From: "Hamaker, Charles" <[log in to unmask]> Date: Thu, 26 Apr 2012 23:58:07 +0000 Comparative use/cost per use data across similar institutions could be a very useful indicator of value and useful point for negotiating with vendors and publishers. As it is we are at the mercy of whatever metric a provider chooses to use be it FTES, the commonly used Tier system, etc. We are at the mercy of some times egregious pricing practices from what I have seen of comparative data. Comparable data from peer institutions would help in negotiating value for money. If a non-engineering institution is paying 10 times the cost per use as a University with a College of Engineeringfor an engineering database, that's helpful data to have in making a local decision about value for money paid. We are also at the mercy of "value" judgments from our main patrons who don't care much to be told their particular favorite database with its local use puts the non-engineering institution in the top quartile of cost per use-to give an example. So I guess I have to disagree with Fred. In practice, CPU data is a valuable indicator that when we get down to trying to figure out (as many of us are right now) what is the best value for money can be one of several factors to be included in the decision making process. Chuck Hamaker ________________________________________ From: Frederick Friend <[log in to unmask]> Date: Thu, 26 Apr 2012 12:54:46 +0100 I agree with Joan Stein's review of the work that has been carried out over many years. We have learned a great deal about both usage and value, and I found the ACRL report particularly valuable in directing attention to institutional definitions of what constitutes value. It is in the institutional context that usage statistics can be of value, and if there is an answer to Ann's question it has to be reached institution by institution, defining the factors which apply in each institution on issues such as proportion of current and past acquisitions which are electronic or paper. The risk of inaccuracy comes when usage statistics are cumulated, e.g. to say that "US libraries are now gaining greater value from their electronic resources than from their paper resources", because value cannot be cumulated in the same way as usage statistics. Inaccuracy also creeps in when financial information is combined with usage statistics and treated as an indication of value. A statement such as "US libraries are only paying x cents for each item downloaded" is meaningless as a measure of value in isolation from the institutional and user environments which give content its value. An interesting discussion! Fred Friend Honorary Director Scholarly Communication UCL