Re: Open access to research worth A3 1.5bn a year

[heatherm@eln.bc.ca]

In response to Phil Davis' post of Sept. 29:

For the most up to date information on the OA impact advantage, please see
Steve Hitchock's excellent bibliography of the many, many studies
replicating this effect, at:
http://opcit.eprints.org/oacitation-biblio.html

Phil pointed out that the JEP study, quoted by Peter Banks, failed to
support the hypothesis of the impact advantage of OA.  This single study
is not significant, particularly when, as Phil points out, selection bias
alone could explain the anomalous results.

As Stevan Harnad has pointed out on American Scientist Open Access Forum,
the JEP article quoted by Peter Banks is dated - another potential source
of error in the current environment.  That is, early on, when fewer
articles were available as open access, fewer people would have been
looking for OA copies.

As Phil points out, the OA impact advantage may not be a fixed figure.  
Hence, the current average of about 50% - 250%.  It is possible that there
will always be differences in this figure, due to different reading and
citation patterns in different disciplines, and for different journals.  
It is demonstrating the consistency of the OA impact advantage effect that
is important, not the specific percentage increase in citations.

My prediction is that the OA impact advantage will gradually rise over
time. That is, the more people are aware of OA, the more they will seek OA
articles.  As David Stern pointed out in another message today, a link
resolver may not integrate open access articles seamlessly; there is work
to be done here.  Once this technical work is done, it seems reasonable to
predict that the OA impact advantage will rise yet again.

In time, too, it seems likely that usage of non-OA articles will decrease,
simply because they are less useful.  Readers will (and should) begin to
expect clickable references.  Non-OA articles will be clickable for some
readers, but not others.

Some of us have already gone past this point already.  If there are two
articles demonstrating the same phenomenon, and you only need one to
demonstrate a point, which will you pick, the OA article, or the non-OA
article?  If you care about your readers, you'll pick the OA version.

best wishes,

Heather Morrison
http://poeticeconomics.blogspot.com