College and Research Libraries SUE PEASE AND MARY NOEL GOUKE Patterns of Use in an Online Catalog and a Card Catalog Will using an online catalog institute a change in patrons' catalog use pat- terns? In a study at the Ohio State University, success of patrons in finding ti- tles in two department library card catalogs was compared with success in finding the same titles in the online catalog of three and one-half million rec- ords. During the study, information on patron use patterns was also obtained. In a later study, the success rate in searching titles of their own choosing in the online and card catalogs was measured. Results show that 90 percent of online users preferred the online to the card catalog; 64 percent of those who tried the online catalog switched to its use; patrons used the online catalog more often than the card catalog; and patrons' use of the card catalog decreased, but did not end. INTRODUCTION Academic libraries establishing an online catalog in the near future (particularly those in larger institutions) must determine how an online system will affect patterns of catalog use. If dual systems are in effect, how will pre- vious catalog use patternS change? Will the online catalog be used to the same degree as the card catalog or will card catalog use de- crease as online use increases? If patrons switch back and forth between systems, will this interfere with their ability to use either system? Will the online system become a sub- stitute for the card catalog or will patrons still depend upon the card catalog to search for certain items? Extensive research has been done on cata- log use1 and concern is being focused on users of both card and online catalogs. Tag- liacozzo and Kochen, authors of a major cat- alog use study, summarize these concerns: Though many would agree that the manipulative acts required to use automated information re- Sue Pease is reference librarian, Commerce Li- brary , and Librarian for labor education and re- search service, and Mary Noel Gouke is reference librarian, Education/Psychology Library, Ohio State University, Columbus. trieval systems may be different from those re- quired to use traditional information systems, al- most everybody seems to accept the assumption that the same basic cognitive processes are involved in both cases. Is this assumption justified?2 The authors did an experimental study of patron success and failure on known-item searches in the online and card catalogs at the Ohio State University libraries (in press, journal of Academic Librarianship, July 1982). Data analyzed in this article are taken from a section on the questionnaire relating to comparative use of the online and card catalog by patrons. Data relate to their class, sex, and major, length of experience with the online system, and success and failure rates in both the online and card catalogs. In addition to the experimental study de- scribed above, further research on actual searches brought to the library by patrons themselves has provided additional use data that will also be included. The Ohio State University libraries pro- vided an excellent setting for this research, since dual catalog systems have been in effect for the entire collection of three and one-half million volumes for ten years. These systems include a Union card catalog and an online catalog. CoNTENTS OF THE ONLINE CATALOG The online catalog contains: (1) a title and I 279 280 I College & Research Libraries • july 1982 (2) an author record for every item repre- sented in the Union card catalog. (Analytics are not included.) It also contains records for the following items not in the Union card cat- alog: 1. I terns on order and in processing 2. Books owned by the State Library of Ohio 3. Special microform collections a. ERIC documents b. Human Relations Area File c. Wing's Short Title Catalog (1641-1700) POINTS OF AccESs TO REcoRDs IN THE ONLINE CATALOG Authors- Author records for items cata- loged by title main entry, or j9int authors be- yond the first author, were not included ini- tially in the online catalog. (They have been included since 1977.) Titles- A title entry is included for every record in the online catalog (including some items that do not have a title entry in the Union card catalog). Added Entries- Added entries were not included initially in the online catalog. (They have been included since 1977.) Cross-References- Cross-references were not included initially in the online catalog. (They have been included since 1977.) Subject- Subject access using LC subject headings is available for all records cataloged since 1977. Shelf List Position- A Union shelf list search may be done by call number revealing items adjacent to that call number in the Union shelf list. FoRM OF REcoRDs Each item in the online catalog has a one- line entry, giving author, title, and date of publication (or starting date for serials). The one-line entry gives access to a fuller circula- tion record, which gives the above informa- tion, plus the following: (1) call number, (2) library location(s), (3) availability for use and loan period, and (4) bound serial hold- ings, including indication of microform cop- ies. In addition to this shorter circulation record, a full bibliographic record may also be re- quested for all items cataloged since 1977. The online catalog may be accessed by us- ing the following comma.nds: title (TLS), au- thor (AUT), author/title (ATS), subject (SIS), call number search (DSC), and shelf position search (SPS). Title searches, with which the authors' experimental study was primarily concerned, are done by using the TLS command, followed by a slash and an algorithm of four letters of the first word, fol- lowed by five letters of the second word. (Very common words that are not used are listed on a stop list posted on each terminal.) Patrons have had access to the online catalog for five years, and thirty-eight public termi- nals are available in the Main Library, with twenty-nine of these accessible on the first floor. There has been no queuing problem. One or more terminals are available in all de- partment libraries. A traditional card catalog system also is in operation with the Union card catalog in the Main Library and individual card catalogs in each department library. Lack of Boolean logic in this system makes it atypical of new systems such as RLIN, · WLN, etc., and findings in regard to pa- trons' ability with this system may not have relevance to systems in the future. The in- creased searching power of Boolean logic would, in cases of incorrect or incomplete in- formation, lead to increased retrieval on the part of librarians and other skilled searchers. However, as far as the authors are aware, there are no studies showing that the average academic library patron would be able to take full advantage of this increased search- ing capability. In the authors' Departmental Libraries Experimental Study, there was a range of twenty-eight percentage points be- tween the success of the patron sample and the success of a skilled librarian in using the online catalog. Patrons were not able to uti- lize fully the searching power of the online catalog as compared with the skilled searcher. This indicates that the average user of the online catalog may have greater diffi- culty with a more complex system offering more options and utilizing Boolean logic. METHOD The Department Libraries Experimental Study was conducted in two of the largest de- partmental libraries in the Ohio State Uni- versity Library System: th~ Commerce Li- brary and Education Library. Success of patrons in finding titles in the two depart- ment library card catalogs was compared with success in finding the same titles in the online catalog of three and one-half million records. Using random selection, fifty-two titles were drawn from each library card catalog and each title was verified in the online cata- log to assure its existence in the system. The 104 patrons (52 in each library) in the study were chosen at predetermined random times from those using the public terminals in each library. This was done in order to assure that the patron had some familiarity with the on- line catalog. Patrons were selected by random methods, and length of experience with the online cat- alog varied from several days to five years. The sample contained both experienced and inexperienced users. Each patron was asked if he or she would participate in a brief exper- iment on improving use of the online catalog. Mter agreeing to participate, the patron then checked off information on a brief question- naire. If the patron was to begin at the card cata- log, she/he was taken from the terminal to the card catalog. If to begin at the terminal, she/he remained there and was given a folded title sheet and requested to locate the title in the online system. A five-minute time limit was set for the search and for writing the call number on the sheet. For the next step, the title sheet was turned over and a search was made for the same title in the card catalog using the same time limit. The re- verse of this procedure was made with the next patron and this rotation continued for the duration of the project. The time taken searching each title, the patron's search patterns and algorithms used from the terminal screen, and his or her search patterns in the card catalog were re- corded. MAIN LIBRARY " AcTUAL" SEARCHES STUDY Mter the completion of the experimental study in the department libraries described above, a study of actual card catalog use was done in the Main Library. Two forms on which patrons could record their own searches at the online and card catalogs were prepared and pretested. Extensive pretesting of the forms was done until the authors were Patterns of Use I 281 satisfied that accurate records of known-item and subject searches were being obtained. Forms were passed out to patrons in the Main Library over a four-week period. Partici- pants were chosen from those using the four IBM 3278-2 terminals located in the lobby with the Union card catalog. Patrons depos- ited forms in a marked box before leaving the area. A determined effort was made to re- trieve each form handed out. The return rate of forms was around 95 percent. USE DATA FROM ABOVE Two STUDIES In the Department Libraries Experimen- tal Study, the brief questionnaire filled out by each patron before beginning to search re- quested: 1. the patron's class or status 2. major area of study 3. how long he/ she had used the online catalog 4. frequency of use of: a. the online catalog per month b. the card catalog per month In the Main Library Actual Searches Study, patrons using the online and card cat- alogs gave information on class, major area, their use of both catalogs, and the reason for this use. RESULTS Table 1 shows the classes of patrons in the authors' two studies, the Department Li- braries Experimental Study and the Main Li- brary Actual Searches Study. A comparison is made with the Specht study done at the University of Illinois using the same online system. 3 Table 2 shows the length of patrons' online experience in the Department Li- braries Experimental Study and the Main Li- brary Actual Searches Study. The study indicates that the proportion of graduate and undergraduate users at the on- line catalog varies in different libraries. Comparison was made in the Main Library where samples were taken at both the online and card catalog. At the Main Library, pa- trons using the card catalog were of a lower class level (freshmen, sophomore) than those using the online catalog. PREFERENCE FOR THE ONLINE CATALOG Four studies of online users have now been published. 4 Three of these studies asked pa- 282 I College & Research Libraries· july 1982 TABLE I CLASS OF PATRONS IN THREE ONLINE CATALOG STUDIES DepartmeJl t Libraries Main Library Found at Experimental Study Actual Study Online Catalog Found At Undergrds Found At Undergrds Found At Undergrds Specht Study Online Catalog And Online Catalog And Card And Illinois Terminals Grads Terminals Grads Catalog Grads Main Library Number Percent (%) Number Percent ( %) Number Percent (%) (%) Freshman 5 5 4 Ill Sophomores 4 4 6 16 ) Juniors 16 16 49 7 18l Seniors 24 24 8 21 Grads 51 51 51 13 34 Totals 100 100 38 100 TABLE2 LENGTH OF pATRONS' EXPERIENCE WITH THE ONLINE CATALOG 1 Month or Less More than 1 Month and Less than 1 Year 1 Year or More Totals Department Libraries Experimental Study Number Percent 21 21 32 31 49 48 102 100 Main Library Actual Study Number Percent 14 33 12 29 16 38 42 100 trons for their preference, and results indi- cated a preference for the online catalog by users (see table 3). These studies indicate that users will ac- cept a catalog in the online format. For more data on reasons for this preference, see Moore. 5 The question as to what extent users will accept the online catalog as a substitute for the card catalog is not answered. Perhaps user behavior in a system such as that of the Ohio State University libraries, where dual systems have been available for some time, will indicate the extent to which this online catalog has become a substitute for the card catalog. PATRONS FouND AT THE ONLINE CATALOG This online catalog has certain advantages over the card catalog, including greater cur- rency (four to six weeks for main entries), availability of circulation information, com- plete bound serials holdings , additional in- dividual records for each item in certain microform series that are not analyzed indi- 5 8 15 24 66 7 11 75 54 20 32 34 16 25 25 46 63 100 vidually in the card catalog, and greater ease of use. Based on these advantages and on pa- trons' preference, indications are that almost all patrons who try the online catalog will change over to using it more often than the card catalog. If we study patrons found at the online catalog and consider their use pat- terns in relation to their expressed prefer- ence, this appears to be so (see table 4). The use patterns reported by those found at the online catalog in the authors' earlier Department Libraries Study parallel the use patterns of those found at the online catalog in the Main Library Actual Searches Study. In evaluating these results, does a sample of online users give a fair representation of all patrons who have tried the online catalog, did not like it, and returned to card catalog use? This theory is examined in the section on TABLE3 PREFERENCE FOR ONLINE CATALOG Dow lin Moore 1. 85.4 % 2. Preferred Choice (w/ subject in all four access- 94 % ) systems surveyed TABLE4 Gouke And Pease Experimental Study 3 . 90 % PREFERENCE AND U s E PATTERNS oF UsERS FouND AT THE ONLINE CATALOG Department Libraries Main Library Prefer Online ( % ) 90 Use Online Catalog More Often ( % ) 86 91 "Patrons Found at the Card Catalog" in this article. The reasons why patrons preferred the on- line catalog to the card catalog were not studied in depth by the authors because the focus of the study was on collecting data on patron success and use patterns, and because this information is being researched and is available in other papers, particularly that of Carol WeissMoore. 5 UsE PATTERNS oF ONLINE CATALOG UsERS: DEPARTMENT LIBRARIES EXPERIMENTAL STUDY 1. Online catalog patrons reported that their use of the card catalog decreased but did not stop altogether. Comparing both libraries, card catalog USES 11 10 9.5 0 o 0 Patterns of Use I 283 use averaged 3.4 uses per month at the time patrons began to use the online catalog. The sample of patrons who had used the online catalog more than one year had a card cata- log rate of only 1.9 uses per month (see figure 1). The degree of the drop in card catalog use may depend upon whether patrons were heavy or light users of the card catalog. If card catalog use was high, as in the Com- merce Library (5.6 uses per month), it dropped sharply and then leveled out. If it was low (1.2 uses per month), as in the Edu- cation Library, it continued low, with a very slight rise (see figure 2). 2. A majority of patrons found at the on- line catalog changed over and began to use the on-line catalog more frequently per 10.8 o• oo oo 9 •oo 0 8.9 8 7 6 5 4 3 2 0 0 • (?07.~ 00. 0 0 0 0 5.100 .0 3.4 • • • • • 1.7 < 1 week ONL 2.0 ••• • • • • • months cc 1.9 • • • ·• years (Length of Patron's Experience) BOTH • • • • 0 0 0 Fig.l Department Libraries Experimental Study Uses Per Month of the Online Catalog and Card Catalog Related to Length of Patron's Experience with the Online Catalog: Mean for Both Libraries 284 I College & Research Libraries • july 1982 USES 11 10 9 9.7 0 8.6~ 0 8.0 8 7 5.6 6 • • 5 • 4_ 3 2.4 • 2.6 2.2 ••• • • • • •• 2 0 ••••• 0 • 0 • 1_ 0 1.3 1.5 1.0 0 < 1 week months years (Length of Patron's Experience) __ ONL • COMMERCE ••• cc 0 EDUCATION Fig. 2 Department Library Experimental Study Uses Per Month of the Online Catalog and Card Catalog Related to Length of Patron's Experience with the Online Catalog: Commerce Library and Education Library month than the card catalog. Figure 1 indicates that sometime during or after the first month of use, online catalog use rose dramatically from 1. 7 uses per month to 7.5 uses per month, or a 341 per- cent rise. Mter the sharp rise in the first year, frequency of online use continued to rise, but more gradually to a mean high point of 8.9 online uses per month after patrons had used the online catalog for more than a year. This early changeover is supported by an- swers to the check question, "Which catalog do you use more often?" After less than a month of use of the online catalog, 55 percent reported using it more often than the card catalog. After some months of use, 84 per- cent reported using it more often than the card catalog. 3. Patrons reported using the online cata- log more often than the card catalog. This is supported by the statistics. Using one class level as an example: gradu- ates beginning to use the online catalog made 4.9 searches per month in the card catalog. Graduates having used the online catalog for more than a year made 10.8 online uses and 2. 7 card catalog uses per month for a total of 13.5 uses per month. (See tables 1 and 2 for a sample description in the Department Li- braries Experimental Study.) Indications are that patrons are accessing the library's holdings records more fre- quently than they did in the past when pa- tron use was limited only to the card catalog. If no card catalog existed, these patrons would probably be accessing the library's on- line holdings more frequently than they cur- rently use the card catalog (see figure 1). This is supported by results from the OSU Poll, winter quarter, 1981 (Library Section), where 363 students at all levels reported more uses per month of the online catalog than of the card catalog. 6 (See table 5.) Patterns of use for patrons found at the on- line catalog in the Main Library Actual Searches Study is shown in table 6. PATRONS FouND AT THE CARD CATALOG: REsmuAL CARD CATALOG UsE Does a group of "drop-out" users who tried the online catalog and then returned to card catalog use exist? If so, perhaps these patrons will not use the online catalog frequently enough to be represented in the sample of on- line users. TABLES osu POLL RESULTS FOR STUDENT SAMPLE MEAN NuMBER oF UsERS PER MoNTH PER PERSON IN ONLINE AND CARD CATALOGS BY CLASS Online Card Catalog Catalog Difference Grads 8.8 6.7 2.1 Seniors 4.3 3.9 .4) ) Juniors 3.9 3.3 .6) ) .85 Sophomores 4.0 2.9 1.1) ) Freshmen 3.6 2.3 1.3) Total 24 .6 19.1 4.9 3.8 Patterns of Use I 285 No study, as yet found, has been made that includes both users found at the terminals and the card catalog. In studies by Specht, 7 Moore, 8 and in the Department Libraries Experimental Study, the only patrons sur- veyed were those found already using the on- line catalog. In the Dow lin study, patrons entering were surveyed and, if they had not previously been introduced to the online cat- alog, were given a lesson in its use. 9 Their preference was based on a very brief ac- quaintance and was not extensive enough to predict their actual use pattern in the future. The question as to whether a group of "drop-out" online users exists and may be found at the card catalog was investigated in the authors' study of actual searches in the Main Library. Both online and card catalog users were surveyed, and use patterns were studied for both groups. A description of patrons found at the card catalog is given in table 7. Table 7 and figure 3 show that the majority of users at the card catalog had been acquainted with the online catalog. Considering the advantages of the online catalog, why were so many online users found at the card catalog? The majority (30 percent) were there because of a general lack of confidence in the online catalog derived from disappointing past experience. How- ever, they did continue to use the online cata- log when they did not find the desired item in the card catalog. Typical comments of this group were: "More familiar with card catalog." "I have greater confidence in the card cat- alog." "I always start with the card catalog." "I couldn't get it to work." TABLE6 Use online only MAIN LIBRARY ACTUAL SEARCHES STUDY PATRONs FouND AT THE ONLINE CATALOG PREDOMINANT PATTERNS OF UsE (No cross-checking card catalog) Use online more often (Will cross-check card catalog if don't find) Usc card catalog more often. In this Rarticular case was at online because .ailed at card catalog se selectively about the same amount Totals Number Percent 6 18 22 67 2 6 3 9 33 100 286 I College & Research Libraries· july 1982 TABLE7 MAIN LIBRARY ACTUAL SEARCHES STUDY PATRONS FouND AT CARD CATALOG PREDoMINANT PATTERNs OF UsE Number Percent Have never used online. Don't know how 35 44 Have used online, but still prefer and use card catalog more often. Will cross-check in online if not found 24 30 (drop-out users) Use online most of time, but came to card catalog because failed to find what was wanted in online 13 16 Use both selectively about same amount. Thought this particular search would be more successful and/or faster at card catalog 6 7.5 At card catalog because used online in past and got poor results. Didn't like and stopped using 2 2.5 (drop-out users) Totals 57% tried On-Line Catalog Fig.3 Main Library Actual Searches Study Percent of Card Catalog Users Surveyed Who Have Tried Online Catalog " Instructions were unclear. " Another group (2.5 percent) was there due to an attitude of strong dislike for the online catalog on the basis of unsatisfactory past ex- periences. " The card catalog for me is very, very use- ful and when I have asked anyone (library staff) to search in the machine for an item, they have never come up with any informa- tion useful to me except to tell me if the book was out. Even periodicals seem difficult to find computer-style. " "The instruction I have received for LCS, both written and oral, has been lousy. LCS has not worked well for me. " "Hate computers. " Other users were there for specific reasons 80 100 related to the particular search they were do- ing. Sixteen percent were there cross- checking because they had failed to find what they wanted in the online catalog. An- other 7.5 percent were there because they were selective users of both catalogs. They came first to the card catalog because they believed it would yield the most satisfactory results on their particular type of search. Typical comments were: ''I'm looking for 1937 material which is not on the computer under subject search." "I felt title given by professor was proba- bly incorrect." "If I look for something specific I go to the computer, otherwise, to card catalog. " "Didn't think one could search by subject on the computer." "Wasn't sure if it was a valid subject head- ing, but I see it is." Several of these comments indicate that when there was doubt about the accuracy of the information in hand, the card catalog was preferred. Further research should be done in this area. DRoP-OuT ONLINE UsERS As can be seen in table 7, when the users of the card catalog in the Main Library were surveyed, the above-mentioned theory was supported and a group of drop-out online us- ers was found that continued to prefer the card catalog and to use it more often. When this group was added to the online users found at the terminals, the online catalog re- mained the most frequently used , but the switchover to the online catalog was less pro- nounced than appeared from a sur~ey of on- line users only. Figure 4 shows the figures for online users found at th e terminals and the Percent who use On-Line more often 100 95 't>-\ 90 "(..,..::> 85 ''? ;:;-. ~~ ~ "(.., ...... ~ 80 ~~ -\ rt- 75 ' -<..: ~~ "(.., 70 ,~'t>~ 65 .<:- 0~~ ~~ ..... ~ 60 • Patterns of Use I 287 ·~ • o• 0 55 o<> '(s ~o. o<> ~c., e.:\ s ~ 50 o<>o. ~e. \~~ o<>o ~~.,\,··\.e.'(~:\o~' 45 . '\ ~ ~~ ~'\ 0. ~ (,~ 40 ~~ '(0. 35 \.~0 (,~ 30 25 • • 20 Main Library Actual Study On-Line Users Found at the 15 Card Ca ta 1 o g 10 5 0 < weeks months years Length of Experience with the On-Line Catalog Fig. 4 Main Library Actual Searches Study Percent of Online Users Who Use Online More Often entire group of online users, including those found at the card catalog. The center dotted line represents all online users , regardless of where found. The graph in figure 5 repre- sents all patrons who have tried the online catalog. The basic reason that this group of drop- out users continued to use the card catalog more frequently was not dislike of the online catalog (25 percent of total drop-out users- table 7) , but lack of confidence in the online catalog and in their ability to use it. They felt the card catalog was faster, easier, and that they were more successful with it. This group of users resembled those who had never used the online catalog in class level, but they differed in one respect. While they preferred to use the card catalog, and expressed reservations about the online cata- log, a large proportion reported they in- tended to use it for a cross-check when the de- sired item was not located (see table 8). These users continued to use the card cata- log: (1) For retrospective material, particu- larly older subjects, as in the case of the pa- tron who was searching for historical source material on the Federal Housing Authority. (2) If too many similar titles came up online. (3) For cross-checking for items not found in the online catalog. These findings are supported by figures from the previously mentioned OSU Poll 288 I College & Research Libraries • july 1982 Fig.5 Main Library Actual Searches Study Preferential Use Pattern of all Patrons Who Have Tried Online Catalog showing that 73 percent of those who had used any catalog in the past quarter had used both online and card catalogs. 1o The cross-checking behavior of online us- ers must be examined more closely to deter- mine if the behavior is adaptive or if it is a se- curity mechanism taken because of a lack of confidence in using the online catalog. In ta- ble 9, 87 percent of those using the online less than a year still cross-checked in the card cat- alog. While the intention to cross-check is re- duced as the user becomes more familiar with the online catalog, 50 percent of those who have used the online catalog more than a year still cross-check. Are these cross-checks necessary, or are they redundant and a waste of time? Results showed that 67 percent of these searches that were unsuccessful in the online catalog were successful in the card catalog. Thirty-three percent were still unsuccessful. (Of those searches that were unsuccessful the second TABLES MAIN LIBRARY ACTUAL SEARCHES STUDY "IF You DoN'T FIND WHAT You ARE SEARCHING FOR, WILL YouUsETHE0NLINECATALOG?" Never used online Knows how to use online, but went to card catalog first Yes ( %) 44 82 No ( % ) 56 18 time, two-thirds were unsuccessful the sec- ond time due to the item not being in the col- lection, and one-third due to user error in the card catalog.) With a 67 percent success rate, cross-checking behavior may be adaptive. Regarding content, the online catalog con- tains many single items from certain micro- form series for which no records exist in the card catalog. The card catalog contains ana- lytics not entered in the online catalog and inaking a cross-check would be adaptive on this basis. With this sample, cross-checking was not justified on the grounds of differing database content. Of those who came to the card catalog after failure in the online cata- log, no items in patrons' actual searches were present in only one catalog. Seventy-six per- cent of items had records in both catalogs and the other 24 percent did not have records in either catalog. Results, shown in table 10, indicated that the reason patrons failed in the online and came to the card catalog were: (1) Need for points of access and full records for retrospec- tive materials (75 percent); (2) Too many matches on titles and corporate authors (17 TABLE9 MAIN LIBRARY ACTUAL SEARCHES STUDY PERCENT OF ONLINE UsERS WHo INTEND TO CROSS-CHECK IN CARD CATALOG IF DESIRED ITEM NoT FouND Used online more than a year Used online less than a year TABLE10 Yes ( %) 50 87 No ( %) 50 13 MAIN LIBRARY ACTUAL SEARCHES STUDY REASONS FOR F AlLURE IN THE ONLINE CATALOG (PATRONS wHO CAME TO THE CARD CATALOG AFTER FAILING IN THE ONLINE) Too many matches Title Corporate Author Joint Author (before 1977) Subjects (before 1977) Wanted full biblio- graphical record (before 1977) Patron error Totals Undetermined Number Percent ( %) 2 17 17 8 6 50 75 2 17 ) 1 8 8 12 100 2 percent); (3) Patron error in online catalog (8 percent). This indicates that those who have never used the online catalog and those who know how to use it but do not cross-check may not be able to gain access to the microform series and other collections that are added to the online catalog but not added to the card cata- log. PATRONS WHo HAVE NEVER TruED THE ONLINE CATALOG Considering those who have never become acquainted with the online catalog, we see that the library's publicity efforts have been relatively successful and that more than half of all users at the card catalog have tried the online catalog (figure 3). This is evidence of the strong effort the library has made to reach all potential users. Table 11 shows the attitudes of those who have never tried the online catalog. Table 12 shows how these attitudes were related to class and sex factors. Results of the Department Libraries Ex- perimental Study indicate that the online catalog and the card catalog are mutually re- inforcing systems. Continued use of the card catalog did not interfere with online catalog success and may even have slightly enhanced success. In fact, "heavy" users of the card cat- alog did slightly better overall than did non- users of the card catalog (see table 13). This can also be seen in figure 6, which shows suc- cess of online users in the card catalog in three different lengths of experience with the online catalog. Here again, card catalog suc- cess is not affected by a long period of pre- dominantly online use. SuMMARY AND CoNCLUSIONs The great majority of patrons found at the online catalog expressed a preference for it over the card catalog. These findings on pref- erence are supported by actual use patterns. More than half of those who had actually tried the online catalog "switched over" to use it more often than the card catalog. This discrepancy between preference and use may be due to the fact that, in most pref- erence studies, only online users have been surveyed. In the authors' Actual Searches Study, users at the card catalog were also sur- veyed and a group of "drop-out" users was Patterns of Use I 289 TABLE 11 MAIN LIBRARY AcruAL SEARCHES STUDY A'ITITUDES OF THOSE WHO HAVE NEVER UsED THE ONLINE CATALOG Number Percent Willingness to try Prefer card catalog, faster, easier Dislike computers; prefer card catalog, don't want to learn TABLE12 8 8 3 42 42 16 MAIN LIBRARY AcruAL SEARCHES STUDY A 'ITITUDES oF UsERS WHo HAVE NEVER TRIED THE ONLINE CATALOG CLASS AND SEX Dislike Prefer Card Computers, Willing Catalog; Avoid Don't Want To Learn Online To Learn (%) (%) (%) Male 75 45 45 Female 25 55 55 Underclass 14 22 Upperclass 57 56 50 Grads 29 22 50 TABLE13 DEPARTMENT LIBRARIES EXPERIMENTAL STUDY SuccESs IN THE ONLINE CATALOG RELATED TO CARD CATALOG UsE (nonuse) ("heavy" use) Use of Card Catalog 0 use per month 5 + uses per month Online Success (%) 64 69 found who had returned to predominantly card catalog use. This indicates that future studies of user behavior in regard to online catalogs should survey all online users in- cluding those found at card catalogs- if use- ful conclusions are to be drawn. · Patrons' use of the online catalog may be based more on their level of success in locat- ing what they needed, rather than on per- sonal preference. In this study, many of those who reported they were not successful with the online catalog tended to drop out and re- turn to predominantly card catalog use, in which they were more successful. The online catalog. did not serve as a com- plete replacement for the card catalog for most users. The great majority of those who had tried the online catalog continued to 290 I College & Research Libraries ·july 1982 PERCENT 100 90 80 50 8.0 7.6 •• • ••••••••• o. 7 . 7 • • ___.---::0 0 ----. 7.5 • 7.3 ··o-· 6.9 5.0 • ·--------5. 3 • • • 8.5 .• p • .. 8.1 < 1 week months years (Length of Experience with On-Line Catalog} -ONL ••• cc • COMMERCE 0 EDUCATION Fig.6 Success in Each Library in the Online and Card Catalog: Education and Commerce Libraries make some use of the card catalog. A large number of those who had switched over to the online catalog- and developed skill- still cross-checked in the card catalog all items they did not find in the online cata- log, due to the lack of approaches for retro- spective materials, particularly older sub- jects, in the online catalog. This indicates that planners of online catalogs need to pro- vide approaches and records equal to the card catalog for all retrospective, as well as current, materials. Another reason for continued card catalog use by online users was the use of certain al- gorithms producing too many matches. This problem may be solved in the future by use of several different algorithms. The patron who fails to reach the same level of success with the online system as in the card catalog is a major concern. The li- brary's educational programs, instructional sessions, and other efforts over the years to acquaint users with the online catalog and encourage its use have been successful, con- sidering the large university enrollment and the constantly changing nature of the univer- sity population. However, even with the most strenuous educational efforts, some li- brary users will not be reached and will never attain a sufficient level of skill with the online system to take full advantage of its use. A revealing fact emerged from the OSU Poll supporting this conclusion, in that 33 percent of online users surveyed had been reached by formal workshops, 22 percent by individual instruction from library staff, and 11 percent by friends and "other methods." However, 34 percent learned from printed instructional material at the terminals. 11 For this reason it is important, in addition to offering formal and informal instruction, to concentrate efforts on providing simpler terminals and instruction sheets both of which are designed to prevent patron errors at the point of use. Perhaps the first step in raising patron success levels on online cata- logs should be to cast aside the comforting as- sumption that patrons will be more success- ful with the online catalog than the card Patterns of Use I 291 catalog. Once we have faced this fundamen- tal problem, the task of studying patron er- rors and designing improved systems and point-of-use instructions can begin in ear- nest. REFERENCES I. John Aubrey, "A Timing Study of the Manual Searching of Catalogs," Library Quarterly 42:399-415 (Oct. 1972); Robert Blackburn, "Two Years With a Closed Catalog," journal of Academic Librarianship 4:424-29 Qan. 1979); Mary Kay Daniels Canning, "The Cat- alog; Its Nature and Prospects," journal of Li- brary Automation 9:48-66 (March 1976); Ruth Rafter, "The Performance of the Card Catalog; A Review of Research," Library Re- search 1:199-222 (Fall1979); Ben-Ami Lipitz, "Catalog Use in a Large Research Library," Library Quarterly 42:129-39 (Oct. 1972); Carl McAllister and John M. Bell, "Human Factors in Design of an Interactive Library System," journal of the American Society for Information Science 22:96-104 (March-April 1971); Don R. Swanson, "Requirements Study for Future Catalogs," Library Quarterly 42:302-15 Quly 1972); Renata Tagliacozzo, Lawrence Rosenberg, and Manfred Kochan, "Access and Recognition: From Users' Data to Catalogue Entries," Journal of Documenta- tion 26:230-49 (Sept. 1970). 2. Renata Tagliacozzo and Manfred Kochan, "Information Seeking Behavior of Catalog Us- ers," Information Storage and Retrieval6:363 (1970). 3. Jerry Specht, "Patron Use of an On-Line Cir- culation System in Known-Item Searching," journal of the American Society for Informa- tion Science 31:335-46 (Sept. 1980). 4. Kenneth Dowlin, "Users Prefer Online," Li- brary ]ournal105:1595 (Aug. 1980); Mary N. Gouke and Sue Pease, "Title Searches in an On-Line Catalog and a Card Catalog: A Com- parative Study of Patron Success in Two Li- braries," journal of Academic Librarianship (in press); Carol Weiss Moore, "Users Reac- tions to On-Line Catalogues: An Exploratory Study" (Prepared under a grant from the So- cial Sciences and Humanities Research Coun- cil, Toronto, Canada); Specht, "Patron Use of an On-Line Circulation System,·· p .335-46: 5. Moore, "Users Reactions to On-Line Cata- logues." 6. Ohio State University, Behavioral Science Laboratory, "OSU Student Poll, Library Sec- tion, Winter 1981." 7. Specht, "Patron Use of an On-Line Circula- tion System," p.336. 8. Moore, "Users Reactions to On-Line Cata- logues." 9. Dowlin, "Users Prefer Online," p.1595. 10. Ohio State University, "OSU Student Poll, Li- brary Section, Autumn 1980." 11. Ibid.