It’s that time of year again…the end of the fiscal year, time to turn in statistics of all types for the Annual Report. If only it were that easy….

Database statistics are supposed to be useful measures of the amount of use a database is getting. In theory, this is a great idea. We could measure how many people are using our expensive resources and make all kinds of decisions and assumptions based on those numbers. But…each vendor seems to measure something different. Or at least use different terminology for what they are measuring. Some of them don’t even explain their terminology. Which is fine, for individual statistics. But when all of these statistics have to be compared to one another and put into a single spreadsheet….well, things get tricky. I’m still trying to sort out what columns I can/should have for comparing all of these statistics. Some of the databases measure by the minute (for streaming audio) and some measure by session, or search, or page view, or “usage”.

To make matters worse(?), some of the vendors haven’t even posted their statistics for May yet. It’s June 13. I have no idea when the stats will be posted. There’s not really a hint on the vendor website – they say vague things like “10-15 days after the first of the month”

There are some standards available for the vendors to use for their statistics – ICOLC and COUNTER, but it seems that many vendors are not complying with either one. I realize it takes a lot of time and effort (and money, of course), but it really would make things a lot easier for their users. Of course, there are questions about how meaningful the current statistics are anyway, with the advent of federated searching and other advances. This is covered in Measurement of Use of Electronic Resources: Advances in Use Statistics and Innovations in Resource Functionality, by Deborah D Blecic, Joan B Fiscella and Stephen E Wiberley, Jr. in College and Research Libraries in January 2007. They point out that the guidelines have made it easier to collect data and helped remove some of the doubts about the reliability of statistical data. But they also argue that the meanings of the words sessions and searches (the common terms used in measuring statistics) have been changing as the use and functionality of electronic resources changes.

So it seems that database usage statistics will continue to be nebulous and difficult. At least until all the vendors are compliant with one or both standards. Which may be a very long while.