Survey of Electronic Journals in OCLC: The Extent and Quality of Cataloging

Taemin Kim Park
Serials Department
Indiana University Libraries
Bloomington, IN

Abstract

Electronic journals are becoming a major challenge for academic and research libraries. The Directory of Electronic Journals, Newsletters, and Academic Discussion Lists 1995 lists nearly 700 titles of e-serials. The Electronic version of the Directory is also available on the Internet (gopher://arl.cni.org/11/scomm/edir/edir95). New titles (newly created or newly discovered on the net) between annual edition of the Directory are announced electronically on the NewJour Listserv. This new phenomenon provides challenges, including bibliographic control and access.

This study addresses the bibliographic control and access of electronic journals available on the network in terms of the extent and the quality of cataloging by national bibliographic utilities and local libraries. The titles which were selected were peer-reviewed or had ISSNs from section 1: electronic journals, magazines, and zines, and section 2, electronic newsletters and other e-serials of the Directory of Electronic Journals, Newsletters, and Academic Discussion Lists 1995.

Each title in the sample was searched in OCLC to determine its availability in a national database as well as its holdings libraries. The extent of cataloging, the level of description of cataloging, the source of cataloging, or CONSER members, and the degree of subject analysis--classification numbers and/or subject headings--were investigated. Bibliographic access at the local library level was examined by accessing the holdings libraries' OPACs through the Internet and by examining the means of locating them at the local level.

Managing e-journals is a really important challenge for academic libraries. The significance of this study lies in its examination of the current state of bibliographic control and access of electronic journals in national bibliographic utilities and local libraries.

Introduction

Electronic journals are being increasingly added to national bibliographic databases and local library OPACs (online public access catalogs). Parang and Saunders (1994a) reported that the first edition of Directory of Electronic Journals, Newsletters, and Academic Discussion Lists contains about 30 journals while the 1995 ed. of the Directory shows over 675 electronic journals, newsletters, and other titles. The Chronicle of Higher Education reported 6 peer-reviewed e-journals in 1991 (Wilson, 1991), and today almost 141 of the 675 journals and newsletters are peer-reviewed according to this author's calculation.

There have been ongoing discussions and projects by the American Library Association, the Library of Congress, and national bibliographic utilities to learn and to provide effective bibliographic control and access to these new information resources. As are most other serials, e-journals are published with bibliographically-dynamic relations. For example, they can appear as printed counterparts, or as title changes of printed journals, or as new titles. Electronic journals assume a variety of forms, including diskette, CD-ROM, and networked. They are also called direct-access computer file serials and remote-access computer file serials respectively in the cataloging community.

The purpose of this study is to examine the degree and extent of bibliographic data of selected electronic journals in the national bibliographic utilities, and the means to provide access to them at individual libraries. This study's examination will be limited to networked e-journals.

Literature Review

The literature review for this study was supported by searching the CARL Uncover, LISA, ERIC, and by browsing current journals in print as well as in electronic format in the LIS literature. Discussions on appropriate Listservs (e.g., Serialst, Conserlst) were also monitored. The ARL's (Association of Research Libraries) survey results on electronic journals revealed that the matter of technical compatibility or incompatibility was the most common factor in the selection decisions of e-journals at individual libraries (Parang and Saunders, 1994a). Forty-three percent of ARL libraries have different criteria for selecting e-journals than those which are used for the selection of printed journals while 54% were following the same criteria (Parang and Saunders, 1994b). McKnight (1993) noted the importance of users in the success of e-journals by observing that "simply being electronic is not enough--users will only adopt electronic journals if they offer at least as much as the paper version, preferably more, including offering such important quality control mechanisms as the peer review or refereeing system" (p. 10). ISI (Institute for Scientific Information) announced a plan to include electronic journals in their citation index publications (1994).

In the discussion about cataloging Internet resources, Caplan (1994) summarized the major activities in terms of how the library communities--ALA, LC, and OCLC--have been working together to provide bibliographic access to electronic information resources. In the spring of 1992, OCLC's Internet Resource Project, ALA/ALCTS/LITA/RASD/MARBI (Committee on Representation in Machine-Readable Form of Bibliographic Information), the Library of Congress, and the Online audiovisual catalogers initiated a cataloging experiment involving remote-access electronic data resources. Resources were limited to text files of electronic books, journals, newsletters, reports. Thirty volunteer catalogers participated in this cataloging experiment project and cataloged a sample of 300 data resources following AACR2r, using the USMARC computer files format. Results of this project suggested some cataloging guidelines and recommendations needed to make changes to the computer files format, as MARBI Proposal 93-4, "Changes to the USMARC Bibliographic Format" to accommodate online information resources. One of the major changes resulting from MARBI Proposal 93-4 was the development of the new field 856 of the USMARC formats--Electronic Location and Access. The field contains the information necessary to locate and access Internet resources. Basically, field 856 provides the access methods (e-mail, ftp, telnet, etc.) needed to locate the remote access resources. More recent use of URLs in this field allows direct hpertext links in Web-based OPACs to the Internet resources. The state of ongoing changes is the main characteristics of cataloging the network-based resources. MARBI Proposal no.96-1, "Changes to field 856 in the USMARC formats" is already under discussion (Guenther, 1995). Erik Jul (1995) reported the OCLC Internet cataloging project, "Building a Catalog of Internet-Accessible Materials," which is a cooperative project among libraries worldwide to identify, select, and catalog computer files available via the Internet. Some 2500 records for internet resources are being cataloged for the Intercat catalog project.

Another interesting example of cooperation at the regional level, CIC/EJC (Committee On Institutional Cooperation/ Electronic Journals Collection) is coordinated and managed by the librarians of the CIC member universities, and the staffs of CICNet, and CIC (1995). According to CIC/EJC's home page, the EJC aims to be an authoritative source for electronic research and academic serials publication, by incorporating all freely distributed scholarly e serials. More than 800 electronic serials are archived on the CICNet gopher.

Electronic journal projects at individual libraries presented new challenges during the '90s. Virginia Polytechnic Institute and State University's electronic journal project was one of the first e-journal collections at a specific library. Gail McMillan (1991) described the broad issues of managing and processing electronic journals--file transfer problems (text vs. graphics), storage and access options, internal serials control, cataloging treatment, and holdings information. In the examination of issues of access and bibliographic control of electronic journals, Woodward and McKnight (1995) noted that libraries need to provide access to local collections as well as to materials outside their own libraries' walls. One noted difficulty from the user's perspective was the visibility of electronic journals, because many titles are published by specific discussion lists and bulletin boards. In a study of choice of formats for cataloging non-textual serials, Cecilia Leathem (1994) investigated a sample of 120 serial titles in audiovisual media, sound recordings, computer files, scores, and maps, by searching them in the OCLC database. The result of her study indicated that the format preference for bibliographic records by the library type showed that academic libraries chose the serials format for 59% of their cataloging records, followed by national libraries (50%), special libraries (14%), and public/school libraries (12%).

The first phase of USMARC Format integration was completed in early 1995. OCLC's Technical Bulletin (1995) explains that "format integration is the validation of all data elements for all forms of material, regardless of the record type (Type) and bibliographic level (Bib Lvl)." Phase 2 of format integration is involved with additions to and changes in the fixed fields and control fields and was implemented in March 1996. It allows flexibility within a selected workform to describe items with characteristics of more than one format.

Methods

A purposeful sample of electronic serial titles was selected from the Directory of Electronic Journals, Newsletters, and Academic Discussion Lists, 1995. This annual directory, which first appeared in 1991, provides access to academically-related serials on the Internet (Introduction page). The idea of using a random sampling procedure was rejected due to the nature of the Directory, which lists electronic journals, magazines, newsletters, zines, and other serials. Viewing the unstable nature of networked information, a more unconventional approach, has been taken in the sampling procedure. For example, the titles in the zines are mostly targeted for discussions among a very specialized groups of people, rather than as a main channel of knowledge dispersed in the scholarly community. It is more frequently assumed that journals which are peer-reviewed or refereed will be more likely to be selected for libraries' collections. A title having an ISSN implies a publisher's intent or commitment to publish continuously. Based on those assumptions, titles which are peer-reviewed or have ISSNs were considered for this study, in conformance with normal practice in the selection of library materials. Woodward and McKnight (1995,p.73) also supported those assumptions by noting that some electronic titles are not considered electronic journals because "non-peer-reviewed electronic communications exist primarily for the benefit of the individual researcher or research group ... where ideas can be put forward for discussion..." The possible bias resulting from these characteristics of purposeful sampling will be discussed in the implications section of this paper.

A total of 254 titles which were peer-reviewed or had ISSNs was selected from the 675 titles listed in section 1: Electronic journals, magazines, and zines, and section 2: Electronic newsletters and other e-serials of the Directory. One hundred twenty-three titles were peer-reviewed and 119 titles carried ISSNs from section 1; 18 titles were peer-reviewed and 77 titles had ISSNs from section 2. Each title from the total of 254 titles was searched in the OCLC database on December 26, 1995. This author searched them all for consistency and accuracy of the search results. If a record was found, the full record was printed. If any library's holdings were attached, the list of holding locations was printed as well.

Research Questions And Preliminary Findings

Bibliographic control and access of remote access (networked) e-journals were examined from several aspects. To gather data which was parallel to previous studies, research questions frequently raised regarding the bibliographic control of monographs (questions 1 and 2 from Leazer and Rohdy, 1995) were examined in this study. These included:

  1. the extent of bibliographic control
  2. the quality of bibliographic control
  3. the degree of bibliographic control at local libraries.
The extent of bibliographic control was based on the hit rate of the records found in the OCLC database. The quality of bibliographic control was measured by examining several elements of the bibliographic records.

Bibliographic control of remote access electronic journals by bibliographic utility

The success rate of locating a bibliographic record in OCLC was 46% (118 titles of 254 titles). All records in this sample were cataloged in the serials format except six records which were in the computer files format. Table A shows the original source of bibliographic records: the Library of Congress, other national agencies (NSDP), and OCLC member libraries. The result indicated that 62% of the records were created by NSDP (National Serials Data Program) which is the US agency of the International Serials Data System. NSDP is also a member of CONSER. The OCLC member libraries contributed 38% of the original cataloging records. It is not surprising to note that the National Serials Data Program is the major creator of original records of remote access e-serials in the OCLC database because they are also the US agency which provides ISSNs.

The degree of completeness of the bibliographic records is displayed in Table B. Sixty-one percentage of the records had full-level cataloging description input by OCLC members and/or CONSER members including the Library of Congress. The majority of records that were less than full-level were minimal-level records which were likely lacking subject headings and were prepublication records from NSDP. The quality of remote access e-serials records in OCLC is relatively high when one considers that 73% of the records were created, authenticated, and modified by CONSER members.

About 44% of the bibliographic records had LC classification data (Table C). The assignment of classification numbers to remote access serials may be perceived as less important because there is no need to provide shelving locations locally. However, a larger proportion of the bibliographic records (66%) had LC subject headings (Table D). Examination of field 856 of cataloging records--electronic location and access--illustrated various methods of accessing and locating remote electronic resources. Gopher was the most common access method, followed by e-mail, http, and ftp. About 58% of remote access journals were held by more than one OCLC member library--as many as 37 libraries held a single title. Analysis by publisher type indicated that university-based publishers (42%) were the leading producer of remote access serials. These included academic departments, research centers and laboratories affiliated with universities. This group was followed by commercial publishers (29%), other professional associations and museums (19%), and government bodies and individuals (10%).

Bibliographic control of remote access electronic journals by local libraries

For the degree of bibliographic control at local libraries, a sample of bibliographic records which had holdings attached was selected for examination. The OCLC holding symbols were checked to identify the names of the libraries, and a sample of the libraries' OPACs was accessed via Internet catalogs.

A preliminary analysis showed that there are different levels of access provided by local libraries. There are libraries providing local call numbers for remote access e-serials so that all the libraries' collections would be provided for the subject access. Other libraries did not provide call numbers for the remote access e-serials because there is no reason to maintain shelving locations for those materials. Libraries add local notes to direct the user to the electronic resources--to provide access to the journal issues themselves. There are many variations in notes, as well as means of access at local OPACs. Several models are identified from a small sample of titles searched in the local libraries.

Model A

Model B Model C Model D Local libraries offered several options which provided means of locating remote access electronic serials. Library users need to move around several screens in order to view the journal itself. Each library may have different levels of control based on an individual library's cataloging policy, design of library resources, and computing resources.

Implications and Significance

This study provides a snapshot of bibliographic control of remote access electronic serials in the OCLC database. The paper illustrates some descriptive data on various aspects of bibliographic control of electronic journals: the extent of bibliographic control, the quality of bibliographic control, and access control at bibliographic utilities and local libraries. The findings of this study are not generalizable due to the use of the purposeful sampling rather than random sampling. The selected titles may not be typical electronic serials and are more homogeneous among networked electronic serials because the titles are limited by either ISSN or having been peer-reviewed. There are also other directories which list electronic journal resources.

For comparison, Leazer and Rohdy (1995) studied several aspects of bibliographic control for general library material and foreign monographs received at a university library. The success rate of locating bibliographic records for general research material in OCLC was about 92%; the percentage of LC-created records was 75% (38% for foreign materials); 83% (86% for foreign materials) were full-level records; and 90% (23% for foreign materials) of bibliographic records had LC classifications and 90% (12% for foreign materials) had LC subject headings. The bibliographic control of foreign research material was significantly poorer compared to that of the general research materials.

The result of my study demonstrated that the success rate of locating a bibliographic record of a remote access electronic serial in OCLC was 46%; 62% of the cataloging records originated at NSDP, and 37% from member libraries. Sixty-one percentage were full-level records contributed by CONSER members and OCLC member libraries. Seventy-three percentage were CONSER authenticated records. For the aspect of subject access, 44% had LC classification information and 66% had LC subject headings. The extent of bibliographic control of remote access electronic serials in the national database was relatively low, at least compared to that of general research materials. The quality of e-serials records seemed to be high, considering that 73% were CONSER records. However, the extent and the quality of bibliographic control of e-serials in the OCLC database can not be meaningfully addressed due to the lack of other comparable studies.

One reason might be that this information resource is still new to cataloging agencies and communities. Another reason may lie in the difficulties of evaluating and managing electronic information resources--what titles should be selected and cataloged. Continuous cooperation among ALA, LC, and national bibliographic utilities will bring more remote access resources into national databases and local OPACs.

Local libraries need to provide a clear path to guide the user to the point where the remote access electronic journals can be easily viewed. As long as the local note is clear and easy to navigate, the matter of connecting to local servers, or to cooperative servers (e.g., CICNet), or to each file's URLs may be of less concern in terms of access. Fully cataloged e-serials in the national database would be more accessible and beneficial to end users, with local libraries' efforts of quality control, in terms of designing easy navigation and maintaining the correct location information for networked publications.

This study is significant because it addresses issues and problems of the bibliographic control of remote access electronic serials in bibliographic utilities and local libraries. My study presents descriptive data in relation to the extent, the quality, and the local control of bibliographic information of those new resources. This study demonstrates the potential issues of accessibility, as well as the acceptance of this new format of information resources by end users. The question of how the library will handle or manage electronic journals is increasingly providing us new challenges and opportunities.

References

CIC Electronic journals collection (1995, June 6) Available http://ejournals.cic.net/.

Caplan, P. (1994). Controlling E-journals: The Internet resources project, cataloging guidelines, and USMARC. Serials Librarian, 24, 3/4, 103-111.

Guenther, R. (1995, Dec. 5). MARBI Proposal No. 96-1. [Discussion]. INTERCAT [Online]. Available e-mail: Listserv@OCLC.ORG

Indexing electronic journals (1994 Oct.). Online Libraries and Microcomputers, 12(10), 3.

Jul, E. (1995). Building a catalog of Internet-accessible materials. Available http://www.oclc.org/oclc/man/catproj/catcall.htm

King, L.A. and Kovacs, D. (1995). Directory of Electronic Journals, Newsletters, and Academic Discussion Lists. Washington, DC. : Office of Scientific and Academic Pub., Association of Research Libraries.

Leathem, C.A. (1994). An examination of choice of formats for cataloging nontextual serials. Serials Review, 59-67.

Leazer, G.H. and Rohdy, M. (1995) The bibliographic control of foreign monographs: A review and baseline study. Library Resources & Technical Services, 39, 29-42.

McMillan G. (1991 winter). Electronic journals: Considerations for the present and future. Serials Review, 77-86.

McKnight, C. (1993 Jan.). Electronic journals--past, present... and future? ASLIB Proceedings, 45(1), 7-10.

Parang, E. and Saunders, L. (1994a). Electronic Journals in ARL Libraries: Issues and Trends. Spec Kit 202. Washington, DC : Association of Research Libraries.

Parang, E. and Saunders, L. (1994b). Electronic Journals in ARL Libraries: Policies and Procedures. Spec Kit 201. Washington, DC : Association of Research Libraries.

Technical Bulletin 206: Format integration phase 1 (9501 rev.) Available http://www.oclc.org/oclc/tb/tb206/tb2061.htm

Wilson, D.L. Testing time of electronic journals (1991). Chronicle of Higher Education, Sept 11, A22-A24.

Woodward, H. and McKnight, C. (1995). Electronic journals: Issues of access and bibliographic control. Serials Review, 21 (2), 71-78.

Table A: Source (original input) of bibliographic record

LC or NLC (1)
NSDP 62% (44)
Member 38% (73)
CONSER *73% (86)
n=118
LC (Library of Congress)
NLC (National Library of Canada)
Member (OCLC Member libraries)
NSDP (National Serials Data Program)
CONSER (Cooperative Online Serials Program)
*total records contributed by CONSER members

Table B: Level of bibliographic description

Full 61% (72)
Less-than-full 39% (46)
n=118

Table C: Presence of Library of Congress classification

Records with LCC 37 % (44)
Records without LCC 63 % (74)
n=118

Table D: Presence of Library of Congress Subject Headings

Records with LCSH 66% (78)
Records without LCSH 34% (40)
n=118

© 1996, American Society for Information Science. Permission to copy and distribute this document is hereby granted provided that this copyright notice is retained on all copies and that copies are not altered.