College and Research Libraries LAURIE S. LINSLEY Academic Libraries in an Interlibrary Loan Network This paper presents the results of a survey of the users of the Florida Library Information Network (FLIN), Florida s statewide multi type interlibrary loan network. Responses to the survey provide a profile of the various types of li- ~rari~ that participate in FLIN. T~e paper centers upon the role of academic hbranes, the heamest users of ILL m the state. Interlibrary loan staffing, re- quest fulfillment, OCLC utilization, network performance, and other factors are d'lScussed. Examination of these aspects of interlibrary loan provides the basis for evaluation of the critical role of academic libraries in a statewide in- terlibrary loan network. INTRODUCTION The State Library of Florida recently un- dertook a project with the University of Cen- tral Florida to examine interlibrary loan (ILL) among participants in the Florida Li- brary Information Network (FLIN). The purpose of the Florida Interlibrary Loan Im- provement Project (FILIP) was to examine and analyze the ILL patterns and problems in FLIN and to make recommendations for improvements. In order to conduct this ex- amination, it was necessary to obtain a great deal of information about the nature of cur- rent statewide ILL activity. A survey was se- lected as the most appropriate vehicle to gather this information. BACKGROUND The Florida Library Information Net- w_ork is a centralized multitype ILL network that serves approximately 530 academic, public, school, and special libraries in Flor- ida. FLIN is a centralized network, begun in 1968 and headed by the State Library of Laurie S. Linsley is coordinator, Interlibrary Loan Service, University of Central Florida. Linsley served as project director of the Florida In- terlibrary Loan Improvement Project. FILIP was funded by a grant from the Library Services and Construction Act, and administered through the Division of Library Services, Department of State, State of Florida. 292 I Florida, with the four public libraries in Jacksonville, Miami, Orlando, and Tampa serving as resource centers. The backup aca- demic resource libraries are the nine state university system libraries and one indepen- dent academic library. The network was re- cently restructured, and these changes will be described later. The description of the network outlined in this paper represents the structure at the time the study was conducted. The state library is the bibliographic center of the network and handles more than 70,000 requests annually. All network users except the fourteen re- source libraries send their requests directly to the state library for processing. Requests are filled from the state library collection first. The remaining requests are then sent unveri- fied via closed-circuit teletype to each of the four public libraries, which serve as resource centers for the network. Unfilled requests of a research nature are then verified and for- warded in turn to each of the ten academic li- braries, following prescribed protocols. More than half of the libraries in Florida have book collections of fewer than 20 000 volumes. Many of these libraries hav~ ex- tremely limited verification tools. Network users are encouraged to provide basic verifi- cation whenever possible, but this is not a re- quirement. The state library makes no at- tempt to verify requests sent to the public library resource centers. Requests not sup- plied by one of these libraries are then veri- fied before being referred to one of the ten ac- ademic libraries or out of state. The public and academic resource libraries do not have verification responsibility. These libraries are required only to supply the material if possible. PREVIOUS STUDIES While numerous studies of various aspects of ILL have been conducted, there are no re- ports of interlibrary loan surveys that exactly parallel this study. There are a variety of re- ports of statewide ILL studies and network- ing activities. 1 A dissertation concerning ILL among Tennessee libraries is also notewor- thy. 2 In this dissertation, Dianna Lynne Smith examined ILL patterns among Ten- nessee's academic, public, and special li- braries. Her study assessed the volume and characteristics of the ILL requests initiated and received by these libraries in Tennessee. Stevens and Smith summarized the Smith dissertation and reported on the volume of ILL by type of library, characteristics of ma- terials requested, and success rate. 3 ILL SURVEY Each of the FLIN interlibrary loan li- braries was surveyed regarding its ILL activ- ity, and a total of 372 responded. Seventy percent of the libraries responded, a high re- sponse rate for a detailed survey of this na- ture. Responses were received from forty- four of the college and university libraries, with a 73 percent response rate. While these academic libraries collectively represent only 12 percent of the respondents, they lead the other libraries in book collection size, ILL · staffing, utilization of ILL tools, and annual mean number of ILL requests borrowed and lent. It is the purpose ot this paper to examine the unique role of these academic libraries in Florida's interlibrary loan network. All responses were categorized according to one of these nine types of libraries: college or university; community or junior college; corporate; government or other nonprofit; military base; public; school (K-12); state in- stitution (correctional or state hospital); other (special libraries such as art, hospital, museum, and various types of nonprofit li- braries). The libraries were polled about such items Academic Libraries I 293 as book collection size, interlibrary loan staffing, volume of ILL, OCLC utilization, verification tools, and FLIN performance. (The full final report, which includes a copy of the survey instrument and raw scores, is available through ERIC.) 4 In the tables, n signifies the number of libraries _ that re- sponded to a particular question. For exam- ple, "n = 341" means that 341 libraries re- sponded to that particular question. When a zero response was received, "-" is recorded instead of a percent. A zero percent ("0% ") means that the raw score was rounded to zero percent. The figures generally do not to- tal100 percent because of rounding. BooK CoLLECTION SIZE Responses to the first few questions of the survey provide a basic profile of the respon- dents. Most of the libraries (87 percent) have book collections (excluding periodicals) of under 100,000 volumes. College and univer- sity library book collections are the largest: more than half of the college and university libraries that responded to this question have book collections of more than 100,000 vol- umes. As can be seen in table 1, college and university library book collections are also the most evenly distributed, with the greatest percentage falling in the 100,000- 199,999-volume range. Only two other types of libraries report having book collec- tions in the 500,000 + -volume range, and in both types the incidence of this size collection is much lower than in college and university libraries. ILL STAFFING PATTERNS Staffing patterns vary widely among the libraries. Generally speaking, the larger the library, the greater the number of hours and personnel devoted to ILL activities. The amount of staff time devoted to ILL activi- ties is far greater in college and university li- braries than in any other type of library. The next closest type of library in terms of time spent on ILL is the public library, which spends less than one-third as much staff time on ILL activities. In college and university li- braries, librarians are devoting about half as much time to ILL as paraprofessionals. Col- lege and university libraries are the only type of library that reports a total of the equiva- lent of more than one full-time staff member 294 I College & Research Libraries· july 1982 TABLE 1 SrzE OF BooK CoLLECTIONS OF RESPONDENTS BY LIBRARY TYPE (n = 370) Coli/ C omm !Jr Book Volumes Univ Coli (Total Cases) (44) (38) Less than 12 % 16 % 20 ,000 20,000- 16 45 49,999 50,000- 16 29 99 ,999 100,000- 28 10 199,999 200,000- 19 499,999 500,000+ 9 devoted to ILL. (See table 2.) ILL BoRROWING Corp Govt (12) (43) 83 % 86 % 17 14 The survey reveals that by far the greatest level of borrowing is done by the college and university libraries and the public libraries. The form of material being requested (photo- copied articles and books) is generally what would be expected. College and university li- braries and some of the special libraries (cor- porate and government) borrow more photo- copied articles than books. Community colleges, public libraries, and school libraries, and some of the special li- braries (military, state institution, and other) borrow more books than articles: the total mean number of books borrowed is almost twice that of photocopies. The fill rate for both books and photocopies for these li- State Mil Public School Inst Other Total (6) (110) (40) (32) (47) (372) 17 % 32 % 73 % 97 % 87 % 52 % 83 32 25 9 23 21 2 3 12 8 2 7 3 2 3 4 0 2 braries is more than 80 percent. The fill rate for books requested by college and university libraries is relatively low, only 7 4 percent. This coufd be explained by the tight deadlines imposed by course require- ments in colleges and universities. Public li- brary patrons are generally willing to wait longer for materials than are college and uni- versity faculty and students. (See table 3.) The survey reveals some particularly inter- esting information about ILL borrowing from the academic libraries. Almost half (44 percent) of all the libraries report that the majority of requested materials are received from college or university libraries. This is understandable, since these libraries have the largest book collections in Florida. They also participate heavily in OCLC, so that knowledge of their holdings is more easily as- TABLE2 INTERLIBRARY LoAN STAFFING PATTERNs BY LIBRARY TYPE Coli/ Comm / State Total/ Personnel/Type Univ JrColl Corp Govt Mil Public School Inst Other Mean (Total Cases) Mean#of (44) (38) (12) (43) (6) (110) (40) (32) (47) (372) Hrs/Wk Hrs/Wk Hrs/Wk Hrs/Wk Hrs/Wk Hrs/Wk Hrs/Wk Hrs/Wk Hrs/Wk Hrs/Wk Hrs./Wk Spent on ILL By Librarian 9.2 2.2 3.3 3.5 .4 4.7 1.1 2.6 6.4 4.3 (n = 315) By Para prof. 18.7 1.4 6.0 3.3 1.8 5.0 .0 2.1 1.4 5.0 (n =275) By Clerk 10.8 1.4 .0 1.9 .3 4.9 .2 4.4 4.6 4.2 (n =279) By Other Assts . 13.5 .6 .0 .8 .0 1.0 .0 2.0 .3 2.6 (n = 272) Total Hrs/Wk 52.2 5.6 9.3 9.5 2.5 15.6 1.3 11.1 12.7 Academic Libraries I 295 TABLE3 INTERLIBRARY LoAN BoRROWING STATISTics/MosT RECENT REPoRTING YEAR BY LIBRARY TYPE Col!/ Comm / State Total / Univ Jr Coli Corp Govt Mil Public School Inst Other Mean (Total Cases) Mean#of photocopied articles borrowed (44) (38) (12) (43) (6) (110) (40) (32) (47) (372) 409 45 223 185 3 14 4 6 131 103 (n = 294) Fill rate - photocopied articles 85 % 83 % 94 % 88 % 50 % 84 % 82 % 82 % 97 % 86 % (n = 199) Mean#of bks borrowed (n = 303) 332 48 19 40 49 378 24 131 225 205 Fill rate- bks (n = 243) 74 % 86 % 91 % 84 % 97 % 78 % 82 % 76 % 91 82 % TABLE4 INTERLIBRARY LoAN BoRROWING BY TYPE OF REQuESTING LIBRARY Coli/ Comm l Supplying Libraries Univ JrColl Corp Govt (Total Cases) bt4) (38) (12) (43) Coll/Univ 1 % 76 % 67 % 43 % Commi]I Coli 5 Corp 17 Govt 8 8 40 Mil Pub 3 3 Sch State Inst. Other* 8 8 14 •The State Library of Florida was specified in most cases . certained. Figures reported in table 4 also substantiate the common belief that libraries tend to borrow from like types of libraries. ILL LENDING As anticipated, the survey reveals that the greatest level of lending is achieved by col- lege and university libraries and the public li- braries. The form of material being lent (photocopied articles versus books) is quite evenly balanced in college and university li- braries. Only two types of libraries, corpo- rate and state institution, lend more photo- copies than books, but both of these types of libraries report very low means. The mean fill rates for lending are somewhat lower than those for borrowing, although the lend- ing fill rate for books by the academic li- braries is higher than borrowing (79 percent versus 7 4 percent). (See table 5.) State Mil Public School Inst Other Total (6) (110) (40) (32) (47) (372) 25 % 17% 43% 7 % 46 % 44% 1 1 1 25 8 7 14 3 11 25 0 25 55 33 50 9 26 13 1 4 0 20 3 25 43 16 ToTAL ILL VoLUME We have been discussing ILL borrowing and lending in terms of means. This does not show the entire picture, because the forty- four college and university libraries repre- sent only 12 percent of the respondents. In terms of total volume, what types of libraries lead in interlibrary loan borrowing and lend- ing? The 110 public libraries borrow and lend more items than any other type of li- brary. The college and university libraries follow closely in second place. Together, the college and university libraries and public li- braries borrow 66 percent of the items and lend 88 percent of the items. The volume of lending is almost perfectly balanced between these two net lenders. The ratio of borrowing to lending in Florida's college and university libraries ( 1:1. 6) is comparable to that re- 296 I College & Research Libraries ·July 1982 ported by the Association of College and Re- than half of the total subsystem usage is ac- search Libraries (1: 1.5). 5 (See table 6.) counted for by college and university li- OCLC ILL SuBSYSTEM braries. A few more libraries are filling re- quests through the subsystem than are using The libraries were queried regarding the it for initiating requests (see table 7). Note use of the OCLC Interlibrary Loan Subsys- that percentages are based on number of re- tern. For both borrowing and lending, more sponses to that particular question. TABLES INTERLIBRARY LoAN LENDING STATISTicS/MosT RECENT REPORTING YEAR Coli / Comm l State Total/ Univ JrColl Corp Govt Mil Public School Inst Other Mean (Total Cases) (44) (38) (12) (43) (6) (llO) (40) (32) (47) (372) Mean no. of 637 8 7 101 2 20 7 5 87 164 photocopied articles lent (n = 168) Fill rate- 74% 73% 85% 96% 89% 78% 80% 100% 78% 80% photocopied articles lent (n = 128) Mean no. of 639 33 3 ll2 ll ll36 9 0 203 460 bkslent (n = 146) Fill rate- 79% 84% 88% 75% 54% 77% 87% 68% 78% bkslent (n = 121) TABLE6 INTERLIBRARY LoAN VoLUME/MosT REcENT REPORTING YEAR BY LIBRARY TYPE Coli/ Comml State Univ JrColl Corp Govt Mil Public School Inst Other Total (Total Cases) (44) (38) (12) (43) (6) (llO) (40) (32) (47) (372) Items borrowed 25,132 3,300 2,898 7,441 207 35,637 768 3,782 13,027 92,192 Items lent 40,184 887 74 4,ll1 53 40,356 150 31 5,908 91,754 TABLE7 UsE OF OCLC ILL SuBSYSTEM BY LIBRARY TYPE Coli / Comml State Total/ Univ JrColl Corp Govt Mil Public School Inst · Other Mean (Total Cases) (44) (38) (12) (43) (6) (llO) (40) (32) (47) (372) Send reqs. via 4% 1% 0% 1% 1% 1% 8% OCLCILL Sub. (n = 341) Have OCLC, but 3 2 0% 3 0% 2 ll do not send via Sub. (n = 341) Do Not Have 5 8 3 10 1 27 10% 8 9 81 OCLC (n = 341) Fill reqs. via 9 2 2 2 1 17 OCLCILL Sub. (n = 180) Have OCLC, but 3 5 2 2 13 do not fill via Sub. (n = 180) Do Not Have 5 7 2 16 2 19 8 3 9 71 OCLC (n = 180) Table 8 details further the use of OCLC in college and university libraries. Percentages in this table are based on responses of only this type of library and reflect the portion of libraries as a percentage of the total number of libraries (forty-four) in order to provide an even clearer picture of OCLC usage in the academic libraries. Almost a third of the aca- demics are sending their requests through the subsystem , while a few more (36 percent) are filling requests through the subsystem. A rel- atively low percentage of academics , 11 per- cent, are not filling requests received through the subsystem. College and university li- braries utilize the subsystem more than any other type of library .which has access to it. At the time the FILIP survey was con- ducted, the OCLC Interlibrary Loan Sub- system was not an integral part of the state- wide interlibrary loan network. Individual libraries sent requests or responded to re- TABLES UsE OF OCLC ILL Su BSYsTEMI'AcADEMIC LI~RARIES BY LIBRARY TYPE Send reqs. via OCLC ILL Sub. (n = 43) Have OCLC, but do not send via Sub. (n = 43) Do not have OCLC (n = 43) Fill reqs . via OCLC ILL Sub. (n = 31) Have OCLC, but do not fill via Sub . (n = 31) Do not have OCLC (n = 31) Colli Univ (44) 32 % 25 41 36 11 23 Academic Libraries I 297 quests via the subsystem as they wished, but FLIN did not utilize the subsystem for refer- rals. A major recommendation that resulted from this project was a trial test of the OCLC Interlibrary Loan System in order to deter- mine its usefulness for FLIN referrals. This will be described in the "Project Recommen- dations" section, below. FLORIDA COM CAT AND FULS A major verification tool for books is Flor- ida COMCAT. This is a microfiche listing that contains the catalog holdings of major Florida libraries. Only 7 percent of the FLIN libraries use this valuable location and verifi- cation tool. Two-thirds of the libraries do not even own Florida COMCAT. The Florida Union List of Serials (FULS) contains serials locations for many Florida libraries. Only a third of the libraries use this tool for interli- brary loan, and almost half of the libraries report that they do not own FULS. (See table 9.) Florida libraries are not taking full advan- tage of the basic verification and location tools for the state. What circumstances have led to this situation? Prior to this study, it had been assumed that the majority of FLIN us- ers owned Florida COMCAT and used it for verification and location. However, actual usage was found to be exceedingly low. Li- braries who have access to OCLC have little need for Florida COMCAT, since the same holdings information is available via OCLC. The. greatest value of Florida COM CAT is for the state's smaller libraries. This micro- TABLE9 INTERLIBRARY LOAN VERIFICATIONILOCATION TOOLS BY LIBRARY TYPE Coil / Comm / State Uni v JrColl Corp Govt Mil Public Sch ool Inst Other Total (Total Cases) (44) (38) (12) (43) (6) (110) (40) (32) (47) (372) Use Fla . COM CAT 3 % 1 % 0 % 3 % 0 % 7 % for ILL (n = 346) Have Fla. COMCAT, 4 4 2 % 0 % 8 3 % 6 % 4 28 but do not use for ILL (n = 346) Do not own Fla. 5 6 2 9 21 8 11 7 66 COM CAT (n = 346) UseFULSfor 10 7 2 5 4 4 33 ILL (n = 347) Have FULS, but 2 2 11 3 3 2 26 do not use for ILL (n = 347) Do not own FULS I 5 16 7 6 4 42 (n = 347) 298 I College & Research Libraries· july 1982 fiche can be utilized for bibliographic verifi- cation of books by title. This simple verifica- tion step makes ILL requests channeled through the State Library of Florida much more efficacious. The state library, which serves as the bibliographic center of FLIN, can then immediately forward a request to a library that owns the book. A library that has a local cooperative agreement can use Flor- ida COMCAT to find a nearby location as a borrowing source, and then send a request directly to a nearby library which owns it. The FILIP survey had two questions relat- ing to Florida COM CAT, as well as a glos- sary that provided a brief definition for each term used in the survey. A number of re- sponding libraries indicated that they had never heard of Florida COMCAT before and requested information about obtaining a copy of it. The FILIP survey itself therefore served as an educational tool. The smallest li- braries in Florida often do not have micro- fiche readers and therefore they cannot use Florida COMCAT. The groups of libraries described above comprise the Florida COM- CAT nonusers. The utilization of FULS is concerned with very different circumstances. Expressed in sheer volume, Florida COMCAT is more valuable to Florida libraries than FULS, be- cause FLIN libraries borrow twice as many books as photocopies. Since the most current edition is on microfiche, the same restrictions apply to this tool as to Florida COMCAT. The results of the FILIP survey indicate that Florida libraries are not taking full ad- vantage of these excellent regional verifica- tion and location tools. Complete utilization of these basic ILL tools could have a tremen- dous positive impact on the future success of ILL requests for FLIN libraries. FLIN PERFORMANCE The Florida libraries were asked to com- ment on the weaknesses of the network in a narrative portion of the survey. The com- plaint that was recurringly voiced by all types of librarians was summed up by one public librarian: ''They are so SLOW!" Li- brarians seemed well aware of the contin- ually growing number of requests that the state library must process with no increase in staff. Lack of communication with the net- work users was also evidenced by the respon- dents' requests for regular meetings and up- dates on network policy. While recognizing its shortcomings, most of the librarians ex- pressed gratitude for the existence of the ILL network. The majority of Florida librarians (67 percent) rate the overall performance of the Florida Library Information Network as excellent or good. A small percentage of li- brarians rate the network as adequate, fair, or poor. (See table 10.) PROJECT REcoMMENDA noNs Major recommendations of the project were: 1. The State Library of Florida will conduct a trial test of the OCLC Interlibrary Loan Subsystem for the referral of ILL requests in the Florida Library Information Net- work. The state library will analyze and evaluate the results of this test. 2. The State Library of Florida will widely publicize the availability of Florida COMCAT and encourage the utilization of Florida COMCAT as a location and verification tool. 3. The State Library of Florida will phase out the TWX and teletype linkage with the FLIN resource libraries. The state li- brary will encourage utilization of the OCLC Interlibrary Loan Subsystem for sending and receiving requests. PRESENT NETWORK STRUCTURE The State Library of Florida has made sig- nificant alterations in the Florida Library Information Network, based upon the rec- ommendations of this FILIP study, as well as the recommendations of a study conducted by RobertS. Gorin and Ronald A. Kanen of the state library staff. 6 The latter study com- pared the use of OCLC, TWX, the U.S. Postal Service, and closed-circuit teletype for the referral of FLIN interlibrary loan re- quests to the fourteen resource libraries dur- ing February and March 1981. State library verification of requests is the major modification that has been made in the ILL network. Finding a location for most book requests ensures greater efficiency and accuracy for referral of ILL requests through the network. Additionally, greater use is be- ing made of the ten academic resource li- braries. Previously, protocol required that requests for certain types of materials not be Academic Libraries I 299 TABLE 10 FLIN PERFORMANCE BY LIBRARY TYPE (n = 310) C oll i Comm !Jr Univ Coil Corp Govt (Total Cases) (44) (38) (12) (43) Excellent 1.6 % 3.5 % 1.0 % 3.2 % Good 2.6 2.9 .6 4.8 Adequate .3 .6 .3 1.0 Fair 1.3 1.9 Poor 1.0 .3 Do not use 5.5 1.6 .6 1.6 forwarded on to the academic resource li- braries. With full verification, this type of re- quest is now referred to an academic resource library whiCh owns the title, if no public li- brary resource center can supply it. 7 Based upon statistics obtained from this study, the State Library of Florida has re- tained the closed-circuit teletype linkage with the four public library resource centers. This was found to be the most effective and efficient method of communication with these libraries. The OCLC Interlibrary Loan Subsystem was chosen as the most effective method for referring requests to the ten aca- demic resource libraries, as well as to out-of- state libraries. The TWX network has been discontinued, as it was found to be the most inefficient method for referring ILL requests. 8 CoNCLUSION The Florida Interlibrary Loan Improve- ment Project, which is the first statewide in- State Mil Public School lust Other Total (6) (110) (40) (32) (47) (372) .3 % 12.6 % 5.8 % 3.9 % 1.0 % 32.9 % .6 15.2 . 1.9 3.5 1.9 34.0 3.2 .3 5.7 1.6 1.3 6.1 .3 .3 1.9 .6 1.3 1.9 .3 5.5 18.9 terlibrary loan study of this depth , provides details on ILL activity among the state's aca- demic, public, school, and special libraries. Participants in the Florida Library Informa- tion Network, Florida's centralized, multi- type ILL network, were polled about such items as ILL staffing, volume, tools, and FLIN performance. While college and uni- versity libraries collectively represent only 12 percent of the respondents, it was quite ap- parent that these forty-four libraries play a unique network role. The academic libraries have the largest book collections and the most ILL staff; their means for borrowing and lending are higher than any other type of library; they lend almost half of the total items loaned in the state; they utilize the OCLC ILL Subsystem more than any other type of library. Academic libraries, while relatively small in number , are vital members of Florida's in- terlibrary loan network. REFERENCES S 2.(7 , 1. Some of these are: Interlibrary Cooperation ; a Wisconsin Plan . Report of the Task Force on In- terlibrary Cooperation and Resource Sharing (Madison: Wisconsin Department of Public In- struction, Division of Library Services, 1976; ERIC Document Reproduction Service, ED 130 666) ; A Study of Interlibrary Loan Trans- actions in the T exas Information Exchange (Austin: Texas Information Exchange, 1976; ERIC Document Reproduction Service, ED 156190); A Study of Organization and Govern- ance of Alabama State Library Systems (Mont- gomery: Alabama Public Library Service, 1978; ERIC Document Reproduction Service, ED 163930). 2. Dianna Lynne Smith , "The Magnitude and Characteristics of Interlibrary Loan Involving Public, Academic, and Special Libraries in Tennessee" (Ph.D. dissertation , Univ. of Illi- nois , 1976) . 3. Rolland E. Stevens and Dianna Lynne Smith , "Interlibrary Loan in Tennessee, " Southeastern Librarian 27 :175-80 (Fall1977). 4. Laurie S. Linsley, Florida Interlibrary Loan Improv ement Project. Final Report (Tallahas- see: Department of State, State Library of Flor- ida , 1980; ERIC Document Reproduction Ser- vice, ED 200 209). 5. " ACRL Statistics Reported, " College & Re- search Libraries News 41:193 Guly/ Aug . 1980). 6. RobertS. Gorin and Ronald A. Kanen , Florida Library Information N etwork Project: A Com- parative Study of OCLC , TWX , U.S . Mail and Closed-Circuit T eletype (Tallahassee: Depart- ment of State, State Library of Florida, 1981). 7. Gorin and Kanen , Florida Library Information N etwork , p.31 - 32. 8. Gorin and Kanen, Florida Library Information Network , p.33.