College and Research Libraries 66 College & Research Libraries January 1984 User Response to and Knowledge about an Online Catalog David Steinberg and Paul Metz BACKGROUND As online catalogs proliferate in li- braries, complementing card catalogs with the ultimate prospect of substituting for them in many cases, the library profes- sion has recognized the importance of de- signing systems that are easily used and that reflect patron needs and expecta- tions. The expression "user friendly," an overnight cliche, refers to one aspect of this concern. However, the cognitive problems of software design-the use of mnemonic commands, graceful transi- tions between screens, and plentiful "help" or tutorial features-represent only one aspect of the system design chal- lenge that will provide users with what · they want and expect. It is equally impor- tant to provide access by the approaches users want to employ. The task of determining this is not easy, for users and librarians do not always speak the same language. For example, when users tell us that they want subject access, it is easy for librarians to narrow the problem to the 6xx fields of the MARC record and to make plans for Boolean or keyword approaches to this information. But as Lee Jones has· argued, what patrons seem to want is an approach to books by their "real" subjects, defined much more broadly. 1 To a patron, the term hydraulic represents a subject and should provide an access path regardless of whether it oc- curs in subject, corporate author, or title fields. Clearly, the art of listening to our patrons requires great sensitivity and awareness. Research on user expectations and de- mands is now conducted regularly and of- fers the promise that systems designers will be able to work from a realistic ap- praisal of the user and the user's preferred strategies. Studies by Moore and by Nor- den and Lawrence have contributed sig- nificantly to what we know by document- ing the popularity of online catalogs, the desirability of subject searching, and the perception by users that insufficient re- call, rather than the lack of precision in searches, is still a problem. 2 Most recently, OCLC, with the support of the Council on Library Resources, conducted a number of concurrent studies of online public access systems (OPACs) in a variety of library settings. These studies add even more to what we know about the users of auto- mated catalogs, though there is much still to learn. .r METHODOLOGY The present study reports the percep- tions of users of another OPAC, the Vir- ginia Tech Library System (VTLS), in re- sponse to a survey administered in spring of 1983. Used at Virginia Tech since 1980, and now at some thirty other libraries as well, VTLS is accessible by all author, title, subject, and added entry fields, as well as by call number, item number, and a full range of bibliographic control numbers, including OCLC number. Because it is an integrated system that encompasses circu- lation control, information on item avail- ability is provided. Records have been in- put for all cataloging since 1980 and for all materials that have circulated since 1975, when a short-record precursor of VTLS was introduced. For serials, holdings are described. For records entered between 1975 and 1980, only limited information is available, and there is no subject access. David Steinberg is a doctoral candidate in education and Paul Metz is user services librarian, both at the Vir- ginia Polytechnic Institute and State University, Blacksburg. Commands are mnemonic ("AI" plus search argument for author searching, "T/" for title, etc.) as are screen flows, in- cluding "NS" and "PS" for next and pre- vious screens. The design principles and search features of VTLS have been more fully described by Uluakar, Pierce, and Chachra. 3 The current study evolved from a pilot study conducted at Virginia Tech in 1982 by John Ulmschneider, then an intern at the Library of Medicine . Ulmschneider had drawn a number of his questionnaire items from the CLR/OCLC instrument. His research instrument was modified by the deletion of a small number of items and the addition of a short battery of ques- tions designed to assess user knowledge of the system. This portion of the instru- ment was intended to help discover the problems of user education the library needed to address and to identify those forms of access that users expect to find in an online catalog and that a truly user- . oriented system would attempt to pro- vide. The sample group for the study com- prised eighty-five system users . The sam- ple was chosen by the senior author, who was stationed near two of the high-traffic computer catalog terminals situated be- tween the card catalog and general refer- ence desk. Questionnaires were also com- pleted by patrons using other terminals in the main library, and in the branch library for the College of Veterinary Medicine. In- dividuals were asked to complete a ques- tionnaire after finishing a search. All but four subjects were willing to respond, yielding a response rate of 95 .5 percent . All eight of the university's colleges were represented in the sample. The sample population consisted of forty-seven un- dergraduate students, twenty-three grad- uate students, seven faculty members, four members of the university staff, and four nonuniversity patrons . · FINDINGS In many ways the findings replicate results obtained elsewhere and increase the likelihood that various OP AC studies may be generalized for other settings. As in other studies, most patrons (58.8 per- Research Notes 67 cent) learned to use the system from printed instructions; the next most fre- quent sources of information were library staff (14.1 percent) and the help feature (10.6 percent) . As in other studies, most patrons were extremely satisfied with the system, as 78.8 percent considered their searches to have been very or somewhat successful and 81.2 percent considered the system easier to use than the card cata- log. This preference for the online catalog over the card catalog was most pro- nounced in searching for journal titles or for books by personal authors . That 50.6 percent found the computer catalog easier for journal searches, while only 7.1 per- cent preferred the card .catalog, may be due to an aversion to searching through the large number of cards frequently en- countered in card catalogs for the works of corporate authors or for periodicals with common titles. The findings also support previous re- search on the popularity of subj~ct search- ing. Patrons were asked what they knew DISORDERED MATERIALS: SCIENCE AND TECHNOLOGY Selected Papers by S.R.Ovshinsky Edited by: David Adler Professor of Electrical Engineering Massachusetts Institute ofTechnology Publi shed on th e occasio n of Sta n O vs hin sky's 60 th bir th day, th is se lection o f over 35 of his pape rs from 1968 to 1981 covers muc h of his work in amo rp h ou s ma teria ls rang ing from sw it c hin g and memory to e nergy conve rs ion app li ca tions su ch as amorp h ous a ll oy photovoltaics. T h is book is re levant not on ly to those who wis h to un ders tan d the bas ics of a morp hous materia ls bu fa lso to th ose interested in the deve lopment of new synt het ic materials . 1982. 296 pp .. Cloth. ISBN 0-91033 1-00-6 US$30 .00 1 ··;::'~"1JII Amo rphous Institute Press ,_,~~.,~ ... 1050 East Square Lake Road ~.;~,;.~ •• Bloomfield Hills . Mich igan 48013, &.t'"da U.S .A . 68 College & Research Libraries before they initiated their searches and what they had entered. In each case, sub- jects were ranked first, followed by title, author, and call number. The preference for title searching over author searching is consistent with Norden and Lawrence's findings, even though they report that us- ers of card catalogs prefer the author ap- proach. According to Norden and Law- rence, users tend to bring greater knowl- edge of titles than they do of authors to the catalog, but are discouraged by the diffi- culties of title searching in card catalogs. 4 The great popularity of subject searching has been attested to in all previous studies of online catalogs, even though subject searches were found to be infrequent in Lipetz' classic study of the use of card cat- alogs.5 In respect, it is likely that the Lipetz study may have been misinterpreted by many as an indication of lack of desire for subject access, rather than the difficulty of obtaining subject access. The popularity of the subject approach to materials in on- line catalogs probably reflects the ease of searching but also-and more ominously-unfounded optimism about subject searching, a point to which we will return. Respondents were also asked a battery of questions about what they were looking for in their most recent search. Again, the desire to obtain materials on a given sub- ject was clearly paramount, followed by the desire to check on the availability of materials. The results of this part of the questionnaire are displayed in table 1. It should be noted that most patrons an- swered yes to several questions, indica- ting multiple goals in searching, or a defi- nition of the term ''most recent search,'' which referred more to a search session than to a specific bibliographic inquiry. The Virginia Tech and State University li- braries furnish a periodicals notebook that provides call numbers by author and title in a simple layout; the availability of this tool may have caused less searching than might be found in another periodicals set- ting. . The short quiz about the system in- cluded at the beginning of the question- naire represents the new feature that this study introduces to the literature on on- line catalogs. In order to prevent guess- January 1984 work by respondents accustomed to out- witting multiple-choice tests, a "don't know" option was provided for each question. Questions asked about the sys- tem are reproduced in table 2, along with the distribution of responses. The correct response to each item is in italics. For the first item two responses were considered correct because while there are item rec- ords for about 50 percent of the library's volumes, volumes represented either by item records or by serials holdings records constitute closer to 75 percent of the collec- tion. · TABLE 1 PURPOSE OF MOST RECENT SEARCH Purpose Percent Reporting ''Find the call number of a spe- cific book" 49.4 ''Find the call number of a sr,e- cific journal or magazine ' 21.2 "Find what back issues of a specific [journal] or maga- zme the hbrary had" 12.9 "Find works by a specific au- thor" 36.5 "See if a book was checked out or on reserve" 57.7 ''Find a book on a specific sub- ject" 69.4 N = 85 for all items . The word journal was inadvertently omitted from the third item . In some respects, the responses to the quiz about VTLS features revealed a level of understanding that is probably no worse than would be found if patrons' perceptions of the card catalog were as- sessed. More than 80 percent of patrons show some understanding of how author and title searches are formulated. Nearly all know that the system can indicate whether books are available (not surpris- ingly, since this was the second-most- sought type of information), and 71.8 per- cent know that periodical holdings information may also be accessed. Serious public misconceptions are ap- parent from those items addressing sys- tem coverage and the utility and means of . subject searching. More than one-third (34.1 percent) of the respondents had no idea what percentage of the collection was represented in VTLS, and another 9.4 per- Research Notes 69 TABLE 2 PATRON RESPONSES TO VTLS QUIZ 1. The percentage of the Library's titles available through the computer catalog is: a. 100% (3.5% of respondents) b. 75% (43.5%) c. 50% (13.0%) d. 25% (5.9%) e . DK (34.1%) 2. If a book has ever been checked out, it is mote likely to be in the computer catalog than if it has never been checked out. a. True b. False c. DK (57.6%) (11.8%) (30.6%) 3. The computer catalog was: a. Devefoped at Virginia Tech and is complete (1.2%) b. Developed at Virginia Tech and is still being improved (43.5%) c. Purchased, and is complete (1.2%) d. Purchased, and is being improved (14.1 %) e . DK · (40.0%) 4. The computer catalog can tell me what volumes or years of a periodical the library has . a. True (71.8%) b. False (9.4%) c. DK (18.8%) 5. The computer catalog will tell me whether or not a book is in use. a . True (94.1%) b. False (1.2%) c. DK (4.7%) 6. In order to search the system for a given title, I always have to know the complete title . a . True (10.6%) b. False (85 .9%) c. DK . (3.5%) 7. The more of an author's name I enter into the computer catalog, the more I narrow my search to get just the author I want. a. True (82.3%) b . False (10.6%) c. DK (5.9%) d. Missing · (1.2%) 8. All of the books in the computer catalog can be found through a subject approach. a. True (32.9%) b. False (41.2%) c. DK (25.9%) 9. It is possible to search the computer catalog by using subject terms other than the Library of Con- gress subject headings used in the card catalog. a . True · (29.4%) b. False (28.2%) c. DK (42.4%) N = 85 for all ite ms . cent guessed wrong. Despite an active program of bibliographic instruction for freshmen and select other groups, dis- claimers about subject searching in all printed instructions, and the cautions of reference librarians assisting users, nearly .one-third of users believe that all materials in the system are accessible by subject, · while only 41.2 percent know positively that this is not true. Worse yet, only 28.2 percent are aware that subject searching is possible only by the use of the same terms represented in the card catalog. That such misconceptions translate into practice is shown by the analysis: forty-five patrons reported that they had searched by sub- ject, whereas only twenty-two had used terms from the Library of Congress Sub- ject Headings. The public's perceptions about subject searching seem to represent the epitome of wishful thinking. Eager to find informa- tion by subject, patrons rashly assume that if this search approach is offered, ma- terials must be available by following it. Nor den and Lawrence also found that subject searching at Ohio State became very popular as soon as it was introduced, 70 College & Research Libraries despite soft-pedaling by librarians aware of its limitations. 6 It is not difficult to extract policy implica- tions from these data. First, they demon- strate the magnitude of the user- education task by showing that even where very active programs exist, serious misconceptions can exist. The damage may not be so severe as long as dual sys- tems are maintained (even the current sample of VTLS users reported that they used the card catalog somewhat more January 1984 than the automated system). However, damage can become greater if card cata- logs are closed or frozen. A second policy implication is that if patrons are so intent on subject searching that they become prey to optimistic delusions about its util- ity, wise librarians and system designers will be equally eager to provide both data and access points to make these percep- tions realistic. Perhaps it is better to quash a happy delusion than to encourage it, but better still to make it true. REFERENCES 1. Keith W. Russell, ed., Subject Access: Report of a Meeting Sponsored by the Council on Library Resources (Washington, D.C.: Council on Library Resources, 1982), p.7. 2. Carole Weiss Moore, "User Reactions to Online Catalogs: An Exploratory Study," College & Re- search Libraries 42: 295-302 (July 1981); and David J. Norden and Gail Herndon Lawrence, "Public Terminal Use in an Online Catalog: Some Preliminary Results," College & Research Libraries 42: 308-16 (July 1981). 3. Tamer Uluakar, Anton R. Pierce, and Vinod Chachra, "Design Principles for -a Comprehensive Library System," Journal of Library Automation 14, no.2: 78-89 (June 1981). 4. Norden and Lawrence, "Public Terminal Use," p .315 . 5. Ben-Ami Lipetz, "Catalog Use in a Large Research Library," Library Quarterly 42: 129-39 (Jan . 1972). 6. Norden and Lawrence, "Public Terminal Use," p.314. - ~ ' t J : I Acquisition Librarians ..... t I I I : l I II I· f Now is the time to try the best source for: e any book in print e ACCURATE INVOICING e meaningful reports ACCURACY IN BILLING & SHIP- PING..... to the specifications you desire. Invoices are arranged alpha - betically by author or title, or numer - ically by purchase order. Separate billing available upon request . • I, I ' . • rush order service the e competitive discounts BOO;Ir' ' e plus many other services ~ I ! CAL\~~~~~~~~~ODAY !.!!!!:~~wl» I ANY BOOK IN PRINT SINCE 1M2 1 1 • In Canada & Michigan ~~N~sE~~L~H~~~CGH~G~~R;~r50 l.i(" CALL COLLECT(517)849-21~ SANIB9-Jssg ~