially instructive. It is ever so
> important that other institutions contemplating costs and
> strategies learn from QUT, so they can emulate its success
> and also spare themselves the time and effort and expense
> that were expended on options that did not work or proved unnecessary.
>
> My own reading of the QUT history is this: that at first QUT
> did too little, and then it did too much. All that are really
> needed are the following six components, the first three (*)
> being long-term ones and the last three being only for the
> start-up years:
>
> *(1) An OAI-compliant IR, set up and maintained
>
> *(2) An official institutional policy requirement to deposit the
> final accepted, peer-reviewed draft (not the publisher's PDF) as
> an institutional record-keeping matter: a fulfillment condition
> for annual review, for research assessment, and for standard CV
> creation/submission
>
> *(3) Impact incentives (clear, easily accessible and displayable
> download/citation statistics, clear information on the OA citation
> advantage and its link to promotion and research income; possibly
> some initial token start-up incentives offered by the
> institution too,
> monetary or material)
>
> (4) Activist librarian support for approaching authors to elicit
> the final draft for depositing, and even doing the
> deposit for them
> if necessary (needed for initial start-up years only)
>
> (5) Allowing restricted-access deposits as an option, and offering
> to take any copyright concerns off the author's hands, with the
> library doing any necessary checking or follow-up
>
> (6) Minimizing all copyright checking and follow-up, by making
> open access the default access-setting for all deposits, pending
> any request-to-remove from the publisher, and restricted access
> reserved only for special cases (with the automatic email-eprint
> option as a back-up)
>
> My guess is that if the optimal strategy is adopted from the
> outset, skipping the false starts, the start-up cost will
> prove much lower. The long-term maintenance costs, once the
> depositing practice is familiar and habitual, will be lower still.
>
> I suggest that the architects of the few successful,
> experienced *OA* IRs to date (CERN, Minho, Southampton.
> SURF/DARE) make a strategic document summarising the
> components that were essential to their success, with
> supporting data and cost/time estimates for start-up and
> long-term, based on their actual data and experience It would
> be a good information source to synchronise with the RCUK
> forthcoming policy announcement.
>
> (Of course, these considerations are all institution-based:
> Once funder self-archiving mandates such as RCUK's are
> adopted, they will greatly speed up and simplify the start-up
> process for the institutions.
> The Wellcome Trust's policy is probably helping already.)
>
> Stevan Harnad
>
Received on Thu Jan 19 2006 - 16:31:11 GMT