College and Research Libraries GARY A. GOLDEN, SUSAN U. GOLDEN, AND REBECCA T. LENZINI Patron Approaches to Serials: A User Study Although many catalog use studies have been reported, those limited to patron success with locating serials have been uncommon. This study, conducted at a separate serial card catalog in a major research library, measures the success of more than jour hundred patrons in the bibliographic retrieval of serials. The authors interviewed patrons and then analyzed the data in an attempt to determine how patrons approach a card catalog when searching for serials, their success rate, and reasons for their successes and failures. BIBLIOGRAPHIC coNTROL of serials, whether manl!al or automated, has long plagued li- brarians, and evaluation of user success with this control is seldom, if ever, undertaken. While many studies have been reported on patron usage of card catalogs, none of the published user studies has been limited to pa- tron access to serials within a card catalog. Studies involving serials have focused pre- dominantly on usage of actual items and ti- tles, with an emphasis on collection develop- ment and control, rather than on biblio- graphic retrieval of those serial titles. 1 One study conducted by Peterson did investigate patterns of serial usage according to type of patron and type of citation; however, he did not query specifically the success with which a patron located any serial bibliographi- cally. 2 The recent Murfin study on periodical retrievability focused primarily on locating a volume on the shelf, but did summarize user success with the "periodical directory." Mur- fin found that "only 50 percent of those who used the directory were able to use it cor- rectly."3 One explanation for the small number of usage studies limited to serials may be the wide variance in the handling of serial files. Computer-produced serial lists in printout, Gary A. Golden is documents librarian; Susan U. Golden is automated records maintenance coor- dinator; and Rebecca T. Lenzini is assistant auto- mated records librarian, all at University Library, University of Illinois, Urbana-Champaign. 22 I microfilm, or book format are not uncom- mon, nor are cardex title entry files or inter- nal files of one sort or another serviced by library personnel. A separate serial card caUl- log that is accessible to the public and in- cludes main entries, added entries, cross- references, and holdings invites a study of its users, their approaches to serials, and their successes and failures in finding them. This study was conducted at such a catalog, lo- cated at the University of Illinois, Urbana campus, where nearly 100,000 serial titles are included in the serial catalog. During an age of rapid automation one may question the usefulness of another card catalog study, even a study limited to a previ- ously unexplored area. However, we cannot hope to develop successful interactive online catalogs without a thorough understanding of the usage made of our present manual files. In her ar~icle "The Performance of Card Cat- alogs: A Review of Research," Hafter sum- marized current thinking by saying, "There appears to be a feeling that on-line systems can and should be designed by analyzing pa- tron behavior at the card catalog. "4 Those developing online catalogs must know the major access points needed to ensure the highest probability of user success. It is doubtful that any online catalog will have the necessary access points to ensure 100 percent user success. Trade-offs will undoubtedly come because of costs of central memory · core. Important access points may not be in- cluded due to lack of awareness on the part of ·the librarian: In fact, we may well discover that the most important access points have not been included in our traditional card cat- alogs and need to be uncovered for inclusion in future catalogs, whatever their form. The research reported here is a first at- tempt at an exploratory study to determine ·how patrons who are looking specifically for serial publications approach a card catalog. Successful serial searches were analyzed to determine what factors contributed to the patrons' success, while failures were ana- lyzed to determine what, if anything, could be done to improve the probability of success. The problems of frequent name changes, form of catalog entry versus the patron's bib- liographic citation, and the use of cross- references were explored to judge their rele- vance to serial bibliographic retrievability, as was the impact of user instruction. The probable influence of AACR2 and increased title entry of serials was also investigated. METHODOLOGY ""' The methodology selected for any user study has a profound effect on the data gener- ated. As Hafter observed, "Almost all of the catalog use surveys are flawed by inadequate sampling procedures.'' 5 Lancaster devoted an entire chapter to "Studies of Catalog Use" in The Measurement and Evaluation of Li- brary Services and notes that "the most valu- able studies of catalog use have been con- ducted through interviews with users at the time they search the catalog" rather than by survey afterward. 6 The problems of con- structing and administering the interviews are well covered by Hafter. 7 With the interview approach in mind, a questionnaire was designed and pretested on thirty library patrons. Discussion between the researchers and with the University of Il- linois Survey Research Laboratory led to the questionnaire employed in this study (see ap- pendix 1). The testing instrument was de- signed to be straightforward, unambiguous, and easily coded. Days and times for the ad- ministration of the questionnaire were se- lected randomly between March and May of 1980, a period including both heavy and light use. A minimum of four hours were selected randomly for each test day and the question- naire was administered over the total hour selected. A total of twenty-five weekdays and Patron Approach to Serials I 23 ten weekend days using the hours between 9 a.m. and 11:59 p.m. produced a sample size of 452 of which 445 were usable (see table 1 for the composition of the sample). During the designated testing hour, pa- trons who approached the serial card catalog and pulled out a drawer to begin a search were asked to participate in the study. Only eighteen people elected not to answer the questions, a high rate of success no doubt due to the unique cooperativeness of a college campus. Patrons not employing the serial card catalog in an attempt to locate serial publications were not interviewed. Each participant was questioned and the appropri- ate answers were circled by the interviewer. The average interview time was approxi- mately five or six minutes. Any interviewer must guard against bias- ing the sample in the selection of subjects for interview. Friendly faces or slower users are more likely to be selected for interviewing if no control mechanism is employed. To en- sure random selection of subjects within any given hour, interviewers approached the first person to pull out a drawer after the start of the hour. Upon completion of that interview, the next patron to approach the file and pull out a drawer was questioned. Thus, users al- ready in place at the start of an hour were discounted, as were patrons who approached the file while an interview was in . progress. Unbiased selection of candidates was en- sured. Anywhere from two to nine interviews were conducted during the hour. Each pa- tron was observed until the conclusion of the serial card catalog search. For example, if patrons were referred elsewhere in the serial card catalog, the interviewer noted this and observed the second or even third search. By its nature, any obtrusive study lends it- self to some interviewer bias. By only ap- proaching a patron after he/she had already selected the drawer (and had therefore for- TABLE 1 CHARACTERISTICS OF THE SAMPLE Number Percent Freshmen/ sophomores Juniors/ seniors Graduate students Faculty/staff Other• Total 109 121 153 39 23 445 •Includes visitors and students from other schools . 25 27 34 9 5 100 24 I College & Research Libraries • January 1982 mula ted an initial search strategy)' the inter- viewer did not influence this initial search strategy. Also, no guidance on where to look next, what entry to use next, or how to cor- rect spelling errors was given the patron until the questionnaire was completed. These at- tempts to limit interviewer intrusion helped to minimize the adverse effect of being ob- served on the user's performance. REsULTS AND ANALYSIS The strategy used by patrons when search- ing a card catalog for serial publications and their success or failure in finding those publi- cations within a catalog will be developed using inferential and descriptive statistics. Success is defined in terms of bibliographic success, that is, the patron locating the entry in the catalog for the item he/she wanted. If the catalog contained the entry the patron desired but not the specific issue of that publi- cation, this was counted as a successful search since the failure was the fault of the library's collection and not the user's ability to employ the catalog. Patron failure was defined as the inability to find an entry in the catalog to match the citation in hand. Failures were an- . alyzed to discover sources of problems and their possible solutions. No attempts were made to determine if the patron actually re- trieved the item after using the catalog or if that jtem was actually on the shelf in the li- brary. The serial card catalog was employed by patrons to locate what the researchers de- fined as four types of serials: journals and magazines, proceedings, annual reports, and government documents (see table 2). English was the predominant language of the publications sought, with only twenty- seven of the serials searched (6 percent) writ- ten in other languages. Except for three an- nual reports and two proceedings, all the foreign language publications were for jour- nals or magazines. TABLE2 TYPE OF SERIAL SEARCHED Number Percent Journal or magazine 374 84 Proceeding 19 4 Annual Report 23 5 Government document 29 7 Total 445 100 Corresponding to the high num her of jour- nals and magazines searched, 43 percent of the sample stated that they had used an in- dex, abstract, or bibliography to obtain their citation (see table 3). Forty-three different indexes and abstracts were used. The Readers' Guide was the citation source for fifty-four magazines (28 percent of the maga- zines searched), while the Business Periodi- cals Index was a distant second citation source for only seventeen magazines (9 per- cent). Bibliographies and footnotes in both journals and books accounted for 29 percent (or 127 <;:itations) . Forty graduate students and faculty members had citations that were generated via computer-assisted literature search of some database. This nearly equaled the forty-two (mostly undergraduate) stu- dents who had used a class reading list for their citations. The diversity in the sources of patron cita- . tions did not seem to influence the overall success rate. Over 83 percent of the searches were successful in matching a bibliographic entry to a catalog entry (see table 4). Approx- imately 72 percent of those successful searches found the cataloging entry for the item sought in the first drawer selected from the serial catalog. Thirty-six patrons who were unsuccessful using the first drawer they selected persisted in conducting a second, third, and in one case, a fourth search. In most of these self-directed multiple catalog searches, the patron had made a mistake in the initial drawer selection because he was unaware of the drawer's alphabetical limita- tions (e.g., he wanted New York while the drawer chosen covered New to New T). Other multiple searches were necessary due TABLE3 SouRCE OF CITATION Class reading list Index, abstract, or bibliography Bibliography in book Bibliography in journal Footnote in book Footnote in journal Online literature search Other• Blankt Total Number Percent 42 192 36 30 30 31 40 20 24 445 9 43 8 7 7 7 9 " 4 5 100 "Includes recommendations by friends or teachers and looking items up for other people . tlncludes those who did not remember or were not sure . Patron Approach to Serials I 25 TABLE4 SuccESs RATE AND CATALOG AccESs POINTS Number User entdw same as catalog's entry 319 User con ucted a second search 25 User conducted a third search 10 User conducted a fourth search 1 Catalog had a cross-reference from the user's first entry 7 Catalog had a cross-reference from the user's second entry 4 Total 366 to spelling errors by the patron or because the patron was not sure of the form of the entry in the catalog. The cross-reference system in the catalog was not an important factor as only eleven patrons (3 percent of the sample) en- countered a cross-reference. Those cross- references employed were from an abbrevia- tion to a .complete entry in the catalog (nine cases) or because of a name change (two cases). Table 4 also shows that most of the patrons who were successful used a title entry. There were 303 successful title searches versus only fifty-eight successful searches by corporate entry. Twenty-four of twenty-seven patrons who conducted second or third searches had initially searched by a title that did not war- rant a title-added entry under rules prior to AACR2 (i.e., proceedings or annual report of a corporate body). These patrons were forced to look under corporate entries. AACR2, with its emphasis on title entry, would have allowed twenty searches to be successful without the need for a second or third search. Sixty-three of the successful searchers came to the catalog with only an abbreviated form of the entry. Most of the abbreviations were of the type one might find in an index or ab- stract (e.g., Rev Soc Stud for The Review of Social Studies), but some patrons employed a type of mnemonic memory device of their own creation (e.g., PSQ for Political Science Quarterly or USN for U.S. Neivs and World Report). The use of an abbreviated entry and the subsequent transposition to a complete cataloging entry was employed by all groups of serial catalog users. The actual incidence of associating abbreviated entries with com- plete cataloging entries is probably much higher as there was no way of ascertaining the use of this phenomenon by patrons who did not have a written citation (25 percent of Percent Title Corp. Author 72 298 21 6 5 20 2 4 6 .2 0 1 2 3 4 1 3 2 83 313 54 the sample). If this phenomenon is proven to exist on a very large scale, future interactive online catalogs might choose to incorporate these abbreviations into their searching strat- egy, or perhaps offer a method of mapping the user's abbreviation to the correct entry. Eighty-three percent of the searches stud- ied were successful. Numerous factors con- tribute to the success of any one search in a card catalog. In an attempt to isolate some of those factors in this study, four general hy- potheses were tested using the chi-square sta- tistic calculated via The Statistical Package for the Social Sciences. It could be hypothesized that a patron who frequently uses a catalog might be more effi- cient than a patron who uses it infrequently. A frequent user should be more accustomed to filing quirks and the general makeup of that catalog. Table 5 shows the relationship between the amount of serial catalog usage and whether or not a search was successful. The chi-square test shows that the success rate of the frequent catalog user was not sig- nificantly different (sig. = .10) than the suc- cess rate of the infrequent catalog user in this study. By itself, the number of times a patron used the serial catalog was not a determinant of whether the search would be successful. Another factor that might influence the success of any catalog search is whether or not the patron has written down the citation. A TABLES UsE AND CATALOG SuccESs Frequent* Seldomt Total Entry Found 167 (46%) 199 (54%) 366 .x• = .01 df = 1 sig. - .10 Entry Not Found 31 (48%) 33 (52%) 64 • Frequent includes daily or once/twice a week usage. TSeldom includes none, once, or monthly usage. Total 198 232 430 26 I College & Research Libraries • January 1982 written citation precludes an incorrect entry due to forgotten or transposed words and al- lows the patron to concentrate on under- standing the arrangement of the catalog and matching the catalog entry. The hypot?esis to be tested is that the patron who has wntten down a citation will be more likely to con- duct a successful search than the patron who has not. Inspection of table 6 shows that t~ere was no relationship between whether a cita- tion was written (significant at .99level) and the success of a search. When a patron asks a librarian or some other library staff member if the library owns a specific serial, he/she is sometimes referred to the catalog to check under a specified en- try. At Illinois, the referral might also ha:e been from the main card catalog to the senal card catalog. Patrons who had been referred might have been expected to come to a cata- log with a more correct or complete entry (i.e., they are told what to look under), thereby improving their chances for success. Table 7 shows the relationship between a pa- tron who has been referred to the catalog and . the success of his search. The chi-square test shows that the success rate for those who were referred was not significantly different (sig. = .65) than the success rate for those pa- trons not referred. Librarians, instructors, and one's own friends all might attempt to provide instruc- tion in how to use a card catalog. This in- struction might range from a formal class dis- cussion to a librarian offering hands-on instruction at the file. It might be hypothe- sized that students who have had any type of instruction in how to use the serial catalog would be more successful than students who have not had any type of instruction. Table 8 shows that there was some relationship be- tween instruction (sig. = .07), but it was very weak (phi= .09). Four factors were tested to see if any one TABLE6 WrurrEN CITATIONS AND SuccESs RATES Entry Found Citation written 246 (70% l Citation Xeroxed 13 (4% Citation not written 93 (26% Total 352 x'= .0196 dj=2 sig. = .99 Note that one cell equals 3. Entry Not Found Total 52 (69%l 298 3 (4% 16 20 (27% 113 75 427 TABLE7 SuccESs AND REFERRAL TO SERIAL REcoRD Referred Not referred Total Entry Found 142 (40%) 214 (60%) 256 x'= .1944 dj= 1 sig. = .65 Had instruction No instruction Total Entry Found 104(32%) 213 (68%) 317 x• = 3.148 df = 1 sig. = .07 phi= .09 Entry Not Found 33 (43%) 43 (57%) 76 Entry Total 175 257 432 Not Found Total 14 (21%) 53 (79%) 67 118 266 384 significantly influenced the success of a given search in the serial card catalog: (1) fre- quency of serial catalog use, (2) written cita- tions, (3) referral to the serial catalog, and ( 4) previous serial catalog instruction. None of these factors alone were significant indicators of whether a search would be successful. However, due to the interrelationships of all four factors, it is possible, for example, that a person with previous instruction was also re- ferred to the serial catalog, and due to a con- founding effect, one factor influenced the other. It is also possible that the presence of instruction or referral or previous usage var- ied greatly from one patron to the next. How- ever, through the use of partial correlation analysis, one is able to see the effect of one of these factors on the independent variable suc- cess, assuming other things are equal. . Table 9 is a matrix of partial correlations. Squaring the partial correlations (figure in parentheses) gives the proportions of the vari- ance in successful searches that can be ex- plained by each of the dependent factors. In- struction and frequency of use accounted for the most variance, 23 and 16 percent, respec- tively, while a written citation and referral to the catalog accounted for only 3 and 9 per- cent, respectively. Although no one factor significantly influenced the probable success of a search, all these factors together ac- counted for approximately 50 percent of the variance in successful searches. FAILED CATALOG SEARCHES Successful catalog searches and the factors Patron Approach to Serials I 27 TABLE9 pARTIAL CORRELATIONS Successful F~n~~~~y Search Record Use Successful search 1.00 .47 (22%) Frequency of serial .37 1.00 record use Written citation .17 Referred .31 Previous instruction .39 contributing to them present only a partial analysis of card catalog usage. The reasons for unsuccessful searches need to be explored. Three causes for an unsuccessful catalog search are (1) collection failure, where the library does not own the publication and no entry could be expected to be found in the catalog; (2) catalog failure, where the mate- rial being sought is in fact owned by the li- brary, but the catalog fails to inform the pa- tron of this fact; and (3) user failure, where the patron has an entry that is in the catalog in the proper place, but the patron has failed to find it. An analysis of the latter two types of failures offers a true diagnostic evaluation of a catalog and is needed to improve our future catalogs. Table 10 shows the number of unsuccessful searches and the three reasons for failure. An unsuccessful search was one in which the user failed to match a citation to an entry in the serial catalog. The twenty-four serials not owned by the library were checked in New Serials Titles, Ulrich s International Periodi- cals Directory, and other reference tools to determine if the patron was seeking material that in fact existed in print. Twenty titles were verified as correct in spelling and exist- ing in print. Four titles could not be found to exist in print in the form of the citation the patron presented at the serial catalog. Since libraries do not own everything in print, TABLElO uNSUCCESSFUL SERIAL RECORD SEARCHES Number Percent* Not owned by library 24 5 In serial record, 39 7 but not found In main card catalog, but not in serial record 16 6 Total 79 18 *Figures are rounded off. Written Citation .19(3%) 1.00 Referred .31 (9%) 1.00 Previous Instruction .41 (16%) 1.00 there will always be collection failures, but the relatively small level at Illinois is indica- tive of the size of the collection. The thirty-nine serials that were in these- rial catalog but were not found represented either a user or catalog failure. As table 11 demonstrates, 49 percent of the failed searches were because of user failures. The primary type of patron failure recorded was caused by the patron failing to locate a cita- tion that had in fact an exact match in the serial catalog. Nineteen undergraduate stu- dents and graduate students fell into this cat- egory. Because of the very small sample size of user failures, no detailed statistical analysis could be reasonably attempted. However, one possible contributing factor to these failed searches might have been the user's in- experience with the catalog. Sixteen of these twenty-two patrons stated that they had never before consulted the serial catalog. Pa- tron carelessness and nervousness caused by being observed might also have contributed to user failure. The second type of failure was caused by incomplete entry and accounted for only 9 percent of the searching failures. One typical example of these failures was American Hos- pital Statistics instead of Hospital Statistics. The users committing this type of failure did so because of a "sloppy citation," one in which they did not write down the complete entry. Interestingly, all five of these patrons had written down the sources of their cita- TABLE 11 pATRON AND CATALOG F AlLURES Percent of Number Failed Searches* Patron missed entry 22 40 Patron had incomplete entry 5 9 Serial record failures 28 51 *Does not include collection failure. 28 I College & Research Libraries • january 1982 tions and could have gone back to those sources to correct their citation. Errors in the serial catalog accounted for 51 percent (or eighteen) of the unsuccessful searches. Two reasons for this type of failure were misfiled cards (two cases) in the serial catalog, and serials that had a card in the main card catalog, but not in the serial cata- log (sixteen cases). Because of the number of different filers in the serial catalog and the inclusion of different filing rules at different stages of the serial catalog development, mis- takes due to filing error are inevitable. The sixteen serials found in the main card catalog but not found in the serial catalog were caused by local problems associated with the creation of the serial card catalog. Disturb- ingly, only four of the sixteen patrons indi- cated to the interviewer that they would check the main card catalog after their serial catalog search failed. Multiplicity of catalog files within a library and the failure of link- age from one to another directly caused twelve failures in locating a serial publica- tion owned by the library. CoNcLusioN The patrons sampled were successful in 84 percent of their searches. Almost 72 percent of these successful searches had no difficulty using a traditional card catalog as they matched their bibliographic citation to the correct catalog entry in the first drawer se- lected. Patron persistence by conducting sec- ond and third catalog searches resulted in the remaining 12 percent of the successful searches. This finding substantiates Rafter's conclusion that "users have a very high suc- cess rate at the catalog. "8 Implementation of title access to serials under AACR2 would mean that twenty out of forty-seven second and third searches (43 percent) would have been unnecessary. An interesting phenomenon was employed in 17 percent of the successful searches. In these searches, sixty-three patrons ap- proached the serial catalog with only an ab- breviated form of the entry they were search- ing and were still able to conduct a successful search. Since online catalogs have the poten- tial for access via an abbreviation or key- word, more study of this phenomenon is nec- essary, including how these abbreviations are derived and used. In an attempt to determine the causes for our high success rate, four factors were iso- lated and tested. Frequency of serial catalog usage, written citations, referral to the serial catalog from another source, and previous instruction were by themselves not signifi- cant indicators of whether an individual search would be successful. However, as a group these factors accounted for almost 50 percent of the variance in a successful search. There are several questions that the study did not answer. For instance, was the high success rate of the sample due to the high number of journals and magazines sought (as opposed to continuation-type entries)? Is the high rate of journal usage indicative of the type of serial searching done in most cata- logs? Do card catalog users and users of present online catalogs search for serials in a manner consistent with our sample's search- ing patterns? These questions need to be ad- dressed in future use studies to determine what factors contribute to patron success or failure in searching a catalog for a serial pub- lication. As is the case in many user studies, more questions were raised than answered by this research. With the current trend toward au- tomated catalogs in many college and univer- sity libraries, the need to know the methods of searching our present catalogs for serial publications becomes imperative. The re- search reported here was the first step in un- · derstanding our present system. More studies are needed so that the catalog of the future will be responsive to as many patron de- mands as possible. REFERENCES 1. The following are typical of these serial usage studies: Roger R. Flynn, "University of Pitts- burgh Study of Journal Usage: A Summary Re- port," Serials Librarian 4:125-33 (Fall 1979); Dianne C. Langlois and Jeanne Von Schulz, "Journal Usage Survey: Method and Applica- tions," Special Libraries 64:239-44 (May-June 1973); Maurice B. Line, "Rank Lists Based on Citations and Library Uses as Indicators of Journal Usage in Individual Libraries ," Collec- tion Management 2:313-16 (Winter 1978); Pauline A. Scales, "Citation Analysis as Indica- tors of the Use of Serials: A Comparison of Ranked Title Lists Produced by Citation Count- ing and from Use Data," Journal of Documen- tation 32:17-25 (March 1976); Edward T. Shearin, Jr . , "Journal Usage Survey," North Carolina Libraries 34:9-10 (Summer 1976); and Audrey Sylvia Tobias, "Yule Curve De- scribing Periodical Citations by Freshmen: Es- sential Tool or Abstract Frill?" Journal of Aca- demic Librarianship 1:14-16 (March 1975). 2. Stephen L. Peterson , "Patterns of Use of Period- ical Literature," College & Research Libraries 30:422-30 (Sept. 1969). 3. Marjorie E . Murfin, "The Myth of Accessibility: Patron Approach to Serials I 29 Frustration and Failure in Retrieving Periodi- cals ," Journal of Academic Librarianship 6:18 (March 1980). 4. Ruth Hafter, "The Performance of Card Cata- logs: A Review of Research," Library Research 1:202 (Fall1979). 5. Ibid. , p .203. 6. F. W. Lancaster , TheMeasurementandEvalu- ation of Library Services (Washington, D.C.: Information Resources Press , 1977), p .20 . 7. Hafter, "The Performance of Card Catalogs," p.204-5. 8. Ibid ., p .217. APPENDix I: SPECIAL Usrn STuDY Quest. # Date Time Initials 1. What is your association with the university? Freshman . . . ........ . . . .... . .. . ..... 1 Sophomore ... . ... . .... . . . .. . .... . ... 2 Junior .. . . . .... ... .. . . . . . ......... . . 3 Senior .... . ...... . . . . . .. . ........ . .. 4 Graduate student ................ . . . .. 5 Staff . . . .. .. . .. . . . . .. .. .. . .... . .... . 6 Other 7 Faculty ....... . . . . . .. .. .. . ...... . ... 8 Dept. _ ___________ _ 2. What is your major field of study? (State) _ __________ _ 3. Have you ever used this serial record before? Yes (Go to Q.4) .. . . . . . ...... . . . ...... 1 No (Go to Q.7) . .. .. .. .. .. . ... . .. . .... 2 4 . Approximately how many times do you use this serial record? Daily ........... . .. . . . .... . . . .... . .. 1 Once or twice a week . . ..... . . .. . . ..... 2 Once a month . . . . ... . . ...... . . .. . . .. . 3 Only once before ......... . ......... . .4 Other . . . . . .. . ...... .. .. ... . . .. . .... 5 5. Have you ever been shown how to use this se- rial record? Yes (Go to Q .6) . ... . . .... . . . . . . ...... 1 No (Go to Q .7) . . . ..... ... . .. · .. . .. . ... 2 6. Who gave you this explanation? (Circle all that pertain) Librarian .. . ... ..... .. ..... ......... 1 Friend .......... . . .. ... . .. . ... . .. . .. 2 Teacher . ... . ...... .. ..... ... ..... . .. 3 Other 4 7. 8. Were you referred to the serial record from another source or catalog? Yes (Go to Q.8) .. ... . . . . ..... . .. . .... 1 No (Go to Q.9) .. . . . . . ....... . ........ 2 Were you referred from (Circle all that per- tain) Circulation desk ..... . ................ 1 Reference room ....... . . . .. . ......... 2 Information desk . . . . . . .. . ...... . ..... 3 Undergraduate library .. . .... . . .. .. . . .4 Main card catalog ... . . . .... . ..... . ... 5 Shelflist ... .. . . . ........... . . .. ...... 6 Other 7 9. What are you looking for today? (Record cita- tion exactly) 1 2 3 4 10. Language of item searched? English . . . ..... .. ... . . . .. . .. . ....... 1 German . . ... . ........... . . . .... . ... 2 French ........ . . . . . . . .. . .. . .. . ..... 3 Russian . ... ...... . .................. 4 Spanish . . . ... . . ............. . .. . .. . . 5 Other 6 11. Was the citation in number 9 written down? Yes .. . . . ... . ...... .. ..... . .......... 1 Xeroxed . ............. . .......... . .. . 2 No .... .... . . ........ . . .. . . .. . ... . .. 3 12. Where did you get this citation? Class reading list . . .......... . ....... 01 Index/ Abstract . . . . ...... ....... . . . .. 02 Bibliography . . . . . . . .... . . .... .... .. 03 Bibliography in book ........... ...... 04 Bibliography in journal article ..... . .. . 05 Footnote in book . .. .. .. . .. . .. .. . . ... 06 Footnote in journal article .. . . .. . .. . . . 07 Online literature search ..... . .. . . .. . . 08 Recommended by teacher or friend ................. . . .. .... 09 LCS . . ........... . .. . . . .. . . . .. . . .. . 10 Other 11 30 I College & Research Libraries • January 1982 13. What is the name of this source? Written ................ . ........ .... 1 Xeroxed . .. ..... ..... .. . ............. 2 Orally stated . . .. . .. . ... . ...... . ..... 3 Don't remember. . .. .. . ...... . .. .. ... .4 14 . Have you ever seen or used this material before today? Yes ......... . ... . .. . .. . ... .. ... . .... 1 No .. . .. . ...... . ... . .. .............. 2 15. How did the user look up item? (List ap- proaches) Same as Number 9 ........... . .. (xref?) 1 Other entries: (xref?) 2 ___________ (xref?) 3 ____________ (xref?) 4 16. Did you find what you were looking for? Yes (GotoQ.18) .. ... . .. . . .......... . 1 No (Go to Q.17) ....... . .. . .. .. ....... 2 17. What will you do next? (Circl~ all that apply) Forget about item , abandon search .................... 1 Contact interlibrary loan ... .. . ........ 2 Ask someone for help ............. ..... 3 CheckLCS .................... . . . .. . 4 Nothing today, come back again . . .. .. .. . .. .... .... 5 Try main card catalog ..... . .... . ..... . 6 Other 7 18 . What will you do with this information? Retrieve piece . ...... . .............. . . 1 CheckLCS .. . ..... . .. . .. . .. . .. . ..... 2 Other 3 19. Have you participated in this study before? Yes .. .. - · .......... . ..... . ........ . .. . . 1 No .... . .............. . ...... . ........ . . 2 Don't remember ... . .. . .. . .. .. ........... 3 Think so . .. . .. . ..... . . .. .. .. .......... .. 4 '