On Wed, 13 Feb 2006, Sarah Kaufman wrote:

> having spoken to academics within this institution, it has become
> apparent that potential depositors may be wary of depositing into a
> digital repository as they fear that a repository that includes pre-prints
> may not appear 'credible'.
> 
> Has anyone else dealt with this sort of concern, and how you responded
> to those that have voiced this concern? Do any repositories exclude
> items that have not gone through the peer-review process? If you accept
> items that have not gone through the peer-review process, do you apply
> any forms of quality control on the item?

This can save people a lot of time that will otherwise be wasted re-inventing
this superfluous wheel:

(1) The right way to make the distinction between published, peer-reviewed
material and unpublished material is the classical way: by *tagging* it as such.

(2) The IR softwares have tags for peer-reviewed articles as well as for
unrefereed preprints.

(3) The scholarly/scientific community is quite aware of this distinction;
it has already been dealing with it for years in the paper medium,
in the form of published articles versus unpublished drafts.

(4) An IR is not a publication venue -- it is a means for providing
*access* to published -- and, if the author wishes, unpublished -- work.

(5) Any user who wishes to reserve their time and reading to
peer-reviewed, published work can do so; they need only note the tags (is
it "peer-reviewed"?  is it "published"? what *journal* is it published in?)

(6) Disciplines differ in the degree to which they use pre-referring preprints:
physics relies heavily on them, biology less. This is a choice for researchers
to make, both as authors (deciding what to deposit) and as users (deciding what
to read).

(7) This decision cannot and should not be made a priori by IR managers.
An IR deposit is not a publication, any more than a mailed first draft 
on paper is. It is a decision by an author to provide, and by a user to
use, a document.

(8) The most absurd thing of all would be to institute IR-level system
of "quality" control: Leave that to the peer specialists and the journals. 
IRs are just access-providers.

(9) It can and should, however, be decided whether an IR is for research
output only (documents and data, whether pre- or post-peer-review)
or it is also for non-research output (e.g., teaching materials). Some
IRs that are sectored by subject matter will also want to decide what
disciplines they are catering for.

(10) The right thing to tell naive researchers who have never self-archived or
never use and OA IR, is that an OA IR is neither a publication nor a library
catalogue of publications: It is a means for researchers to maximize access to
their research output, both before and after peer-reviewed publication.

See the well-worn self-archiving FAQs on these questions:

    http://www.eprints.org/openaccess/self-faq/#What-is-Eprint
    http://www.eprints.org/openaccess/self-faq/#7.Peer
    http://www.eprints.org/openaccess/self-faq/#5.Certification
    http://www.eprints.org/openaccess/self-faq/#6.Evaluation
    http://www.eprints.org/openaccess/self-faq/#2.Authentication
    http://www.eprints.org/openaccess/self-faq/#3.Corruption
    http://www.eprints.org/openaccess/self-faq/#23.Version

Stevan Harnad