1.8.6 Digital library studies

Mellon report

This was a study prepared for the Andrew W. Mellon Foundation and completed in 1994. The study, outlined in [Mel, 1992] and summarised in [Okerson, 1996], used Association of Research Libraries (ARL) statistics to study 24 major US research libraries. The study examined acquisitions expenditure and patterns, increases in serials prices, and possible technological solutions.

Core Project

The Chemistry Online Retrieval Experiment (CORE) project was carried out at Cornell University from 1990 through 1995. It was a collaborative effort of Bellcore (Bell Communications Research), Mann Library (Cornell), the American Chemical Society and the OCLC Online Computer Library Center, Inc. From the perspective of the Mann library, the primary goal of CORE was to "create a system with a substantial depth and breadth within a single discipline (chemistry) - a critical mass" [Entlich, 1995, p. 114]. CORE ended up processing over 400,000 page images from twenty ACS journals. The product of this processing was both bitmap and SGML versions of the pages together with hypertext citation links, searching and support for graphics and special symbols. In fact, the average article included:

• text from two different sources
• two full-page bitmaps/page (at 100 and 300 dpi)
• each figure in a separate file
• thumbnail reductions of the figures
• equations in a separate file

CORE included a range of investigations carried out as part of the project. Three studies in particular are relevant to this thesis.

The first was intended to examine the interaction between scholars and the journal literature [Olsen, 1994]. The broad questions were:

What are the scholars' purposes in using the literature and what are the actual processes carried out by scholars with journal literature which must be accommodated by the electronic version of journals? [Olsen, 1994, p. 7].

This study was based on interviews with forty eight scholars (sixteen each from chemistry, sociology and English) based at Cornell University and the University of Pennsylvania. These were not scholars who had been exposed to the CORE system and had in fact never previously used electronic journals. The questions dealt with:

• reasons and techniques for locating journal literature
• methods of reading this literature
• when and where the reading too place
• which literature was most useful
• the advantages and disadvantages of printed journal literature compared with the electronic journal
• general demographics
• computer use.

The second study focused on users' interaction with the e-journal interface [Stewart, 1996]. The data was derived from "open-ended, hour-long interviews with thirty-nine (sic.) users of the Chemistry Online Retrieval Experiment (CORE) at Cornell University" [Stewart, 1996, p. 340]. Unfortunately only eight of the thirty nine subjects were faculty members - the bulk of the remainder were graduate students; this can be assumed to have affected both the age profile and response pattern. The questions used built on the system functions identified by scholars in the earlier research [Olsen, 1994] as essential in an electronic system. Subjects were asked to evaluate the importance of features for selecting optimal research literature and reading that literature. They were also asked about selected characteristics of an ideal system and impact of an electronic system on reading productivity.

The third study [Entlich et al., 1996] examined the system logs and used online questionnaires to supplement the interview data reported in [Stewart, 1996]. Of the 161 accounts distributed for the system only 75 came into active use. Comparison of the logs with the fact to face interviews enabled the researchers to compare the users stated desires and intentions with their actual behaviour. The online questionnaires were introduced during the final six months of the trial and consisted of four randomly selected multiple choice questions presented at the end of a user's session.

Project ELVYN

Project ELVYN (allegedly from ELectronic Versions - whY Not) was an initiative to take a specific new print journal, Modelling and Simulation in Materials Science and Engineering (MSMSE) from the Institute of Physics Publishing and deliver it electronically [McKnight et al., 1994], [Rowland et al., 1995]. The project was particularly concerned with the problems of transmitting a journal from publisher to reader and the costs associated with this [Meadows et al., 1995]. No journal useability studies were undertaken.

TULIP

The University Licensing Program (TULIP) started in early 1991 and concluded at the end of 1995 [Elsevier Science, 1996]. Its participants were Elsevier Science and nine US universities: Carnegie Mellon, Cornell, Georgia Institute of Technology, M.I.T., University of California, University of Michigan, University of Tennessee, University of Washington, and Virginia Polytechnic and State University (Virginia Tech.). The goal was to test systems for networked delivery of journal content to the user's desktop as well as the use of these systems. Elsevier delivered the content (scanned page images, dirty OCR, and raw ASCII full text) to the universities who then or adapted their own systems to manage and deliver this content. The focus of the research was on "technical issues, on user behaviour and on organization and economic questions" [Elsevier Science, 1996, p. 7].

The user behaviour research was both quantitative and qualitative. The quantitative work was based on logfile analysis from the sites. The qualitative research was based on four focus groups with around ten graduate students each and one to one interviews with twenty six university faculty. The questions dealt with:

• current information needs patterns
• access to and use of electronic information
• a detailed evaluation of TULIP
• professional journal usage.

ARL SPEC Survey

Early in 1994 a Systems and Procedures Exchange Center (SPEC) survey was sent to 119 members of the Association of Research Libraries (ARL) in the U.S. The study aimed to determine "the policies and practices of ARL libraries in selecting, acquiring and providing access to electronic journals" [Parang and Saunders, 1994a, p. 1]. A total of 77 questionnaires were returned (65% response rate). The survey also sought to identify issues and trends in the ways ARL libraries were planning to deal with electronic journals [Parang and Saunders, 1994b].