College and Research Libraries Letters Directors of University Libraries To the Editor: Several articles have appeared recently on the careers of directors of libraries (see McAnally and Downs 'in C&RL of March 1973, Cohn in C&RL of March 1976, and Parsons in Wilson Library Bulletin of April 1976). In this connection, the statistics that follow should shed additional light on the subject of early withdrawal from adminis- tration. Shown below are percentages relating to four groups of directors who withdrew from administration prior to age sixty-five (most universities require retirement at this age). For example, of those directors in office in 1956, 31.5 percent had in the period 1956- 72 left administration prior to age sixty-five. The directors studied were those who were responsible for academic libraries within the Association of Research Libraries (ARL). 1926-42 10-20 percent 1956-72 31.5 percent 1956-63 21 percent 1968-75 26 percent Looking first at the two shorter periods, it should be noted that the increase in the percentage of those leaving early in the period 1968-75 probably reflects the de- cline in fiscal support that began to be felt in 1967. Still other reasons can be adduced, but some of the more serious ones, such as computerization, were felt in the period 1956-63 as well. Furthermore, in both periods directors sometimes withdrew early for reasons having little or nothing to do with administrative tensions. As of 1956 there were thirty-eight aca- demic libraries in the ARL. With respect to the directors of these, by 1972 twelve of these had left administration prior to age sixty-five; another twelve were still in office in 1972; and fourteen had dropped out at age sixty-five or older. The twelve still in office in 1972 comprised a durable cohort: five had served between eighteen and twen- COLLEGE & RESEARCH LIBRARIES ty years, another five between twenty-one and twenty-eight years, while two had been in office for thirty-five years. These figures relate only to the years served as director. Even among those who left prior to age sixty-five in the period 1956-72 (it was in 1972 that McAnally was gathering his evi- dence) , several directors had served long periods: five had been directors for fourteen to nineteen years, and three between twen- ty-two and twenty-five years. The remain- ing four were less durable, having served ten or fewer years; all four of these dropped out prior to 1967. In the years before 1945, pressures were less intense than in the postwar years ( ab- sent were computerization, area studies, and the demands for greater productivity) . Besides, the work ethic was still widely re- spected, and few gave serious thought to early retirement. Thus it should come as no surprise that the percentage dropping out prior to age sixty-five would be smaller than in the later period. Though the precise fig- ure is somewhat in doubt, it is clear that no fewer than 10 percent and no more than 20 percent were involved, with the latter figure being the more probable one. To some persons 20 percent may appear to be surprisingly large for these earlier years, but the explanation seems obvious: admin- istrative pressures are not the only clue to early departure from administration.-Lou- is Kaplan, Professor Emeritus, The Univer- sity of Wisconsin-Madison. ARL Statistics To the Editor: The ARL Office is currently gathering the statistics from our member institutions for our two statistical reports, 0 so some of the problems noted by Mr. Piternick (C&RL, Sept. 1977) are very much on our minds. 0 ARL Salary Survey, 1976-77, to be pub- lished in October 1977; ARL Statistics 1976- 77 to be published in December 1977. I 521 522 1 College & Research Libraries • November 1977 Immediately after the publication of the ARL Statistics 1975-76, a statement echo- ing the concerns expressed by Mr. Piter- nick was issued from this office regarding the comments in the introduction to the published survey. These comments ana- lyzed the ability of ARL libraries to de- velop their resources and services. In pointing out the problems with the analysis, we included a table that indicated the change in ARL university members since 1968-69. For . those readers interested in performing additional analysis using ARL data, Table I below reflects changes in the ARL membership in the past eight years. TABLE 1 AssociATION OF REsEARCH LIBRARIES MEMBERSHIP Year 1968-69 1969-70 1970- 71 1971-72 1972-73 1973-74 1974-7.5 1975-76 Number of Members 76 76 78 78 81 82 88 94 Institutions Added Howard; Rice Arizona State; Calif., San Diego; Calif., Santa Barbara Brigham Young; Kent State (St. Louis dropped membership) Colorado State; Emory; Houston; South Carolina; SUNY Albany ; SUNY Stony Brook Hawaii; McMaster; Miami; Queen's VPI; Western Ontario In terms of Mr. Piternick' s other com- ments, while we would reinforce his con- cerns that those using the data need to understand their limitations, at the same time we would emphasize more strongly the potential usefulness of the data for analyz- ing the riature and conditions of the re- sources of the nation's great research libraries. One need only refer to the sig- nificant study by William Baumol and Matityahu Marcus, Economics of Academic Libraries (Washington, D.C. : American Council on Education, 1976), to appreciate the important insights that can be achieved by a thoughtful and intelligent use of statistical data such as those published by ARL. The ARL Committee on Statistics is cur- rently investigating other methods of gath- ering and analyzing ARL data, including adapting some of the statistical models used by Baumol and Marcus. We hope to issue a special report on this later this year.- Suzanne 0. Frankie, Associate Executive Director, Association of Research Libraries, Washington, D.C. Response To the Editor: I think only a few points need be made in response to ARL' s comments: I. The "statement echoing the concerns . . . " referred to was not known to me at the time the article was written. ARL has kindly provided me with a copy-it does little more in this connection than call at- tention to changes in ARL membership made during the last decade and hence applies to only one of the many criticisms made by me. 2. ARL statistics do indeed constitute an important body of data and have been used, as I mentioned, in a number of sig- nificant studies among which, of course, that of Baumol and Marcus belongs. It is inevitable and desirable that they be so used, and thus that their quality be as high as is achievable . I have not advocated dis- continuing the publication of ARL Statis- tics, only .that they be improved. 3. What I am advocating is simply that ARL recognize and call attention to the obvious limitations in the accuracy and comparability of their data in the publica- tion of the figures and in treatment of them in making comparisons and in drawing con- clusions. In most cases this means a reduc- tion in the number of significant figures used in the publication of data, the avoid- ance of rank-orders assigned on the basis of insignificant differences, and the use of more appropriate analytical methods in drawing conclusions as to institutional dif- ferences and common trends. 4. It is gratifying to learn that ARL is currently investigating alternative methods of gathering and analyzing statistical data. -George Piternick, Professor, School of Li- brarianship, University of British Columbia, Vancouver. The Atkinson Report To the Editor: I read with considerable interest Mr. De Gennaro's review of Capital Provision for University Libraries: Report of a Working Party [the Atkinson Report] (C&RL, July 1977). There has been considerable discus- sion about this document in Australia, pri- marily due to attempts by a number of university administrators to apply the con- clusions of the document in the Australian environment. I draw the attention of your readers to several articles by Mr. Harrison Bryan, uni- versity librarian, University of Sydney, on this particular report. These have appeared in the Australian Library Journal and in Australian Academic and Research Librar- ies. He has done a detailed analysis in rela- tion to applying the recommendations of the report both to British libraries and to Australian libraries. I think that it is essential that American librarians not only be aware of this report and its implications but of reasoned argu- ments in relation to these views. The mea- sures involved in this document ignore many important aspects of colleges and university libraries. It is an attempt to de- velop standards based on student numbers ignoring the requirements for breadth of teaching and level of courses. These are es- sential elements in considering university library provision.-M. Jacob, Director, User Services Division, Ohio College Library Center, Columbus, Ohio. Collection Development To the Editor: As a fellow bibliometrician, I would like to add a few. comments to Jim Baughman's article in the May 1977 issue of C&RL as implementation tips for collection develop- ment librarians inclined to try the tech- niques recommended. Data quality control, specifically the ini- tial choice of the bibliography used in the study, is of paramount importance. The scope of the indexing or abstracting service in terms of periodicals covered and of in- dexing policy determines how the litera- ture is defined. Next, inclusion or exclusion of specific data in the analysis may affect the results. For example, Baughman select- Letters I 523 ~d entries under 309 subject headings, that IS 6 percent of the total of 4,995 subject headings, for the purpose of investigating "ma-cro relationships." While his rationale appears sound, can it not be the case that journal articles on specific topics represent- ing the less "productive subject" areas ( op- erationally defined as entries under subject headings with less than eleven entries) are indeed the literature of new, often inter- disciplinary fields of inquiry, which become linking agents between disciplines? Data processing techniques to implement Coffman's indirect method (a general meth- od, and not a technique by itself) may in- clude clustering techniques, often available as canned programs in libraries of local computer centers. Consultation with com- puter scientists or computing consultants on campus may be fruitful to select a tech- nique, which preserves the properties need- ed for finding relationships between data units, as recommended by J ardine.l The canned program selected may add refine- ments in arithmetics, a detail important from an engineering standpoint but not elaborated in Coffman's general formula for establishing communication. In case widespread interest manifests it- self among working librarians in biblio- metric experiments, a formal study and discussion group within ACRL with liaison to the special interest group (SIC) on. The- oretical Foundations of Information Science of the American Society for Information Science should be established. Thus we ~ay organize to cope with a micro-problem m the interdisciplinary macro-ocean.-Su- san V. Emerson, Visiting Assistant Profes- sor, School of Librarianship and the Li- brary, University of Oregon. REFERENCE 1. N. Jardine, "Algorithms, Methods and Mod- els in the Simplification of Complex Data,'' Computer ]ournall3:116-11 (Feb. 1970). To the Editor: James C. Baughman's comments on cita- tion patterns in the May 1977 C&RL re- quire some modification. What he identifies as a "serial" apparently excludes most mon- ographic series, since the latter are rarely identified in the footnotes of scholarly jour- nals. For example, in the articles of the 524 I College & Research Libraries • November 1977 197 4 issues of the American Historical Re- view, about 11 percent of the monographs cited turned out to be in either a traced or numbered series (excluding publishers' trade series). Further, if one limits one's consideration to the post-1955 monographic citations (which will account for the bulk of today' s acquisitions), almost 15 percent are from a monographic series. Thus, add- ing monographs in series to periodicals and other identifiable serials gives a total of more than 30 percent, which is substantial- ly larger than the 23 percent reported on page 246 of Baughman's article. The latter, however, is a more useful figure (when qualified), since a monographic series can be acquired as individual volumes.-]oseph ]. Lauer, History Librarian, SUNY at Buf- falo Libraries. Response To the Editor: It is with gratitude that I acknowledge the letters submitted by Professor Emerson and Mr. Lauer. Their comments display interest and insight. In response to Professor Emerson's letter, I enthusiastically concur with her call for the creation of a discussion group within ACRL with liaison to the special interest group within ASIS. Currently, Library Lit- erature does not list the term "collection development" as an independent subject heading. Realizing that collection develop- ment is a major problem, one that will be- come more acute as our body of literature increases disproportionately to fiscal abun- dance, the profession should engage in a more serious determined effort toward problem solution in this area. While the ALA Handbook does not list a collection development committee within ACRL, such a committee could serve many useful pur- poses, including stimulating research in the structuralist and other approaches, co- ordinating activities, providing leadership, and promulgating policy guidelines based upon systematic knowledge. Hopefully, ACRL will assume a leadership role in this important, emerging aspect of library man- agement. Professor Emerson's cautionary statement concerning general methods, and algorithms to implement these methods, should be carefully considered by investigators to in- sure the integrity of their future studies in- volving clustering methods. As Dr. Emerson implies, there are constraints in any study; in this instance the Social Sciences Index ( SSI) is the major constraint. The 309 sub- ject headings contain 36 percent of the en- tries from SSI (Volume 1, 1974-75). The decision to use 309 subject headings was based on the notion of high frequency and low frequency; i.e., subject headings with eleven or more entries were determined to be high frequency (more productive) sub- ject areas. The rationale for using the high frequency subject areas is based on the hy- pothesis that the more productive subject areas would reveal the current state of af- fairs in subject relationships, hence, current established interdisciplinary relationships. Dr. Emerson's question, asking if "the less 'productive subject' areas . . . are in- deed the literature of new, often interdisci- plinary fields of inquiry" and may "become linking agents between disciplines" not only represents an important problem in the growth of subject literatures but also sug- gests a relevant hypothesis that needs further investigation in the social science literature. In short, journal literature per se is a relevant topic for productive research. With respect to Mr. Lauer's letter, I am grateful for his clarification of the term "serial" and heartily concur with his state- ment that "the latter [the 23 percent fig- ure] . . . is a more useful figure (when qualified), since a monographic series can be acquired as individual volumes." It is my hope that others will realize that the structuralist approach, with its orienta- tion toward discovering underlying relation- ships, is a valid and important method leading to systematically based collection development.-]ames C. Baughman, Asso- ciate Professor, School of Library Science, Simmons College, Boston, Massachusetts. Editor's Note: The Resources Section of the Resources and Technical Services Division includes a Collection Development Com- mittee (which has as one of its functions the development of "guidelines for the for- mulation of selection parameters") and two discussion groups for chief collection de·- velopment officers, one for large research libraries and a second for medium-sized re- search libraries. WHO'S WHO IN HEALTH CARE Who are the government and industry officials formulating health policy and providing for the nation's health carel Who are the key executives in the nation's hospitals and other health facilitiesl Wflo are the leaders in .health education who .influence the entire health care professionl The answers to these questions, and many more, will be · found in Who's Who in Health Care, the First Edition of which will be published .in October, 1977_. EXCERPJ' OF TYPICAL BIOGRAPHY. Kerr Lorin E, physician, occupational . hlth. b. July 5, 1909, Toledo OH; m Gol deceased; children - John W, Susan a Judith K Valentic. MD, U of Mi • . 1 MSPH, 1939; intern Toledo Hosp, 1935'-36, res, 1936-37; res Nell'opsychiat I U of MI. 1937. Dir Eknau of Med Relief, 1937-38; acting supt Municipal Toledo Hlth Dept, 1937-38; asst dir Ingham Co Hlth Dept and Ml Student Trc Cntr, Mason MI. 1940; dir Iron Co Hlth Dept Stambaugh MI. 1940-41; COn:lm L Co Hlth Dept Obertin OH, 1941-44, sr surg USPHS, 1944-48; asst to e)(ec ~ l.HWA Welfare and Retirement Fund, 1~48-69; dir dept of Occup Htth United COMPREHENSIVE Who's Who in Health Care, 81/2" X 11" trim size and hard bound, will profile more than 8,000 health industry profes- sionals including medical school educators, health research- ers, senior government officials, hospital and nursing home administrators, leaders in public health, HMO and PSRO directors, foundation, pharmaceutical and insurance execu- tives and many others. EASY TO USE Who's Who in Health Care is conveniently arranged alpha- betically and is cross referenced by name, state, current affilia- tion and professional field. INVALUABLE Who's Who in Health Care is a reference source that no library, professional office or school should be without. Pre- paid price of $60.00 includes ·shipping and handling. If not completely satisfied, return within 10 days for a full refund. r-------------------------------~-~~~ HANOVER PUBLICATIONS, INC. I 200 Park Avenue • Suite 303E New York, New York 10017 Gentlemen: Please reserve --- copy(s) of Who's Who in Health Care, First Edition, at $60.00 a copy. 0 Enclosed is my check or purchase order in the amount of >--------- 0 Please charge to my credit card account (please check): 0 Master Charge 0 BankAmericard My card number is _______________ and expires on ------ (Month) (Year) Credit Card Signature ______________________ _ Please ship my copy(s) of Who's Who in Health Care to: Name ___________________________________________________ __ Street Address·--------------------------------------------------- City _____________________ State __________ Zip ____ _ l---------------------~--~----------j From Australia ... DESIGN FOR DIVERSITY Library Services for Higher Education and Research in Australia Harrison Bryon and Gordon Greenwood, Editors 790 pp., 16 B&W photos. Index, footnotes, appendices 0-7022-1314-4 $39.95 A serious and scholarly examination of the most significant sector of the Australian library world. · Education at the college level is a major growth industry in Australia, as in all expanding societies. In Design For Diversity, Australia's leading librarians from 11 major institutions explore the library conse- quences of rapid development in this field. Harrison Bryan is Director of the University of Sydney library, Australia's oldest, largest, and most heavily used university library. He has played a leading part in the development of the profession of librarianship in Australia, and helped bring to Australian libraries the international recognition they deserve. Gordon Greenwood has headed the Department of History at the University of Queensland since 1949. A distinguished historian with an international reputation, he is long time Chairman of the Library Com- mittee of his university and Foundation Chairman of the Commission on Advanced Education. Table of Contents The Development of Academic & Research Libraries in Australia Provision & Planning at the National Level The State Libraries Libraries in Colleges of Advanced Education Library Resources for Teaching and Research in the Humanities & Social Sciences Scientific Information Resources, Services & Needs Archives & Research Library Cooperation: Reality & Limitations Direct Service to Readers The Library as Educator Library? Resource Materials Center? Library-College? Some Aspects of Administration and Management A Decade and a Half of Library Building (1959-73) Automation & Technology Today & Tomorrow Education and Training Staff Policy Formation & Library Experimentation in Australia & Other Western Societies THE UNIVERSITY OF QUEENSLAND PRESS c/o Technlcallmpex Corporation, 5 South Union Street, Lawrence, Mass. 01843