TAKING NOTE: ASSESSING THE PERFORMANCE OF REFERENCE SERVICE IN ACADEMIC MUSIC LIBRARIES.
A PROGRESS REPORT Higher education at the turn of the century is faced with dwindling resources and an increased emphasis on accountability. Together with our counterparts in general libraries, music librarians must justify the importance of programs, collections, facilities, and staff with more than vague statements or an emotional defense of the art of music and the preservation of the context in which it was created. We must be able to answer the question "How well are you doing your job?" with measurable information. Although it is fairly easy to collect and report statistics on many aspects of our work (e.g., collection size, circulation, and gate count), it is much more difficult to measure accurately our success in one of the most important services offered by music libraries: answering reference questions. In 1994, the Reference Performance Subcommittee of the Music Library Association developed a plan to evaluate reference service in academic music libraries. [1] Although there have been numerous studies of the effectiveness of general reference departments in central academic and public libraries, [2] no separate study has been made of music reference service performance at academic branch libraries (or subordinate reference service areas in departments of centralized libraries). Reference service in academic branch libraries can vary considerably from the central library model. Branch libraries tend to rely heavily on student and paraprofessional employees [3] to refer, or even to answer, reference questions. Historically, student employees have seldom been involved in reference work in centralized libraries, and the role of paraprofessionals in reference service has also been significantly smaller. [4] In addition, reference work in music libraries may be more complicated than in other subject-specific and general libraries due to the nature of our materials (e.g., multiple physical manifestations of musical compositions in various sound recording, video recording, and score formats, and publication of works in collections and series) and how they are cataloged (using complex uniform tides and form-genre subject headings). Successful music reference work demands thorough knowledge of both the content and function of library catalogs and major reference tools in music, certainly more than can be expected of most student employees. With these characteristics of music library reference service in mind, the Reference Performance Subcommittee developed three primary research objectives: 1) to study the reference performance of student and paraprofessional employees in academic music libraries in relation to that of music librarians; 2) to compare reference performance measures of success in academic music libraries to those in general academic library reference departments; and 3) to locate the top-scoring academic music libraries so that the positive elements of their reference service procedures could be studied, described, and disseminated to the profession through publication and conference presentations. In this progress report, we begin to address the first two research objectives of our study. The third objective will be addressed more thoroughly in future publications; nevertheless, some "best practices" at individual institutions emerge in our commentary concerning the first two objectives. With funding from the Music Library Association and several universities, [5] the subcommittee began to evaluate academic music reference service areas using the Wisconsin-Ohio Reference Evaluation Program (WOREP) as its survey instrument. WOREP was originally developed in the early 1980s by Marjorie E. Murfin of the Ohio State University Libraries and Charles A. Bunge of the University of Wisconsin-Madison School of Library and Information Studies. [6] W. Michael Havener of the University of Rhode Island has since joined Marjorie Murfin in leading the study. WOREP provides libraries with a valid and reliable tool for measuring their effectiveness in answering reference questions. Libraries use a well-tested survey instrument that gathers information from both patrons and library personnel about multiple variables concerning reference transactions. [7] Each survey instrument consists of two pages, one to be completed by the patron, the second to be completed by the library employee. The form is perforated, a nd the two pages are separated at the beginning of the reference transaction. Information from both halves of the survey instrument is reunited during statistical analysis in order to provide a unified record of each reference transaction. At the end of the study, a computer-generated report of the statistical analysis is printed and sent to the administrator of each participating library. These reports form the basis for our study. Data gathered includes perceptions of question-answering success from both library employees and patrons; environmental considerations (e.g., was the library busy?); patrons' perceptions of quality of service (differentiated from question-answering success or failure); and communication difficulties. In order to encourage participation in the WOREP music study, the Music Library Association provided funding to subsidize the cost for thirty music libraries to participate in the study. As of this writing, sixteen libraries in the United States and Canada have completed the WOREP music study. This report is intended as a mid-point progress report on the project to note several trends observed in the first sets of data. The focus of this report is on the nature of music reference work, and therefore will only consider results from the thirteen libraries that collected data on a sufficiently large percentage of music-related questions. [8] Eleven of the thirteen libraries were separate branch facilities, and two libraries were subordinate music or fine arts units housed within a central library facility. The institutions surveyed varied greatly. Collections ranged in size from approximately 10,000 volumes to more than 500,000 volumes, and total students enrolled in music programs ranged from fewer than fifty to more than one thousand. All but two of the participating libraries were at institutions that offered graduate degrees in music. Most reference-question statistics cited in this article exclude patron requests for specific items answered by searching the library catalog. [9] (Any figures that include specific-item searches will be noted as such.) Although such searches in music can be extraordinarily complex, it was necessary to remove these transactions from consideration in order to compare music library statistics with corresponding figures for general reference departments in the WOREP database. The designers of WOREP based their decision to remove specific-item queries on the theory that successful performance on such questions tended to depend at least as much on the size of the library's collection as on the skill of the person providing reference service. WOREP results from 1983 to 2001 support this theory. [10] More complex specific-item requests for smaller items in larger publications (e.g., particular articles or compositions within a collection) are, however, included in all reference-transaction figures in this article. Since the participating music libraries were chosen through a volunteer and not a random or systematic selection process, they do not represent the entire population of music reference service points. These composite figures should therefore be considered as descriptive statistics and not as statistically valid predictions for values in the population of all academic music libraries. At the same time, they begin to tell a fascinating story about our work environments. COMPARISON OF MUSIC LIBRARIES Several factors appear to have influenced success rates [11] of the individual music libraries. Each of the thirteen music libraries in the study had at least one negative environmental factor with which to contend, related either to staffing, physical location or arrangement, or adequacy of its collection. Several of these factors appear to have had a strong effect and will be closely monitored as more libraries participate in the study. For purposes of comparison, the music libraries in this study are divided into two roughly equal groups in order to fall either above or below the average exact-success rate (61.36%) for all music libraries in the study. The seven libraries with exact-success rates above the average of 61.36% will subsequently be referred to as group A, and the six with rates below the average will be referred to as group B (see table 1). Neither size nor the existence of a graduate program appeared to have an impact upon the overall performance of the libraries surveyed. Libraries in group A included both the largest and smallest collection size and student enrollment. Both libraries without graduate programs were in this group. Staffing Levels The music libraries surveyed differed markedly in staff size, ranging from those with only one librarian and student employees to those staffed by several librarians and multiple paraprofessionals in addition to student employees. The libraries also varied widely in their use of non-librarians to answer reference questions. The two top-scoring libraries used student employees least often to answer reference questions (3.03% and 3.95%, respectively), although they differed considerably in their use of paraprofessionals (15.15% and 30.26%, respectively). Librarians answered more than 50% of reference questions in eight of the thirteen libraries, but this in itself did not ensure above-average success-three of these libraries fell into group B. What caused this? Other negative factors seemed to be at work in these three libraries. In one, the music collection was split between two different buildings. Another had by far the largest number of undergraduate and graduate music students enrolled and the highest percentage (65.38%) of questions perceived to be moderately difficult or difficult.'2 In the third library, all three of the primary public service employees (librarian and paraprofessional) were relatively new to the library's environment and may have been hampered by a lack of experience in their respective positions- less than two years each. By far the most consistent trend was the negative effect that using student employees in reference work had on success rates. It appears that the more that students are involved in reference service, the greater the likelihood that the library will fall into group B. The three libraries using student employees to answer the majority of their reference questions'3 fell to the bottom of group B.14 Only two of the libraries in group A used student employees to answer reference questions more than 30% of the time.[15] Neither this study nor the general WOREP study distinguishes between graduate and undergraduate student employees. A future study of success rates by educational level of student employees would provide additional insight into the reference process. Since an effective reference interview is also an essential part of any successful reference transaction, methods of training students for reference work should be an important component of future research. Subject Background of Staff Whether or not student employees were music majors appeared to have little influence upon their success rates (see table 2). For example, of the three libraries that used student employees to answer reference questions most often, the two that used music majors almost exclusively (libraries 8 and 11) received lower scores for student success than the third, which employed only a small minority of music majors (library 2). Paraprofessional success also fluctuated dramatically despite music background. [16] Experience of Staff in Their Current Positions Libraries in group A tended to be staffed by more experienced librarians and paraprofessionals. In group B libraries, employees had significantly less experience in their current positions; the few librarians and paraprofessionals who were more experienced in their positions appeared to spend less time doing reference work, at least as indicated by their job tides. Experienced librarians in group B libraries tended to be heads of libraries that were staffed by more than one librarian, and experienced paraprofessionals in this group did not hold jobs with titles that indicated positions primarily involved in public service. In group B libraries, no public service librarian had occupied a current position for more than four years. All but two librarians in this group had held their current positions for only one or two years. In group A libraries, however, only one librarian was new on the job. The rest all had more than five years of experience, and the average number of years of experience for the eleven lib rarians in this group was 9.8 years. The situation was similar with paraprofessionals. In group B libraries, no public service paraprofessional possessed more than three years of experience in the current position, while in group A libraries, only two reported less than three years of experience. Other Factors An important component of the library employees' portion of the WOREP survey consists of identifying various environmental factors that may affect the success of the reference transaction. In addition to showing the results for each factor marked, the WOREP statistical profile report contains a section that aims to synthesize these concerns under the heading "factors involved in less than successful questions." It is not surprising that when the list for each library was scanned for factors recorded in five or more reference transactions, the two highest-scoring libraries reported the fewest negative factors. Some negative aspects seemed to relate to specific elements in each library's physical configuration. Negative factors for one library with a split collection (scores and recordings housed in the music library, books housed in another building) reflected this problem. [17] In addition, libraries utilizing students to answer a high percentage of questions always exhibited more than one factor characteris tic of a lack of depth of knowledge. [18] COMPARISON WITH GENERAL LIBRARY REFERENCE DEPARTMENTS The environment in any reference area is unique to that setting and cannot be easily compared to another situation. Nevertheless, this study hopes to measure general characteristics of reference service in music settings and determine how they mirror or differ from our general-library counterparts. This report notes differences and similarities between the two settings at this point in the study. Staffing Levels Of the music libraries included in this report, none had a librarian on duty at all times the library was open. These libraries, most of which were open more than seventy hours per week, ranged from a high of the equivalent of 2.1 full-time librarians (or 2.1 FTE) involved in public service to a low of 1 FTE, with the average music library having 1.5 ETE librarians. [19] In most cases, librarians were available only during standard weekday business hours. Evening and weekend hours in these libraries were variously covered by paraprofessional and student employees, who not only serve as the "front line" for many libraries but may actually provide the bulk of available reference service at some libraries. It is interesting to note that the librarians and paraprofessionals in the music library environment outscored their general library counterparts, while student employees ranked lower (see table 3). Music libraries, however, tend to rely much more heavily upon paraprofessional and student employees to answer reference questions. At reference service areas in general libraries, patrons received help from librarians 77.3% of the time, from paraprofessionals 10.7% of the time, and from student employees 7.6% of the time. Patrons in music libraries were assisted by librarians only 41.5% of the time, by paraprofessionals 23.3% of the time, and by students 33.8% of the time. In light of our findings regarding success and patron satisfaction (see table 3), this heavy reliance on student employees may have a negative effect on the overall success of answering reference questions in music libraries. Types of Questions Asked Although data from specific-item questions are typically excluded from most WOREP analysis, it is interesting to note that the total percentage of these questions answered in music libraries was much higher than that in general libraries. In music libraries, 38.48% of all reference transactions were requests for specific items, in comparison to 13.36% of all reference questions asked in general libraries. It is easy to surmise the reasons for this contrast; as mentioned above, music presents a daunting array of physical and bibliographical manifestations with which any patron must contend. Even the most sophisticated library user may need help in locating a known item at any given time, and specific-item queries often make up the bulk of patron questions at the undergraduate level. The differences in success rates, however, did not differ greatly when specific items were removed from consideration. Patrons reported finding exactly what they were looking for 63.89% of the time when specific items were included in the set of all reference questions under analysis, and 62.25% when they were removed. Similarly, 76.16% of patrons reported that they found approximately what they were looking for when specific items were included, and 75.50% when they were removed. Closer examination of reference performance concerning questions that do not pertain to specific items reveals several significant differences between music libraries and general library reference departments. As might be expected, arts and humanities majors comprise the majority (82.99%) of music library patrons in our study, but only 24.04% of patrons in general libraries. This disciplinary focus of patron interest creates a unique balance in the types of questions asked and, subsequently, in tools used to answer those questions. For example, 12.84% of all questions in music libraries related to finding a smaller item in a larger publication (e.g., a song within a collection, or a selection on a recording) versus only 1.97% in general reference departments. More than six times as many music patrons (5.09%) approached their information need by requesting something by a particular author/composer/performer; in general libraries, this approach constituted only .75% of all questions. In contrast, factual infor mation comprised a smaller percentage of queries in music libraries: 5.33% of music patrons versus 10.4% of general patrons requested short answer facts, and 3.31% of music patrons versus 10.05% of general patrons desired facts and statistics. Interestingly, a smaller percentage of music patrons requested an explanation of a particular source (4.44% v. 11.78%), or had very broad or vague questions, such as "I would like something/anything on a topic" (5.41% v. 13.6%). [20] There was also lower interest by music patrons in the analysis of, or trends in, the discipline, as compared to those questions asked in general libraries (1.45% v. 5.61% of questions). Types of Sources Used The types of sources used to answer music questions are, of course, related not only to the types of questions asked, but also to the knowledge base of specialized employees and the physical configuration of music libraries. Questions answered by the following sources varied most dramatically from their general library counterparts (see table 4). Less frequent use of periodical indexes may be a reflection of differences in disciplinary research methods between music and disciplines in the natural or social sciences. The dramatic use of the catalog for more than 72% of all queries may again point to the difficulty patrons face as they navigate through the maze of prolific composers and languages, generic and uniform titles, and physical manifestations in which any work may be available. Other library catalogs (e.g., WorldCat: The OCLC Online Union Catalog) were used more than twice as often in music libraries as opposed to general reference departments (9.08% v. 3.54%) to answer questions. In contrast, other electronic resources, including the Web, were used less than half as often in music libraries (8.31% v. 26.25%). Two other dramatic and consistently different statistics were the music library employees' reliance on their own knowledge (32.31% v. 12.75%) and the circulating collection (13.08% v. 3.68%) when answering questions. [21] One could surmise that employees of a specialized library would naturally tend to rely more upon their own knowledge due to their depth of subject expertise as well as the more concentrated subject focus of questions. Use of the circulating collection may be a combined result of the type of queries (e.g., smaller item within a collective volume), the physical proximity to the circulating collection that most music librarians enjoy, and a more intimate knowle dge of these specialized collections. The circulating collection acts as an extension of the reference collection in many cases. Despite perceived levels of "busyness" music library employees may feel during any given day, both library employees and patrons reported that the music reference area is a less busy environment than its general library counterpart. For those patrons who reported finding exactly what they wanted, library employees related that music reference service was busy 12.25% of the time, whereas general reference employees reported being busy 25.73% of the time; patrons found a similar, although elevated, contrast, with 29.39% of music patrons finding the reference employees busy, as compared to 39.85% of patrons in general libraries. Music library employees tended to spend slightly more time per patron, though they often consulted a comparable number of sources per patron (see tables 5 and 6). CONCLUSION These findings, however incomplete, begin to paint a fascinating picture of the unique reference environment of the music library. Most of the trends identified in this article have been consistent throughout the study to date, and it will be interesting to see if they remain so as we add more libraries to the growing WOREP database. At the present time, it appears that the performance of librarians and paraprofessionals in music reference areas is slightly more effective than that at general library reference service points; in contrast, student employee success rates are not as high--a serious concern for branch libraries often staffed solely with student help, especially during busy evening and weekend hours. Greater employee experience, particularly at the librarian and paraprofessional levels, appears to be a consistently positive factor in reference success. When compared with general-reference departments, both questions and answers in music exhibit some significant differences. Music questions focus much more heavily on locating specific, known items, and music reference employees rely on their local catalogs, the circulating collection, and their own knowledge much more frequently than librarians in general reference situations. It will be interesting to observe the progress of this study as more music libraries participate in the WOREP project--remembering in the process that each of these statistics represents the dedication of human beings serving those who feel the need to come a little bit closer to the intangible magic of music. Beth Christensen is music librarian at St. Olaf College, Northfield, Minnesota; Mary Du Mont is music librarian at Rice University, Houston, Texas; and Alan Green is head of the Music and Dance Library at Ohio State University, Columbus, Ohio. (1.) Since 1994, the following individuals have contributed significantly to the work of the Reference Performance Subcommittee research group: Alan Green, principal investigator (Ohio State University), Beth Christensen (St. Olaf College). Mary Du Mont (Rice University), Allie Wise Gaudy (Western Illinois University), Lynn Cullickson (Northwestern University), Martin Jenkins (Wright State University), Ruthann McTyre (University of Iowa), Charles Reynolds (University of Michigan) Tracey Rudnick (University of Connecticut), Alicia Snee (California State University, Sacramento), and Philip Vandermeer (University of Maryland, College Park). (2.) For an overview of the extensive research in reference measurement and evaluation, see Marjorie E. Murfin and Lubomyr R. Wynar, Reference Service: An Annotated Bibliographic Guide (Littleton, Colo.: Libraries Unlimited, 1977), 188-213; Marjorie E. Murfin and Lubomyr R. Wynar, Reference Service: An Annotated Bibliographic Guide, Supplement 1976-1982 (Littleton, Colo.: Libraries Unlimited, 1984), 189-215; and Frances Benham, "Reference Effectiveness," in The Reference Assessment Manual (Ann Arbor, Mich.: Pierian Press, 1995), 99-112. (3.) The term "paraprofessional" will be used throughout this article to describe employees holding regular, non-student, non-librarian positions. When paraprofessional and student employees are combined as a class distinct from librarians, the term "non-librarians" will be used. (4.) It should be noted, however, that a recent study of longitudinal data from the Wisconsin-Ohio Reference Evaluation Program (WOREP) shows a marked increase in the percentage of overall transactions handled by non-librarian employees, from 11.63% in 1983 to 32.14% in 1997. See W. Michael Havener and Marjorie E. Murfin, "Cronbach Revisited: Positive Bias, A Powerful Enemy to Validity in Library Services," in Proceedings of the 2nd Northumbria International Conference on Performance Measurement in Libraries and Information Services, held at Longhirst Management and Training Centre, Longhirst Hall, Northumberland, England, 7 to 11 September 1997 (Newcastle upon Tyne, England: Information North for the Department of Information and Library Management, University of Northumbria at Newcastle, 1998), 148-49. (5.) Ohio State University, the University of Michigan, Kent State University, and Baylor University have contributed funding for the project. (6.) Charles Albert Bunge, "Factors Related to Reference Question Answering Success: The Development of a Data-Gathering Form," RQ 24 (summer 1985): 482-86. See also Marjorie E. Murfin, "Evaluation of Reference Service by User Report of Success." The Reference Librarian, nos. 49-50 (1995): 229-41. (7.) The original WOREP survey instrument, designed for use in general reference departments. has been modified slightly for this study to measure more precisely activities specific to music libraries. See the appendix for a copy of the music survey instrument. The original WOREP survey instrument and documentation of its testing for validity and reliability are published in Marjorie E. Murfin and Gary M. Gugelchuk, "Development and Testing of a Reference Transaction Assessment Instrument," College & Research Libraries 48 (July 1987): 314-38. (8.) This report includes only those libraries that collected data on at least one hundred reference transactions and in which more than 75% of the questions asked were music related. It considers results from 1,335 matched sets of patron and library-employee survey forms (a 91.93% return rate). Because the database of participating libraries is currently small, it is difficult to provide collection and staffing details without jeopardizing the anonymity of these libraries; however, all participating libraries are listed on our Web site, "MLA/WOREP Music Reference Performance Study," CIC Music Libraries, 22 January 1999, http://www.cic.uiuc.edu/music/mla.ref/index.html/ (accessed 22 April 2001). (9.) For the class of questions excluded from most WOREP reference analysis, see the library-employee portion of the survey instrument, box 1, item A-1, in the appendix. (10.) According to Marjorie E. Murfin, "Since the purpose of the WOREP was to provide an assessment of the skill of reference staff in assisting patrons to find information, it was decided to measure the two categories [i.e., specific item transactions and requests for subject information] separately. Analysis would then be concentrated on requests for subject information which would be a more accurate reflection of staff skill. It would be a measure less subject to the influence of collection size [than if specific item requests were included in our calculations). This has proved to be the case as the data below show. Thirty-four small libraries (under 500,000 volumes) had a success rate of 55.7% on specific title questions and 47 large libraries (one million volumes plus) had a success rate of 61.0%. In contrast, success rates on requests for subject information were 58.0% and 56.7% respectively and clearly were not dependent on collection size." Marjorie E. Murfin, letter to Alan Green, 6 April 2001. (11.) Two different categories of success rates are used in this article. "Exact-success rate" refers to the percentage of patrons who reported finding exactly what was wanted ("Yes, just what I wanted") and were satisfied. See questions 1, 3, and 4 on the patron portion of the survey instrument in the appendix. "Approximate success" includes transactions where neither the library employee nor the patron marked "no" or "not found" for the success and satisfaction variables. (12.) See box 6 of the library-employee portion of the survey instrument in the appendix. (13.) 73.08%, 71%, and 69.57% of questions. (14.) In these three libraries, paraprofessional employees were minimally involved in answering reference questions-a possible additional contributing factor to lower success rates. (15.) These two libraries had '40.32% and 36.51% of questions handled by students. (16.) Of the music libraries studied, 89.5% of all music librarians and 37.5% of all paraprofessional music library employees had graduate degrees in music; 62.5% of all paraprofessionals had at least a bachelor's degree in music. Unfortunately, comparisons with general reference department counterparts are not possible since the general WOREP study does not collect information about either the educational backgrounds or the length of employment experience (see next section) of staff. (17.) Library employees' comments included "collection weak," "one source used," 'need books in another area" and "patron says 'not enough.'" (18.) Library employees' comments included "one source used," "time spent = 0-3 minutes," and 'direct and suggest only." (19.) This number reflects the total FTE of librarians whose primary responsibility is reference service. It does not distinguish between Lime at a separate reference desk and time simply "on duty" within a music library, since conditions vary widely among libraries: similarly, it does not account for times, such as meetings, when the public service music librarian is inevitably pulled away from the music library reference area. (20.) See box 1, item D2a on the employee portion of the survey in the appendix. (21.) The use of one's own knowledge to answer questions was most dramatic when comparing levels of library employees at both music and general libraries. Music librarians used their own knowledge 31.76% of the time versus 12.81% in general reference departments; paraprofessionals in music libraries reported using their own knowledge 33.06% versus 9.14% of general-library paraprofessionals; and students in music libraries relied on their own knowledge 32.22% versus 17.9% in genera! libraries. TABLE 1. Libraries' exact-success rates (%) Group A 1. 86.36 2. 71.05 3. 69.12 4. 68.25 5. 63.08 6. 62.90 7. 61.73 Group B 8. 58.70 9. 56.92 10. 53.85 11. 52.56 12. 52.46 13. 40.68 TABLE 2. Subject background of student and paraprofessional staff Student # of Students Library * exact success (%) majoring in music 1. (Group A [down arrow]) 54.55 7/14 2. 50 7/8 ** 3. 100 4/10 4. 58.82 9/15 5. 65.22 19/19 6. 61.54 0/11 7. 62.5 1/9 8. (Group B [down arrow]) 54.84 9/10 9. 20 5/9 10. 62.5 17/18 11. 60 7/7 12. 63.64 2/11 13. 83.33 2/2 # of Graduate Library * students out of total 1. (Group A [down arrow]) 2 2. 1 3. 0 4. 3 5. 0 6. 0 7. 0 8. (Group B [down arrow]) 3 9. 0 10. 8 11. 1 12. 2 13. 2 Paraprofessional # of Paraprofessional Library * exact success (%) with music background 1. (Group A [down arrow]) 50 3/3 2. 77.78 0/4 3. 78.26 0/1 4. 81.25 1/2 5. 87.5 2/2 6. 73.33 1/1 7. 60 0/1 8. (Group B [down arrow]) 63.64 2/2 9. N/A N/A 10. 53.85 2/3 11. 22.22 0/1 12. 0 0/1 13. 21.05 5/5 (*)In both tables, libraries are again ranked by overall success rates, as given in table 1. (**)7 out of 8 total. TABLE 3. Comparison of success and patron satisfaction by type of library employee (%) * Found exactly what was wanted and were satisfied: Professional music librarians 65.67 General librarians 60.78 Music library paraprofessionals 59.40 General library paraprofessionals 58.90 Music library student staff 56.22 General library student staff 61.48 * Found approximately what was wanted and were satisfied: Professional music librarians 79.29 General librarians 71.84 Music library paraprofessionals 74.44 General library paraprofessionals 69.91 Music library student staff 68.11 General library student staff 72.30 TABLE 4. Types of Sources Used (%) General reference Music libraries departments Periodical indexes 12.31 28.48 Reference books 37.23 41.42 Local library catalog 72.92 34.00 Other library catalog (OCLC, etc.) 9.08 3.54 Other electronic resource (Web, CD-ROM) 8.31 26.25 Own knowledge 32.31 12.75 In-house tools 5.08 6.96 Vertical files .92 1.37 Circulating books, periodicals, etc. 13.08 3.68 TABLE 5. Employee estimate of time spent per patron (%) Music libraries General reference departments 0-3 minutes Students 44.81 42.43 Paraprofessionals 45.90 44.05 Librarians 28.19 33.55 All employees 36.42 35.65 3-5 minutes Students 33.88 37.25 Paraprofessionals 35.25 36.35 Librarians 30.56 42.63 All employees 32.56 41.58 5-15 minutes Students 18.58 18.53 Paraprofessionals 13.93 17.43 Librarians 32.05 21.23 All employees 24.54 20.36 more than 15 minutes Students 3.28 1.99 Paraprofessionals 4.92 2.84 Librarians 10.09 3.11 All employees 7.10 2.91 TABLE 6. Number of sources consulted by percentage of transactions Music libraries General reference departments 1 source Students 55.77 59.72 Paraprofessionals 44.44 55.50 Librarians 30.72 42.40 All employees 40.49 45.13 2 sources Students 26.28 24.82 Paraprofessionals 31.48 22.96 Librarians 26.47 30.19 All employees 27.23 28.89 3 sources Students 8.97 10.77 Paraprofessionals 19.44 15.25 Librarians 19.61 14.14 All employees 16.58 14.15 4 sources Students 4.49 3.04 Paraprofessionals 1.85 3.30 Librarians 8.82 6.20 All employees 6.28 5.63 5 sources Students 4.49 1.64 Paraprofessionals 2.78 2.99 Librarians 14.38 7.07 All employees 9.42 6.19 |