Re: Open Access Advantage (or Not!)

[JOHANNES VELTEROP <velteropvonleyden@btinternet.com>]

Joe is right. The OA-advantage is just a visibility/ 
discoverability advantage. However, a few major publishers have 
seen some very strong correlations: between downloads and 
citations, and between visibility and downloads (e.g. articles in 
issues that are being made freely available, generally for 
promotional purposes, usually attract appreciably more downloads 
than comparable articles, e.g. in other, non-free issues of the 
same journal). This may well mean that a visibility/ 
discoverability advantage translates into a citation advantage.

More citations obviously doesn't imply better scientific quality. 
If that were the case, just opening an article up would improve 
its quality. (Though a case could be made that 'quality' includes 
an article's usefulness to science and society as a whole, and 
that does increase just by being open). The citation advantage - 
and impact factor advantage for a journal with OA articles - can 
only be temporary and vanishes when the majority of the research 
literature is OA.

Findability matters. I fully agree. Do publishers not pay 
attention to SEO? Some do, some don't. Most make at least 
abstracts and metadata available to search engines and indexing 
services, in order to increase findability.

The publishing process is author-driven, rather than 
reader-driven, and Joe is right to say that "...the author should 
be challenging publishers on different metrics. Who is your 
search-engine marketing team?  How many page views does your site 
get?  How does your Web presence compare to the competition's?" 
The economics of the traditional publishing model, however, are 
based on a reader-driven presumption. Open access publishing on 
the other hand, with author-side payment ('author-side' instead 
of 'author', as they shouldn't be expected to pay out of their 
own pockets - just like readers in the traditional model aren't 
paying for subscriptions out of their own pockets), restores the 
link between the economic underpinning of journal publishing and 
the forces that drive it: the author-side need for 'official' 
publication, with all that comes with it (an article that may 
otherwise be very good, yet is not officially published, is 
pretty much ignored by Academia's assessment ! and reward 
structures, for instance).

When Joe says that "Open access is a poorly thought-out, 
amateurish, and risky strategy", I don't quite know what he's 
referring to. Open access per se? The benefits of free, universal 
access to scientific research results for the worldwide science 
community are blindingly obvious. Open access publishing? 
Probably much better thought-out than the traditional, 
historically and 'organically' grown subscription model ever was. 
Self-archiving (making use of the tremendous benefits of a formal 
journal publishing system, yet without taking responsibility for 
the costs incurred in maintaining that system)? He's right that 
that might prevail if traditional publishers don't begin to take 
the internet seriously. Publishers should facilitate full open 
access and give authors the option to publish in that way. They 
are not, however, in any position to impose open access. That's 
for the funders/backers of science, an increasing number of who 
are seeing that publishing is integral to do! ing research, and 
thus the cost of publishing is integral to the cost of doing 
research, making OA possible and economically feasible in the 
process.

Jan Velterop

___

"Joseph J. Esposito" <espositoj@gmail.com> wrote:

I am not competent to assess the discussion between Stevan 
Harnad, Phil Davis, Peter Banks, and others concerning what is 
called the Open Access Advantage, which sounds oddly to my ear 
like a frequent flier program. I do wish to clarify two domains 
that are being confused in this discussion, unless it is I who is 
befuddled by the terms.

I take it that the discussants use "citation" in the formal sense
in which one author cites another in a paper. It should be
self-evident that Open Access cannot have anything to do with
citations, whether they are great or few in number. OA simply
means a user can read material through the mediation of a Web
browser without having to pay for it or having someone else
(e.g., a librarian) pay for it. A citation requires a positive
action on the part of an author. An author can cite a paper that
is OA or one that is "toll-access." Indeed, presumably sometimes
authors cite papers they have not read at all. If all the
world's papers were OA, and every researcher read every one of
them in his or her field, it is theoretically possible that not
one citation would result from it. Authors cite articles because
they provide value to the authors' own work. Open Access has
nothing to do with it and therefore, if there is an OA Advantage,
it must lie elsewhere.

The OA Advantage, if it exists, lies not in citations but in
findability. How can a researcher cite an article that he or she
does not even know exists? I happen to believe that at this time
the likelihood of a researcher not knowing about a meritorious
article that could be of value to his or her work is highly
improbable, but I don't wish to argue the point here.
Researchers find articles because they see them cited, because
colleagues recommend them, because they use insitutional or
product- or publisher-specific search engines, or because they
use a "universal" (that is, publicly available) search engine
such as Google. Open Access only pertains to the universal
search engines. All other ways of finding articles have nothing
to do with OA and thus cannot yield an OA Advantage. OA, thus, is
a means to market articles (that is, call attention to them) to
Google and its kin, and any OA Advantage lies in Google-like
findability, not in increased citations. The term of art for
this is search-engine optimization. But even a well-SEO'd
article will not yield any citations if other authors don't
choose to cite the article. OA can bring a researcher to the
foyer, but it is no guarantee of a dance.

[SNIP]