College and Research Libraries College Libraries and Resource Sharing: Testing a Compact Disc Union Catalog Charles T. Townley In 1990, the seventeen collegiate members of the Associated College Libraries of Central Pennsylvania tested a compact disc union catalog to determine its effect on resource sharing in a college library consortium. This article reports the outcomes of the test in the areas of bibliographic quality, user evaluation, public relations strategies, and operating guidelines. Desired enhancements and additional research needs are identified. Recommendations are made for the use of compact disc union catalogs in college library consortia. ollege libraries form networks and consortia to improve local services through coordinated action. Nowhere is this more true than in resource sharing. Alone, a college library cannot afford enhanced col- lections in more than a few subject areas. Yet, by banding together, college library collections can form a virtual research li- brary to the benefit of faculty and students at all participating institutions. The practical problem has been how to identify, request, and deliver desired material among participating libraries in an effective manner. This can be likened to building a three-legged stool for re- source sharing where all three legs are necessary for effectiveness. Over the years college library consortia have developed several means for sharing re- sources, including union card catalogs and serials lists for identification, inter- library loan for requests, and the use of mail, reciprocal borrowing, and delivery services for exchanging rna terials.1 Recent technological advances create both opportunities and challenges for re- source sharing among college libraries. Machine-readable bibliographic records and electronic public-access catalogs im- prove local access but are difficult and expensive to link together with a com- mon command language. Enhancements in telecommunications make contact possible, but most college library consortia cannot afford dedicated telecommunica- tions among members. And national net- works, like Internet, currently do not reach many smaller colleges. Delivery services can be expensive, and telefacsimile qu- ality can be inadequate, particularly for tables and illustrations. CD-ROM technology offers convenient and inexpensive local copies of a machine- readable union catalog within a college library consortium. Updated on a peri- odic basis from member library records, a CD-ROM union catalog offers the potential to achieve bibliographic con- trol necessary for effective college li- brary resource sharing. CD-ROM union catalogs at each participating institution can be used for identifying needed mate- rial, initiating requests, and encouraging Charles T. Townley is Dean of the University Library at New Mexico State University, Las Cruces, New Mexico 88003-QD06. 405 406 College & Research Libraries use of preferred delivery techniques. This article reports on a CD-ROM union catalog, named C. D. Cat, developed by the seventeen members of the As- soda ted College Libraries of Central Pennsylvania (ACLCP). 2 Reported are outcomes of the project in the areas of bibliographic quality, user evaluation, public relations strategies, and guide- lines for continuing operations. Addi- tional research needs are identified and desired enhancements described. Rec- ommendations are made for the use of compact disc union catalogs in other col- lege library consortia. BACKGROUND The Associated College Libraries of Central Pennsylvania is a consortium of seventeen college libraries, the Dickin- son Law School, and the State Library of Pennsylvania. Now celebrating its twenty- fifth anniversary, ACLCP has long been a leader in college library cooperation, attracting more than $1,000,000 in grants and projects during its existence. In 1968, it undertook the formation of Interli- brary Delivery Service, now a separately incorporated service transferring mate- rials among more than 150 Pennsylvania libraries.3 In 1983, ACLCP organized tel- efacsimile communications among aca- demic libraries throughout Pennsylvania. These two services compose two legs of the resource-sharing stool for academic li- braries in central Pennsylvania! In 1988, two external events en- couraged ACLCP to add bibliographic control as the third leg of an effective September 1992 resource-sharing stool. First, through the success of Access Pennsylvania (a union catalog of school, public, and publicly supported academic libraries), CD-ROM technology demonstrated that it could create a union catalog at low cost.5 Sec- ond, funding became available through Title 11-D of the Higher Education Act to help defray the cost of developing and testing a CD-ROM union catalog among the seventeen college libraries. ACLCP prepared and received funding for a $134,000 proposal based on the Intel- ligent Catalog product developed by the Library Corporation. ACLCP personnel spent the first eighteen months of the project in planning and development ac- tivities described in a previous article.6 In January 1990, each ACLCP col- legiate library received an Intelligent Catalog station and a test CD-ROM disc containing 1,086,778 bibliographic rec- ords from seven of the seventeen partici- pating libraries (see table 1 ). ACLCP personnel used the catalog to test a new linked-and-merged database design that joined records with the same Library of Congress card and/ or OCLC control number. The test reduced the number of entries by 35 percent to 709,523 master records. Authority control was not at- tempted, but Library of Congress cross- references were included. The Intelligent Catalog software permitted searching, using traditional author, title, and sub- ject approaches, Boolean (keyword) ) searching, and several artificial intel- ligence techniques, including mapping and readers' adviser services. Users TABLEt TEST DATABASE STATISTICS Member Library No. of Records Used No. of Master Records Bucknell University 316,880 206,848 Dickinson College 221,573 98,854 Elizabethtown College (partial) 7,879 2,954 Franklin & Marshall College 213,587 213,541 Harrisburg Area Community College 64,600 29,484 Juniata College 70,607 30,414 Kutztown University 191,652 127,520 Total 1,086,778 709,523 College Libraries and Resource Sharing 407 could manipulate records they selected, leave notes for library staff, and print out their results. GOALS, OBJECI1VES, TESTING The ACLCP union catalog seeks to complete a bibliographic resource-shar- ing system among the member libraries of the Associated College Libraries by providing bibliographic control through a CD-ROM union catalog. Objectives in- clude: 1. Assuring adequate bibliographic quality to permit identification of materials held in member libraries; 2. Providing user-friendly access to li- brary users; 3. Developing user-education pro- grams that result in desired user behavior; 4. Generating user support adequate to assure continued production of the catalog. The test period extended from January to June 1990. Each participating library provided a minimum of six weeks of public access during the test period. Each ACLCP library helped evaluate C. D. Cat and its success in achieving union cata- log objectives. BIBLIOGRAPHIC QUALITY The project advisory committee named an evaluation committee comprised of in- dividual members of ACLCP with ex- perience in technical services and automation to evaluate the bibliographic quality of bibliographic database. 7 This committee solicited comments from all project participants. The committee evaluated this information and made recommendations to the vendor. The evaluation committee found the database capable of identifying materials held in member libraries. The committee found the linked-and-merged design a very effective means of controlling mul- tiple entries without the cost of a fully de-duped database. While duplicate re- cords did occur because some records did not contain the linking Library of Congress card or OCLC control number, many records were merged using this process. Indeed, the number of entries in the database was reduced by 35 percent from 1,086,778 records to 709,523 records by using the linked-and-merged design. The danger of false merges-when match- ing records do not contain correct LCCN or OCLC control numbers-proved to be an insignificant issue. Some 497,641 (70 per- cent) records showed only one location, suggesting the uniqueness of college li- brary collections and the strength to be gained when cooperating college librar- ies develop a common source of biblio- graphic information. The role of public relations and bibliographic instruction should be to guide responsible use of the consortially created research library. Several troublesome bibliographic problems did occur in the test database. One member's location stamps were in- correctly read so that all holdings were placed in a branch library. A number of minor call number variations, which might make it difficult to identify shelf locations from C. D. Cat data alone, were identified. The committee identified several sorting problems with the title fields. The vendor did not follow the requested order for entering member li- brary databases, which resulted in less complete bibliographic records serving as a master record from time to time. The evaluation committee observed that the current database organization is inadequate in its handling of articles in foreign languages. The committee believed that if catalogers go to the trouble of pro- viding indica tors for articles in foreign lan- guages, the vendor should be able to read and ignore them for filing purposes. As an unexpected bonus, the creation of the catalog has enabled member li- braries to identify and resolve a poten- tially disastrous problem with the hex code used for diacritical marks in foreign language material. The regional provider of data from a national bibliographic database had inadvertently set its hex code incorrectly when creating databases for member libraries. This resulted in 408 College & Research Libraries spellings of prelude, for example, as prbelude. An analysis of records in C. D. Cat by personnel from Shippensburg University identified the source of this problem, which is now being corrected with the full cooperation of the regional provider. The evaluation committee has ap- proved the bibliographic quality of the C. D. Cat database for use in ACLCP. Committee members recommend the linked and merged design for operational use as an effective and inexpensive means of identifying bibliographic resources in a college library union catalog. USER EVALUATION A local coordinator named by each li- brary director organized user evaluation at member libraries. Each local coordina- tor was trained on project objectives and guidelines, equipment, and use of the database at the beginning of the test pe- riod. Each local coordinator prepared a strategy for introducing the catalog in- tended to meet the needs of the local institution. Some coordinators empha- sized direct use of C. D. Cat on a refer- ence referral basis. Others developed bibliographic instruction programs, usually supporting courses with signifi- cant research assignments. Local coordi- nators kept a diary in which they indicated observed strengths and weak- nesses of C. D. Cat as well as their sug- gestions regarding the catalog and its improvement. Students, staff, and faculty who used the test database were asked to complete an evaluation questionnaire regarding its usefulness. These questionnaires were collected and analyzed by project personnel, and the results are incor- porated in this article. Local coordina- tors discussed suggestions of particular merit at a summer workshop where they shared their experiences and plans. Users reported using the catalog for the full range of subjects common to lib- eral arts colleges (see table 2). Historical topics were most popular, followed by arts and sciences. The catalog was useful for most subjects, including professional topics like business and health. September 1992 TABLE2 MOST SEARCHED TOPICS OFC. D. CAT Topic Frequency % History 42 19.4 Arts 28 12.9 Sciences 26 12.0 Literature 25 11.5 Business 20 9.2 Religion 20 9.2 Political science 16 7.5 Health 15 6.9 Psychology 14 6.5 Sociology 11 5.1 Total 217 ~e ease and speed of the catalog are its most popular features (see table 3). Screens are easily read and the keyboard is simple. Most users come with a known item or subject and are pleased to be able to find it quickly. The value-added fea- tures, including keyword searching and indexing, and artificial intelligence, like maps and first page of popular works, drew mildly positive responses from the small number of users who explored these features. Many respondents indicated that there are no bad features in the union catalog. They cited as a negative feature that the test database only contained holdings from seven of the seventeen libraries. Staff users expressed dissatis- faction with slow speed when searching three or more keywords, which can take more than a minute. Users reported overall success in using the compact disc catalog. Some 93 percent of users report finding useful information. Even when known item searches are carried out, users succeed 85 percent of the time. Several local coordina- tors and users prefer C. D. Cat features to their local public access catalog. User behavior during the test period is considered predictive in estimating growth in delivery services that will be associated with the operational version of C. D. Cat (see table 4). During the test College Libraries and Resource Sharing 409 TABLE3 EVALUATION OF C. D. CAT'S FEATURES Feature Best features of C. D. Cat: Easy, user-friendly Fast Access to other libraries (locations) Access to more material (quantity) Multiple search modes (keyword) Printer Technical features (first page, etc.) Prompts, help screens, phone Displays and keyboard Indexing of additional fields Total Worst features of C. D. Cat: None Slow Incomplete holdings Books not in this library No periodicals Training/hard to learn Cannot see what you are typing Multiple entries Prompts/help Incomplete printer citations Keyboard too sensitive No logoff Unanticipated screen change Total period, direct reciprocal borrowing and personal use of other libraries were strongly encouraged, but use of this service by these patrons was disappoint- ingly low (12 percent). Usage patterns do not appear to differ significantly, based on whether users learned about the database in bibliographical instruction or through a reference referral. Interli- brary loan remains the primary means of obtaining materials not available locally (28 percent). Allowing for the novelty of C. D. Cat during the test period, the data suggest that one hundred sessions on the Frequency 153 75 54 45 25 18 12 9 8 6 405 46 14 13 10 10 10 8 7 7 7 6 4 3 145 % 37.8 18.5 13.3 11.1 6.2 4.4 3.0 2.2 2.0 1.5 31.7 10.0 9.7 7.0 7.0 7.0 5.5 4.8 4.8 4.8 4.1 2.8 2.1 CD-ROM catalog will generate at least fifty interlibrary loan requests, or one- half of a request per use (see table 5). Several ACLCP libraries have decided to use C. D. Cat as a high quality backup for their local OPAC. During the test pe- riod, these libraries.had occasions to test its adequacy when their local system was down. All reported it to be successful. Indeed, 31 percent of public users indicated they found material in the local library that was previously unknown. At least one li- brary has ordered an additional C. D. Cat station for local OPAC backup purposes. 410 College & Research Libraries TABLE4 OVERALL USER SUCCESS Yes No Found useful information 282 20 Found known item 45 8 % 93.4 84.9 PUBLIC RELATIONS AND BIBLIOGRAPHIC INSTRUCTION September 1992 cal borrowing or valuing quality of ref- erences, rather than quantity that can reduce delivery demand. Local coordinators made several rec- ommendations regarding changes to in- troductory screens so users can quickly determine that C. D. Cat is indeed a union catalog. Four ACLCP libraries also use the Intelligent Catalog tech- nology for their local public access cata- logs, and their users have particular Local coordinators and other interested problems distinguishing between the ACLCP members, including representa- local and union catalogs. Another area of tives from the reference, bibliographic in- concern is that printouts provided by the struction, and circulation committees, met machine should carry a banner indica t- in June 1990 to review the recommenda- ing the availability of direct reciprocal tions of the technical committee, local borrowing to discourage interlibrary coordinator diaries, and individual user loan requests. data regarding the catalog. Local coordi- Addressing bibliographic instruction, nators developed public relations strate- local coordinators recommend that C. D. gies and bibliographic instruction models Cat be presented in faculty workshops at that would guide users in desired direc- each college. They also suggest develop- tions. Local coordinators concurred with ing and using a common descriptive bro- the technical committee that biblio- chure in local workshops and bibliographic graphic quality was excellent. They con- instruction. They feel C. D. Cat is most firmed that the database had been well useful in advanced research courses and received among most participants. in support of faculty research. Reference librarians recommend several Reference librarians, in general, do not changes to enhance public access. They like the artificial intelligence features of believe that the reference librarian must C. D. Cat. The mapping feature that pro- introduce and interpret the union cata- vides access to geographically related log. This introduction is especially im- topics is not detailed enough. The read- portant in college environments where ers' adviser is for current best-sellers not connectivity with other colleges is usu- academic works. Local coordinators ally limited to social and sporting events believe the components developed as a and does not include academic contacts part of Intelligent Catalog software are like cross-listing of courses or shared aimed at a school or pubic library computer resources. Further, it is con- audience and should be disabled or re- sidered important to guide users into placed with information appropriate for desired behavior patterns, like recipro- a collegiate audience. Behavior Used local library Used direct borrowing Used other library Requested ILL Nothing Other Totals TABLES USER BEHAVIOR 7 Test Ubraries No. % 48 30.9 9 5.8 9 5.8 44 28.4 19 12.3 26 16.8 155 No. NA 21 21 157 67 47 313 All Ubraries % NA 6.7 6.7 50.2 21.4 15.0 College Libraries and Resource Sharing 411 Finally, local coordinators believe that the compact disc catalog sells itself. Once users are aware of the catalog, they are interested in using it. The role of public rela- tions and bibliographic instruction should be to guide responsible use of the consortially created research library as a supplement for each library's local collection. Interlibrary loan personnel are under- standably concerned about the potential impact of C. D. Cat. With each two ses- sions on C. D. Cat predicted to generate one interlibrary loan request, interlibrary loan personnel believe the ACLCP library directors must provide additional support for interlibrary loan operations. They en- courage reference librarians to counsel students to use the local library and recip- rocal borrowing before requesting interli- brary loans. Interlibrary loan personnel also suggest that it might become neces- sary to limit the number of interlibrary loan requests from students, to en- courage awareness of the differences be- tween quality and quantity in terms of overly large bibliographies, or to follow the lead of one ACLCP library and charge fees for interlibrary loan copies. Interlibrary loan personnel think it is essential that the OCLC control number and/ or the LCCN number be provided as a part of the C. D. Cat printout to expedite the interlibrary loan process used by most ACLCP libraries. They also look forward to the creation of an inter- library loan form on the C. D. Cat equip- ment that can be completed by users, down-loaded, and held for review and processing by interlibrary loan person- nel. Both features will help in coping with increased demand. Reciprocal borrowing needs additional encouragement and publicity. Signs on C. D. Cat stations and printout banners should encourage reciprocal borrowing. Bibliographic instruction should tout the advantages of using other libraries in person. One possible method of encourag- ing direct borrowing is to provide some form of reinforcement, possibly through gift or discount certificates, for those people who use it. While one concern is that a student without transportation is unable to use reciprocal borrowing in an effective way, it has been noted that most can find transportation for social and ath- letic events. Finally, librarians believe they have to encourage reciprocal borrowing by offering to call ahead to check on the availability of desired material until availa- bility information is accessible on C. D. Cat, possibly through a dial-access modem to member library circulation systems. CONTINUING OPERATIONS Based on the positive assessments of bibliographic quality and user accep- tance as well as on the development' of promising public relations techniques that will guide use into desired areas, the General Policy Committee has endorsed the implementation of C. D. Cat on a continuing basis. Participating libraries will pay $550 per year for equipment maintenance, staff coordination, and two CD-ROM databases each year. A C. D. Cat director will coordinate training, evalua- tion, and submission of local tapes. An advisory committee will continue to moni- tor the project and undertake substantive reviews every three years. DESIRED ENHANCEMENTS College libraries recommend several enhancements to make compact disc technology more effective in college li- brary consortia. Interlibrary loan per- sonnel believe college users are capable of creating and storing their own interli- brary loan requests directly on the C. D. Cat workstation. This will save valuable staff typing and verification time. In ad- dition, ACLCP wants to add a telefac- simile or electronic mail card to the C. D. Cat station so that libraries can communi- cate interlibrary loan requests directly with other members without having to leave the union catalog workstation. The consortium wants to add an al- ready existing union list of periodicals to the catalog to reflect the complete serials collections of the participating libraries. At present, some local catalogs provide information on periodicals and others do not. Some show holdings and others do not. Using the union list as the only source of bibliographic information for se- rials will eliminate this inequity in records. 412 College & Research Libraries Finally, ACLCP should take advantage of the presence of the union catalog to introduce articulated collection develop- ment among member libraries. While some duplication is desirable to support local instructional needs, C. D. Cat per- mits much better articulation for pur- chasing esoteric materials. NEEDED RESEARCH Additional research is required in creating components of artificial intel- ligence that are meaningful for academic library users. As a discipline, artificial intelligence has much to offer libraries in access bibliographic information. Stu- dents need guidance to appropriate in- formation and an opportunity to become excited by the labyrinth of knowledge available in academic libraries. A readers' adviser function that could guide fresh- men and sophomores to basic paper topics would be useful. Faculty need to be guided to the best research in their fields of study. Reference to related subject headings or related keywords would be helpful for faculty. Efforts to introduce artificial intelligence features into public catalogs should continue. User behavior also needs to be better understood. Appropriate materials to sup- port instruction need to be developed. Strategies that encourage direct reciprocal borrowing should be developed and evaluated for their effectiveness. Efforts stressing quality as well as quantity of information need to be developed. The effects of a union catalog on distance learning, student recruiting, and library use should be analyzed. Costs and benefits of union catalogs versus linked systems and common com- mand languages should· continue to be analyzed. ACLCP members are inclined to see linked systems as the long-term solu- tion to full resource sharing. But signifi- cant additional funding and research are needed. In the meantime, protocols for col- lection articulation can be developed, tested, and applied in a CD-ROM catalog environment. Perhaps availability infor- mation can be delivered from local cata- logs on demand through modems. Periodical and serial subscriptions may September 1992 constitute a good starting point for ar- ticulated collection development. RECOMMENDATIONS I C. D. Cat represents the completion of a three-legged stool for resource sharing among the member libraries of ACLCP. Materials can be identified in C. D. Cat, solicited through interlibrary loan, and delivered through already existing delivery services, reciprocal borrowing, or telefacsimile. While much enhance- ment is possible and desired, ACLCP member libraries now provide virtual research library access to their collegiate communities and are well positioned to initiate articulated collection develop- ment among themselves. The successful test ••• indicates that CD-ROM technology can effectively address the resource-sharing needs of college library consortia. This successful test of C. D. Cat indi- cates that CD-ROM technology can ef- fectively address the resource-sharing needs of college library consortia. The quality of the catalog as an accurate bib- liographic resource is assured. It is both useful and popular for identifying needed materials. In addition, ACLCP has found it helpful in showing areas for improve- ment of local databases and OPACS. Pub- lic relations and administrative guidelines offer promise for encouraging use that is appropriate and desired. College library consortia wanting to encourage resource sharing and lacking the funding for dedicated telecom- munications, linked systems, and/ or a common command language would do well to consider a compact disc union catalog. Such a catalog, located at each member library, demonstrates the effi- cacy of cooperation and the ability of col- lege libraries to identify and deliver desired information. Appropriately de- signed for consortium needs and abilities, a catalog station, located at a member li- brary, visually demonstrates how effective resource sharing can be. College Libraries and Resource Sharing 413 REFERENCES AND NOTES 1. Norman D. Stevens, "Library Networks and Resource Sharing in the United States: An Historical Overview," American Society for Information Science Journal 31:405-12 (Nov. 1980). 2. Member libraries include Albright College, Bucknell University, Dickinson College, Dickinson School of Law (not participating in this project), Elizabethtown College, Franklin and Marshall College, Gettysburg College, Harrisburg Area Community College, Juniata College, Kutztown University, Lebanon Valley College, Messiah Col- lege, Millersville University, The Pennsylvania State University at Harrisburg, Ship- pensburg University, State Library of Pennsylvania (not participating in this project), Susquehanna University, Wilson College, and York College of Pennsylvania. 3. Charles H. Ness, "Interlibrary Loan Developments: Pennsylvania," RQ 7114-116 (Spring 1968); Dwight Huseman, "Access to Materials in Pennsylvania: Interlibrary Delivery," Wilson Library Bulletin 59:262-63 (Dec. 1984). 4. Mark Wilson, "How to Set Up a Telefacsimile Network-The Pennsylvania Libraries Experience," Online 12:15-25 (May 1988); Charles Peguese, "Telefacsimile, The Penn- sylvania Experience: A State Library's Perspective," in Research Access through New Technology ed. Mary Jackson (New York: AMS, 1989), p.88-89. 5. Doris M. Epler and Richard E. Cassel, "ACCESS PENNSYLVANIA: A CD-ROM Database Project," Library Hi-Tech 5:81-92 (Fall1987). 6. Charles T. Townley, "The Value-added Compact Disc, Union Catalog of the Associated College Libraries of Central Pennsylvania: A Progress Report," College & Research Libraries News 51:835-39 (Oct. 1990). 7. Members of the Evaluation Committee include Robert Gimmi (Shippensburg); Cathryn Hintze (Harrisburg Area Community College); Soo Lee (Messiah); Michael Lynch (Bucknell); Kristin Senecal (Dickinson); and James Fogarty (ACCESS Pennsyl- vania). IN FORTHCOMING ISSUES OF COLLEGE & RESEARCH LIBRARIES An Authority Control Alternative for Small Colleges Joan M. Bechtel Information Delivery Strategies and the Rural Student Sharon M. West On the Merits of Direct Observation of Periodical Usage: An Empirical Study Marifran Bustion Literature Reviews and Inaccurate Referencing: An Exploratory Study of Academic Librarians Peter Hernon and Cheryl Metoyer-Duran Job Satisfaction Among Support Staff in Twelve Academic Libraries in Ohio Coleen Parmer and Dennis East The Organizational Misfits Patricia A. Suozzi and Sandra Kerbel Research Notes Conference Proceedings in Physics H. H. Barschall and W. Haeberli THE CORNERSTONE OF YOUR SCI ~TECH INFORMATION NEEDS: SciTEcH Reference Plus,.. Sci TECH Reference Plus"' is the instant answer when you and your pacrons have sci-tech questions. '"'"' )fJII hatt! ll "t 11//Ht'd tlll)fhmg ·· - llt.l t.n "t t.pp. Rt.,t..tn.h .ltll.ll)c.\t.lt,pnllnr lihr.tn.tn Covering everything from books and periodicals, to small R&D, labs and mammoth industrial testing facilities, this singular CD-ROM information retrieval tool gives you instant access to these cornerstone references: • all sci-tech citations from both Books in Print and Ulrich's International Perodicials Directory • all of American Men & Women of Science • all of the Directory of American Research & Technology • all of the Corporate Technology Directory (Corptech) You get complete, creative concrol over a broad speCtrum of records on sci-tech publications, personnel, and industry activities. Explore a single database or review the complete "mega" file based on any combination of 27 criteria, and use full Boolean logic. It takes only seconds to. .. • find practically every sci-tech book on virtually any sci-tech topic • retrieve citations for every sci-tech periodical, annual, and irregular published around the world • learn about the research and development capabilities of industrial organizations across North America • locate experts in all the physical, biological, and related sciences-from Acoustics to Zoology • plug into vast amounts of corporate and industry high-tech research Order your copy of this vital resource today! SaTECH Reference Plus"' • $995 Call 1-800-323-3288 Bowker Electronic Publishing 121 Chanlon Road • New Providence, tqJ 07974