Abstracts

LRTS v 38, no. 2, April 1994

A Citation Comparison of Sourcebooks for Audiovisuals to AVLINE Records: Access and the Chief Source of Information

Katherine Hart Weimer

To catalog audiovisuals, information must be taken from the title screens. Often these data are not consistent with data on the labels, containers, and accompanying written materials. The chief source concept for audiovisuals was examined through a comparison of citations from fully cataloged audiovisual records and their corresponding citations from bibliographic sourcebooks. Specific bibliographic elements were evaluated regarding access and descriptive cataloging. The comparison revealed much similarity with title and other title information; however, series, producer, credits, and date are more completely described in the fully cataloged record than in the sourcebook citation. There was no evidence to support cataloging using only eye-readable materials. Cataloging using the chief source of information consistently provides more bibliographic data. The arrangement and completeness of the sourcebooks were also evaluated.

Empirical Analysis of Literature Loss

Charles A. Schwartz

Recent technological innovations -- the OCLC Online Computer Library Center's Online Union Catalog coupled with the OCLC/AMIGOS collection analysis system -- provide the means to assess the declining ability of the nation's research libraries to maintain comprehensive book collections. An application of the new technology is presented in the form of a model of total book publication output and the aggregate acquisitions of 71 research libraries for a selected field over the past decade. The usefulness of such modeling is discussed in terms of generating knowledge that, while low on precision, has broad reliability and relevance.

Collection- or Archival-Level Description for Monograph Collections

Richard Saunders

Patrons seeking monographs in collections rather than as discrete items are frustrated when libraries do not keep some collections together. It is often useful for patrons to have access to groups of books related by provenance, typically in a special collections setting. The author proposes several practical methods to describe book collections in a variety of settings, so that library users can assemble and browse a list of titles received from a significant source, as part of a thematic collection, or from a descriptive catalog.

Adult Fiction in Medium-Sized U.S. Public Libraries: A Survey

James H. Sweetland

Adult fiction collections and collecting practices in 116 medium-sized U.S. public libraries were surveyed in the spring of 1990. Adult fiction represents about 26% of the typical library's holdings, with about 66% of this being "light" or entertainment fiction. The average library acquires 1,462 new titles and adds 157 more by gift in a given year, and selectors prefer to obtain new adult fiction in hard covers. For hardcover purchase, the typical library relies most heavily on the library and trade review media, followed by bestseller lists and user requests; for paperbacks, the typical library relies primarily on user requests, followed by library or trade reviews, and best-seller lists. In addition to these traditional methods, use of book clubs and similar automatic ordering systems, as well the reliance on book jobbers' lists, appears to be increasing.

Library Processing Practices by Discipline: Are Some Books More Equal than Others?

Mary Page and Melinda Ann Reagor

This study is a test of the performance of research libraries in providing timely technical services processing for materials in different disciplines. The researchers use cataloging dates from bibliographic records in the Research Libraries Information Network (RLIN) to measure the difference in processing times among selected research libraries with centralized technical services operations. A random sample of monographic titles from the 1985 catalogs of publishers of high-quality research materials is coded for the broad areas of humanities, social sciences, and science; these titles are then searched in RLIN to identify cataloging dates. The cataloging date is an imperfect indicator of availability, but it is a strong measure of overall technical processing performance. The average processing time per title is used to test whether processing times in this sample vary significantly by discipline. It was found that science materials were processed more slowly (mean 152.9 days} than materials in the social sciences (mean 130.7 days} and the humanities (mean 100.7 days).