Modeling an Academic Approval Program.

A model for core-collection development appropriate for large and medium-sized research libraries is proposed. A strategy of mechanical selection is suggested that will ensure the quality of core selection as well as release selectors from the burden of core selection so they might spend more time identifying difficult materials. We read with interest a LRTS article by Charles W. Brownson entitled " Mechanical Selection, " which investigates the usefulness of mechanical versus expert selection and argues for more quantification in selection practice.(1) Before that, a CR and we offer below an explanation of how that principle-a corollary to the 80/20 rule-might be used in collection development to model an academic approval program. A 1969 article by Richard W. Trueswell offers the clearest statement of "the 80/20 rule"-a characteristic of inventory in business that approximately 80 percent of the number of transactions taken from a warehouse represents about 20 percent of the items stocked. With graphs of circulation versus holdings that show circulation taken from several libraries, the author illustrates the similarities between business-inventory holdings and book circulation, and concludes that some of the techniques used for managing business inventories are applicable to libraries.' Our analysis of the "Outstanding Academic Books" that appeared in the May issues of Choice during a recent ten-year period suggests a similar 80/20 pattern: 80 percent of these titles are by 20 percent of the publishers. This 80/20 core is split into disparate halves: 2,315 Outstanding Academic Books, " primarily in the humanities and the social and behavioral sciences, have been produced by forty-seven university presses; while 2,258 "Outstanding Academic Books," primarily in reference and in science and technology, have been produced by eighty-three commercial publishers. The average yearly output of the university presses of the core is 3,5 89 titles, while that of the commercial publishers is 1 1,464. The challenge for a medium-sized research library, which cannot afford to gather the entire 80/20 core mechanically, but still wishes to bring in a substantial portion of the commercial half of the core through these means, is to choose between one of three possible strategies: (1) gather those publishers that provide the best ratio of "Outstanding Academic Books" to total titles issued; (2) gather those publishers that provide the most "Outstanding Academic Books"; or (3) gather those publishers whose "Outstanding Academic Books" categories, together with those of the university core, provide an even coverage so that collection balance is maintained. This article suggests methods and estimates costs for each of those strategies. While these are not the only means of getting useful material into the library, we believe this information can be used for a mechanical selection of a substantial portion of the domestic mainstream, which Schmidt spoke of capturing with a publisher-based approval plan. The efficiency of selectors will be improved by shifting the burden of core selection away from expertise in order to permit their knowledge of subject areas to be spent identifying fugitive and difficult materials. METHODOLOGY To prove the 80/20 corollary, we built a DBASE file indexing the "Outstanding Academic Books" from the May issues of Choice based on Koenig's network model . …