Re: high priced journals

[David Goodman <dgoodman@phoenix.Princeton.EDU>]

Note carefully the caveat, properly included by Dan, that this measures
only research use of titles by the faculty. It grossly underestimates the
use of review titles, undergraduate-interest titles, and so on. At at many
research oriented universities, the research done by the graduate students
and post docs is as important as that done by the faculty--and not all
faculty automatically add their names to their students' papers.

Some libraries are, of course, able to measure actual physical use. In the
absence of that, citation data is an acceptable surrogate if you keep the
limitations in mind.

Of course, many expensive research titles are used, but not cited. To take
an example, we spend very much more than $10,000 for Web of Science
(=Science citation Index), but not one person here has ever cited it.

Note also that Dan is speaking about a single field. It is not possible to
use this method validly across different subject fields. For example, I
collect for a dept of Ecology and a dept of Molecular biology, and the
citation density is very different. The unthinking use of this formula
would lead to my dropping all the ecology journals almost without
exception. This technique is not of help in balancing expenditures between
fields.

On the other hand, ignoring actual use and actual citations produces very
poor quality collections. A selector who does not select on the basis of
use will probably do as well by selecting by throwing darts at a list of
titles. A selector who selects on the basis of faculty opinion-- rather
than actual use-- will have a collection that reflects the previous
generation's use.

There was a time, though, when major research libraries could purchase not
just the titles needed by the current users, but those titles of potential
use in the future. Although I was not biology librarian here then, I can
tell from our holdings that we were still doing this in biology during the
60s very thoroughly across a very wide field. Since then we have
progressively been cutting back, and I have now reached the point where I
no longer fully support all current research from our own university's
holdings.

David Goodman, Princeton University Biology Library				
dgoodman@princeton.edu            609-258-3235

________________

On Wed, 3 Jan 2001, Daniel Feenberg wrote:

> I suggest looking at both the cost and the benefit side of the equation.
> Count the citations made by your faculty in their work, by journal. You
> don't need a citation index, just the pile of working papers your dean
> collects each year to demonstrate faculty output. A clerk can take each
> paper and put a tick mark on a list of titles you subscribe to for each
> reference in each paper.
> 
> Then divide the journal cost by the number of citations to articles in
> that journal. This gives the cost per citation for each journal. Then list
> the journals in increasing order of cost per citation. Now go down the
> list untill you run out of money. Unsubscribe to the remaining journals.
> 
> This list of journals is optimal in the sense that it maximizes the
> probability that someone looking up a reference in a paper by one of your
> faculty will find it in your library.
> 
> It won't work for small faculties, or large fields, because the
> statistical variation will be to large. But if you include several years,
> that might help, and you could include course reading lists (perhaps
> counting those with a higher weight) and other material as well.
> 
> I admit I have only made such a list once (for the Economics faculty at
> Princeton University) and the librarian did not act upon it. However, some
> of the most expensive items (including a service that cost $10,000 in
> 1978) showed zero citations and should certainly have been dropped.
> 
> Daniel Feenberg
> ___
> 
>  On Wed, 3 Jan 2001 Publishingnut@aol.com wrote:
> 
> > I found a list of the 100 most expensive journals and I wanted to share it
> > with all of you. You can do a variety of searches at
> > http://db.arl.org/journals/ based on your specific needs. When forced to
> > subscribe to a journal such as Brain Research for $16,344 a library must
> > cancel other subscriptions. What should a library do when forced to
> > subscribe to journals that are beyond their budgets?