On 12 Oct 2008, at 13:54, Andrew A. Adams wrote:
The main driver for this seems to be the REF and the need to
potentially track all the
output of our researchers. At this stage our PVC(Research) is still
somewhat
unsure of the nature of the non-technical elements of an IR, i.e.
about the
language of and necessity for a deposit mandate. I therefore need to
make a
decisive pitch for a mandate.
...
We have found the UK's REF to be a very powerful driver for IRs but it
can drive itself in the wrong direction - at a tangent to OA - and can
result in a repository full of metadata (a pseudo-CRIS) testifying to
the facts of publication but denuded of the publications themselves.
I'd like to use Andrew's pitch for OA IRs as an opportunity to recount
the argument for OA from a solely REF/RAE perspective; all of Andrew's
arguments are taken as read but the following is for managers/
administrators whose responsibility is focused on implementing
research assessment.
In my experience the pertinent issues that make OA relevant to
research assessment are as follows
(a) RAE / REF requires Universities to present a case for their
excellence based on evidence - especially research outputs.
(b) however, most University VCs, heads of department or even research
group leaders (ie research managers) simply have very little idea
about what research their staff are conducting.
(c) most UK universities do not have anything approaching a
comprehensive collection of (or even list of) the research outputs
that they have produced
(d) the reason for this is that we have outsourced our intellectual
assets to the publishing industry
(e) to obtain a list of our research outputs we can deal with one or
other of the secondary publishers, but this source of information is
both incomplete and difficult to interpret - and likely to become
increasingly costly.
(f) the only other alternative is to begin to collect information
about our own research outputs and activities using an IR and a CRIS
The REF now puts us in a crisis of institutional knowledge management.
Either we buy our way out of the problem every year (a partial
solution however much we pay) or we take responsibility for our own
intellectual assets and start an IR and mandate everyone to enter all
their research outputs and other pertinent evidence of research
activities.
The $64,000 question is why an IR with full texts instead of a
CRIS? The key is that only with access to the texts can an
institution run its own assessment procedures - appointing its own
panels of experts to evaluate the performance of its departments and
schools. This was a problem that we faced in preparing for the RAE -
the national funding authority made its own licensing arrangements for
its own processes but left institutions unable to prepare by running
their own pilots. In my school, our repository provided the full texts
we needed to pass on to our own expert review panels. Even though the
REF is likely to make substantial use of metrics, it will not be
possible to completely abstract away from the research outputs
themselves.
Of course, having a copy of the full texts for OA (or ID-OA) also
serves many, many other purposes including improved scholarly
communication, citation enhancement, publicity, profile raising,
institutional marketing and teaching.
- Consequence: all university procedures which involve publications
should
draw their information from the repository, particularly promotion and
incentive procedures.
I would like to particularly highlight this recommendation of Andrew's
- I think it is a crucial and simple part of the enforcement of any
mandate. I think it is the reason why the mandate for my school (ECS,
Southampton) was so successful. I wonder how many of the mandates
listed in ROARMAP come with such a clause?
--
Les Carr
