ELIS_OTDCF_v21no3.doc by Norm Medeiros Coordinator for Bibliographic and Digital Services Haverford College Haverford, PA Electronic Resource Usage Statistics: The Challenge and the Promise ___________________________________________________________________________________________________ {A published version of this article appears in the 21:3 (2005) issue of OCLC Systems & Services.} “I ain't what I used to be, but who the hell is?” – Dizzy Dean ABSTRACT This article discusses the challenge of maintaining meaningful usage statistics for electronic resources. The article notes the potential value of longitudinal data at the institution level that would help libraries make renewal and cancellation decisions. It describes a proposal to integrate usage statistics with other criteria in order to develop a decision support mechanism. KEYWORDS Electronic resource usage statistics; e-resources; Caryn Anderson; Andrew Nagy; Tim McGeary This past Christmas, my mother-in-law gave me a book of quotes from the great Peter Drucker. Like the tear-away desktop calendars my wonderful mother-in-law is also fond of gifting me, the Drucker book is designed to impart one iota of wisdom per day, a ration that fulfills my mind’s capacity for such things. Although there are no usage prescriptions for either the Drucker compilation or the calendar (“Amazing but True each are meant to be read at the start of one’s day. Because I’ve positioned the calendar in line with my office window – a window that receives many exasperated stares, usually following receipt of a budget-crushing journal invoice – the calendar never falls far behind the actual date. Occasionally I catch myself lapsing by a day or two, but a quick rip of the pages gets me synchronized. The Drucker text, however, I find much more difficult to keep current. For one, the book isn’t as obvious as the calendar; it rests near other books and papers that I generally ignore. Moreover, unlike the calendar, the dates don’t stare angrily at me, as if saying, “January 7th was a week ago. Get with the program.” Yet when I do open Drucker’s book of “insight and motivation,” it provides me with thoughts that resonate. I think the March 2nd anecdote is especially revealing of some library applications (Drucker, 2004): “The test of an innovation is whether it creates value. Innovation means the creation of new values and new satisfaction for the customer. A novelty only creates amusement. Yet, again and again, managements decide to innovate for no other reason than that they are bored with doing the same thing or making the same product day in and day out. The test of an innovation, as well as the test of ‘quality,’ is not ‘Do we like it?’ It is ‘Do customers want it and will they pay for it?’” I’ve had the good fortune to meet recently with a small group of individuals from Villanova University, Simmons College, and Lehigh University to discuss the need for an application that will help libraries manage e -resource usage statistics. The catalyst for this discussion was the recognition that these statistics could be incredibly valuable to libraries if harnessed in a meaningful and at least partially-automated way. The scope of our initial discussion has since grown into a much larger framework, all of which I believe to be an “innovation” rather than merely a “novelty” to classify using Drucker’s terms. A look to the past often lends insight into today’s issues. In researching how periodical usage statistics were managed in past decades, I came across an article by Robert Broadus that discusses the value of use studies, but cautions that these studies “measure not what should have been used, but what was used” (Broadus, 1985). Large packages of e-journals that provide access to formally-unavailable e-journals – unavailable because the library recognized that the title was not relevant to the curriculum, pertinent to faculty research, or of academic value – often receive usage because they’re just a click away. These uses of convenience, unfortunately, can be neither counted nor prevented. Broadus continues his piece by positing that a well-performed use study should predict future use of periodicals in a library. For instance, if Journal X is only marginally used during years one -through-three of a use study, and the faculty and curriculum in the discipline to which Journal X is aligned remain constant, it’s reasonable to assume year-four use of Journal X will remain low. Likewise, high use of Journal Y throughout a three-year period should result in continued high use of Journal Y in year-four of the study, given no changes in the faculty and curriculum of the discipline to which Journal Y is aligned. Evidence from the journal study I administer is consistent with this theory. Certainly there are instances where spikes in usage are consequential of a class assignment or other one-time need, but over the course of several years’ study, usage trends have remained fairly steady. Broadus raises a question, however, for which little research has been done; that is, how consistent are journal uses between similar libraries? If Journal X maintains low usage in my liberal arts college library in Pennsylvania, does this journal have similarly low usage in liberal arts colleges elsewhere in the States? Phil Davis provides some insight with his look at the Northeast Regional Libraries’ (NERL) use of the Academic Ideal e -journal package (Davis, 2002). Davis found that the research and medical institutions within NERL during the two years studied tended to use the same group of e -journals most frequently. On the other hand, undergraduate institutions tended to show little similarity in their uses of e-journals within the Ideal stable. Further study substantiating Davis’ findings would be of value to collection development officers. The development work of Caryn Anderson (Simmons), Andrew Nagy (Villanova), and Tim McGeary (Lehigh) mentioned above will fill a void in the e -resources spectrum. Although the Digital Library Federation (DLF) Electronic Resource Management Initiative’s (ERMI) functional specifications accommodate both metadata about the availability, frequency, and location of usage statistics, as well as the actual storage of usage statistics, it’s unlikely vendors building e-resource systems will soon begin work on this important, but glamourless feature. The Anderson/Nagy/McGeary model would incorporate usage statistics into a larger framework that would include elements such as price, impact factor, and faculty interest. The result would be a decision support mechanism that could communicate with library management and electronic resource systems. It’s a powerful idea that I hope will acquire the credentials of the DLF or another funding agency so that this work can be realized. It’s only fitting to end this column the way it began, with a serving of wisdom from Peter Drucker (Drucker, 2004): “Everything improved or new needs first to be tested on a small scale; that is, it needs to be piloted. The way to do this is to find somebody within the enterprise who really wants the new. Everything new gets into trouble. And then it needs a champion. It needs somebody who says, ‘I am going to make this succeed,’ and who then goes to work on it. … If the pilot test is successful – it finds the problems nobody anticipated but also finds the opportunities that nobody anticipated, whether in terms of design, or market, or service – the risk of change is usually quite small.” Dated March 11th; read July 14th. REFERENCES Broadus, R.N. (1985). “The measurement of periodicals use,” Serials Review, vol. 11, p. 57- 61. Davis, P.M. (2002). “Patterns of electronic journal usage: Challenging the composition of geographic consortia,” College & Research Libraries, v. 63, no. 6, p. 484-497. Drucker, P.F. (2004). The Daily Drucker: 366 Days of Insight and Motivation for Getting the Right Things Done. New York, NY: HarperBusiness.