Last words (I hope) on EBSCO/ProQuest study

["Donnie Curtis" <dcurtis@admin.unr.edu>]

Thanks to Ann for the acknowledgement and to those who have contacted me
about my study of content in certain EBSCO and ProQuest databases. A
discussion list such as liblicense has advantages over traditional
publication outlets in letting us share time-sensitive information and get
immediate feedback. This has helped me realize that I need to make some
more comments and a correction. I hope that if some of you are using the
data to make decisions on database subscriptions you will keep in mind
that database content and other features can change radically in a short
period of time (especially this time of year right before ALA), and also
that as much as I tried to standardize the information provided by both
companies, they do present it differently, and some aspects cannot be
accurately compared.

It has been pointed out that the EBSCO lists I was using do not have as
many titles as the lists that are now posted at:
http://www.epnet.com/maglists/maglist.htm 

The Academic Search Elite full text list now has 1,530 titles, and the
Academic Search Premier list has 3,175 titles. It's pretty easy to
download the Excel file and sort it if you want to know how many of those
titles are peer reviewed or embargoed. It has also been pointed out that
some of the titles on those lists (and the lists I used) are not actually
available yet in the databases. I didn't pay much attention to that in my
study, but I probably should have. Upcoming content is really not the same
as current content, if you are capturing a snapshot of what is in
competing databases at a certain point in time.

The correction I need to make to my summary that was posted to liblicense
concerns the section

>Full text ends, but the journal continues, and some full text is still
included in the database:
>   PQRL -- 101 (6%)
>   ASE -- 7 (<.5%)
>   ASP -- 4 (<.5%)

For that part of the study I took a shortcut and used numbers I had gotten
from Sam Brooks, derived from careful research by the EBSCO marketing
group. However, I didn't realize that those numbers represented only the
peer reviewed journals. So I went back to the ProQuest file and discovered
that there were 162 journals where full text has ended but indexing
continues (a sure sign that the journal is still around). Sam has informed
me that there are at least another 16 titles that are still active but are
no longer being indexed in PQRL, but that the numbers I have used for ASE
and ASP in this category are accurate. Doing further research in that area
would be very time consuming, so I am unable to go beyond this right now.
177 journals would be 10.6% of the journals in the PQRL database.

The bigger picture here is that instituting embargoes and halting full
text coverage are two ways that aggregators can voluntarily or
involuntarily restrict the provision of current issues of journals in
their databases. An important task for us is to keep an eye on the extent
of both practices if we want to be aware of what we are subscribing to. We
should also look at new titles being added (quantity, quality,
suitability, currency), which I did not do in this study. And as someone
suggested, we should do qualitative studies of the results of typical
searches our users would perform in the databases. I hope that for simple
(but detailed) content analysis my study can serve as a baseline for
future observations of these three databases, and that others might extend
such studies to other comparable databases and that we will see studies of
the usability aspects.

Donnie Curtis