Re: OA Impact Advantage = EA + (AA) + (QB) + QA + (CA) + UA

[Stevan Harnad <harnad@ecs.soton.ac.uk>]

On Fri, 30 Sep 2005, Mcsean, Tony (ELS) wrote:

> Another complication in trying to arrive at a like-for-like comparison
> is which of their articles authors choose to self-archive. Maybe I'm
> just a cynical old git, but would be surprising to me if authors weren't
> at least slightly more likely to self-archive their best work and less
> likely to be bothered with their more humdrum output.  You might expect
> these to be more heavily cited however they were made available.

Not cynical at all. There *are* indications of a self-selection Quality
Bias (QB) on the part of authors, towards preferentially self-archiving
their better articles (as well as for the better authors to self-archive
their articles). (It would be quite surprising if there were not.) As I
think I mentioned before, Michael Kurtz has written about this in
astrophysics.  (See Steve Hitchcock's bibliography for references:
http://opcit.eprints.org/oacitation-biblio.html ).

However, QB is not the only factor in the OA citation advantage (and, I'll
bet, not the biggest one):

   "OA Impact Advantage = EA + (AA) + (QB) + QA + (CA) + UA"
   http://openaccess.eprints.org/index.php?/archives/2005/09/17/C1.html

What will of course sort out the relative size of the components that
contribute to the current 50%-250% self-archiving advantage will be the
very process of moving upward from the current c. 15% self-archiving rate
to 100%.

    http://www.crsc.uqam.ca/lab/chawki/graphes/EtudeImpact.htm

(It can't be QB as we get closer to 100%. as most work cannot, by
definition, be best work!) We already know (again from Michael Kurtz's
work in astrophysics) that at 100% OA there is still the Usage Advantage
(3 times as many downloads) and the Early Advantage (a permanent increase
in total citations, the early the paper is made OA).

Michael also reports that in a 100% OA literature, reference lists are not
longer (in fact, they are even a bit shorter!). But what 100% OA does
provide is is QA, the Quality Advantage, levelling the field, so that
authors can select the best and most relevant articles to use and cite,
rather than merely the best and most relevant *amongst the ones that
happen to be published in the fraction of the total number of journals
that their institutions happens to be able to afford*.

That last bit, in a nutshell, is the real essence of OA: Allowing work to
achieve its rightful impact on the strength of its quality/merit alone,
not constrained by the extraneous factor of accessibility/affordability.

Stevan Harnad

On Fri, 30 Sep 2005, Mcsean, Tony (ELS) wrote:

> I'd just like to add a footnote to Phil's characteristically thoughtful
> comments.  Another complication in trying to arrive at a like-for-like
> comparison is which of their articles authors choose to self-archive.  
> Maybe I'm just a cynical old git, but would be surprising to me if authors
> weren't at least slightly more likely to self-archive their best work and
> less likely to be bothered with their more humdrum output.  You might
> expect these to be more heavily cited however they were made available.
> 
> It's a bit embarrasing to be contributing yet another anecdotal/unproven
> hypothesis to the list, but in my own defence I can say that I'm not
> saying (a) that this definitely happens, (b) it's impossible to quantify
> for the purposes of analysis or (c) that if it does happen you can't come
> up with a methodology to compensate.  I'm just saying it's a factor that
> needs looking at and to me it doesn't look easy to fathom and that the
> various conclusions we currently have to hand may be interesting but are
> not even close to definitive.
> 
> Tony McSe�n
> Director of Library Relations
> Elsevier