College and Research Libraries Research Notes Book Availability at the University of California, Santa Cruz Terry Ellen Ferland Margaret G. Robin~on A standardized methodology developed by Paul Kantor for the Association of Research Li- braries was used to investigate book availability at the McHenry Library, University of Califor- nia, Santa Cruz. The 61 percent success rate experienced by Santa Cruz users compares well with the 50 percent figure frequently found at academic libraries . Most books not found by McHenry users were unavailable for a few, clearly identifiable reasons: the library's failure to acquire an item, combined with prior charge- out of an item to another library user, ac- counted for over 60 percent of the not-found material. Availability studies at large academic li- braries indicate that users fail to locate promptly about 50 percent of the materials they seek. The present study was con- ducted to document the rate of user suc- cess and failure in finding known items at a medium-sized, open-stack academic li- brary. A standardized methodology de- veloped for the Association of Research Libraries was used . Results were expected to shed light on availability problems, sug- gest areas that need improvement, and provide data for comparison of local per- formance with that of similar institutions . SETTING McHenry Library is the main library serving the Santa Cruz campus of the Uni- versity of California. The academic com- munity consists of 6,600 undergraduate students, 600 graduate students, and 500 faculty members. McHenry houses ap- proximately 585,000 volumes in the social sciences and humanities, as well as sub- stantial backruns of serial titles that the smaller, overcrowded science library can no longer accommodate. Bibliographic ac- cess to the collections is provided by a mi- crofiche catalog of UCSC holdings, sup- plemented monthly and cumulated annually, and an online catalog that in- cludes most UCSC monographs pub- lished after 1973, as well as partial hold- ings of other UC campuses. A computer- ized circulation system was installed in 1980. Automated circulation functions are well controlled. Security is provided by staff at a turnstile exit rather than by an electronic detection system, and has been widely, if subjectively, perceived to be an area where more control is needed. Shelv- ing backlogs have typically accumulated at peak service periods in the academic year. Some staff who have frequent con- Terry Ellen Ferl is Principal Cataloger, and Margaret G. Robinson is Head, Reference Services, University of California, Santa Cruz, California 95064 . The authors gratefully acknowledge receipt of a University of California affirmative action award in support of this study. 501 502 College & Research Libraries tact with users view these backlogs as a major impediment to user satisfaction. 1 METHODOLOGY The techniques used in the present study are based on those published in Paul Kantor's Objective Performance Mea- sures for Academic and Research Libraries. 2 In Kantor's manual, availability is defined as a measure of the extent to which patron needs for specific documents are promptly satisfied. The data for the study-a sample of several hundred citations-are collected from actual user searches for specific desired items. Items reported by users as "found" and "not found" are tallied. The gross measure of availability (MA V) is thus a fraction: MA V = the number of items found -;- the num- ber of items sought. To obtain more de- tailed information, the not-found items are traced promptly through the library system and assigned to one of five condi- tion categories: DACQ the book is not in the collection DCA T = the book is in the collection but not found in the catalog by the user DCIRC the book is charged out to another user DUB the book is not on the shelf and not charged out DUSER = the book is in the right place on the shelf but the user overlooked it The prefix D stands for "disservice"; an item that falls in a particular category is considered a "disservice event" attribut- able to failure related to that category. UCSC Study The present study examined several subcategories of the five principal mea- sures of availability. Kantor notes that while subdivisions of the main categories will not prove statistically significant with a sample of 400 to 500 items, they can be "useful for thinking about what is going on. " 3 Because many performance subcat- egories were being examined, a tracing form separate from the user-survey form was designed (see figures 1 and 2). 4 Serials were excluded from this project since bib- liographic access for them varies consider- September 1986 ably from that for books. Four student assistants with previous li- brary experience were employed to collect the data, primarily throughout the month of November 1984, a period of high library activity in the academic cycle. A total of 33 1 /4 hours was used to distribute 363 forms to users, an average of 11 forms per hour. Distribution was done during the peak-use hours between 10 a.m. and 3 p.m. and was always curtailed early enough for student assistants to complete the tracing of the returned forms. There- turn rate of 42.9 percent yielded 156 valid forms with 408 usable citations, in accord with Kantor's estimates. Of the total forms distributed, 207 were either not re- turned or not usable. The unusable forms were eliminated primarily because the item sought was a serial. The more detailed tracing process in the present study required, on the average, about twenty-five minutes per citation. Returned forms were hatched and traced as quickly as possible, usually within two hours of collection of the data. Groups of items being traced were rigorously ex- cluded from the study if the tracing pro- cess was interrupted, for example, by on- line system failure or student assistant failure to complete all tracing steps for each item in the group. Written guidelines were developed for the use of student assistants unfamiliar with the automated circulation system or other internal files. Questions were occa- sionally raised by the student assistants regarding the interpretation of user re- sponses, e.g., What exactly constitutes an incorrectly transcribed call number? or, What if the book sought by the user is on the shelf but only very slightly out of call number order? Project directors needed to be available to handle such questions and assure that citations were not rendered unusable because of delay in tracing. Distribution of the survey forms to users took three times longer than expected, due mainly to institutional size and the fact that many entrants were not looking for library materials but were instead visit- ing administrative and instructional units located in the library building. Kantor's estimates on distribution time apply to larger academic libraries with heavier cat- Research Notes 503 Are You Finding the Books You Want? We are studying the availability of books in McHenry Library. Please help us by using this form for scratch paper when you look for books. Write in the blocks below the authors and titles of all the books you want to find . If you locate any of these books in the catalog or MELVYL, write the complete call number in the column so labeled. If for any reason you don't find the books you want, put an X in the column headed "Can't Find." Please leave this form at the Exit Desk at the end of your visit. Thank you for your cooperation. Check Status: 0 UC undergraduate 0 UC grad student 0 UC faculty 0 Other Author and Title Call Number Can't Find . FIGURE 1 User Survey Form alog use. Another factor that slowed the distribution rate was the considerable number of users seeking subject informa- tion rather than known items. RESULTS One hundred and forty-five users par- ticipated in the survey. Among those re- porting their status, 70.4 percent were un- dergraduates, 10.6 percent graduate students, 5 percent faculty members, and 13.6 percent campus (but not library) staff or community members. Not surpris- ingly, graduate students were dispropor- tionately represented compared to their numbers in the total academic commu- 504 College & Research Libraries September 1986 Tracer _________ _ Date _________ _ Complete 0 FiU out one of these forms for each citation on the User's List that has a check in the can't find box. Go through each step for each citation until instructed to "Check the COMPLETE box." (A tracing form may be "Complete" at any point between steps I and 10, depending on the status of the item you are tracing.) Author ____________________________________________________________________________ _ Title ------------------------------------------------------------------------------- Call Number _________________________________________________________________________ _ Reference Area I. If you can't read the citation, check line I and check the Complete box. 2. Look up the citation in each part of the catalog and record the full UCSC call number exactly as you find it: Fiche Cumulation------------------------------- Fiche Supplement------------------------------- Online Catalog ------------------------------ If the user recorded a call number, check here __ and continue to 3. 1. __ _ If the user didn't record a call number and you found one, check line 2 and check the Complete box. 2. __ _ If you don't find the citation in any catalog, check here __ and continue to 4. 3. If the call number recorded by the user is incorrect or incomplete, check line 3 and check the Complete box. If the user's call number is correct, check here __ and continue to 5. Acquisitions Area 3. __ 4. Look up the citation by title in the Order File, check the appropriate line and check the Complete box: 4(a) Not in Order File 4(a) __ 4(b) On order but not yet received 4(b) __ 4(c) Received · 4(c) __ 4(d) In accessioned backlog 4(d) __ 4(e) Order cancelled 4(e) __ Bib Records Area 5. Check Shelf List and record number of copies on line 5. If more than one copy, account for each one as you proceed through steps 6- 10. Circulation Area 6. Look up citation by call number in automated circulation system (CLSI) and if book is charged out, check the appropriate line and check the Complete box: 6(a) User (6 letters, 3 numbers, or 9 numbers) 6(b) Reserves (RB**MCH*, RB**SCI*) 6(c) Interlibrary Loan (ILLUCB, ILLUCD, etc.) 6(d) Stored at NRLF 6(e) Library process (Code: -------. If CLSI says "On Shelf," check here _ and continue to 7. If title record is not yet in CLSI, check line 6(f) and continue to 7. 1·. Look up citation by call number in the paper Circ files . If there, check the appropriate line and check the Complete box. If not, check here _ and continue to 8. 7(a) Bindery File 7(b) In process for storage 7(c) File of items to be keyed in CLSI Stack Area/Circ Area 8. If the book is on the shelf in its proper location, check line 8 and check the Complete box. 9. Look for the book in the following shelving locations arranged by call number. Check the appropriate line if found, and check the Complete box. 9(a) Booktrucks on appropriate floor 9(b) Sorting shelves on appropriate floor 9(c) Booktrucks outside of Circ 9(d) Sorting shelves in Circ 10. If citation is still unaccounted for, check line 10 and check the Complete box. FIGURE2 Book Availability Tracing Form 5. __ 8. 6(a)_ 6(b)_ 6(c) _ 6(d)_ 6(e) _ 6(f) _ 7(a)_ 7(b)_ 7(c) _ 9(a)_ 9(b)_ 9(c) _ 9(d)_ 10. __ nity. Project results could not be analyzed by user status because of the small num- ber of survey participants in most individ- ual user categories. The results are given in tables 1 and 2. 5 Among 408 books sought, 61 percent were found by users. Of the 39 percent not found, 35.5 percent were already charged out, 24.6 percent were never acquired by the library, 11.7 percent were unac- counted for, 8.8 percent had call numbers incorrectly copied by the user, 5.3 percent were in place on the shelf but not found by the user, and 4.4 percent represented cata- log entries that the user failed to locate. These factors collectively accounted for 90.3 percent of the items not found. The remaining 9. 7 percent were in other cate- gories that were individually not statisti- cally significant. TABLE 1 PERFORMANCE DATA TOTALS Items sought Items found Items not found Number Percent 408 250 158 100.0 61.3 38.7 Research Notes 505 ANALYSIS Results may also be expressed in a branching diagram, which shows the rela- tionship between independent perfor- mance components that contribute collec- tively to the outcome of a user's search for a known item. The figures for the branch- ing diagram were obtained by transferring the raw data in a prescribed manner to an availability analysis form reproduced in Kantor's manual (figure 3). The branching diagram (figure 4) is interpreted as fol- lows. Of the total of 408 books sought, 368 had been acquired by the library, a perfor- mance of 90.1 percent. Of those 368, the users failed to find 21 books listed in the catalog, a performance of 94.2 percent. Of 346 books located in the catalog, 63 were already checked out, a performance of 81.9 percent. The investigators could not account for 25 books that users weren't able to find, a performance of 91.0 per- cent, . and users overlooked 9 books on the shelves, a performance of 96.7 percent. These factors combined to depress book availability-the chance of finding a spe- cific book-to 61 percent. TABLE2 Olegible citation Never ordered PERFORMANCE DATA FOR ITEMS NOT FOUND Ordered but not received Order canceled Catalog entry not located by user Call number copied incorrectly by user Charged out to user Charged out to reserve desk Charged out to interlibrary loan Charged out to regional storage facility Charged out to library unit or process Title not yet loaded in circulation database Received and in process In accessioned backlog At bindery In process for regional $torage facility Item not yet added to circulation database On booktruck in public area r- On sorting shelf m public area On booktruck in circulation work area On sorting shelf in public area Unaccounted for On shelf but not found by user Total Number* 1 39 1 0 7 14 561/6 1112 0 0 41/2 0 0 0 0 0 31/3 2 1 0 1f2 18112 81/2 158 *When multiple copies of a work exist in the collection, the copy traced is represented by a fraction : 1/3 oneof3copies, 1/6 copies, etc. · Percent .6 24.6 .6 0.0 4.4 8.8 35.5 .9 0.0 0.0 2.8 0.0 0.0 0.0 0.0 0.0 2.1 1.2 .6 0.0 .3 11.7 5.3 100.0 oneo£6 506 College & Research Libraries September 1986 Availability Analysis CORREcnON FACTOR = (All reponed lailur .. )/(Analy•ed Failur") (7) = !51 7 ..J.!Z._ List cmy CODftDieDI breakdOWD ol D-ACQ A = l:...LH Mev C.r '"l'. fU.i,., J Form Subtotals B Corrected dis- c On_ petite D (20) ~ J.LL.% MAV-ACQ f_ List cmy CODftllieDI breakdown olD-CAT ~ i J (21) ~ ..H.1t_ %. MAY-CAT ~_u_% MAV-UI r .! --(5)-- (10) _j_£_:_ ~0:;!:.: ... (6) I 57 0-USER B List Dil prompt oatlslactlou (I) l J.[o Failure. - TANTALUS IMC. UBRARY MANAGEMENT CONSULTANTS 3:251 ORMOND ROA D • Cl.£VELAND, OHIO +4 118 • C21 61321 ntJ 2140 W ROA D • SUIT£ 218 • CLEVELAND. OHIO +4 118 © 1983 by Tanta lus Inc . Copy (25) EDler the product ol all lift ,, % r;m- FIGURE3 Availability Analysis Form CONCLUSIONS As a young institution with recently de- veloped collections, adequate shelf space, and smoothly functioning automated sys- tems, McHenry Library displayed a high document-availability rate of 61 percent. This compares well with the 50-60 percent MA V range that Kantor characterizes as "quite typical" of research libraries con- ducting similar studies and with the "lower values" that, he adds, "are often found at quite respectable institutions. " 6 Materials sought unsuccessfully by McHenry users were unavailable for a re- markably small number of reasons: 6 out of 22 traced factors accounted for 90.3 per- cent of unavailable items. Such potential institutional bottlenecks as processing shelves in the acquisitions and cataloging departments, accessioned backlog, the bindery, and the circulation department's numerous sorting and holding shelves proved to be virtually insignificant barri- ers to the satisfaction of user needs for known items. Two factors-the library's failure to acquire an item, and prior charge-out of an item to another library user-combined to account for more than 60 percent of unavailable materials. Because UCSC is a research-oriented in- stitution with less than comprehensive collections, because UCSC faculty and graduate students have traditionally de- pended heavily on borrowing from neigh- boring UC Berkeley's vast collections, and 250 61% Prompt Satisfaction Research Notes 507 Requested DACQ (40.2) DLIB (25.4) FIGURE 4 Branching Diagram because primary access to McHenry's col- lections is provided by an online union catalog of UC-wide holdings, we were not surprised by the high incidence of users seeking material not owned by the library. We were, however, surprised by the large number of users wanting materials al- ready in circulation. While we were pleased to uncover this problem, we have not found an obvious solution. Kantor identifies duplication of materials and ad- justments to the library's loan period as the most common approaches to lowering high scores in the DCIRC category. 7 Some McHenry materials unavailable because of circulation status had already been ac- quired in multiple copy. Our automated circulation system cannot currently pro- duce listings of specific books based on the number of times they have circulated-a listing that could be very useful in deter- ' · mining what needs to be duplicated. Ad- ditionally, the existing two-week under- graduate loan period is too short to permit further reduction, and the substantially more generous faculty loan period is de- termined by an essentially administrative process unlikely to be j.nfiuenced by an ob- jective performance study alone. The two highest performance scores were user related. Remarkably few users failed to locate the desired items in the cat- alog: only 4.4 percent of not-found items were unavailable for this reason. We were surprised, however, that a small but sig- nificant number of items (8.8 percent) were unavailable because users copied the call numbers incorrectly, often reversing key numerical elements. Finally, user fail- ure to locate the desired item when it was on the shelf accounted for a low 5.3 per- cent of unavailable items. We have con- veyed these discoveries about user failure to McHenry's library instruction coordi- nator for integration into our active user- education program. We were pleased with the compara- tively low percentage of materials that re- mained unaccounted for after analysis of results: 11.7 percent of the not-found items, representing only 4.5 percent of all items sought by users. This figure is par- ticularly impressive, given McHenry Li- brary's lack of an electronic security sy_s- 508 College & Research Libraries tern and the fact that no comprehensive inventory has been conducted since the late sixties. We attribute our low unaccounted-for score partly to the greater control provided by our auto- mated circulation system. The system tracks materials charged out to fifty differ- ent institutional categories (e.g., new book shelf, reserves, bindery, cataloging revision, lost and billed). The status of books charged out to these special catego- ries is readily found by users who ask at the circulation desk. We also attribute the low figure to the more detailed tracing process developed for this project. Be- cause we traced many subcategories of the main availability categories, we were able to identify factors that were not contribut- ing to the unaccounted-for category. The few books unaccounted for in the project must have been either in use in the library; on their way to a controlled location; mis- shelved; or, missing but not known by the library to be so. The capability of narrow- ing the range of unaccounted-for materi- als leads us to recommend use of the ex- panded tracing process wherever feasible-despite the fact that it is time- consuming, and that results recorded in September 1986 many of the additional subcategories may not prove statistically significant. Our experience investigating book availability at Santa Cruz underscores the need for and usefulness of an objective study. Widely held assumptions-most notably, that periodic shelving backlogs constitute a major problem for academic li- brary users-were unsubstantiated, at least insofar as users' needs for known books are concerned. The research results suggest instead that our collection build- ing policies and procedures and our li- brary instruction program should be re- viewed with the goal of improving availability. These results are consistent with Kantor's comment that, because of the absence of accepted standards, objec- tive performance studies simply "point out bottlenecks ... monitor changes from year to year ... [and] pinpoint areas where an effort at imr,rovement will give the greatest payoff." We look forward to reading the results of comparable avail- ability studies conducted at similar institu- tions that will help us further interpret our results and use them to improve service to the academic community. REFERENCES AND NOTES 1. Statistics and conditions described are those of Fall-Winter 1984. 2. Paul B. Kantor, Objective Performance Measures for Academic and Research Libraries (Washington, 0. C. : Assn. of Research Libraries, 1984). 3. Ibid., p.54. 4. The user survey form was adapted from one developed for an availability study conducted at the University of California, San Diego. The authors wish to thank Jill Fatzer, now at Ohio State Uni- versity, and Judith Herschman of UCSD for their advice in conducting such studies. 5. The sampling techniques used and the relationship between sample size and reliability of results- the subtle concept of confidence levels-are fully explained by Kantor, Objective Performance Measures, p .17-23. 6. Kantor, Objective Performance Measures, p.49. 7. Ibid., p.50. 8. Ibid., p.49. Introducing a unique collection of never-before-published historical records. UNPUBLISHED US. SENATE COM I INGS Despite their value, transcripts of many U.S. congressional hearings have never been printed and made available to the public. Until now. CIS has uncovered thousands of long-buried Senate hearings transcripts, and has prepared a major microfiche collection of these materials. A detailed, easy-to-use index makes the collection fully accessible . Spanning the mid-1800s through 1964, CIS Un- published U.S. Senate Committee Hearings bridges important gaps in U.S. history. Voices from the past Listen in as the story of America's past is told by: - • Joseph McCarthy • Nelson Rockefeller • Margaret Chase Smith • John F. Kennedy • Douglas MacArthur • John Foster Dulles • Jimmy Hoffa • Edward R. Murrow • Howard Hughes ... and thousands more. Fresh historical insights With this exciting new documents collection, you'll get: • a new and unique "sense of the times" for more than 100 years of history • new additions to the legislative histories of dozens of key laws • a behind-the-scenes look at important con- gressional investigations • revealing glimpses of some of the nation's most intriguing public officials. Write or call for free information For more information on this important new collection call CIS toll-free to request a free brochure: 800-638-8380 Or fill in and mail the coupon below. 0 Please rush me more information on the CIS Index to Unpublished U.S. Senate Com- mittee Hearings and companion microfiche collection. Name Dept. Organization Address City / State / Zip MAIL TO: 111!1 Congressional Information Service, Inc. 4520 East-West Hwy., Suite 800-C Bethesda, MD 20814-3389 Tel. 301-654-1550 On microfilm ... from Research Publications The Golden Age Spanish Theology, History and Literature from 1472 to 1700 Spanish Rare Books of the Golden Age offers the opportunity to research the wide variety of literature written in Spain or composed in Spanish from 14 7 2 to 1 700. Spanish Rare Books of the Golden Age includes drama, works on jewish and Christian theology, medicine, and classi.cal texts written in Latin and vernacular translations. Spanish Rare Books of the Golden Age represents the writings of Boscan, Luis de Leon, Borja, Cervantes, Aleman, and others. Based on a collection at the Library of the University of Illinois, Spanish Rare Books of the Golden Age encompasses the development of prose and poetic styles from a period rich in religious, literary, and political works. Divided into units of approximately 50 reels each, the standing order unit price is only $3,000 per unit (over 1 5% off the individual unit price). Prices for U.S. and Canada only. Shipping and handling charges will be added. To order, or for more information on Spanish Rare Books of the Golden Age call or write: Research Publications 12 Lunar Drive/Drawer AB VVoodbridge,Ct06515 (103) 397-1600 Toll-free: 1-800-RfACH-RP TWX: 71 0-465-6345 FAX: 103-397-3893 Outside North and South America P.O. Box45 Reading, RG 1 8HF England nL: 0734-583147 nLEX: 848336NADL G research publications®