College and Research Libraries Evaluating the Conspectus Approach for Smaller Library Collections Larry R. Oberg The conspectus method of collection evaluation has been successfully used since the late 1970s by the large member libraries of the Research Libraries Group and the Association of Research Libraries. Today, the Library and Information Resources for the Northwest (LIRN) and other conspectus-based collection assessment projects are demonstrating the value of this approach to smaller university, college, and public libraries as well. This article reviews the LIRN project, · the LIRN modifications to the original RLG instrument, and the conspectus process. It also discusses conspectus methodology problems, staff time costs, and the value of the completed conspectus to an individual library or group of libraries. ne of the most vexing problems facing academic librarians, in- deed all librarians, is the lack of adequate, reliable fools for eval- uating our collections and services. We measure what is easiest to measure, Goldhor, notes, most often process or in- put, not output variables. 1 Many of our traditional measures have been subjective and impressionistic and remain so today. For example, we deter- mine the quality of our reference services by measuring patron satisfaction, the number of scheduled desk hours, and staff educational level and years of experi- ence. Yet these determinants are ancillary to what we really want to know, which is, How accurate are our responses to patron questions? We have not done much better evaluat- ing our collections. We seem unable to compare them to the total universe of in- formation or with collections in similarly sized institutions. Often, we do not know how well they match the curriculum and meet the needs of students and faculty. In the face of proliferating new publica- tions, severe inflationary pressures, and increasingly sophisticated competition at budget time, we cannot effectively com- pete for static or dwindling funds by argu- ing that we need more money because we have too few books or by reminding our provosts that the library is the heart of the institution. Instead, we must seek out, de- velop, and utilize measures that demon- strate objectively, both to administrators and to ourselves, how well we are fulfill- ing our role and mission. Faculty and administrators, if they are to support our programs, must be reassured that requests for increased funding derive from a systematic, sustained planning and evaluation process. We must convince them that we are spending their money wisely by measuring, in other than subjec- tive terms, the collection and service im- provements resulting from higher-level funding. One method of collection evaluation, Larry R. Oberg is Director of Libraries at Albion College, Albion, Michigan 49224. This article expands upon a paper read at the Liberal Arts College Library Directors' Meeting, Grinnell College, Grinnell, Iowa, October 19, 1987. 187 188 College & Research Libraries the conspectus, fulfills at least some of our needs. This paper describes the conspec- tus collection evaluation process, dis- cusses its problems and cost in staff time, and demonstrates its usefulness to smaller college, university, and public libraries. THE CONSPECTUS The conspectus is a collection assess- ment method that maps subject strengths and weaknesses within an individual li- brary, a consortium of libraries, or a geo- graphical region using standardized crite- ria and descriptions . The dictionary definition of conspectus is a survey or syn- opsis of a subject, but in library parlance it refers to an assessment methodology de- veloped in the late 1970s by Paul Mosher of Stanford University and other Research Libraries Group (RLG) collection develop- ment officers for use by RLG librarians. 2 "The conspectus was designed to produce comparable data to facilitate collection coordination among the large RLG libraries." The conspectus was designed to pro- duce comparable data to facilitate collec- tion coordination among the large RLG li- braries. Mosher recalls that the goal of the original development group was to de- scribe "all major U.S. research libraries as part of the largest scholarly research re- source collection the world has ever known." 3 Since its introduction by RLG, the con- spectus has been adopted by the North American Collections Inventory Project (NCIP). Developed in 1983 by the Associa- tion of Research Libraries' Office of Man- agement Studies, NCIP is using the con- spectus to generate data for an online inventory of North American research col- lections to assist scholars in finding the re- search materials they need. 4 Conspectus methodology is used by North American academic and public li- braries of all sizes and, increasingly, by May 1988 foreign libraries as well. In addition to in- dividual institutions such as Albion Col- lege Library, groups using the conspectus in the United States include the Alaska Statewide Inventory Project, the Illinois Statewide Collection Development Proj- ect, New York's METRO (Metropolitan Reference and Research Library Agency), Idaho's VALNet (Valley Library Net- work), 5 the Boston Library Consortium's Collection Analysis Project, and the Li- brary and Information Resources for the Northwest (URN), a regional project es- tablished in 1984 and underwritten by the Fred Meyer Charitable Trust, a private foundation located in Portland, Oregon. 6 LIRN URN's thrust was fourfold. It sought to assess the depth and quality of the infor- mation resources available in the region; establish a shared database for manipulat- ing, analyzing, and displaying the infor- mation generated by the assessment; fos- ter and encourage cooperation and resource sharing among all types and, im- portantly, all sizes of libraries; and finally, establish a technologically advanced re- gional document delivery system for cost- effective information exchange. The first charge to those of us who par- ticipated in the URN project was to locate or develop a methodology to assess the collections of large research institutions, small colleges, and special and public li- braries of all sizes. The only instrument that held promise of fulfilling our needs was the RLG conspectus. It was modified to more precisely describe the collections of the many small libraries of the region that were expected to participate. The URN version is called the Pacific North- west Conspectus. Today, the LIRN assessment program is moving toward completion. More than 210 Pacific Northwest libraries, both large and small, academic, special, and public, are finishing or have finished at least some of the twenty-four basic subject divisions of the conspectus. 7 The data collected are reported to the Pacific Northwest Con- spectus Database operated by the Oregon State Library Foundation in Salem. Tex- tual and graphic reports comparing collec- tions at varying levels of specificity by li- brary location, type, budget, funding source, school enrollment, etc. are pro- duced in batch mode. 8 THE ASSESSMENT PROCESS The Library of Congress classification system (LC) forms the general framework of the original conspectus, although URN has developed a Dewey-based version as well. Subsets of the collection-subjects- are evaluated and ranked on a scale of 0 to 5. The Pacific Northwest Conspectus be- gins with a subject-by-subject examina- tion of a library's collection. Each subject or subset of a subject is given three sepa- rate numerical ratings that rank the exist- ing collection level (CL); the acquisitions, or monetary, commitment (AC); and the target level or collection goal (GL). The ac- quisitions commitment and the collection goal may be higher or lower than the cur- rent collection level. The subject rankings are amplified by language codes: E, predominately En- glish; F, selected foreign-language titles; W, wide selection of foreign-language ti- tles and; Y, primarily foreign-language ti- tles. Detailing language coverage adds a significant dimension to the description of a collection. The alphanumeric values as- signed to each subject, for example, 3AF, are then recorded on the conspectus work sheets. Completion of the work sheets requires concise comments to describe strengths, weaknesses, and other characteristics of the cataloged collection that are not brought out by the LC or Dewey classifica- tion systems. Examples of these com- ments are "strong in pre-1960s materi- als," "good representation of 19th- century authors, 11 and "have all editions of the International Encyclopedia of the Social Sciences. II The comments are in- cluded in the note section of the work sheets. Materials not classified in the LC or Dewey systems, for example, theses, gov- ernment documents, journals, newspa- pers, special collections, microforms, films, electronic databases, etc., are also described. Examples of these notations Evaluating the Conspectus Approach 189 are "strong in U.S. government docu- ments, 11 "weak journal collection, II and ''have excellent uncataloged radical cul- ture collection from the 1960s. II In the Pa- cific Northwest Database, the comments, which may be up to ninety characters in length, are retrieved in textual reports (see figures 1-3). EXAMINING THE COLLECTION For the most part, the methods used to examine collections are traditional. They include shelf scanning, list checking, the compilation of statistics, and citation- reference studies. But any method that sheds light on the depth of a collection may be employed. 9 Shelf scanning, a lost art for many librar- ians, produces immediate and tangible results by revealing broken runs of serials, little or heavily used materials, a lack of up-to-date materials, and other lacunae or strengths. It is accomplished quickly and easily and builds upon the subject exper- tise of teaching faculty and librarians. List checking is the most time-honored method of collection evaluation. Special- ized and selective lists are published in most subject areas by accrediting bodies, professional associations, researchers, bibliographers of all stripes, librarians, and, of course, the American Library As- sociation. A major tool for this type of as- sessment becomes available in 1988 when ALA publishes the third edition of Books for College Libraries. The subject experts most likely to be pressed into service by academic librari- ans are the teaching faculty. On most col- lege campuses, teaching faculty share se- lection responsibilities with librarians and, in any event, are familiar with the bibliographical resources of their subject specialties. The compilation of statistics and various numerical counts provides useful infor- mation on the quality of a collection. Ex- penditures, the number of volumes added per year, shelflist counts, the relative strength and use of the book and the peri- odicals collections, and an analysis of in- terlibrary loan and circulation patterns are examples of statistics usefully gathered for conspectus-based collection assessments. Library: n· . 1V1S10D: CHEMISTRY Date• ! Bv· LC Llll£ DIVISION, CATEGORIES end SUBJECTS COLLECT! 011 & LANGUAGE COD£$ I:UH .IUIIIER 1:1. AC GL !XMIENTS CHEOOO _, r DIVISION I CHEMISTRY ' l (ALWAYS IN CAPS) GD1•69 CH£001 Chentl a try GD11·23 CHEDD2 History, Biography GD23.3·39 CHED03 Alche.y CID40·49 CHE004 Stucly & Teaching GD71-145 CH£005 Anelytlcal ChPistry ~ 1 Categor~ (always underlined) J ....... -GD81·95 CHED06 OU.lftltlw Analysis (Organic & lnorpnlc) QD96 CH!DOT Spectroscopy (Applied) GD101s121 CH£008 Qulntftatlw Anelyals (OriiWIIC & lnorpnlc) """" .. Subjects l ' GD115 CHED09 £lectroch•lcal Anelyals GD117.~ CH£010 Chr-tography GD117. T4 CHE011 Thei'Ml Anelyals GD146·196 CHED12 Inorganic Chentlstry GD161-169 CHE013 Inorganic Ch•. · AI !.aline & Alka. Earth & Metals GD172.T6 CH£014 Inorganic Ch•lstry • Trinaltlonal Metals GD189·193 CH£015 Inorganic Chea~istry • Salts GD196 CHE016 Inorganic Polymers GD241·449 CHE017 Organic- Chemistry