College and Research Libraries Use of Series Title Authority Cross-References at a Large University Library Henry H. McCurley Jr. and Elizabeth J. Weisbrod The purpose of this study is to determine whether users of the Ralph Brown Draughon Library at Auburn University employ the cross-refer- ences provided by the series title authority file in their searches. The authors present and discuss the results of a transaction log analysis. The analysis focuses attention on the usefulness of cross-references from series title authority records, because only searches that required such cross-references appear in the transaction log report. The study indicates that patrons do use the cross-references gathered by series title authority records. s the module on series and se- ries tracings in the CONSER ·Cataloging Manual indicates, "when monograph or serial catalogers are asked what they consider to be the most difficult aspect of catalog- ing, invariably the answer is 'series."!} The quest for bibliographic control of monographic series may be as quixotic as the quest for the Holy Grail. It is perhaps even more futile, because in the Grail quest, it will be recalled, those armored in purity of heart had some chance of suc- cess. Unfortunately, in dealing with monographic series, purity of heart does not suffice. Those seeking to manage monographs in series must avail them- selves of different arms. The best means of attaining bibliographic control over monographic series may be some form of title authority file. A well-designed series title authority file can greatly improve patron access to monographs in series by providing a consistent, established form of the title, clarifying the treatment of the series, and giving cross-references to li- brary users. Cross-references, a feature of many card catalogs and increasingly of online cata- logs, direct the patron to the "established" form of a given heading, whether name, subject, or title. The established form of a heading is not always an obvious choice. Authors use different names, subject dis- ciplines have many different terminolo- gies, and publishers, editors, and authors often vary the title of a series, serial, or even an individual work over time. In a card file system, it is often possible to find what one is searching for, eventually, by approxi- mation. This is not the case in an online system, where even a slight deviation from an established form can result in no hits and leave a searcher with no clues on how Henry H . McCurley Jr. is a Serials Cataloger at the Ralph Brown Draughon Library, Auburn University; e-mail: hmccurle@lib.auburn.edu. Elizabeth J. Weisbrod is a Cataloger at the Ralph Brown Draughon Library, Auburn University; e-mail: eweisbro@lib.auburn.Iib. The authors would like to thank Kathy Ford and Peggy Seaquist for technical assistance. 285 286 College & Research Libraries to proceed. This article examines use of cross-references for series title searching. This study is based on the transaction logs of the Auburn University Libraries, a member of ARL, with holdings of more than two million physical volumes, two million items in microformat, and one- and-a-half million government publica- tions, serving a student population of more than 3,000 graduate students and 18,000 undergraduates. The main library is the Ralph Brown Draughon (RBD) Li- brary, with branches at the College of Vet- erinary Medicine and the School of Ar- chitecture. Its holdings support more than 300 ·degree programs, including ninety- six doctoral programs, particularly in the sciences and technology. The library uses OCLC as its primary bibliographic util- ity and has been a NOTIS library since 1984. In addition to the cataloging, acquisi- tions, circulation, and other modules, NOTIS contains a relatively sophisticated authority module. Since implementing NOTIS, the RBD has invested a great deal of time and effort in making its authority file an important aid to patron access. Until recently, attention was directed pri- marily to name authority, both personal and corporate, and subject authority.2 In recent years, however, greater attention has been paid to title authority, for both uniform titles and, in particular, series titles. 3 A project to establish a series author- ity file online began in January 1992, and by the end of 1994, the file contained more than 26,000 records. The benefits of this authority file to cataloging personnel were clear: a file of unambiguous series decisions immediately accessible to ev- eryone in technical services, authority records that could not be misfiled (a seri- ous problem with the existing card file), virtual elimination of duplicate records, a system of cross-references between vari- ous forms of the series title, and series titles uniformly established in AACR2 form. May 1996 Although the technical services divi- sion derived many benefits from the project, the most visible ad vantage to public users lay in the system of cross- references. The series authority records provide references from invalid title vari- ants to the established form of the series title, or to earlier and later forms of the title. There was some question, though, as to whether these cross-references were ever utilized. They do not display with keyword searching, only through title or author searching. Most staff, however, indicated that they generally used key- word rather than title to search for se- ries, and encouraged patrons to do like- wise. Although reports on OPAC searching activity are generated monthly by the library's Automation Department, these reports normally do not describe cross- reference usage. This paper employs the results of a special transaction log report to examine use of the cross-reference fea- ture of series authority records by cata- log users. Description of Authority File As of May 1994, the RBD had 24,224 se- ries authority records online and slightly more than half of these records (12,385, or 51.13%) contained cross-references. This number includes authority records for traced and untraced series, and mul- tipart items established under title. All of the cross-references appear to technical services users of NOTIS, but cross-refer- ences only show on the OPAC for series and multipart items that are traced. Cross-references from invalid title vari- ants are created by the 410 and 430 fields on the authority records, and cross-refer- ences from earlier and later forms of the series titles are created by the 510 and 530 fields. A search for any of these forms will cause a message to appear on the OPAC with instructions to search under the valid form of the title or to search also under another form (or forms) of the title (see figure 1). Use of Series Title Authority Cross-References 287 FIGURE 1 Example of Series Authority Record LTAU DONE NBF1975 NOTIS CATALOGING Y06E AU- NBF1975 FMT ART z DT 04/14/92 RIDT 05/02/95 STAT mm ElL n SRC LANG ROM ¢ MOD UNIQUE n GOVT ¢ S/SYS n D/1 n SUBD NUM a S/TYP z NAME c SUBJ a SER c KIND a H/ESTAB a T/EVAL a IP a RULES c 010: : Ian 42019881 040: : Ia DLC lc DLC ld DLC 130: 0: Ia Progress in water technology. 410/1:20: Ia International Association on Water Pollution Research. It Progress in water technology 530/1: 0: lw b Ia Water science and technology 642/1: : Ia v. 12, nr. 215 DLC 643/1: : Ia Oxford Ia New York lb Pergamon Press 644/1: : Ia f 15 DLC 645/1: : Ia t 15 DLC 646/1: : Ia s 15 DLC 670/1: : Ia Seminar on Eutrophication of Deep Lakes. Eutrophication of deep lakes, 1980. 670/2: : Ia Other Series (Serial); numbered 670/3: : Ia ADD & TRACE; CLASS SEPARATELY 670/4: : Ia Marc for wk cat aau3018 The NOTIS OPAC has a feature that "redirects" a user's search from an ear- lier to a later form or a later to an earlier form, or from an invalid title variant to the established form. When a user re- ceives a "search also" or "search also un- der" message, he or she may select the number of the cross-reference from the screen, which will then redirect the search and retrieve titles under the established form of the title. If, for example, a patron searches under the heading "Marriage and Family Review" (which is an invalid variant because the established form FIGURE2 Example of Cross-Reference for Series Title Search Request: T=MARRIAGE AND FAMILY REVIEW AU Books, Journals, Docs Search Results: 1 Entry Found Title Index MARRIAGE AND FAMILY REVIEW 1 *Search Under: MARRIAGE FAMILY REVIEW STArt over HELp OTHer options Type number to display record NEXT COMMAND: 288 College & Research Libraries contains an ampersand), the result will be the message shown in figure 2. Rather than retyping the search, the patron has only to select the cross-reference by typ- ing "1" in order to be redirected immedi- ately to the titles that fall under the es- Clearly, the cross-references pro- vided by the series authority records are potentially very useful to those who are searching for titles in monographic series but are having difficulty because of variants. tablished form of the series heading, as shown in figure 3. Clearly, the cross-references provided by the series authority records are poten- tially very useful to those who are search- ing for titles in monographic series but are having difficulty because of variants. However, the extent to which they were actually being used was not at all clear. May 1996 Literature Review Although no studies concerning title or series title cros.s-reference use were found, the literature contains a number of studies examining the value of subject and name cross-references. Noelle Van Pulis and Lorene Ludy examined subject searching at the Ohio State University Libraries in 1988.4 The libraries' online system, LCS, contained the machine- readable form of the Library of Congress Subject Headings (LCSH) which provided "see" and "see also" references, classifi- cation numbers, and scope notes for sub- ject headings used in the catalog. At this time, LCS did not have keyword search- ing capabilities. The study found that subject search terms matched LCSH vo- cabulary about 80 percent of the time and that 14 percent of these terms matched see references. Overall, cross-references directed users to appropriate terminol- ogy in 6 percent of all subject searches. Other studies found cross-references less FIGURE3 Results of Series Title Redirect Search Request: T=MARRIAGE FAMILY REVIEW Search Results: 34 Entries Found MARRIAGE FAMILY REVIEW AIDS AND FAMILIES (AU) AU Books, Journals, Docs Title Index 2 ALTERNATIVE HEALTH MAINTENANCE AND HEALING S (AU) 3 ALTERNATIVES TO TRADITIONAL FAMILY LIVING (AU) 4 AMERICAN FAMILIES AND THE FUTURE ANALYSES OF (AU) 5 CHARYBDIS COMPLEX REDEMPTION OF REJECTED MAR (AU) 6 CHILDHOOD DISABILITY AND FAMILY SYSTEMS (AU) 7 CORPORATIONS BUSINESSES AND FAMILIES (AU) 8 CROSS CULTURAL PERSPECTIVES ON FAMILIES WORK (AU) 9 CULTS AND THE FAMILY (AU) 10 DEVIANCE AND THE FAMILY (AU) 11 EXEMPLARY SOCIAL INTERVENTION PROGRAMS FORM (AU) ------------- CONTINUED on next page -- STArt over HELp OTHer options Type number to display record FORward page NEXT COMMAND: Use of Series Title Authority Cross-References 289 effective than keyword searching or other system features in resolving problem searches. In 1984, Arlene G. Taylor investigated what effect a name authority file would have had on "no- hit" author searches on a NOTIS system at Northwestern University. 5 At that time, Northwestern did not have an online authority file. Using actual searches from transaction logs, she matched users' input against the LC au- thority file and, for those names that LC had not established, against authority records established according to LC prac- tice. She found that only 6.4 percent of no-hit author searches would have re- sulted in a successful search if the cross-references had been present. She suggested that several system modifi- cations would have been more helpful than cross-references. In 1992, Joan Cherry studied "zero-hit" subject searches at the University of Toronto Libraries. 6 The libraries' OPAC did not contain subject cross-references, al- though title, keyword title, and keyword subject searching were available to users. She found "that keyword subject, key- word title, or title searches using the origi- nal query from the user's zero-hit subject search were as fruitful or more fruitful than new searches constructed from cross-references provided by LCSH."7 Al- though these other types of search strate- gies would have produced successful re- sults, they were seldom used. She sug- gested that educating users about LCSH or providing cross-references on the OPAC would not help produce success- ful outcomes for most zero-hit subject or name searches. Conversely, Alexis Jamie- son, Elizabeth Dolan, and Luc Declerck found that more than 50 percent of name see references and nearly 70 percent of see references for subjects on LC authority records could not be matched with cata- log records using keyword searching. 8 They concluded that keyword searching was not an adequate substitute for a cross- reference structure. Perhaps cross-refer- ences, although less sophisticated than other approaches, provide an essential link for some searchers. Limitations of the Study This study provides a count of instances in which a user actively followed a cross- reference through the redirect system. The authors cannot know for certain, though, precisely how much the cross-references are used or whether they are used in search of series. For example, a searcher may have used a cross-reference but by- passed the redirect system by retyping the search using the established heading rather than selecting the number of the cross-reference. This sort of usage would not have been included in the study. Con- versely, some redirects may have been counted that were not the result of series title searching. For instance, a user may have encountered a cross-reference for a series while browsing a corporate author display or searching for a book with a similar title. The user may not have been searching for that particular series, but selected the cross-reference out of curios- ity (or confusion). Transaction logs do not reveal why a user chose a particular cross- reference or for what materials the user was actually searching. Methodology The library's transaction logs record all instances of redirects from cross-refer- ences. Using these logs, the library's Au- tomation Department produced a report that located and printed out all instances in which a user selected a title cross-ref- erence for the period January 1, 1992, through May 31, 1994. As previously mentioned, January 1992 marks the be- ginning of the project to load the series authority file. Some of the cross-refer- ences selected were for uniform titles (Gawain and the Green Knight being a particular favorite). Once these were eliminated, the remainder consisted of 2,793 redirects of series titles or multipart items entered under title. In other words, 290 College & Research Libraries May 1996 TABLEt Series Title Redirects 1992 1993 1994 Authority Authority Authority Redirects Records Redirects Records Redirects Records Month Used Loaded* Used Loaded Used Loaded January 0 227 134 8,082 131 11,890 February 7 1,188 139 8,990 125 11,978 March 24 1,726 139 10,343 120 12,199 April 40 2,251 144 10,871 101 12,287 May 67 2,837 177 11,126 119 12,385 June 37 3,681 121 11,234 July 77 4,172 129 11,344 August 72 4,875 125 11,449 September 57 5,315 98 11,511 October 87 6,008 158 11,591 November 121 6,670 136 11,668 December 51 7,720 57 11,810 Note : Auburn University's quarters begin in January, late March, mid-June, and late September. *Figures in "Authority Records Loaded" column include only series authority records with cross- references. the report recorded each instance in which a user, after receiving a search un- der or search also under display, entered the line number of the established form of a title and was redirected to the titles that fall under the established form, or to an earlier or later title of the series head- ing. The report recorded only searches on the public catalog, not those made on the technical services portion of NOTIS. It should be emphasized that these 2,793 searches do not represent the total number of series title searches that oc- curred during the period under exami- nation. Rather, they represent only those series title searches that satisfied the very specific conditions necessary to produce a cross-reference from a series title author- ity record: (1) a series authority record for the title must have been online at the time of the search, and (2) the search entered must have been for either an invalid form of the title that appeared on the authority record for that title (eliminating, for ex- ample, typographical errors) or an earlier or later form of the title. Results The report provides a picture of the types and numbers of title cross-references em- ployed by users of Auburn University's online catalog. Two areas were examined: the distribution of redirects over the twenty-nine-month period under exami- nation and the frequency of their use, and the subject areas in which series titles were redirected through cross-references.9 Distribution The report located 2,793 redirects distrib- uted over a twenty-nine-month period. As table 1 shows, the number of redirects per month increased steadily as the project to load the series authority file progressed. The numbers also rise and fall in accordance with Auburn's academic quarters (see table 1). Although the numbers of redirects steadily increased through 1993, the us- age appears to have dropped somewhat in the first five months of 1994. Although many explanations are possible, one rea- son may be that the number of keyword Use of Series Title Authority Cross-References 291 searches began to increase as the number of title searches decreased proportionally. As stated earlier, the cross-references are not available through keyword searching, only title or author searching. Frequency of Use The report also indicated the number of times each title was redirected through a cross-reference. As table 2 shows, the number of searches per title ranged from one, in the majority of cases, to 115 (see table 2). As shown, a large number of titles were redirected through a cross-reference only once. On the one hand, a small num- ber of titles accounted for a large number of the redirects recorded in the study. Forty-one of the titles (4.45%) were redi- rected a total of 944 times, or 33.8 percent of the total. On the other hand, 470 titles (51.03%) were redirected only once dur- ing the 29-month period. Some of the dis- parity in cross-reference usage may be the result of series use. Heavily used series titles were probably redirected more fre- quently than lightly used ones. In the case of some of the titles that show up fre- TABLE2 quently in this study, the authors know, from discussion with reference personnel, that they are heavily used. However, series use alone may not completely account for the large number of redirects that some titles produced. For Some of the disparity in cross- reference usage may be the result of series use. instance, some series are routinely cited by only their series title and volume num- ber, whereas others are commonly cited by the individual volume title. Those cited by series title would have more opportu- nities to bring a cross-reference into play than those that are not. Other situations, such as a series title similar to that of a monograph or a series title also used as a subject heading, also may have contrib- uted to some titles' use. Another factor may be the series title itself. The titles that accounted for most of the cross-reference usage were searched frequently, but the search matched a cross- reference rather than the form of title se- Freguency Distribution of Redirects lected by the cataloger. The series titles in ques- tion, then, must present some special difficulties for library patrons. This points to a difficulty with a number of series titles: the title is constructed by the cataloger and is not an intuitive choice for searchers. Series titles consisting of ini- tialisms, series indexed under a subseries title, or series titles that ap- pear in several different forms may lead users to search under a cross- reference. Number of Re- Number of directs per Title Titles 1 470 2 180 3 93 4 58 5 23 6 25 7 13 8 10 9 8 10 8 11 6 12 3 13 3 14 1 15 1 Number of Re- directs per Title 16 20 22 23 24 26 29 30 35 39 45 48 49 70 115 Number of Titles 1 2 1 1 3 1 1 1 2 1 1 1 1 1 1 Quirks of the online system account for a number of the redirects. 292 College & Research Libraries Although the reason for following a cross-reference is not always apparent, in the case of the most frequently ap- pearing title, Marriage & Family Review, the reason for the difficulty is clear. The title contains an ampersand, which is not searchable on NOTIS. In the absence of a cross-reference from the series au- thority record, the patron would have had to know to search "Marriage Fam- ily Review." Certainly this is an instance in which the authority record provided a useful cross-reference. Subject Areas This study included 921 different series and multipart item titles. For purposes of analysis, the authors divided the redirects into three general subject areas: humani- ties (Library of Congress classes A, B [ex- cept BF], M, N, P, and Z), social sciences (classes BF, C, D, E, F, G, H, J, K, and L), and science and technology (classes Q, R, S, and T). 10 A few additional titles were for government publications with SuDocs call numbers, while other titles covered such a range of classifications that they could not be characterized as subject spe- cific. As table 3 shows, the largest num- ber of titles were in science and technol- ogy. In terms of the number of redirects, series in the areas of science and technol- ogy also predominated. As table 4 shows, ·over half the searches resulting in redi- rects were for series titles in science and technology. TABLE3 Number of Series Titles Appearing in Study Subject Area Count Percent Humanities 191 20.74% Social sciences 217 23.56 Science & tech. 436 47.34 Gov't. docs. 17 1.85 No specific subject 60 6.51 Total 921 100.00 May 1996 TABLE4 Redirects by Subject Area No. cross- Subject Area refs used Percent Humanities 430 15.40% Social sciences 588 21.05 Science & tech. 1499 53.67 Gov't. docs. 129 4.62 No specific subject 147 5.26 Total 2,793 100.00 As shown in table 5, eleven series titles were redirected more than twenty-five times during the period of study. Once again, science and technology series were the most numerous with eight titles, while two titles were government documents and one was from the social sciences. In- terestingly, there are two instances in which both the earlier and later forms of a series title appear in the list. Research Monograph (National Institute on Drug Abuse) and NIDA Research Monograph are related titles, as are Water Science and Tech- nology and Progress in Water Technology. For purposes of this study, the authors counted titles separately rather than as one series. Several factors may account for the heavy cross-reference usage in the science and technology series. Series titles in the sciences may be more complex than in other subject areas, involving complicated relationships between main and subseries. Perhaps publishing practices in the sciences produce more series titles than other areas, and thus provide more opportunities for cross-reference use. Another reason may lie in how the scien- tific literature is cited. If works in a series are commonly cited by series title and volume number rather than by individual title, more cross-references to the series title may be used. Other factors such as teaching practices in the sciences and the overall structure of the literature also may play a role. Locally, several factors at Au- burn University may contribute to the predominance of science and technology Use of Series Title Authority Cross-References 293 TABLES Series Titles Redirected Mor.e Than 25 Times No. of Redirects Series Title Subject 26 29 30 35 Ellis Horwood Series in Biochemical Pharmacology Research Monograph (National Institute on Drug Abuse) Family Therapy Collections Sci. & tech. Document Sci. & tech. Sci. & tech. Proceedings of SPIE-The International Society for Optical Engineering 35 39 45 48 49 70 ASTM Special Technical Publication Monograph Series (American Chemical Society) Transportation Research Record Sci. & tech. Sci. & tech. Sci. & tech. Sci. & tech. Sci. & tech. Document Social sciences Water Science and Technology Progress in Water Technology NIDA Research Monograph 115 Marriage & Family Review titles. Over half of Auburn's students (51%) are enrolled in degree programs in the sci- ences and technology, as compared to 38 percent in the social sciences and only 11 percent in the humanities. Because of the enrollment, series in the sciences and tech- nology may see more use than those in other subject areas. Finally, the makeup of the local catalog may play a role. Although the number of series authority records in each subject area is not known, perhaps the online catalog simply contains more authority records for science and technol- ogy series than for other subject areas. Conclusion The report indicated that cross-references for series titles are used. The number of redirects rose as the number of cross-ref- erences in the database increased, indi- cating that patrons and staff not only en- countered the cross-references, but ac- tively followed them. Most titles appear- ing in this study were redirected by a cross-reference only once during the pe- riod of this study. However, several titles were redirected a large number of times, perhaps because of heavy use of these- ries, frequent citation by series title in the literature, or a difficult or complex estab- lished form of the title. Overall, cross-references for series in the sciences and technology were more heavily used than those for other subject areas. Science and technology series ac- counted for nearly half the titles in the study and were redirected through cross- references many more times than titles in other subject areas. Although Auburn University has an emphasis in this area, other factors, such as the complexity of series titles, the structure of the literature, publishing practices, or even teaching methods in the sciences, also may have contributed to the predominance of sci- ence and technology. Further investigation may provide a more complete picture of cross-reference usage. Studies into how much series title searching occurs, what sorts of series searches users employ, and the types of series titles that benefit from cross-refer- ences may help determine how library catalogs can better serve users. Notes 1. Jean L. Hirons, CONSER Cataloging Manual, Module 12 (Washington, D.C.: Library of Congress, Cataloging Distribution Service, 1993), 3. 294 College & Research Libraries May 1996 2. Helen Goldman and David M. Smith, "Name Authority in a NOTIS Environment-Au- burn University Libraries," Cataloging & Classification Quarterly 9 (1989): 121-31. 3. Henry H. McCurley Jr., "Implementation of an Online Series Authority File at Auburn University," Cataloging & Classification Quarterly 18 (1993): 41-58; also, Henry H . McCurley Jr., "The Benefits of Online Series Authority Control," Technical Services Quarterly 11 (1994): 33-50. 4. Noelle Van Pulis and Lorene E. Ludy, "Subject Searching in an Online Catalog with Au- thority Control," College & Research Libraries 49 (Nov. 1988): 523-33. 5. Arlene G. Taylor, "Authority Files in Online Catalogs: An Investigation of Their Value," Cataloging & Classification Quarterly 4 (spring 1984): 1-17. 6. Joan M. Cherry, "Improving Subject Access in OPACs: An Exploratory Study of Conver- sion of Users' Queries," Journal of Academic Librarianship 18 (May 1992): 95-99. 7. Ibid., 99. 8. Alexis J. Jamieson, Elizabeth Dolan, and Luc Declerck, "Keyword Searching vs. Authority Control in an Online Catalog," Journal of Academic Librarianship 12 (Nov. 1986): 277-83. 9. The authors had hoped to provide some comparison between the number of patron searches and the number of staff searches, but identifying the type of searcher was not possible. Of the searches, 32.76 percent are known to have been made from library staff terminals located at reference desks, work areas, and in technical services, while more than half of the searches (50.81 %) were made from public terminals or by dialing into the library. (The remaining 16.43 percent of the searches were recorded as network searches, which could have been made by public or staff users.) Although the percentage of searches originating from the public terminals suggests that patrons are using the cross-references, some of the searches could have been made by library personnel assisting users. 10. Auburn University's main library (the Ralph Brown Draughon Library) organizes its collection using these broad subject categories. The second floor holds humanities materials; the third floor, social sciences materials; and the fourth floor, science and technology materials. TICLES IN PHYSICS A N f \ V n 0 C U .tv\ E NT n E Ll V E R Y S E R V I C E r R 0 M I II I A ,\\ I R I CAN I N S TIT lJ fl 0 F P II Y SICS AN 11 I B S CO doc ARTICLES IN PHYSICS gives you immediate !l«fH to more than 30,000 scientific and technical publications from AlP, its Member Societies, and other leading scientific publishers worldwide. Articles In Physics is also an excellent source for material in disciplines other than the physical sciences. Backed by EBSCOdoc's extensive collection and its worldwide resources, researchers and librarians around the world can obtain articles rapidly via mail, fax, or Internet fax. Turnaround time for in-house titles is 24 hours. CALL TOLL-FREE 800-480-PHYS international +0 1-41 5-259-6002 fax +01-41 5-259-6044 e-mail articles@aip.org URL: http://www.aip.org/articles.html OR WRITE TO: Articles In Physics • 1722 Gilbreth Road • Burlingame, CA 9401 0