Statistics on ejournal usage
Dan Lester 04 May 2000 13:43 UTC
This reply is in response to the question below.
>We would like to count these titles also and
>have heard that some libraries create an intervening web page between
>their own library's web page and the actual journal. The URL of this
>web page is used in the 856 field of bib records and simply takes the
>patron to a virtually blank web page which counts the hit and then takes
>the user to the e-journal site itself. If any library has tackled this
>approach to gather use statistics, could you please explain how this
>intervening web page was created? We use Endeavor if that makes any
>difference.
The answer:
We do just what you're asking for at Boise State University. Each
ejournal has a URL of this format placed in the web catalog:
http://library.boisestate.edu/ej/1234-5678.htm (This is not a
real example)
The URL ends with the page for the ISSN of the ejournal in question.
In addition to the URLs being placed in the 856 field in the bibliographic
record, they are also (at least so far) listed on separate pages on the
website. You can get to them by going to:
http://library.boisestate.edu/ejournals/
Each page has a link to the appropriate page for the ejournal.
Besides some formulaic text (check some out to see), the page contains
the following in the <head> of the page:
<meta http-equiv="Refresh" content="1;
URL=http://www.somepublisher.com/journal/name/">
Of course the actual URL is substituted for the example above.
If you want the page to delay more than one second before refreshing
to the journal URL, change the 1 to a 2 or 5 or whatever. Do NOT
change it to 0 (zero), as that will keep patrons from being able to go
back to the previous page. No one notices or cares about a one second
delay, but it does give them enough time to go backwards if they wish.
See how it works by going to the library pages cited above. Of course
you may not be able to get to all the journals we subscribe to.
What are the flaws in this? It only tracks journals that people go to
via the catalog or the website. If professor Jones bookmarks his
favorite journal on his computer, you won't count those. But, no
library statistics are perfect, and we think this is the best way
reasonably available. And, at least all the stats are comparable.
How do you get the numbers? Run your log analysis software on just
the pages in the directory containing the intermediate referral pages
and see how many uses there are for each page. Bingo, a report. I do
make sure to include the journal title in the <head> of each page so
that the report is more meaningful than one containing just the URL
and ISSN. WebTrends makes beautiful and useful reports, but there are
other packages available.
The system you use for your catalog should make no difference. One
other advantage to this system is that the catalogers NEVER HAVE TO
CHANGE THE URL in the catalog. It does mean the person maintaining
webpages may have to make changes (I just made the ones for all of the
IOP journals today). But it is much easier and quicker to make them on a
template-based webpage than it is in the catalog.
I'll be glad to answer more questions on or off list.
cheers
dan
--
Dan Lester dan@RiverOfData.com
3577 East Pecan, Boise, ID 83716-7115 USA
www.riverofdata.com www.postcard.org www.gailndan.com
