General
A study reported in 1981 of OCLC’s online union catalog to investigate the scope of its music holdings by checking 4 major lists of materials (“Books Recently Published” column of Notes; “Music Received” column of Notes; “New Listings” column of the Schwann-1 Record and Tape Guide; and A Basic Music Library: Essential Scores and Books) in the catalog showed that, of the 317 books listed in the December 1979 “Books Recently Published” column and searched 1 year later, 312 (98.42%) were found in the OCLC union catalog, while of the 287 books listed in the September 1980 “Books Recently Published” column and searched 7 months later, 272 (94.78%) were found in the OCLC union catalog. (Source)
Ibid…. showed that, of the 420 items listed in the December 1979 “Music Received” column and searched 1 year later, 255 (60.71%) were found in the OCLC union catalog, while of the 351 items listed in the September 1980 “Music Received” column and searched 1-2 months later, 109 (31.05%) were found in the OCLC union catalog. (Source)
Ibid…. showed that, of the 276 items listed in the December 1979 “New Listings” column and searched 14 months later, 162 (58.7%) were found in the OCLC union catalog, while of the 282 items listed in the September 1980 “New Listings” column and searched 1-2 months later, 64 (22.7%) were found in the OCLC union catalog. (Source)
Ibid…. showed that, of the 941 items listed in A Basic Music Library, 861 (91.5%) were found, including 231 (94.29%) of the study scores, 195 (96.53%) of the performing editions, and 114 (82.01 %) of the instrumental methods and studies. Further, “a large number of specific editions not found in the search” were represented in the OCLC union catalog by other editions. (Source)
Academic
A 1967 study comparing a dictionary catalog at 1 university with a divided catalog (author/title and subject) at another university by using undergraduates to search entries showed thatthere was no statistically significant difference in average success rates between the 2 catalogs in subject searching or known-item searching. (Source)
A survey of 12 university libraries (11 responding) during a 3-month period in 1965 and a survey of 11 public libraries during a similar period in 1967 to determine the kinds of reference assistance needed at the card catalog (647 problems reported with 284 from university libraries and 363 from public libraries), reported together in 1968, showed thatthe 3 most frequent problems with the card catalog in university libraries were subject headings (18% of university total), filing arrangement (17% of university total), and “see” or “see also” references (15% of the university total). In public libraries the 3 most frequent problems were filing arrangement (23% of public total), call number (15% of public total), and subject headings (13% of public total). (Source)
A 1967-70 study in the Main Library at Yale University involving 2,100 interviews at the card catalog during a 1-year period showed that84% of the card catalog searches were successful, that 5% of the searches were unsuccessful for the patron even though library staff later located the document in the card catalog, that 10% of the searches were for documents that probably exist but that were not in the card catalog at the time, and that 1 % of the searches were for documents too vaguely or inaccurately described to follow up on. (Source)
Ibid…. showed thatthe 4 most popular approaches to searching the card catalog were: author name (personal or corporate, 62.0%), title (28.5%), subject (4.5%), and editor (4.0%). (Source)
A study reported in 1970 at the University of Chicago concerning the use of nonstandard catalog access points (104 individuals selected and read 440 books in the general area of psychology that they had not previously read and then were tested on what they remembered about the book 2 weeks later) showed thatstandard author/title/subject information sufficient to look the book up in a conventional card catalog was recalled in only 18% of the cases. (Source)
A survey in 1976 of 999 library users at San Jose State University Library showed thatthe biggest barrier to locating library materials was difficulties in using the card catalog and finding desired items already in circulation. 42% of the items not found by patrons in the card catalog were in the card catalog, and 42% of the books not located on the shelves were in circulation or in the Reserve Book Room. (Source)
Ibid…. showed thatthere was a 77.52% success rate in locating desired items in the card catalog and a 76% success rate in locating the book in the bookstacks, for an overall success rate of 58.9%. (Source)
A study reported in 1977 at the University of California, Berkeley, comparing the importance of subject familiarity (knowledge of a specified academic field) versus catalog familiarity (knowledge of the structure of the Library of Congress subject headings) in thinking of an appropriate term for subject searching in the library catalog when presented an abstract and book title and involving 22 psychology students (subject familiarity), 22 economics students (subject familiarity), and 17 library students (catalog familiarity), showed thatcatalog familiarity was more important than subject familiarity to a statistically significant degreee in selecting an appropriate subject term [significance level not given]. Specifically, scores for library students in determining subject terms for psychology and economics materials were higher than scores for students in those disciplines determining subject terms in their own discipline. (Source)
Ibid…. showed that, although the difference was not statistically significant, psychology and economics students scored higher in selecting appropriate subject terms for the opposite discipline than they did in their own discipline. (Source)
Ibid…. showed thatstudents were more likely to achieve the illusion of success in using the library catalog than real success. This was demonstrated by the discrepancy between basic matching (exact matches between subject heading assigned by the library to a particular item and the subject heading chosen by the student) and “existence matching” (a marginal match in which the subject heading chosen by the student existed in the catalog but was not the subject heading assigned by the library). Specifically, the number of terms that constituted basic matches versus existence matches for the 3 groups were as follows:
economics students 21% basic; 62% existence matches
psychology students 22% basic; 60% existence matches
library students 35% basic; 64% existence matches (Source)
A 1978 study at Ohio State University Library involving 2 full days’ transactions in the OCLC data base for each of the OSU terminals (1,153 searches of the data base) showed thatof 158 searches by the public the 2 most frequent types of searches were by name/title (77 or 48.7% of the searches by the public of which 22 or 28.6% were successful) and by title (44 or 27.8% of the searches by the public of which 20 or 45.4% were successful). (Source)
A 1979 study at the University of Illinois, Urbana, involving patron use of an online circulation system for known-item searching (interviews with 240 faculty, staff, student, and visiting patrons conducting 310 searches), showed that, of 235 known-item searches, 8% of the location searches (call number previously looked up in card catalog and departmental library location sought only) and 16% of the original searches (online search was first attempt to find out if the library had an item) failed because of an error in using LCS. (Source)
Ibid…. showed thatthe length of 156 successful known-item searches was as follows:
0.2 minutes or less 16 (10%) searches
1.0 minutes or less 82 (53%) searches
2.0 minutes or less 117 (75%) searches
3.7 minutes or less 139 (89%) searches
7.5 minutes or less 156 (100%) searches (Source)
Ibid…. showed thatof 222 searches there was no statistically significant difference in success rate between more experienced users (5 or more previous uses) of the online system and less experienced users (less than 5 previous uses) of the system. Specifically, both groups had the same number of unsuccessful searches (33), while the more experienced patrons had 83 successful searches compared to 73 for the less experienced patrons. However, experienced patrons did complete their successful searches more quickly than inexperienced patrons to a statistically significant degree with a correlation of r = .24 (significant at the .01 level). (Source)
Ibid…. showed that, of 120 original known-item searches (original search was the first attempt to find out whether the library had the item desired), the library owned the item in 91% of the cases, the patron could find the item on LCS (Library Circulation System, an online system) in 66% of the cases, and the copy was uncharged in 90% of the cases. (Source)
A 1980 study of patron use of the serial card catalog at the University of Illinois, Urbana (sample size: 452 patrons; usable responses: 445 patrons) involving faculty, students, and staff showed that366 (83%) of the searches undertaken by patrons in the serial card catalog were successful, i.e., a citation was matched to a catalog entry. Further, the success rate of the frequent catalog user (daily or once/twice per week) was not statistically significantly better than the success rate of the infrequent catalog user. Specifically, 167 (46%) of the frequent catalog users and 199 (54%) of the infrequent catalog users were successful in their searches. (Source)
Ibid…. showed thatof 427 searches there was no statistically significant difference in success rates between patrons who wrote their citations down (or Xeroxed them) and those who did not. For example, 246 (70%) of the patrons who found their citations in the card catalog had written them down compared to 52 (69%) of the patrons who did not find their citations in the card catalog but who had written them down. Conversely, 93 (26%) of the patrons found their citations in the card catalog without writing them down compared to 20 (27%) of the patrons who did not find their citations in the card catalog but also did not write the citation down. (Source)
A comparison reported in 1982 of title search capability on an online catalog vs. card catalog at Ohio State University showed thatpatrons were more successful in using departmental card catalogs (1/4 to 1/2 million cards) than in using a central online catalog (3.5 million records). The same held true though to a lesser degree for a skilled librarian. Success rate in the card catalog ran 9-20% higher for patron groups and 4% higher for an experienced librarian. (Source)
Ibid…. showed thatthe percentage of patron failure in the following problem areas for the online catalog was:
title appearing as subject 75%
hyphenated words 67%
words not on stoplist 50%
all words in title on stoplist 50%
too many matches 43%
words on stoplist 41%
person’s name begins title 40%
abbreviations 40%
foreign titles 33%
1-word titles 25% (Source)
Public
A survey of 12 university libraries (11 responding) during a 3-month period in 1965 and a survey of 11 public libraries during a similar period in 1967 to determine the kinds of reference assistance needed at the card catalog (647 problems reported with 284 from university libraries and 363 from public libraries), reported together in 1968, showed thatthe 3 most frequent problems with the card catalog in university libraries were subject headings (18% of university total), filing arrangement (17% of university total), and “see” or “see also” references (15% of the university total). In public libraries the 3 most frequent problems were filing arrangement (23% of public total), call number (15% of public total), and subject headings (13% of public total). (Source)
A 1973 study in the Burnaby Public Library (British Columbia) involving patron use of the card catalog (survey size: 367 patrons) showed that63% of the respondents were successful in finding the book for which they were looking after 1 search, while 289 (78%) patrons ultimately had “some measure of success” in using the catalog. (Source)
