2006 was the first year that my library gathered usage statistics, thanks to project counter (counter). counter is a cooperative initiative that provides librarians with usage statistics for the electronic journals to which they subscribe. the list of vendors providing usage statistics in this way is growing, which is good, because it means that i can now compare the amount of usage of one journal to another. before counter, vendors all had their own definitions of a “use,” making comparisons across journals impossible. still, about 20% of the electronic journals to which we subscribe do not provide statistics using the standard, so we don’t have a complete picture of how our entire electronic journal collection is being used.
in order to evaluate usage across journals i go to each vendor’s website, log in with an administrative username/password and download a file. in gathering 2007 statistics i went to 15 vendors’ sites. we then merge all those separate files into one large file and sort by title for an alphabetical list.
as cumbersome as putting the master title list together is, the evaluation of this title list is where things get tricky. we get access to many of our electronic journals from more than one vendor. one vendor may supply an archive and another vendor may provide current material. the counter standard does not dictate how the journal title information must be entered into the title field. this means that the vendor that provides the archive may call the title “the journal of the …” and the vendor that provides current material may call it “journal of the …”. these display in my merged title list as two different titles and if i don’t catch what happened i’ll only see partial usage for that journal, as the rest of the usage is hidden under another title. this is disastrous when you’re looking at a title list of over 3000 titles and you’re trying to count how much usage “the journal of the …” got.
one report i pulled put my university’s subscriber number after the title in the title field, so the problem isn’t as easily solved as a macro to delete those stop words (the, a, an) from the beginning of titles. a friend in the industry is also concerned about the way platform data are entered into counter reports, so this problem extends beyond my concerns about the title and really encompasses all the data entered into these reports.
what needs to happen now is some standardization of the way data are entered into counter reports. the upcoming protocol, sushi, won’t resolve this problem, as i was hoping it would. sushi is just a mechanized call and respond protocol, so it will go out and gather the files from the different vendors for me (which is nice!) but i will all still be left with dirty data.
if you’re currently gathering electronic journal usage statistics like i am, and you’re interested in working with me to lobby to standardize the way data are entered into counter statistics reports, drop me a line or leave a comment.
