Re: Decision making by Libraries on serials and monographs and useage (re puzzled by self-archiving thread)

[Phil Davis <pmd8@cornell.edu>]

Joe and Sally describe why OA may (potentially) affect some 
journals more than others.  The marginal effect of OA on elite 
journals, which by definition have wide circulation, would 
theoretically be lower than less-recognized titles.

Then what explains the fact that Eysenbach [1] reported an OA 
citation advantage for PNAS -- one of the most prestigious and 
highly-circulated multidisciplinary journals in the sciences? 
It could mean one of two things:

First, there may be a cumulative advantage for those journals 
that are already highly read and cited.  OA may not provide any 
more access to these articles, but it may provide additional 
access points or quality cues to readers.  Those who believe that 
OA democratizes access to information would be fooling 
themselves; the real effect is to increase elitism.

Alternatively, it could be that in a world of complex 
relationships, OA researchers are conflating causes and effects 
and reporting spurious relationships.  For instance, it just so 
happens that author-supported OA articles published in PNAS (in 
Eysenbach's sample) were more than twice as likely to be featured 
on the front cover of the journal, were nearly twice as likely to 
be picked up by the media, and when cited reached, on average, 
nearly twice as many news outlets as subscription-based articles 
[3].  It would have been interesting to test whether these 
variables explain the OA advantage reported in PNAS, but the 
author declined my request for access to his dataset for 
empirical verification.

In order to do publish one's results in a prestigious 
peer-reviewed journal indexed by ISI's Web of Science, one is 
likely to be located at an institution with both access to 
resources for doing research as well as adequate access to the 
literature. Considering that PNAS has one of the highest 
circulation rates of any scientific journals, that it provides 
free and immediate access to readers in developing countries, 
that it employs a tiered pricing model that allows small 
institutions to pay a fraction of the cost as large institutions, 
and that sharing electronic copies of articles is easier than 
ever, it is difficult to believe that immediate access in a 
journal that provides free access to all of its articles after 
six months would result in a citation advantage.

Our multi-journal, multi-discipline randomized controlled study 
of OA publishing should allow us to tease out the complex factors 
that lead one to read and cite articles.  Until then, "Science 
must not suffer itself to become the handmaiden of theology or 
economy of state." [4]

Phil Davis


Notes:

1. Eysenbach, G. (2006). Citation Advantage of Open Access Articles. PLoS
Biology, 4(5). http://dx.doi.org/10.1371/journal.pbio.0040157

2. Davis, Philip.  Citation advantage of Open Access articles likely
explained by quality differential and media effects.  PLoS Biology
(Response to Eysenbach).
http://biology.plosjournals.org/perlserv/?request=read-response&doi=10.1371/journal.pbio.0040157

3. Merton, R.K. Science and the Social Order. Chp 12 in The Sociology of
Science. Normal W. Storer (Ed.)  University of Chicago Press. p.260.


At 07:48 PM 2/1/2007, you wrote:
> Sally is being swift, not slow.  What she means, I believe, is that OA
> adds access only marginally.  So for an elite publication, 99% of the
> potential readers get access through subscriptions; OA would add the other
> 1%.  For a good but not elite publication, 75% of potential readers get
> access through subscriptions; OA adds the other 25%.  For a mediocre
> publication, 50% get access through subscriptions, etc., etc.
>
> As the quality of the publication goes down, the amount of access through
> OA increases; and this rather startling conclusion is based on the
> crackpot notion that librarians are very, very good at what they do and
> acquire the publications of greatest merit for their
> constituency.  (Taking this to its logical extension, a truly useless
> journal would be 100% OA.)  One could argue forever about the percentages
> (and no doubt we will), but the fundamental point remains:  OA is
> marginally beneficial for access, with inferior publications gaining the most.
>
> Joe Esposito