Full-text Databases Ahead–Proceed With Caution

by Rebecca L. Johnson and E. Ann Ford

This morning a college administrator heard on the news about an article, which appeared in a new issue of a scholarly journal. It will provide the support he needs to make a point authoritatively in his presentation to the board tomorrow. He needs to read the entire article, the sooner, the better, but the library doesn’t subscribe to that particular journal. How can he acquire a copy before the end of the day? If he can access the correct full-text database, he may well find it. That could be his answer.

Full-text databases are the hottest products for libraries. The selection is expanding and changing almost weekly. A full-text database can allow you to search an index for articles in journals or newspapers, and then simply call up the entire text of an appropriate article. A version of this has been available to libraries for some years, but there are now many more possibilities. Experience has shown that both students and faculty are pleased with the availability of full-text articles made accessible on their home and office computers. Many students and faculty members have used such databases in libraries and they usually want to continue.

With the increased emphasis on distance learning, libraries are pressed to provide information access to students at remote locations. These databases speed up the access to information, cut down on time spent finding journals and copying articles. Searching most of these databases is both easy and quick. Little instruction is needed unless dealing with a subject-specific database. Vendors post news on their Web pages and may offer a free trial period. These databases generate a great deal of discussion at conferences, on listservs, and in the professional literature.
The academic world is enticed by new possibilities for electronic access to information. Students and faculty as a group are rather sophisticated computer users, who surf the Internet with ease. With the pressure of time, many library users hope to find everything they need for their research without having to go to the library. They want to find articles in an electronic index, then go to the linked electronic full-text and, if an article meets the need, download it or print it to use in their research. Furthermore, they want to be able to have access to such resources while sitting in offices, dorm rooms, or homes. With some caveats, this has become possible.

To explore the possibilities, libraries must assess the needs of its user population to select a full-text database. In preparing this article, we contacted several individuals who are involved in library collection purchasing. We also reviewed several libraries’ electronic resources policies to determine what criteria are successfully used to select full-text databases. We also identified a number of general interest full-text database vendors and sent them a questionnaire asking for specific information concerning their databases. The questionnaire information was compiled in the form of a matrix to allow a quick review of some of the major criteria that are currently being used to compare and contrast the various databases. The following discussion focuses on the full-text databases of general interest, but many specialized databases are also available. Changes are occurring rapidly among the database vendors, as described in a recent Library Journal article, which includes an extensive matrix on databases, many of which are subject specific.

 

Costs
Cost is necessarily the major concern in subscribing to an electronic full-text database. As shown on the matrix, costs vary with the size of the database and the number of users. Some database vendors could not, or would not, give a general cost estimate. All indicated that the cost must be negotiated on a case-by-case basis, varying according to the library, the size of the student body, the type of access and/or the number of simultaneous users. If the college library is a member of a consortium, it will be able to negotiate a lower subscription rate based on the larger user group involved. Because these are recurring annual costs, a continuing source of funding is fundamental. Unless the library materials budget is dramatically increased, the high subscription costs may lead to reallocating funds from some other portion of the budget or cancellation of paper subscriptions. Besides consuming a great deal of staff time, this can be painful since faculty members are rarely sympathetic with a decision to cut back funds in their specialized areas of interest.

 

Licensing
Licensing terms for electronic resources are typically more restrictive than the copyright laws for print materials. For example a license may forbid providing a copy of an article from a database to fill an interlibrary loan request. This restriction, however, may conflict with the library’s commitment to its consortium. Licenses require close examination. No library should sign a license without consulting the institution’s legal staff, in order to be certain the institution can abide by the stated terms, or to negotiate for changes in the terms.

 

Hidden Costs
Beyond the major subscription expenditure there are often hidden costs.

• Staff for Technical Support
All of the vendors contacted indicated that they offer full technical assistance to all subscribers. In practice though, this generally involves identifying one qualified staff member to act as the authorized technical contact. The authorized contact fields all local technical questions, responds to those that can be handled locally, and forwards the technical support question that requires vendor intervention.

Furthermore, systems vary tremendously, and software and hardware requirements demanded by one vendor might be significantly different, or even incompatible, with another vendor’s requirements. If the library wants to provide seamless integration of databases to create a user-friendly interface, systems support is needed to fit a new database into an integrated infrastructure. Vendors are often unable to provide support at this level, and do not accommodate any variations from their stated hardware and software requirements.

• Proxy Server
Another issue is how to provide access to the materials for all faculty, students, and staff regardless of their location, whether from dormitory room, office, or home. All authorized users must be recognizable to the database, which is usually accomplished via the IP address. It is important to ascertain whether or not a student’s IP address will be correctly read if he/she is using a computer off campus. Frequently when students are using a commercial internet provider, such as AOL, that IP address is rejected. This problem may be resolved with the installation of a proxy server to handle the verification, which will require additional money, staff time and expertise.

• Equipment Upgrades
A third hidden cost, not always factored into expenses, involves equipment costs. For example a library may need to upgrade equipment to make the best use of full-text capability (e.g., the image resolution on older computers may be too poor to display the graphics). Or it may need to purchase more printers and better ones.

With careful consideration and the anticipation of concomitant expenditures, a library can avoid unpleasant surprises.

 

Selection Criteria
Each institution has its own unique needs and it will need to answer many questions in order to decide which database is best for an individual campus. Considerations may include:

• How will the acquisition of a full-text database improve the access to information and the quality of education at the institution?
• How much will it cost and how will it be funded on a continuing basis?
• What journals are included? Will all of the journals be of use for the institution’s curriculum?
• What is the frequency of updating? Is the entire text of the periodical present? Are illustrations and graphics included?
• What staff time or skills or training will be needed to implement the database and educate users?
• Is it Web-based? What is the meth-od of authentication of users?
• What additional hardware and software will be needed? What technical support does the vendor provide?
• What are the license terms? How many may simultaneously use the database?

A more extensive list of criteria/ questions is available at http://www.snc.edu/~mankmm/criteria.htm

 

Subject Coverage
Cost considerations aside, the primary significance for the institution will be whether a particular full-text database strengthens the collection, supports the institution’s curriculum, and benefits the faculty and students. How many of the titles for which the database provides full-text coverage does the institution already subscribe to? Is there a strong balance of scholarly journal titles? Some database providers attempt to meet the needs of both public and academic institutions with a single database and the result may be unsatisfactory for both groups.

Another aspect is to review the list of full-text titles, removing the titles the institution already purchases, and then determine which of the remaining additional titles would be useful. Frequently the full-text titles are of limited scope, such as a textile business newsletter. If the institution does not offer courses in those subject areas perhaps there is no need for enough of the additional titles to justify the cost.

An additional element to consider is the years of coverage. How many years are available online? A database may continue to grow or it may retain only the past five years. How often is the database updated, and are the most current issues promptly available? For example, the attached matrix shows that InfoTrac indexes daily but does not provide full-text of articles for one to three weeks after the article is published. Is that time lag acceptable, or might your institution prefer a database that provides full-text more promptly?

Is It Really Full-Text?
How many journals does the database index, and of those, how many are truly full-text? The database provider contracts with each journal publisher for inclusion and the publisher may not permit full-text access. The information in the matrix indicates that frequently less than one-third of the indexed journals are available in full-text. Only Electric Library and Lexis-Nexis are entirely full-text. Many of the journals may be indexed and abstracted back as far as 1980, but full-text coverage may extend back only a few years. Also some of the full-text titles may not include photographs, graphs, charts, letters to the editor, and other such material.

Ownership or Access?
Another question is: Can an institution archive and retain access to the data paid for if the subscription to the full-text database is cancelled? If not, the institution might feel the need to continue to subscribe to the print versions of the journals. When vendors compare the cost of their database to the cost of the individual subscriptions, their assumption is that institutions will cancel their subscription to the printed journals. If the choice is made to cancel print subscriptions it is important to bear in mind that the database information is “leased,” so the institution usually does not retain access or ownership to the years for which it paid. Unless archiving is permitted, at the end of the “leasing” period the institution owns nothing. Under these circumstances the institution may lose several years of backfiles of those journal titles. In order to avoid such a loss to the library’s holdings, many institutions retain both print and electronic versions.

The matrix in this article charts each of the database vendors queried for specific information. It allows the reader to compare various features of each broad-range database. Not all of the important criteria can be shown on this matrix. Questions such as whether or not the journal coverage is a good match for the institution’s collection must be determined for each and every institution. However, basic information regarding number of titles and years of coverage, number of full-text titles, hardware and software requirements, etc., will help make the initial choice of databases to consider.

Conclusion
There is intense pressure on academic institutions to take advantage of the electronic and technological advances and to provide their students and faculty with the latest and best resources. A vendor may imply that their full-text database must be acquired immediately to keep up with other institutions and to save money. Full-text databases can provide excellent access to information, but they are expensive and the decision to acquire one or more requires careful consideration. Review the matrix, and allow staff time to evaluate the databases with the individual institution’s needs in mind. Negotiate for the best licensing terms and subscription costs. In this way a librarian can decide which database will best satisfy the needs of the individual institution.—Rebecca Johnson is Team Leader for Information Service, and E. Ann Ford is Reference Librarian, University of Iowa Libraries.

Further Readings
Tenopir, Carol and Barry, Jeff. “Data dealers face stormy weather,” Library
Journal 123, no.9 (May 15, 1998): 38-46.
Hawbaker, C. Craig, and Wagner, Cynthia K. “Periodical Ownership versus Fulltext Online Access: A Cost-Benefit Analysis,” The Journal of Academic Librarianship 22, no. 2 (March 1996): 105-110.
Crawford, Gregory A., and White, Gary W. “Full-Text and Libraries: Issues and Implications for Libraries, Librarians, and Library Patrons,” Proceedings of the 18th National Online Meeting, May 13-15, 1997, compiled by Martha E. Williams, and Thomas H. Hogan. (Medford, NJ: Learned Information, 1997.)

 

Information on Database

Database Marketing Company Parent company	Cost	Coverage (in yrs)/ How Current	#Journals Indexed /#Full-Text	Full-Text search?	Photos Illus.?	Web Hardware, Software	Search Engine	Right to Retain	Method of User Authentication
InfoTrac SearchBank Expanded Academic ASAP Information Access Co The Thompson Corp.	2 models: unlimited access or simultaneous users	1980-current yrs. Updated every business day: daily sources indexed daily, weekly within 1-3 days of receipt, others prioritzed. Full text available within 3 weeks of publication.	1580 titiles indexed and abstracted /636 Full-text	Yes (in PowerTrac, keywork, text Word field)	Yes	Web access with graphical Web client/RS6000 AIX 3.2.5, DEC VAX VMS 5.5-2, DEC Alpha VMS 6.0, DEC Alpha OSF/1 2.0, HP9000 UX, Sequent Dynix/ptx 1.4, Sun Sparc Solaris * OS, UnixWare 1.1. Min speed of 56 kbs for Internet connection.	InfoTrac search engine, provides two modes of searching, EasyTrac and PowerTrac	No	IP address: Libraries using proxy servers must authenticate patrons; remote access via IAC's Remote Patron Authentication Pgm
ElectricLibrary Infonautics Corp Infonautics Corp.	$1995 per single concurrent user. Cost decrases as vol. increases (pricing on case-by-case basis)	1 to 13 years with average 3 years archive. Content loaded daily: daily sources current within 24-48 hrs, monthly within month, wkly within wk	634 Indexed /634 Full-text. Also includes 65,679 gov /TV/radio transcripts	Yes	Yes	Windows 386 or 3.1, Mac 68020 or better, Internet TCP/IP or dialPPP or SLIP connection	None	No	IP address, remote access or separte passwords
EBSCOhost EBSCO Publishing EBSCO Information Services	contact EBSCO Publishing	1990-current Updated daily	3,200 Indexed & abstracted (2,000 peer-reviewed) /OVER 1,000 Full-text	Yes	Yes	Web-based Designed to work in a variety of library settings, with all major integrated library systems & online public access.	Available platforms: WWW/Windows/ASCII (unix, VMS, dumb terminals).sun Solaris/Mac	CD-ROM backup for minimal cots	IP address, remote access or separate passwords
Periodicals Abstracts Research II Umi Bell & Howell	Varies	Citations for 1986+ /full text for 1992+ Updated daily	1,800 Indexed /700 Full-text	Yes	Yes– for many titiles	Web-based	Netscape 2.02 or higher (or comparable)	No	IP address
Wilson Omnifile V, Full test ed. HW Wilson Co HW Wilson Co	1 user –$8,450 2-4 users –$12,675 5--8 users –$14,788 9-12 users –$19,013 13-20 users –$25,350	Coverage since Aug.1982 Abstracting since 1984. Full-text since Jan. 1995. Weekly updates via Web only, monthly via magnetic tape	1,644 Indexed /499 Full-text	No	No	Web-based	Related to configuration	Yes	IP address or ID/password
LEXIS-NEXIS Congressional Information Service, Inc. same (CIS, Inc)	$9,000/yr for up to 1,500 FTEs. For "charter subscribers": $1.69/student for 500,00 FTEs, $1.20/FTE for 500,001+ students	Coverage specific to source (20+yrs for many news sources, 200+ yrs for Supreme Ct, etc) – see www.cispubs.com Updated several times daily.	5,718 Full-text /Not indexed	No –limitations vary	No	Web-based	None	No	IP address

Library Issues: Briefings for Faculty and Administrators (ISSN 0734-3035) is published bimonthly beginning September 1980 by Mountainside Publishing Co., Inc., 321 S. Main St., #213, Ann Arbor, MI 48104; (734) 662-3925. Library Issues, Vol. 19, no. 3. &COPY;1999 by Mountainside Publishing Co., Inc. Subscriptions: $75/one year; $140/two years. Additional subscriptions to same address $25 each/year. Address all correspondence to Library Issues, P.O. Box 8330, Ann Arbor, MI 48107. (Fax: 734-662-4450; E-mail: apdougherty@CompuServe.com) Subscribers have permission to photocopy articles free of charge for distribution on their own campus. Library Issues is available online with a password at <http://www.LibraryIssues.com>