College and Research Libraries Survey of Online Systems in U.S. Academic Libraries John A. Camp, Grace Agnew, Christina Landram, Jane Richards, and Judith M. Shelton A survey of online systems in U.S. academic libraries was conducted to gather information on acquisitions, serials, circulation, cataloging, interlibrary loan, and integrated systems. Libraries reported on present systems and future plans, methods of financing, and use of systems personnel, as well as backup systems, type of computer, source for system, and functions within systems. Survey results indicate that 15% of libraries have no online systems and that 16.2% plan no additional ones. Cataloging and interlibrary loan units are the most frequently automated. II he predominance of automa- tion as a theme in the library lit- erature attests to the fact that li- braries are turning increasingly to online systems for a variety of func- tions. Joseph R. Matthews documents this in his annual review of the automated li- brary systems marketplace (see table 1). This pattern of growth promises to con- tinue. A factual presentation of current instal- lations and future plans was the goal of this survey. It was limited to online sys- tems. A state-of-the-art review of online systems was also deemed to be important to supply a historical perspective. In view of the bewildering variety of sys- tems and the time and expense of imple- menting them, it would seem to be desir- able for libraries to learn about the choices made by comparable institutions. How- ever, the magnitude of the project pre- cluded an investigation into the more sub- jective motives and factors influencing libraries in their choice of systems. METHODOLOGY AND HYPOTHESES ~ A search of Library Literature and the ERIC and LISA databases led to numerous studies of particular systems or reports of the experiences of individual institutions. We discovered several state surveys, but they did not provide the depth we were seeking. One recent survey was limited to members of the Association of Research Libraries. 1 The present study concerns the extent of online automation in libraries of four-year academic institutions in the United States. Several general hypotheses provided the basis for a questionnaire: 1. Significant online automation is in place in academic libraries. 2. Where online automation does not exist, it is anticipated. 3. Large libraries are more likely than small libraries to be automated. 4. Systems developed in-house will probably be found in large libraries. 5. Funding may come from a variety of sources, but the regular library budget is not a major source. 6. Participation in bibliographic utilities is extensive. An 83-item questionnaire was designed to test these hypotheses. The question- John A. Camp, Grace Agnew, Christina Landram, Jane Richards, and Judith M. Shelton are at the Pullen Library Catalog Department, Georgia State University, Atlanta, Georgia 30303-3081. 339 340 1982 1983 1984 1985 College & Research Libraries TABLE 1 TURNKEY AUTOMATED LIBRARY SYSTEMS New Installations 81 99 167 235 Number of Worldwide Installation 365 475 769 1109 Sources: Joseph R. Matthews, Library Journal , Mar. 1983, May 1984, Apr. 1985, Apr. 1986. naire solicited general information (library size, number of monographic titles ac- quired annually, whether the responding institution belonged to a bibliographic utility, and whether a systems specialist was employed or available) and informa- tion on seven specific online applications: acquisitions, union list of serials, serials control, interlibrary loan, circulation, machine-readable cataloging, and online public access catalogs. For each of these applications, standard comparison infor- mation was requested, such as method of funding, type of computer on which the system is run, type of system (vendor, bib- liographic utility, developed in-house, etc.). This included specific questions about individual systems; for example, the extent of interlibrary loan activity (both borrowing and lending), methods of access for online public catalogs, and spe- cific features of acquisitions systems. Fi- nally, the questionnaire sought informa- tion on integrated systems and their capabilities, including how long systems had been in place, or if not yet imple- mented, the projected timetable for instal- lation. This questionnaire was first tested by surveying online automation in the state of Georgia. Survey results were published in the Georgia Librarian. 2 Although the ba- sic content of the questionnaire remained the same, minor changes in wording or question sequence were made after the Georgia study. The sample population was taken from a Bowker list of academic institutions and was limited to 300 four-year institutions, excluding junior colleges and technical schools. The questionnaire was adminis- tered through a mailing in June and Sep- July 1987 tember 1985. A full 73.7% (221) were re- turned, and 70% of the total sent (210) were determined to be usable . The data collected were tabulated using the fre- quency and cross-tab procedures of the Statistical Package for the Social Sciences, Version 10 (SPSS-X). The significance of relationships in the cross tabs was mea- sured using the chi-square statistic. Statis- tical significance is indicated by the equa- tions P< .001, P< .01, and P< .05, with P < .001 demonstrating the highest degree of significance. 11 One of our basic hypotheses was that size would influence the online automation decisions made by li- braries.'' The survey was designed to reveal the extent of online automation in academic li- braries and to determine the direction aca- demic libraries are taking with respect to online automation. RESPONDENTS' CHARACTERISTICS One of our basic hypotheses was that size would influence the online automa- tion decisions made by libraries. Of the 210 libraries responding, 65.2% (137) are small libraries with 250,000 volumes or fewer; another 26.2% (55) are classed as medium-sized libraries with between 250,000 and 1 million volumes; and 8.6% (18) are large libraries with more than 1 million volumes (see table 2). TABLE2 SIZE OF LIBRARIES RESPONDING TO THE SURVEY BY VOLUMES IN THE COLLECTION (n=210) No . % Small (0- 250,000 volumes) 137 65.2 Medium(250,001-1,000,000volumes) 55 26.2 Large (More than 1,000,000 volumes) 18 8.6 All of the small and nearly all (52) of the medium-sized libraries acquire fewer than 25,000 volumes a year while 3 medium and 15 large libraries acquire more than 25,000 (P < .001). Seven respondents re- Survey of Online Systems 341 TABLE 3 CHARACTERISTICS OF LffiRARIES REPORTING ONLINE SYSTEMS, FOR SELECTED APPLICATIONS Acquisitions (n = 41) no . % Serials Control (n = 26) no . % OPAC (n = 25) no. % Circulation (n = 40) no. % Volumes in the collection 0-250,000 250,001-1,000,000 More than 1,000,000 (Missing observations) Annual monographic acquisitions 0-5,000 5,001-25,000 . 25,001-40,000 More than 40,000 (Missing observations) Serial-type subscriptions 0-5,000 5,001-15,000 More than 15,000 (Missing observations) Annual Circulations 0-50,000 50,001-200,000 200,001-500,000 More than 500,000 (Missing observations) 13 17 10 (1) 16 12 5 7 (1) Note : Percentages may not equal 100% due to rounding. port annual acquisitions in excess of 40,000 titles (see table 3) . More than 80% of the responding li- braries (166) use a bibliographic utility, and 92.9% (156) of these use OCLC . All of the large and 53 of the medium-sized li- braries use a bibliographic utility. Al- though a majority of small libraries do use one, 29.6% (40of135) donot(P< .001) (see table 4). TABLE4 BffiLIOGRAPHIC UTILITY MEMBERSHIP OF LIBRARIES RESPONDING TO THE SURVEY (n = 210) no. % None 42 20.0 OCLC 156 74.3 RLIN 5 2.4 WLN 5 2.4 More than 1 2 1.0 Note : Percentages may not equal 100% due to rounding. Sixty-two responding libraries report having a systems specialist. In 54 (85.7%). of these institutions, this specialist is a member of the library staff. As the size of the library increases, so does the probabil- 31.7 41.5 24.4 39.0 29 .3 12.2 17.1 11 11 4 18 5 2 (1) 42.3 42.3 15.4 69.2 19.2 7.7 2 8.0 18 72.0 5 20.0 6 24.0 13 52.0 3 12.0 3 12.0 10 19 11 13 12 7 6 (2) 25 .0 47 .5 27 .5 32.5 30.0 17.5 15.0 ity of having a systems specialist, ranging from 13.4% (18 of 134) for small libraries to 88.9% (16 of 18) for large libraries (P< .001) . ACQUISITIONS Of the 209 libraries responding to this set of questions, 169 (80.9%) do not have an online acquisitions system. Of the 41li- braries who do (19.5%), 20 use a system supplied through a bibliographic utility. Since 156 of 168 libraries responding to a question on utility membership belong to OCLC, it is not surprising that 75% of those 20 libraries use a system supplied by OCLC. The next most prevalent system is a vendor turnkey system, used by ten li- braries (25%), followed closely by eight li- braries (20%), using a system developed in-house. A greater percentage of large libraries than small ones have an online acquisi- tions system. Of libraries with more than 250,000 volumes, 38% (28 of 73) utilize this type of automation while only 9.6% (13 of 136) of those with fewer than 250,000 have one (P < .001). This is corroborated by a 342 College & Research Libraries comparison based on the number of titles acquired each year. A full100% of those li- braries acquiring more than 40,000 titles per year have an online acquisitions sys- tem while only 11.9% (16 of 135) acquiring 5,000 or fewer titles per year do (P < .001). The most popular type of online acquisi- tions system among large libraries is a vendor turnkey system (50%, or 5 of 10). The most prevalent type in smaller li- braries, however, is an in-house system (42.9% or 6 of 14) (P < .05). Since the largest number (20) of systems are supplied through a bibliographic util- ity, it is not surprising that most online ac- quisitions systems are run on mainframe computers (46.3%, or 19 of 41). Ten micro- computer systems as well as 10 minicom- puter systems were reported. These are most often vendor turnkey systems. In- house systems are most often found de- signed to use mainframes. Regardless of library size or number of titles added annually, the majority of sys- tems (23, or 57.5%) are funded from the regular library budget rather than from special allocation by the institution (5, or 12.5% ), grants (2, or 5% ), the institution's computer center budget (4, or 10%), or a combination of these (6, or 15%). All libraries but one (97.5%) order mate- rials through the system, and this one plans to add this function. In-process con- trol, accounting, and generation of reports are next most common functions of online systems. Thirty-four libraries (85%) cur- rently have these capabilities. Twenty- three libraries (57.5%) have claiming capa- bility online and the same number have , vendor control. Only 12libraries (25%) re- port online interface with vendors, and only 7 systems are expected to add this feature. The least common module is binding, which 5 of 40 (12.5%) report hav- ing. Only 3 institutions plan to add this function to their acquisitions system. If current plans are implemented, the relative position among functions will re- main the same. Perhaps the most notable discovery is that 4 libraries (10%) have no plans to implement an accounting func- tion, and Slibraries (12.5%) do not plan for the generation of reports. Twenty-two of 41 respondents (53.7%) July 1987 maintain no backup for their system. Of those that do, a,backup in hard copy is the most common format (12, or 29.3%). An- other two systems are backed up using mi- crocomputer diskettes. One uses micro- form and 3 use a combination of the above. Apparently, the more titles a li- brary acquires, the less likely it is to have a backup system. Percentages of those without backup range from 11.8% (those ordering 5,000 or fewer titles) to 100% (those ordering over 40,000) (P< .05). ''Ofthose libraries without an online acquisitions system, only 20 (12.1°/o) definitely do not plan to implement one.'' Only 5 libraries (12.5%) have had their online acquisitions systems for more than five years. Twenty-three libraries (57.5%) have had their systems for one to five years and 12 (30%) report having online acquisitions for less than a year. Libraries with in-house systems (3 of 7, or 42.9%; P < .05) are more likely to have had their systems in place for more than five years, and since, as noted above, smaller li- braries are more likely to have in-house systems, it could be assumed that it is the smaller libraries that have had online ac- quisitions systems for the longest period. However, that statistic showed no signifi- cance (P= .4482). Of those libraries without an online ac- quisitions system, only 20 (12.1%) defi- nitely do not plan to implement one. All the largest libraries do plan to have one, and more medium (31 of 38 or 81.8%) than small libraries (48 of 119 or 38.7%) have such plans. (P< .001) Seventy-eight of 87 libraries (89.6%) planning to add such a system will do so within five years. SERIALS CONTROL SYSTEM Only 26 (12.5%) of the 208 responding li- braries have online serials control systems and, surprisingly, no single type predomi- nates. Six (23.1%) are supplied through a commercial service (such as Faxon), 7 sys- terns (26.9%) were developed in-house, and 8 (30.8%) are supplied through a bib- liographic utility. Three systems (11.5%) are vendor turnkey and 2 (7. 7%) are com- bination vendor/in-house. Only 3 of 25 (12%) online serials control systems have been in use for more than five years; 13 (52%) have been in use between one and five years; 9 (36%) were implemented dur- ing the last year. Funding for 19 of the 26 systems (73.1%) came from the library budget. Four (15.4%) libraries reported funding through the computer center budget and 2 (7. 7%) through special allocation from the parent institution. One respondent listed a combination of these sources. The fact that 17 of the 26 ( 65.4%) are run on main- frame computers is not surprising, be- cause in 9 cases (34.6%) a computer is shared with the parent institution, and at 13 institutions the computer is provided by a bibliographic utility, commercial ser- vice, or vendor. Only 4 systems (15.4%) run on computers dedicated solely to the library. Of the 9 system features investigated, none is functional in all 26 of the libraries with online serials control systems. The three most common are check-in (22, or 84.6%}, claiming (19, or 73.1%) and report generation (18, or 69.2%). The other 6 functions are common to slightly more than one-third of the libraries responding: 11libraries report accounting and routing; 10 have ordering and binding modules; 9 have vendor control; and 8 have an online interface with vendors. Even plans for fu- ture development will not result in a com- mon feature for serials control for all 26 li- braries. Even the check-in module was excluded by at least one library from plans for the future as listed. If implementation plans are carried out as reported, the cur- rent three most common functions will continue to be prevalent, in the same or- der: check-in (25 or 96.2%); claiming (22 or 84.6%); and report generation (19 or 73.1%). Serials holdings information is available online to the public in only 8 of the 26 sys- tems (30.8%), but24libraries (92.3%) issue computer-produced serials holdings lists. When the system is down, 11 libraries Survey of Online Systems 343 (42.3%) provide no backup, and of those that do, hard copy is the backup of choice (11). UNION LIST OF SERIALS Seventy-three of the 204 responding li- braries (35.8%) participate in an online union list of serials. More than half ( 44, or 60.3%) of these systems are provided by a bibliographic utility, and vendors provide only 4 (5.5%) of those reported. The regu- lar library budget was the largest single source of funding (22), although 21 li- braries do use some combination of regu- lar budget, special allocation, and/or out- side grant. The most common type of computer cited for union lists is a mainframe (39 of 69). A significant 20 respondents did not know what type of computer is used by their institution. In only 8 of the 58 re- sponses to the question of ownership did the computer belong to just one member. In a majority of cases (41 of 68, or 60.3%) the machine-readable record could be in- put or updated by the local institution. Hard-copy backup outnumbers micro- form two to one (30 to 15) and 9 of 69 sys- tems (13%) use a combination of these two. Nine systems provide no backup. MACHINE-READABLE RECORDS More respondents use online systems for current cataloging than for any other online activity. Just over three-fourths of the respondents, (162 of 209, or 77.55%) have current cataloging records in machine-readable form: the remaining 22.5% do not. Of those who do, 147 (90%) utilize a bibliographic utility for this func- tion. While it is rare for libraries to have all cataloging records in machine-readable form (5 of 190, or 2.6%), it is noteworthy and somewhat surprising that a consider- able number of respondents (25 of 190 or 13.2%) have no cataloging records at all in machine-readable form. Bibliographic utilities still have growth prospects! Current or planned conversion of rec- ords into machine-readable form was re- ported by 150 (76.1%) of 197 respondents. The preferred method of conversion is. through a bibliographic utility (96 of 161 respondents, or 59.6%) and the conver- 344 College & Research Libraries sion will most often be done by in-house personnel. Nearly a quarter of the respon- dents (22.4%) do not know how the con- version will be done, and only 4 libraries (2.5%) are planning to use a vendor. "Only 25 of the _208 libraries re- sponded affirmatively to the ques- tion of whether they have an online public access catalog." There is a distinct positive correlation between the size of a library, as measured by the number of volumes already in the collection (P < .001) or by the number of volumes added annually (P < .001), and whether any of its records are in machine- readable form. When size is defined by the number of volumes in the collection, 66.9% (91 of 137) of small libraries have machine-readable records, while 96.4% (53 of 55) of medium libraries do. All 18 large libraries have records in this form. When size is defined by annual acquisi- tions, the same pattern emerges. Among the 134 libraries adding 5,000 or fewer ti- tles, 91 (67.9%) have machine readable records. For the 57 adding 5,000 to 25,000 titles the percentages rise to 93 (or 53 li- braries), and all18 libraries adding more than 25,000 titles have them. ONLINE PUBLIC ACCESS CATALOGS (OPACs) Only 25 of the 208 libraries responded affirmatively to the question of whether they have an online public access catalog. This means that only slightly over 12% have automated in this way. However, 119 of 183 libraries (65%) currently with- out an OP AC plan to implement one, and the percentage rises in direct proportion to the percentage of machine-readable rec- ords they possess. Two of 42 respondents (4.8%) with 76%-99% of their cataloging records in machine-readable form do not plan an on- line catalog; nor does there seem to be any relationship between the percentage of July 1987 records in machine-readable form and having an OPAC (P= .1125). The fact that only one of the 5 libraries with 100% of their records in machine-readable form actually has an online catalog may simply illustrate that online catalogs are relatively new and still in the planning stage. Vendor turnkey systems are the most prevalent type of catalog (67%, or 16 of 25) followed, in order, by a combination of types, a bibliographic utility or II other, II and in-house systems. Half of these cata- logs have been in place less than a year and only 1 respondent has had an online public access catalog for more than five years. Funding for the online catalog was ac- complished most often through a special allocation from the parent institution. This was true for 11 of 25 respondents (44.0%). The next most common funding arrange- ment was through the regular library budget (6, or 24%), followed by outside grants (4, or 16%), and a combination of sources (4, or 16%). Minicomputers and mainframe systems are used equally for library automation (12 of 25, or 48% for each); only one library has a micro-based system. Thirteen of 25 computers (52%) are dedicated to library use only while 6 computers are shared with parent institutions and 6 computers are provided by vendors. Hard copy is the backup of choice in 11 of 25 cases (44%), followed by 6 respondents who use micro- form (24%), 1 who utilizes microcomputer diskette (4%), and 1 only reporting ''other.'' The most prevalent capability of online public access catalogs is the display of holdings data. A full 92% (23 of 25) have this in place, and the remaining two li- braries plan to implement this feature. Three times as many libraries have online catalog maintenance as have authority control (19 versus 6), although 84% (21 of 25) of the libraries with online catalogs re- port that they will eventually have these functions available. Seventy-six percent of the libraries (19 of 25) either currently have or plan to have cross-references visible to the public, online interface with a biblio- graphic utility, and the capability of gener- ~ I l I 1 ating reports from the system. Cross- references invisible to the public and set logic are or will be implemented in fewer than half the systems: 10 libraries ( 40%) ei- ther have or plan to have the former, while 11 of 24 (45.8%) will eventually possess the latter. Of the 25 responding libraries, only one has all8 systems capabilities and only one has as few as two. Eighteen (72%) have at least 5 capabilities. Libraries were queried about 8 search capabilities possible with an online cata- log: author, title, author/title, subject, keyword, Boolean operators, call num- bers, and other identification numbers (ISBN, LC card number, OCLC record number, etc.). Nine of the 25 libraries (36%) with online catalogs have all 8 search capabilities, while 4 (16%) have only 4 of these features. The other 12 li- braries range between these extremes. All online catalogs search by author and title, while 19 of the 25 (76%) allow users a com- bination author/title search. Twenty (80%) can search by call number and by other identification numbers. Of 183 libraries that currently do not possess an online public access catalog, 119 (65%) plan to acquire one. Another 43 (24.5%) are undecided; and only 21 off the 183 (11.5%) without an online public ac- cess catalog state that they do not plan to have one. Seventy-six of 120 libraries (63.3%) plan to acquire their catalog before five years have elapsed. INTERLIBRARY LOAN (ILL) Of 208 responding libraries, 159 (76.4%) identify potential lenders of interlibrary loan material through an online system. When asked to specify the type of online system used, 6 of 181 respondents (3.3%) indicated a vendor-supplied system, 29 (16%) a library consortium, and 146 (80.7%) a bibliographic utility. Because only 159 respondents stated that they use an online system, the 181 responses to this question indicate that some libraries use more than one online source to identify lenders. Of 170 responding libraries, 142 (83.5%) actually handle interlibrary trans- actions online (borrowing and lending) and again the large majority (131 of 159, or Survey of Online Systems 345 82.4%) specify bibliographic utilities as the means, although more than one online system may be used. The use of an online ILL system increases with the number of items borro~ed annually, but more signif- icantly with the number loaned (P < .05). Hard-copy backup for ILL is used by 137 libraries (83.5%), but 12 libraries (7.3%) have no backup at all. "Nearly one-fifth of the 208 respond- ing libraries (19.2°/o) have an online circulation system. 11 Of 64 libraries without an online interli- brary loan service, 21 (32.8%) plan to im- plement such a system, 20 (31.2%) do not, and 23 (35.9%) are undecided. Fifteen of 22 libraries (68.2%) with such plans will fulfill them within five years. CIRCULATION Nearly one-fifth of the 208 responding libraries (19.2%) have an online circulation system. More than three-fourths of these (77 .5%) have either a vendor turnkey sys- tem (52.5%) or a system developed in- house (25%). Whether a library has an on- line circulation system appears to be directly related to the size of the library. If one measures size in terms of volumes in the collection, 7.4% (10 of 135) of there- sponding small libraries, 34.5% (19 of 55) of the medium, and 61.1% (11 of 18) of the large libraries have online circulation sys- tems (P< .001). Considering the number of items circulated, the percentages range from 10.2% of the responding libraries cir- culating less than 50,000 annually (13 of 128) to 75% (6 of 8) of those circulating more than 500,000 (P < .001). As with online public access catalogs, the greatest single source of funding for online circulation systems is a special allo- cation from the parent institution (17 of 40, or 42.5%). Nine (22.5%) found the money for the system in the regular library budget while another 9 employed a com- bination of funding sources. 346 College & Research Libraries Only 39 libraries responded to a ques- tion on the level of computer used to run the system and more than half (23 or 58.9%) reported using a minicomput~r. Fourteen libraries (35.9%) use a mam- frame system, and only 2 libraries (5%) use a microcomputer. Whether the com- puter is reserved exclusivel:r for the use of the library is related to the SIZe of the com- puter (P < .01) and to the type of circula- tion system in use (P < .01). Thus, bo~h of the microcomputers are used exclusively by the library but only 3 of 14 (~1.4%) mainframes are reserved for the hbrary. Eight of 10 (80%) libraries with systems developed in-house share the computer with the parent institution while 17 of 20 · libraries (85%) with vendor turnkey sys- tems have exclusive use of the computer (P< .01). In response to a question on circulation system capabilities, all but one of the 40 re- sponding libraries (97.5%) currently have a charge-out/charge-in function and can issue overdue notices to patrons. These features will eventually be operational in all systems described. Holds and delin- quent patron status are both currently functional in 35 systems (87.5%); one more library plans to add the latt~r. Thirty-four systems (85%) ge~erate st~hs­ tics and 3 more will add this function. Charging out to reserves and billing are functions of 30 systems (75%). These two functions are now planned for implemen- tation in four libraries. Twenty-nine sys- tems (72.5%) can issue recall notices. Only 23 respondents (57 .5%) have incorporated interlibrary loan capabilities into their sys- tems, and only 3 others plan to add this. Vendor turnkey systems tend to offer more functions than in-house systems. In addition to charge-out/charge-in and overdue notice generation, which are common to all systems, the vendor turn- key systems offer other features by a mar- gin of at least 30% to 50%. Some of these features are recall notices, provided by 95.2% of vendor turnkey systems com- pared with 40% of in-house systems (P < .05), and delinquent patron status, of- fered by 100% of vendor turnkey systems and 70% of in-house systems (P< .01). One hundred eight of the 164 respond- July 1987 ing libraries (65.9%) plan to impleme~t an online circulation system, two-thuds within five years. INTEGRA TED ONLINE SYSTEMS Fewer than 10% of the 201 libraries re- sponding to this section have an inte- grated online system. The 17 responden~s are distributed thus: 6 of 129 small li- braries (4.7%), 9 of 54 medium (16.6%), and 2 of the 18 large libraries responding (11.1%) (P< .05). Ten of these 17, or more than half (58.8%), were vendor turnkey systems. Of the 17 integrated systems reported, 15 (88.2%) have an OPAC module, 13 (76.4%) have a circulation module, 7 (41.2%) have an acquisitions module, 6 (35.3%) have an ILL module, 4 (23.5%) provide for union listing of serials, and 3 (17.6%) include a serials co~trol module: Nine (52.9%) of the libraries began therr integrated systems operation with the OPAC module. Survey findings indicate that implementation of an OP AC module is not related to the percentage of catalog records in machine-readable form; 75% of the 4 responding libraries with 25% or fewer of their records in machine-readable form, and 71.4% of the 7 with 75% or more in machine-readable form implemented an OPAC first. Three (17.6%) systems be- gan with the circulati~n modu~e. Ac_quisi- tions, union list of senals, and mterhbrary loan were each the first step in three cases. Nine libraries have implemented their integrated online systems within the last year and another 8 reported establishing the system during the past five years. All9 . of the functional acquisitions modules were implemented less than a year ago. Of 178 libraries without integrated sys- tems, 96 (53. 9%) plan to implement them. Only 25libraries (14.0%) have no plans to acquire integrated systems, and 57 (32%) are uncertain. Sixty-two of 97 responses (63.9%) indicated plans to implement within five years. CONCLUSIONS This survey revealed that approximately 15% of the libraries responding currently have no online systems. Furthermore, 1 1 . 16.2% have no plans to add more online components. Size is certainly a factor. Of the 22.4% (47 of 210) who either do not plan any more systems or are uncertain, 91.5% (43) are small libraries (P < .001). This survey reveals that the trends ap- parent from the Georgia pretest of the questionnaire are consistent nationwide. Online automation is most extensive in those applications with an external com- ponent; that is, applications where more than one institution is involved: shared cataloging through a bibliographic utility (77.5%, or 162 of 209), interlibrary loan (76.4%, or 159 of 208), or union lists of seri- als (35.8%, or 73 of 204). The less common and more recently automated online ap- plications are those involving internal functions: circulation, acquisitions, serials control, and online public access catalogs (see table 5). Integrated systems are the least numerous of all, reported by seven- teen (8.5%) of 201libraries. As these fig- ures also make clear, there is significant online automation already in place in aca- demic libraries, and our first hypothesis was indeed valid. If implementation plans for online sys- tems are carried out by libraries as re- ported, the automation status in academic libraries will be increased by the following figures: acquisitions, 40.5% (85); serials control, 48.6% (102); online public access Survey of Online Systems 347 catalogs, 57.1% (120); interlibrary loan, 32.8% (21); circulation, 51.4% (108); and integrated systems, 53.9% (96). The hy- pothesis that online automation is antici- pated where it does not presently exist also proved valid. Although a substantial number of systems are reportedly planned within one year, the majority are expected to be operational during the next one to five years. 11The hypothesis that large libraries would be more likely to have the fi- nancial resources and personnel ex- pertise to develop their own online systems was not corroborated by the survey.'' The hypothesis that large libraries would be more likely to have the financial resources and personnel expertise to de- velop their own online systems was not corroborated by the survey. In every in- stance, the data proved statistically insig- nificant; thus, no conclusions could be drawn relating in-house systems and li- brary size. However, the data did support the hypothesis that large libraries are more likely to be automated than smallli- TABLE 5 NUMBER OF LlliRARIES REPORTING ONLINE SYSTEMS OR PLANS FOR SYSTEMS FOR SELECTED APPLICATIONS, WITH TIME FRAMES1 (n = 210) Acquisitions Serials Control OPAC no. % no. % no. % Have an online system 41 19.5 26 12.4 25 11.9 Less than 1 year 122 29.32 cj 34.63 124 48.04 1-5 years 232 56.12 133 so.oJ 114 44.04 More than 5 years 52 12.22 33 11.53 14 4.04 (Missing observations) (12) (13) (14) Plan an online system 85 40.5 102 48 .6 120 57.1 Within 1 year 106 11.86 167 15.77 11' 14.28 1-5 years 686 80.06 787 76.57 768 63.38 More than 5 years ~ 8.26 87 7.87 21' 22.58 Do not klan to install one 20 9.5 22 10.5 21 10.0 Do not now 64 30.5 60 28.6 44 21.0 Notes: 1 percentages may not equal 100% due to rounding 2 n = 41 6 n = 85 3 n = 26 7 n = 102 4 n=25 8 n=120 5 n=40 9 n=108 Circulation no. % 40 19.0 65 15.05 20S 50.05 145 35.05 108 51.4 111 15.11 739 67.69 189 16.11 20 9.5 42 20.0 348 College & Research Libraries braries (P < .001), as table 6 shows. The hypothesis that the regular library budget would not prove a major source of funding for online automation was based on the competition for dollars. In other words, it seemed likely that other sources would have to be found for funding in or- der to maintain services at their then- current levels. These levels may have been maintained, but the hypothesis proved in- valid. The major source of funding for on- line automation in academic libraries has indeed been the regular library budget (see table 7). This was the most frequently cited source in acquisitions systems, serials control, and union lists of serials (22 of 56, or 39.3%), and the second most cited source in circulation systems an:d online public access catalogs. For these last two, special allocations from the parent institu- tion have provided the major source of funding (circulation; OP ACs). This would suggest that institutions do indeed re- July 1987 spond to patron needs. Outside grants were less frequently reported: 9 of 56, or 16.1% of union lists of serials; 4 of 25, or 16% of the OP ACs; 2 of 41, or 4. 9% of ac- quisitions systems; and 2 of 40, or 5% of circulation systems were funded in this manner. The computer center budget of the parent institution was responsible for only 4 of 41, or 9.8% of acquisitions sys- tems, and 4 of 26, or 15.4% of serials con- trols systems. Bibliographic utilities are the largest sin- gle provider of online systems, a finding that supports an initial hypothesis. This is not surprising for machine-readable rec- ords (147 of 162, or 90.7%) and interlibrary loan (146 of 181, or 80.7%), but it is also true for union lists of serials ( 44 of 172, or 61.1%), acquisitions (20 of 44, or 45.4%), and serials control (8 of26, or 30.8%). Ven- dor turnkey systems are more prevalent in circulation modules (21 of 40, or 52.5%), OPACs (16 of 25, or 64%), and integrated systems (10 of 17, or 58.8%). Systems de- TABLE6 EXffiNTOFAUTOMATIONBYLffiRARYSaE 0-250,000 Size by volumes in the collection 250,000-1,000,000 More than 1,000,000 Number of (n = 137) (n =55) (n = 18) s~stems re~orted no . % no . % no . % 0 32 23.4 1 19 13.9 2 3.6 2 48 35.0 15 27.3 1 5.6 3 26 19.0 14 25.5 5 27.8 4 8 5.8 13 23.6 6 33.3 5 4 2.9 5 9.1 5 27.8 6 3 5.5 7 3 5.5 1 5.6 Note : Percentages may not equal100% due to rounding. TABLE 7 FUNDING OF ONLINE SYSTEMS REPORTED, FOR SELECTED APPLICATIONS Acquisitions Serials Control OPAC Circulation (n = 41) (n = 26) (n = 25) (n = 40) no . % no . % no . % no . % Regular library budget 23 56.1 19 73.1 6 24.0 9 22 .5 Special allocation from parent institution 5 12.2 2 7.7 11 44.0 17 42.5 Outside grant 2 4.9 4 16.0 2 5.0 Combuter center budget 4 9.8 4 15.4 1 2.5 Com ination of above 6 14.9 1 3.8 4 16.0 9 22.5 Other 1 2.4 2 5.0 Note : Percentages may not equal100% due to rounding. veloped in-house are the third most com- mon, accounting for 9% (51 of 567) of the total responses. Consortia provide 16% (29 of 181) of interlibrary loan systems (see table 8). The majority of acquisitions systems, union list of serials, and serials control use mainframe systems. Online public access catalogs use mainframes as often as minis. More circulation systems are run on mini- computers than mainframes. Microcom- puter systems were reported least often (see table 9). The preferred backup for each applica- tion is hard copy. Microforms appear to be the second backup of choice. Of note is the fact that 69 respondents to the questions on backup systems listed none at all. This Survey of Online Systems 349 is particularly surprising for online public access catalogs (6 of 25, or 25%) and circu- lation systems (9 of 40, or 22.5%), because of the importance of these systems to pub- lic service (see table 10). In addition to providing a historical ref- erence point in library automation, this study suggests further areas of investiga- tion. One topic that deserves research is the organizational change that accompa- nies or results from automation. Auto- mated systems have required that certain tasks be shifted, with the result that work flows more smoothly and efficiently. Since many tasks can be coordinated through a computer terminal, operations can be performed in different locations just by installing a terminal. Integrated TABLES SOURCE OF ONLINE SYSTEMS REPORTED, FOR SELECTED APPLICATIONS Acquisitions Serials Control OPAC Circulation (n = 41) (n = 26) (n = 25) (n = 40) no . % no . % no . % no . % In-house 8 19.5 7 26.9 1 4.0 10 25.0 Vendor turn-kfc 10 24.4 3 11.5 16 64.0 21 52.5 Combination o above 3 7.3 2 7.7 4 16.0 6 15.0 Bibliographic utility 20 48.8 8 30.8 2 8.0 Other 61 23.11 2 8.0 3 7.5 Notes : 1 Commercial Service, e.g., EBSCO, FAXON, etc . TABLE9 TYPE OF COMPUTER USED FOR SELECTED ONLINE SYSTEMS REPORTED Acquisitions Serials Control OPAC Circulation (n = 41) (n = 26) (n = 25) (n = 40) no . % no . % no. % no . % Micro 10 24.4 4 15.4 1 4.0 2 5.0 Mini 10 24.4 2 7.7 12 48.0 23 57.5 Mainframe 19 46.3 17 65.4 12 48.0 14 35.0 Don't know 2 4.9 3 11.5 1 2.5 TABLE 10 BACKUP FOR SELECTED ONLINE SYSTEMS REPORTED Acquisitions Serials Control OPAC Circulation (n = 41) (n = 26) (n = 25) (n = 40) no. % no . % no . % no . % None 22 53.7 11 42.3 6 24.0 9 22.5 Hardcopy 12 29 .3 11 42.3 11 44 .0 19 47.5 Microform 1 2.4 1 3.8 6 24.0 1 2.5 Micro diskette 2 4.9 2 7.7 1 4.0 5 12.5 Combination 3 7.3 1 3.8 4 10.0 Other 1* 2.4 1 4.0 2 5.0 Note : Percentages may not equal 100% due to rounding . 350 College & Research Libraries systems especially lend themselves to this; it would be useful to see how many libraries bypass the installation of separate systems in favor of an integrated one. The most obvious project would be a follow-up study in five or six years to see if the projected automation actually takes place as reported in the present study. However, it is important to bear in mind, as suggested by the subtitle of several of Matthews' articles, that change may well alter all plans. Rapid changes in hardware and software may make existing systems obsolete. The introduction of new tech- nology such as compact discs will have a July 1987 tremendous impact. This can even extend to organizational changes such as OCLC' s recent announcement of the de- activation of its present acquisitions sys- tem on January 1, 1989, and its replace- ment, already under way, by the ACQ350 system. A new serials control system, the SC350, is available. And there is a forth- coming OCLC 350 Services LAN (Local Area Network), which will allow several users to share the ACQ350 and SC350 sys- tems. So perhaps it is only safe to say that this is the situation at present. The future re- mains uncertain. REFERENCES 1. Maxine K. Sitts, The Automation Inventory of Research Libraries 1985 (Washington, D.C.: Office of Management Studies, Association of Research Libraries, 1985). 2. Judith M. Shelton and others, "Online Systems in Georgia Academic Libraries," Georgia Librarian 22, no.3:58-59 (Aug. 1985).