Re: Dramatic growth of open access

["Peter Banks" <pbanks@diabetes.org>]

I agree with you that articles would be a far more meaningful 
statistic than journals. However, I would refine that further: 
it's the number of high quality articles that matters, whether 
one defines quality by downloads, citations, or some other 
metric. I don't think the data are there to say whether OA is a 
rising tide by any of those measures, as Heather first asserted.

Also, what is an "open access article?" By the criteria of one of 
the OA manifestos, a journal like Diabetes Care is not OA. 
However, Diabetes Care makes more highly cited clinical diabetes 
research freely available than any other endocrinology/metabolism 
journal. I frankly think its policy of making content freely 
available after 3 months is a more meaningful contribution of 
expanded access to scientific information than the many OA 
optional plans floating around. So is Diabetes Care evidence of 
the rising tide of OA--or an indication OA, as defined in Berlin 
or Bethesda or wherever, really isn't needed? What exactly is OA?

Peter Banks
Publisher
American Diabetes Association
Email: pbanks@diabetes.org

>>> matt@biomedcentral.com 04/12/06 8:12 PM >>>

"I don't think the data show that open access continues to grow
dramatically, not in medicine at least"

There are a several reasons that a year by year analysis of the
DOAJ, is not really suitable as a metric for the growth of OA.

Firstly, as has been discussed recently on this list, the DOAJ
does not list the year that a journal went open access. Rather
(as I understand it) it lists the first year for which OA content
is available from the journal concerned. As such, any journals
which have converted to open access, rather than starting as open
access journals, are assigned to the wrong year in the analysis
below, underestimating the number of journals going OA in recent
years, and overestimating the number of OA journals launched in
the past.

(2) The number of open access journals is in any case an poor
proxy for the overall growth of open access publishing. BioMed
Central launched 50+ titles in the year 2000 (contributing to an
apparent peak in new open access journals in that year). That is
the most journals BioMed Central ever launched in a single year.
So did we "peak" in 2000? Hardly. BioMed Central published 5586
peer reviewed open access articles in 2005, compared to 224 in
2000 - a 25-fold increase, and we continue to see very strong
year on year growth.

(3) Looking a the number of journals in the DOAJ fails to account
for the growing take up of optional open access (e.g. as
practiced by PNAS et al.) and also fails to distinguish between
huge open access journals (like NAR) and tiny ones.

A better approach would be to analyse the number of immediate
open access articles published year on year. This is challenging
to do, not least because several years on it is very difficult to
be sure what *was* open access at the moment of publication. But
that is really the metric that counts.

Matthew Cockerill, Ph.D.
Publisher
BioMed Central ( http://www.biomedcentral.com/ )
London, UK

Email: matt@biomedcentral.com