Thursday, August 15, 2002, 2:48:52 PM, you wrote:

>> ***Forgive my ignorance - how do you ascertain
>> usage of the electronic collection?

RA> Good question.  The answer will vary depending on your local system.  For
RA> online databases, we generally have to depend on usage statistics provided
RA> by the publisher.  We are able to track certain kinds of traffic on our
RA> local network (enough to tell us that our online journals get a huge amount
RA> of traffic), but once the patron heads out onto the Web there's not much we
RA> can do to track them.

At Boise State we believe that the vast majority of ejournal use is
through http://tdnet.boisestate.edu, which produces extensive
statistics on accesses to ejournals.  It doesn't indicate how many
articles were viewed or printed from each, but much of that is
available from publishers.  Even more is available from our proxy
server, which handles in-library traffic, and our other proxy server,
which authenticates off campus usage.

Yes, this is somewhat of an "apples and oranges" mishmash of data, but
it all indicates vast amounts of use.

For print and microform journals, there are circulation statistics and
the much more important "pickup statistics" (in house usage).  Every
issue that is picked up is counted by a scanner and totals tabulated
by our circulation system, along with the "outside circulation"
statistics.

RA> Here's one interesting statistic, though, just as an example: from July 1,
RA> 2000 until June 30, 2001, patrons downloaded 29,212 full-text articles from
RA> our Elsevier journal database alone. (More than half of those articles were
RA> from journals to which we had never subscribed in paper.)  Bear in mind that
RA> that represents use of only one of our many online packages.  There's simply
RA> no question as to the relative usage levels for print and online journals at
RA> my institution.

Same in Boise.  We can also get some idea of usage of ejournals by
observing patrons and the usage of paper on our computer printers in
the public area.  When a patron has an option between finding an
article electronically and printing it vs. going to find a paper copy
and then photocopying it, it's a genuine nobrainer.  In fact, many of
the databases provide an option to limit a search to items with full
text.  That is an extremely popular feature with students.  For the
students writing a typical 3-10 page lower division paper requiring
relatively few resources that isn't even a limitation on their
research.  They still have more available than if they searched in
Readers Guide and did it the way most of us learned back in the "olden
days".

RA> I agree.  Of course, that's just what we did count in our usage report on
RA> print journals.  We counted reshelvings, not actual uses; if someone picked
RA> up four issues, looked at only two of them and then put all four on the
RA> reshelving shelves, that counted as four uses.

Actually, all of these caveats relate to ALL library statistics,
whether items circulated, reference questions, door counts, and so on.
Some days I wonder whether we should bother counting anything at all.
Book circulations don't mean anything besides that the book was
checked out, not that it was cited, skimmed, read, or even understood.

The value in ANY statistics is in their relation to other statistics,
and their long term trends and relationships.  The absolute values
mean nothing.  Even if you think you have a million "books" in your
library, you don't.  And of course your count of a million isn't the
same as mine.  And so on, ad nauseam.

cheers

dan

--
Dan Lester, Data Wrangler  dan@RiverOfData.com 208-283-7711
3577 East Pecan, Boise, Idaho  83716-7115 USA
www.riverofdata.com  www.gailndan.com  Stop Global Whining!