Newly Enhanced Registry of Open Access Repositories (ROAR) Stevan Harnad 11 Jun 2005 00:22 UTC

             ** Apologies for Cross-Posting **

Southampton's worldwide Registry of Institutional Open Access Archives
created and maintained by Tim Brody and now in existence for about two
years has just been enhanced (and will soon be enhanced further) by Tim.

The Registry covers the 434 OAI-compliant OA Archives worldwide whose
existence has thus far been made known. As it is intended to cover OA
content (i.e., full-texts of either (1) preprints or (2) postprints of
peer-reviewed journal articles, or (3) dissertations), Archives that
do not provide *any* full-text content at all (only metadata), or that
provide only content of other kinds (internal documents, courseware,
library records, audio, video, software) are not covered -- though
archives of *mixed* content (both OA and non-OA) are covered. The Registry
also covers only archives that are OAI-compliant, hence interoperable,
harvestable using the OAI harvesting protocol.

[The Registry's intentional non-coverage of non-OA Archives is the
reason why it covers only 434 Archives with about 3.5 million records
(estimated from the fact that its 2.5 million records come from the
312/434 Archives that celestial can currently harvest) rather than
the full 487 Archives and 5.5 million records covered by OAIster ]

Southampton's Institutional OA Archives Registry not only tracks the
growth in the number of OA Archives worldwide across time, but it also
tracks the growth in their contents (i.e., the number of records in each
archive, across time), showing growth charts for each Archive, as well
as across Archives.

The size and growth data are also classified by the type of Archive:

    (i) Distributed Institutional/Departmental Pre-/Postprint Archives (212),
    (ii) Central Cross-Research Archives (55)
    (iii) Dissertation Archives (e-theses) (54)

as well as

    (iv) database Archives (e.g. research data) (8)
    (v) e-journal/e-publishing Archives (39)
    (vi) demonstration Archives (not yet operational) (24)
    (vii) "other" Archives (non-OA content of various kinds) (42)

Where the content is mixed, and there is *some* OA content, the Archive
is classified as (i) - (vi).

The Archives are also classified by country and by the kind of software
used to create them.

There are also data on the total number of records in each individual
Archive, the total number of records across all Archives (and Archives
in each category) and the average number of records per Archive and
Archive category.

The record numbers are somewhat deceptive, however, because, as noted,
not all records correspond to OA objects (pre/postprints and theses), and
not all OA records are full-text. In fact, the majority of records are
almost certainly *not* OA full-texts. Estimates of the proportions will,
we hope, soon be available, as Tim works on an algorithm for identifying
generating separate counts for OA full-texts.

In addition, for 122/434 Archives their records cannot yet be
counted and monitored at all because they are not yet harvestable by (Archive managers are strongly urged to
consult the Registry and if their records are not being harvested by
celestial, to contact Tim to remedy this: sometimes the problem is that
they have not provided the right OAI Base URL, or any at all).

The Registry also makes it possible to make some very interesting
projections, predictions and comparisons, but as you make them,
please bear in mind that they are still tempered by one major source
of uncertainty: If two Archives or Archive categories differ in their
proportion of OA full-texts, you may be comparing apples and oranges
when you compare the size and growth curve of their contents!

Needless to say, any OA Archives that are not yet in the Registry are
urged to register at
and to ensure that they are harvestable by celestial.

Currently, 15% of the annual 2.5 million articles published in the world's
24,000 journals have been made OA by being self-archived by their authors.
The Archives also contain dissertations, non-OA material, and non-full-texts
(metadata only), but their overall percentage of full-text OA is probably
higher than 15%, and for some individual Archives and kinds of Archives,
a good deal higher than 15%.

Note also that the two fullest and fastest-growing institutional
archives of all in terms of annual full-text OA content:

    I: the CERN Archive


    II: the Southampton ECS Archive

happen to be the only two Archives at institutions that have adopted
and implemented a mandatory self-archiving policy for all of their own
annual journal article output. Both are now self-archiving over 90%
of their research output.

The implication is obvious: Other institutions should adopt mandatory self-archiving
policies too:

The demonstrated success of the CERN and Southampton policies -- together
with the results of the latest survey of authors across disciplines
and worldwide, which found that 81% of authors respond that  they will
self-archive *willingly* if required (but not if not); 14% say they will
comply reluctantly and only 5% say they will not comply

        Swan, Alma and Brown, Sheridan (2005) Open access self-archiving:
        An author study. Technical Report, Joint Information
        Systems Committee (JISC), UK FE and HE funding councils.

and both CERN and Southampton have services to do the few keystrokes
for those authors who feel they haven't the time:

        Carr, L. and Harnad, S. (2005) Keystroke Economy:
        A Study of the Time and Effort Involved in

        "Let Us Archive It For You!"

So (a) have a look at the new, enhanced Institutional OA Archives
Registry, (b) register your own Institutional Archive if it's not in
there, (c) create one (or persuade your institution to create one)
if you don't have one yet -- it's cheap and easy

       Sponsler. E. & Van de Velde E. (2001)
       Software: A Review.  SPARC E-News, October 2

especially relative to the dramatically increased research impact (and
research impact income) that OA has been demonstrated to provide

and urge your institution to adopt a self-archiving policy if it
hasn't one, preferably one that requires self-archiving rather than
merely encouraging it. What needs to grow is not just the number of
institutions with OA Archives (212) but the number with institutional
OA self-archiving policies:


Stevan Harnad

P.S. I'm not really proposing to call the Registry "ROAR"! The
term was chosen advisedly, however, to register clearly the
fact that the Registry exists, has existed for some time,
and is conveying a steadily growing amount of innovative and
important information. It needs to be built upon, not bypassed.
(Cf. )

A complete Hypermail archive of the ongoing discussion of providing
open access to the peer-reviewed research literature online (1998-2005)
is available at:
        To join or leave the Forum or change your subscription address:
        Post discussion to:

UNIVERSITIES: If you have adopted or plan to adopt an institutional
policy of providing Open Access to your own research article output,
please describe your policy at:

    BOAI-1 ("green"): Publish your article in a suitable toll-access journal
    BOAI-2 ("gold"): Publish your article in a open-access journal if/when
            a suitable one exists.
    in BOTH cases self-archive a supplementary version of your article
            in your institutional repository.