Hi Peter,
I have been working on institutional publication collaboration and have some
data maybe useful to you.
 The proportion is discipline dependent, but university is always the major
contributor:
Computer science (ACM publications from 1950-2010)
University 62%
Research Institutes 5%
Company 19%
Unknown 14% (Missing or unable to recognise)
Pharmacology (ISI from 1973-2011)
Company 13%
Research Institute 28%
University 60%
Material Science (ISI from 1973-2011)
Company 2.5%
Research Institute 10%
University 87.5%
Law (ISI from 1973-2011)
Company 0%
Research Institute 15%
University 85%
Psycology (ISI from 1973-2011)
Company 0%
Research Institute 10%
University 90%
Please not that these proportions are authorship percentages, i.e. 90% of
university means 90% of author is affiliated with an university.
Let me know if you need anything else.
Best
Jiadi
On 02/05/2012 17:57, Stevan Harnad wrote:
Hi Jiadi, could you post your findings on the following to goal,
please:
7 What proportion of publications come from
"Universities" or other
organizations that potentially support
self-archiving infrastructure?
   [I guess about 80%. Publications from
industry, research institutions,
hospitals, field stations, etc. should NOT be
dismissed as irrelevant or
substandard.]
Don't know. Preumably depends on the discipline.
Les Carr's and my student, Jiadi Yao, has some limited data on
the
proportion of university, research institute and corporate
research in
some fields. I will ask him to post. (I agree that refereed
research
from all sources is relevant, and the standard depends on the
peer-review standards of the journal that accepts it.)
Thanks, S
Begin forwarded message:
From: Stevan Harnad <[email protected]>
List-Post: [email protected]
List-Post: [email protected]
Date: May 2, 2012 12:10:39 PM EDT
To: "Global Open Access List (Successor of AmSci)"
<[email protected]>
Subject: Re: What is Green Open Access and how is it practised? Some
questions
On Wed, May 2, 2012 at 3:43 AM, Peter Murray-Rust <[email protected]>
wrote:
If we are being pragmatic, it is necessary to know the
facts on which we
base our strategy.
I will publicly admit that I do not understand the goals
of "Green Open
Access".
 I would like to ask a set of (hopefully simple) factual
questions about
Green. Please humour me by only answering the questions.
I may have wrong
assumptions - that why I am asking.
1 does Green OA require the archival of complete
published fulltext?
   [My assumption is YES]
Yes, the author's refereed final draft or higher.
2 does Green OA depend on publishers agreeing to authors
self-archiving
their manuscripts?
   [My assumption is YES]
No. It depends on authors self-archiving their manuscripts.
3 is there any "official organization" that *formally*
negotiates Green OA
with publishers?
   [My assumption is NO]
No. But the majority of journals (and almost all the top ones)
already
endorse immediate Green OA.
4 what percentage of publishers currently forbid Green
OA as defined in Q1?
   [My assumption is about 40%]
About 35% of journals, about 40% of publishers.
5 How many institutions do Green OA mandates potentially
apply to?
   [I estimate between 1000 and 10000]
All research institutions (universities and research institutes)
worldwide, which is presumably at least as many as the higher
figure...
6 Is there one or more global organization *formally*
coordinating these
institutions?
   [I suspect NO]
No. They are an anarchic network, like the Web. (But the
repositories
are at least OAI-interoperable.)
7 What proportion of publications come from
"Universities" or other
organizations that potentially support self-archiving
infrastructure?
   [I guess about 80%. Publications from industry,
research institutions,
hospitals, field stations, etc. should NOT be dismissed
as irrelevant or
substandard.]
Don't know. Preumably depends on the discipline.
Les Carr's and my student, Jiadi Yao, has some limited data on the
proportion of university, research institute and corporate research
in
some fields. I will ask him to post. (I agree that refereed research
from all sources is relevant, and the standard depends on the
peer-review standards of the journal that accepts it.)
8. How many institutions currently offer Green OA?
   [I think Peter Suber recently suggested about 1500
have repos].
ROAR has the current figures for repositories. (But most are
unmandated, and hence only contain a small percentage of the
full-texts of the institution's total annual refereed research
output.)
9 what is the current full economic cost of a
self-archived manuscript in a
(a) UK University? (b) Elsewhere?
   [see below]
Negligibly small, per paper. And would be even smaller if the
repositories were mandated, hence full, rather than unmandated, and
near empty.
10. Is there any agreed mechanism for (a) humans (b)
machines to tell that
an object in a repo is a Green manuscript?
   [I assume MAYBE for (a) and NO for (b)]
A human can certainly tell, by inspection. No mechanism for machines
yet, as far as I know (though it's feasible). But as long as the
percentage OA is a sparse as it is, it's hardly urgent to develop
such
a mechanism. What's urgent is to mandate deposit.
11. Is there any SIMPLE way of finding all Green
manuscripts across all
repos? Â [I assume NO.]
Not yet. Nor is it urgent, while OA content is so sparse. But it's
feasible. The much more urgent priority is mandating deposit of the
content.
12. How is the compliance of authors in depositing Green
OA measured? By
whom? [I assume this has to be done by an institution
and this requires them
to (a) know how many publications have been published by
"their staff" and
(b) know how many are in the repo. I assume it is the
aggregation of these
figures that gives the "20%" green figure.
Most institutions are not systematically measuring this yet, but
samples have been tested webwide (and not just for repositories, but
also authors' websites) and 20% seems to be a good ballpark figure.
Mandated repositories are more likely to measure annual deposit
rate,
and effectively implemented mandates (like Southampton ECS, QUT,
Minho
and Liege) capture over 70% of their annual refereed research
output.
13 How many institutions know and publish metrics of
Green deposition
including a percentage of the possible?.
For the vast majority of institutions that lack mandates, this
percentage would not be very useful. Yassine Gargouri is doing %OA
analyses for the institutions with effective mandates, and comparing
them with weak and no mandates.
14. The goal of Green OA is, as I understand it, for all
Universities [sic]
to put copies of all their peer-reviewed publications
into a professionally
supported Institutional repository. YES/NO
All research institutions, which includes universities, research
institutes and corporate R&D centres.
15. Can Green OA deliver 100% of the scholarly
literature [sic]?
 [I assume NO].
Yes, if all institutions and funders mandate Green OA.
     If not what is a figure that proponents would feel
represented a major
positive outcome ("success")?
If all institutions and funders adopt (effective) Green OA mandates,
close to 100% of refereed research would be OA. That's the only
outcome worth aiming for.
[*] I think PeterS suggested about 1.5-5 FTEs per IR.
Assume 2, and cost
each at 100K USD Full economic costs. I trawled UK
Universities and found
that they had between 500 and 10000 items. Not all of
these are final
manuscripts - some are theses (although these are so
heterogeneously
archived it's almost impossible to know) and some are
other artifacts.
Assume 1000 deposits per year (and I think that is
optimistic) and you get
over 100USD per manuscript, not including researcher
time. I don't think
that this reduces dramatically by volume as many
manuscripts require
assistance from the repo staff.
Repositories are created and used for other purposes than OA. But
even
if they were just for OA, the cost per paper for an unmandated
repository is uninformative, except to show that a repository is not
worth much unless deposit is mandated.
For a repository that is capturing 100% of its annual research
output,
the cost per paper deposited is negligibly small.
Stevan Harnad
