Databases and electronic journal subscriptions have become a hallmark of academic libraries. Currently, the LLC subscribes to 28 different periodical databases. These databases provide content for over 45,000 journal titles, as well as specialized information specific to individual disciplines. To assess the value of these databases, we measure how they are being used, what content they provide, and how much they cost.
Resource utilization is measured by running usage reports for each of the databases to which we subscribe. These reports illustrate how often a database has been searched, the average length of a viewing session, and how many times an item has been viewed.
Below is a graph relating the combined resource use (item views) for our electronic resources:
These aggregate data show a general trend towards increasing electronic resource use. In terms of individual product subscriptions, JSTOR is a prime example:
But not all of the databases we subscribe to demonstrate the same upward trend. CredoReference, for instance, has remained relatively flat in terms of usage:
And utilization of the Classical Music Library has fallen off entirely:
In general, the use of electronic resources across campus is quite impressive. However, in obtaining these data it is now clear that significant journal duplication exists across databases.
We focused on duplication among the most popular content within our subscriptions, and we defined “popular” as 6 or more uses in the past 2.5 years. On average, our larger database subscriptions have between 40-60% duplication among popular content. If we were to check for duplication among all the content provided, the duplication percentage would likely be much higher. Academic Search Complete, for instance, has approximately 50% duplication among the most viewed items:
Some of the LLC’s more specialized databases, such as BioOne, have a much higher duplication percentage:
ScienceDirect, by comparison, is an extremely unique database due to its low duplication percentage:
Regrettably, some of our resources are rarely used. Approximately 78% of the resources made available have not been used in the past 2.5 years. The most extreme example of this is LexisNexis; 99% of its resources have not been used at all:
Some resources provided by the LLC are highly specialized while others are broader in their scope. Because of this, we assess our resources on a cost-per-user basis. Cost-per-user is calculated by dividing database cost by the total number of potential users. Classical Music Library, for instance, has a much higher cost-per-user rate because it is specific to the Music Department:
SAGE Premier, on the other hand, spans multiple disciplines so its cost-per-user is much lower:
In taking these data together, we now have a much better idea of how our electronic resources are being utilized. It also allows us to pinpoint areas of weakness within both our electronic and physical holdings that can be targeted for development in the future.