READER COMMENTS

The intent of the White Paper on Electronic Journal Usage Statistics is to provide a basis for discussion among publishers, librarians, and aggregators that will lead to effective cooperation in collecting and analyzing usage statistics. CLIR and the author welcome readers' comments on the White Paper and other contributions to the discussion.

Having heard this report was published, I found it on your web site and browsed it. It's full of really good information and adds effectively to the emerging literature on this topic. As usual CLIR is "on point." I've one small disagreement with the conclusion that "no one is working directly with the publishers who have developed to date, understand the variables, and are in a position to provide guidance." It may be that I am misunderstanding the intent of the statement (i.e., "no one"); but from where I sit, it appears that the ICOLC guidelines developed a couple of years ago have been used agressively by numerous individual consortia, working *with publishers* to discuss what is possible, etc. Also, the ICOLC has engaged directly with key vendors on these "standards" and is in process of revising them for futher discussion. Publishers and librarians talk about usage statistics all the time! What is absolutely true, though, is that no one has organized a forum that is devoted specifically and *only* to such e-statistics matters. Judy is right: there is a facilitating opportunity here. Whether or how Professor McClure proposes to do just that under the umbrella of the ARL e-metrics initiative, I'm not quite sure. But I believe that such a meeting is envisioned as part of the ARL project.

Ann Okerson
Sterling Memorial Library
Yale University

Judy Luther has indeed provided a useful description of the current state of affairs in this area but as Ann Okerson points out there are fora where librarians and publishers have discussed these issues and continue to discuss them and further try to refine them.

The most recent ARL Newsletter provides a short summary of the ARL E-Metrics projects and articles on related developments:

• DATA GATHERING PRACTICES IN THE NETWORKED ENVIRONMENT
  http://www.arl.org/newsltr/213/data.html
• ROUND-UP OF OTHER E-METRICS DEVELOPMENTS
  http://www.arl.org/newsltr/213/roundup.html
• RESEARCH LIBRARY SPENDING ON ELECTRONIC SCHOLARLY INFORMATION IS ON THE RISE
  http://www.arl.org/newsltr/213/spend.html

The first article mentioned above is a summary description of Phase One activities from the ARL E-Metrics project under contract with Florida State University's Information Use Management and Policy Institute. The ARL E-Metrics project is directed by Charles R. McClure, Wonsik "Jeff" Shim, and John Carlo Bertot under the leadership of project co-chairs, Sherrie Schmidt, Dean of University Libraries, Arizona State University Library, and Rush Miller, University Librarian and Director. The full description of Phase One report is available at http://www.arl.org/stats/newmeas/emetrics/

The ARL Working Group on Database Vendor Statistics will be meeting at ALA in January 2001 to define a strategy for discussing with vendors and has been trying to establish a relation with the ICOLC 'structure' to work jointly in this area. Sue Phillips, Ann Okerson and Lee Anne George provide the current linkages between the ARL Working Group on Database Vendor Statistics and the ICOLC 'structure.'

Developments on e-statistics matters need to be viewed as one of the areas where libraries are trying to develop better accountability mechanisms to ensure the provision of quality services to their primary constituencies, their users. Both publishers and libraries are under pressures for being more accountable and for providing better services—they are under pressure to 'perform' as the competition from new, exciting and enterprising information.coms is increasing.

However, there are important differences between publishers (for profit and not-for-profit) and libraries. In the mix, the new, exciting and enterprising information.com world is offering yet another level of complexity. Libraries' value is determined by the direct and indirect SOCIAL benefits accrued to its users; not-for-profit publishers are valued both for the quality of their services and for the cost-recovery viability of their operations; for-profit publishers value is determined by the level of profits they are accruing; and, the value of the information.com world is determined by a combination of the venture capital they can secure and the speculation of how quickly they can turn it into profit.

This is said to simply ascertain that beyond the economic complexities of the scholarly information markeplace, there are fundamental value differences between what libraries may need to measure and what publishers may need to measure notwithstanding that they may be offering the same products. As libraries are approaching publishers asking them for various e-statistics that might help them demonstrate the value of the library to the user community, vendors and publishers need to realize that libraries' requests are only a small part of the accountability and service quality mechanisms they need to build into their systems. Publishers and vendors will need to measure more than what libraries will want them to measure for their own accountability and quality assurance processes and purposes.

Libraries have some accountability systems into place with standards such as the ANSI/NISO Z39.7-1995 Library Statistics Standard, ISO2789, and ISO 11620. Libraries are also looking more seriously into developing an understanding of library service quality issues with efforts such as LibQUAL+ and outcomes-based assessment .

NISO is planning a Forum on Library Statistics in February to inform the forthcoming revision of the ANSI/NISO Z39.7-1995 Library Statistics Standard. Vendors and publishers may need to move into the direction of creating their own versions of statistics standards (some in cooperation with libraries, others independent of libraries) to ensure the provision of quality services for their own survival.

Statistics and metrics do not have an inherent value of their own—it is the underlying concepts they are trying to measure that are important. Libraries as well as vendors and publishers need to define the concepts that are important to measure and describe—the concepts that will tell the story of libraries, publishing, scholarly pursuit of knowledge, access to information, etc. —"numbers are just the evidence that will support these stories."

Martha Kyrillidou
Senior Program Officer for Statistics and Measurement
Association of Research Libraries
21 Dupont Circle
Washington DC 20036