Conference Reports

ACRL STS Forum for Science and Technology Library Research

1. Citation Patterns of the Physics Preprint Literature with Special Emphasis on the Preprints Available Electronically
   Greg Youngen, Physics/Astronomy Librarian
   University of Illinois, Urbana-Champaign

   The author gathered data from the Los Alamos National Lab (LANL) High Energy Physics preprint server, the print literature, and Science Citation index to address these questions:

   • Are preprints gaining on conference papers and journal articles as a primary information source?
   • Are the number of citations to preprints growing in proportion to the number of preprints appearing electronically?
   • Do citations to preprints tend to include the electronic preprint number used by electronic preprint servers such as the one at LANL?
   • Are authors of preprints more likely to cite other preprints?
   • Are conference papers more likely to cite preprints than are refereed journal articles?

   Preprints may fit the criteria listed below, which Mr. Youngen cites from an article by David Lim:

   • manuscripts that have been reviewed and accepted for publication.
   • manuscripts that have been submitted for publication but for which a decision has not been made
   • manuscripts that are intended for publication but are being circulated among peers for comment prior to submission for publication

   Youngen focused on the disciplines of physics and astronomy in which 12,000 preprints appear annually, with a growing number only in electronic format. LANL and maintainers of the Stanford Linear Accelerator Laboratory (SLAC) SPIRES database work cooperatively to manage preprints in electronic format, or e-prints.

   The LANL server began loading full-text preprints in 1991. Another source is the Astrophysical Data System at Harvard. That server includes papers which link to LANL.

   Citations to preprints have declined since 1988 while citations to e-prints show a rapid increase since 1992. Citation format is not entirely consistent, but a report number often indicates electronic preprints, while the word, "preprint," appears for paper copies. The journals Astrophysical Journal and Nuclear Phyics B instruct authors not to cite e-prints, but papers in those journals are among those with the highest incidence of such citations.

   The American Physical Society has a policy to cite report numbers rather than calling the documents preprints. A question remains of whether conference papers and e-prints more often cite preprints and e-prints.

   Researchers in physics and astronomy appear to rely on the establishment in their disciplines and judge reliability or potential value of preprints by the names they see on the work.

   When searching ISI's SciSearch to gather citation data, Youngen used the statement, "CW=preprint," to refine his search for the type of publication he sought rather than including phrases such as, "in press," "to appear in...," and other terminology commonly used to refer to papers not yet published in final form.

   Youngen has future plans to do further analysis comparing preprints to their final products in order to monitor change.

   Some examples Youngen mentioned are that the AIP and IOP are posting perprints of articles to appear in journals they publish months before they are printed in final form. By contrast, the New England Journal of Medicine and the American Meterological Society do not accept any paper that has been posted previously on the web.

2. Collection Development Strateies for Spatial and Numeric Data Files
   William H. Walters, Bibliographer and Spatial-Numeric Data Specialist
   Cornell University

   Walters reviewed Cornell University Mann Library's approach to managing spatial data, such as GIS, and numeric data in digital format. Bibliographers at Cornell have developed the genre model for collection development. Four genre areas have been established: full-text, bibliographic, numeric, and spatial data, and a bibliographer specializes in each. The four full time bibliographers work as advocates for the subject areas represented in the thirty academic departments Mann library serves.

   The library's web page, {http://www.mannlib.cornell.edu/}, serves as the primary link to electronic resources and is called a "Gateway" to information on the Internet. Symbols on the page indicate whether sites offer unrestricted access or are available only to members of the Cornell community. Scope notes written by bibliographers add value to sites selected for the pages. The notes are more complete and detailed for spatial data files and web sites.

   Walters attempts to identify publishers, data distributors, and web sites which offer information in areas that support research and teaching. He examines current print holdings to investigate whether electronic equivalents or supplements may be available. Conversely, he reviews current electronic holdings to identify printed equivalents. Faculty suggestions and directories of government agencies often lead to useful sources.

   Census information, agricultural economics statistics, and human genome project data are examples of the kinds of information gathered. Researchers might create digital line graphs using data from several sources overlaid to depict several different characteristics for same geographic area.

   Cornell bibliographers have developed an outline of disciplines identified as "core selection areas" for the library, and these usually correspond to academic departments.

   Evaulation Criteria for selection:

   • absolute value
   • value as component of particluar library collection
   • reliability and ease of access
   • archival value
   • direct costs
   • indirect costs

   Bill referred to the Spatial Data Transfer Standard, SDTS, which has potential to facilitate more creative and widespread use of such data.

   Unresolved issues:

   • preserving and archiving data, "connecting rather than collecting"
   • users' ability to get and use data
   • license restrictions
   • need for clear statement of selector responsibilities