Human and Economic Factors Affecting the Acceptance of Electronic Journals by Readers
Abstract: Human factors that influence the acceptability of electronic journals to users include the interface, the hypertext structure, the searching and browsing functionalities, and the speed of the network response. Economic factors that influence the progress of electronic journals include not only the level of prices charged, but also the pricing structures adopted. Three recent studies of the acceptance of electronic journals by users and other published work are considered. It is concluded that the continued provision of high-quality scholarly publications to readers in the electronic era will require that publishers provide the needed design quality at prices that are significantly lower than those charged today for printed journals.
Résumé: Parmi les influences humaines sur le degré d'acceptabilité de journaux électroniques pour les lecteurs, il y a l'interface, la structure hypertextuelle, la facilité de faire des recherches et d'explorer, et le temps de réponse d'un réseau. Parmi les influences économiques, il y a non seulement le montant chargé pour un journal électronique, mais aussi les structures de prix adoptées. Nous considérons trois études récentes sur l'acceptabilité de journaux électroniques pour les usagers et ces travaux et d'autres études publiées. Nous concluons que les éditeurs de journaux électroniques devront fournir une mise en page comparable à celles de journaux publiés, mais ce à un prix bien plus bas, s'ils veulent continuer à produire des publications savantes de haute qualité pour leurs lecteurs à l'ère électronique.
Introduction
Electronic publication of scholarly research findings has been the subject of a great deal of study, development, and comment over the last few years (Rowland, 1995). Much of this work has been technical and has concerned the provision of workable and trouble-free channels for the dissemination of information in electronic form. There has also been much comment, a large proportion of it taking place through electronic discussion lists and newsgroups, about the attitude of scholars to electronic publication. Most of this comment has, however, been concerned with the scholar as author and in particular has dealt with the acceptability of electronic publications for the purpose of measurement of scholarly output in considerations of job appointments, tenure, promotion, and research funding. Relatively few workers have considered the response of information users to electronic publications; this is unsurprising, because until recently the number of scholarly publications available in electronic form was small, and therefore there was necessarily little usage to observe.
This situation is now changing rapidly. The number of genuinely scholarly electronic periodicals has been charted regularly by such observers as Bailey (1997) and Okerson (1996), and, while still small compared with the total number of printed scholarly journals, the number of titles available free of charge over the Internet is now substantial. Harter & Kim (1996), however, have noted that the average size of these periodicals (in terms of number of articles published per annum) is small. Thus, if one measures publication by number of articles rather than number of periodicals, the electronic proportion becomes smaller.
The major breakthrough, however, has come with the decisions by most major publishers of established printed scholarly journals to make their titles available in parallel electronic and printed forms. Some publishers started their programs of conversion in 1995, more in 1996, and yet more in 1997. In the scientific, technical, and medical (STM) field, Hitchcock, Carr, & Hall (1996) looked at the situation at the close of 1995 and aptly subtitled their report The Calm before the Storm. The ready availability of hundreds of titles in electronic form, including many that are core journals of their respective scholarly disciplines, means that one can now mount realistic user studies of electronic scholarly journals.
Early electronic journal studies, such as the work of Senders (1977) and the BLEND and Quartet studies (Shackel, 1991; Tuck, McKnight, Hayet, & Archer, 1990), foundered largely on the inadequacy of the technical infrastructure available. Too few potential users had easy access to terminals; available networks were inadequate in capacity and quality of transmission; there was too much incompatibility between different software packages. Today, these constraints have been largely overcome, at least in the developed world. Although much debate continues on electronic publishing discussion lists and newsgroups about detailed technical matters, problems on the technical side of electronic publishing may be regarded as soluble where they are not already solved.
Effective availability of electronic journals to users, especially to students and other users in universities, will therefore now depend on non-technical factors: principally the ergonomic acceptability of the interface provided and the economic availability of the journals to these users at an affordable price. These two types of factor are the subject of this paper.
Human factors: User studies of electronic journals
Several research studies on electronic journals have been undertaken at Loughborough University. Three -- ELVYN, InfoTrain, and Café Jus -- were funded by the British Library Research and Development Department (BLR&DD) (now renamed the British Library Research and Innovation Centre) (Meadows, Rowland, & Yates-Mercer, 1997; Rowland, McKnight, & Meadows, 1995; Woodward, McKnight, Meadows, Pritchett, & Rowland, 1997a, 1997b). Another two were the topics of masters' dissertations (Bell, 1996; Falconer, 1995). Two further BLR&DD-funded studies, one at Loughborough University concerning information-seeking behaviour of biological scientists (Rolinson, Meadows, & Smith, 1995) and one at the University of Sheffield (Wood, Ford, Miller, Sobczyk, & Duffin, 1996), which looked inter alia at the electronic information-seeking activities of academics at that university, are also examined.
ELVYN
The ELVYN (Electronic Versions: Why Not?) study (Rowland et al., 1995) focused on the electronic version of a single newly established journal from an established scientific publisher (Institute of Physics Publishing [IoPP]) entitled Modelling and Simulation in Materials Science and Engineering (MSMSE). IoPP wished to test a model that included electronic delivery of files to each individual institution. The journal was mounted on local machines at each participating university, for delivery over their own campus network. Part of the study involved testing usage of the electronic version of MSMSE with groups of materials-science students in each of the participating universities.
One important conclusion from the ELVYN study was that trained information specialists, when contemplating information retrieval, tend to think of a search mode for accessing information -- that is, that the user arrives at the system with a particular information need and requires documents that provide the information. Academics and researchers, on the other hand, tend to assume a browsing mode -- they look quickly through the latest issues of journals, trying to spot interesting articles. Interfaces have to allow for both modes of retrieval.
In ELVYN we interviewed potential users before the electronic journal system was designed, in order to provide facilities that the users wanted. However, even when a major effort had been made to provide the functions that materials scientists said that they wanted, it was difficult to recruit users for the experimental journal. Although the journal was newly launched in printed form, users with a research interest in its subject field had been sought. The conclusion that we drew from this was that to make it worthwhile for users to learn to use an electronic journals system, there needed to be a critical mass of journals available. They did not see the point in climbing the learning curve to use one journal.
ELVYN used several different methods to distribute the data at different universities. The most commonly used was a World Wide Web (WWW) interface, with the papers tagged in hypertext markup language (HTML) that was algorithmically derived from the publisher's standard generalized markup language (SGML) coding. At that time, HTML could not handle special characters (the Greek alphabet and mathematical symbols), so these items had to be converted algorithmically into small in-line graphics. Downloading and display of these graphics slowed down considerably the delivery of files to the user. Furthermore, the early version of HTML in use at that time did not allow very ready provision of tables. The figures in the papers had to be displayed by another viewer program (External Viewer, xv), which had to be launched each time a user wished to view a graphic. All of these problems have, of course, to varying degrees since been solved. But the difficulty encountered in ELVYN in providing the service required by materials scientists did emphasize that scientific users require a solution to the problems of graphics, tables, and special characters before an electronic journal is regarded as satisfactory.
Another university was using a WAIS server at the time. Users could search the full texts but not display their hits; they had to request a printout from the university computer centre. Not surprisingly, at this location "Lack of figures, tables, and equations" was listed as the chief disadvantage, but "Fast searching" as the chief advantage.
InfoTrain
The InfoTrain (Training Electronic Journal) project (Meadows et al., 1997) created an electronic journal jointly between three universities (two in the U.K. and one in Australia), specifically for the purpose of teaching information-science students how to create, how to use, and how to assist clients to use electronic journals. This study was predicated on the assumption that one could not modify or alter the structure of a real electronic journal available on the World Wide Web, but a journal created specifically for teaching purposes could be modified by students themselves in the cause of active learning. Though this was not explicitly a user study, the initial responses of information students to design features of the InfoTrain journal, and the suggestions for improvement that they have made, provided useful data about human factors affecting the acceptability of electronic journals to real users.
One part of the course work undertaken by Loughborough undergraduate information studies students using the InfoTrain electronic journal involved a comparative evaluation of several electronic journals from a design point of view. They were asked to compare the designs used at the other two participating universities (the City University in London and the University of South Australia) with "real" electronic journals, one free and one commercial. Favourable and adverse comments were equal in number for the commercial journal, but were about 2:1 adverse to favourable for the other InfoTrain versions and for the free electronic journal. This latter ratio suggests that the application of professional design skills to an electronic journal does have some beneficial effects. Among the adverse comments were such things as: "Unable to see document as a whole," "Unable to change onscreen text size," "Hard to distinguish detail," which referred to the full-text display, "Problems accessing tables and figures," "Problems locating articles," and "Help function is of little use," which referred to difficulties in hypertext navigation.
Café Jus
Café Jus (Commercial and Free Electronic Journals User Study) is the most recent of these projects (Woodward et al., 1997a, 1997b). It took advantage of the much greater number of electronic journals available over the WWW in 1996 than before and investigated the usage of many different electronic journals, both commercial and free, by academic staff and postgraduate students in numerous departments at Loughborough University. Again its purpose was to cast light on those human factors that might limit or encourage the rapid acceptance of electronic journals by users. Important to this study, as to the earlier ELVYN study, was the fact that the users concerned were not necessarily enthusiasts for, or especially knowledgeable about, electronic journals or indeed information technology itself. They could be regarded as a cross-section of normal, average users of academic information -- albeit located in a university where administrators at the highest levels hold considerable enthusiasm for information technology (IT), networking, and the Internet.
Groups of postgraduate students from six departments spread across engineering, science, and social science disciplines were shown how to access electronic journals, of which over 300 were provided by the university library at the time. They were asked to perform searches on one or more journals with subject coverage that was of real interest to them and then fill in a questionnaire giving their views of the journals (similar to the one used with undergraduates in the InfoTrain study). The most common difficulty was slow access time, especially for commercial publishers' sites that carry a large number of titles. Publishers need to recognize that if access via the Internet is to become a major feature of their product lines, they will need to invest in adequate hardware, software, and network links to provide reasonable speed of access to many users simultaneously. Recent anecdotal evidence, quoted in the general press in the U.K., suggests that Elsevier, at least, has appreciated this point. When launching their ScienceDirect service, which will include all Elsevier's titles and will be at least 10 times larger than any of the other publishers' services currently running, Elsevier admitted that this service when fully operational will be the largest installation linked to the Internet anywhere.
To reach an individual journal article from first logging on to the publisher's home page often involves passing through many navigational screens, and when access is slow this can be very frustrating. Again, as in ELVYN, there seems to be a lack of recognition that users wish to browse as well as search. In some cases, pretty appearance, dictated perhaps by the publisher's marketing department, seems to have taken precedence over efficiency of hypertext navigation (for example, by it not being obvious which colour of text is used for hotspots for hypertext links).
Most of the commercial journals, but few of the free ones, use page-integrity viewers such as Adobe Acrobat or RealPage. These require the users to have a copy of the viewer on their own machine; this is free of charge, but does have to be downloaded from a WWW site and does take up space on the user's machine. Students use open-access laboratories, where there is clearly competition for space on the machines with other more versatile programs such as statistics packages, spreadsheets, and graphics software. Publishers favour these viewers because, while permitting full-text searching, they preserve the design and appearance of the printed page. For users, however, this is not necessarily advantageous. Most machines now have Netscape (if not always the latest version), and so journals presented in HTML format can be read without the need for further software. Furthermore, the different shape of a printed page (vertical format) and the computer screen (horizontal) means that printed page designs, especially two- or three-column layouts, are unpleasant to read on the screen if page integrity is preserved. Many of the adverse comments in the questionnaire survey referred to the failings of Acrobat; its icons are not intuitive, for example, so it was not obvious to students that they could zoom to make the text larger and more legible. The attachment of publishers to page integrity may be misplaced, and may show a lack of appreciation that design criteria for an electronic publication are different from those of a printed one. The importance of really effective and intuitive hypertext link structures needs to be stressed. In this connection the Open Journal project, at Southampton University in the U.K., needs to be mentioned (Hitchcock, 1996). It seeks to provide software that facilitates the provision of the hypertext link structure separately from the information content.
Despite these criticisms, two thirds of the students found the electronic journals easier to access than the print versions. About the same proportion found the print version easier to use. This apparent contradiction is rationalized by suggesting that the hypertext navigational characteristics of electronic journals are appreciated as an aid to searching and browsing, but that once one has arrived at the full text that is wanted, it is more pleasant to read it in the print version.
Users in industry
One of us (Ian Bell) undertook research for a master's dissertation in a major pharmaceutical company, investigating attitudes to and usage of electronic journals by both information staff and laboratory research staff there (Bell, 1996; Bell & Rowland, 1997). This followed up an earlier comparative study (Rolinson et al., 1995) when the usage of IT generally was observed at the same company and compared with usage in other biological research establishments and biological departments in universities in the U.K. This study had demonstrated a considerably greater usage of electronic information sources in the industrial situation than in the academic world or government research laboratories, but was undertaken before many primary research journals were available in electronic form.
Bell used questionnaire surveys with two groups in the company: information services staff and laboratory research workers. He also interviewed managers in both types of department (information and research). The most striking findings were that the research staff were not only better informed about electronic journals and keener to use them than the information services staff assumed, but the researchers were also better informed about electronic journals than information services staff. This was a surprising finding, given that the information services department at the company in question has a high reputation in the U.K. information world. This finding contrasted sharply with the findings of our various user studies at Loughborough University.
In the pharmaceutical industry, it seems, the end users are now ready to use electronic information resources. However, they also had a lot more respect for electronic versions of established printed journals than they did for new electronic-only ones. This was because the recognized quality of the leading journals was of paramount importance. One research worker said, "Greater convenience is not sufficient justification for threatening the quality of our research enterprise." There was agreement among both information and research staff that both categories of worker needed training about electronic journals -- and we in the universities would agree that we have a similar need. Research staff also appreciated the "gatekeeper" role of information staff: the work of evaluating, sorting, and sifting the mass of available information on behalf of the research staff remains a task for information professionals.
Finally, there remained concern about how electronic journal publications will be paid for. Again, industry and academia share this concern, which leads on to the other major topic of this paper: the economics of electronic scholarly journals.
Economic factors
There has been a great deal of controversy about the appropriate level of pricing for scholarly information. Publishers and subscription agents contend that, if electronic journals are to provide a high level of quality, in all senses of that word, then costs cannot be much lower than at present. Even if printing ceases altogether, only the costs of paper stock, print machining, and distribution will be saved, and those costs typically make up less than one-quarter of the price of a scholarly journal. Indeed, publishers maintain that as long as dual publication (electronic and print) continues, their costs will actually be higher than before. Meanwhile, some academics and academic librarians, especially in North America, believe that by adopting a zero-budgeting approach they could devise workable publishing mechanisms using the Internet that need cost no more than 25% to 30% of traditional journal prices (E-journal costs, 1995). Whatever the truth of the matter, there is certainly discontent that publishers make substantial profits out of research results that are provided to them by the academic community free of charge.
These matters will be decided by pragmatic rather than moral arguments. The major groups of customers for scholarly information are found in universities -- academics, research staff, and students. Outside North America, personal purchases of scholarly journals out of one's private finances are few. Most scholars and students depend on their university to provide them with information, either by "holdings" -- actual ownership of books, journals, and CD-ROMs by the university library -- or by "access" -- provision of information services to networked PCs on academics' desks, regardless of where the information may actually be held. The existence of e-mail discussion groups, newsgroups, gophers, and especially the World Wide Web has led to academics and students becoming used to the idea of free information; indeed, academics find that many students now look first at the WWW for all information and have to be reminded of the need to look in printed sources and other electronic sources as well.
For several years now, academic libraries in most developed countries have found their resources under considerable strain. This is due to a number of factors, among them the widespread election of governments committed to lower taxation and lower public expenditure, and the growth of student numbers leading to a greater emphasis on the university library as an undergraduate teaching resource rather than a research resource. In developing countries, of course, resources for university libraries have always been inadequate. Journal subscription prices regularly rise by percentages much greater than the general rate of price inflation. There are a number of reasons for this: journals may publish more pages, leading to higher prices even if the price per page remains the same; new journals are launched, leading to an increase in the cost of maintaining a comprehensive collection in a given subject field; currency exchange rate fluctuations can lead to sharp increases on imported journals; but, above all, there is a vicious spiral whereby as prices rise, sales fall. Fixed costs then have to be spread among fewer sales, leading to further price increases, further falls in circulation, and so on. Whatever the reason, libraries see their costs increasing out of reach of their budgets. Rightly or wrongly, they see electronic publication as a potential solution to this problem.
As noted earlier, most of the free Internet journals are quite small (Harter & Kim, 1996). By contrast, many traditional journals are very large; for example, the Journal of Biological Chemistry, published by the American Society for Biochemistry and Molecular Biology and available on-line since April 1995, publishes over 30,000 pages per annum. The even larger Biochimica et Biophysica Acta is published by Elsevier and will presumably become available in electronic form during 1997 as part of Elsevier's ScienceDirect program to make all of its journals available on the Internet. To publish this number of pages the journals' staffs must have processed an even larger number, since some papers will have been rejected, and these too generate work. It has therefore been argued elsewhere (Rowland, 1996) that the sheer size of such operations necessitates the employment of full-time administrative staff of a reasonable quality, regardless of the medium employed for distribution. Thus, provided that some fundamental pattern of publication not unlike the present one persists, journals are likely to continue to be sold at substantial prices, though not necessarily as high as at present.
There are, however, other possible patterns. It has been suggested many times -- going right back to Bernal (1939) -- that primary dissemination of research results might be achieved by some form of central repository system. This idea was revived in the U.K. more recently by Swinnerton-Dyer (1992), the mathematician who at one time headed the Universities Funding Council. Computer and telecommunications technologies have made this idea more feasible than it was in Bernal's day, when the repository would have had to operate in a manual way. Bernal and Swinnerton-Dyer's ideas assume, of course, that the fundamental purpose of publication is to provide ongoing availability of tested, quality-controlled information. At the other end of the spectrum, one finds the Internet libertarians, who regard any form of refereeing or quality control an affront to free speech and a form of Establishment censorship. They would see free discussion over the Internet as the major form of communication among scholars and believe that the only reason for a refereeing system in the past was the expense of paper publication: material needed to be assessed prior to publication as one could not afford to print everything. Now, one can put everything up on the WWW and allow all scholars -- not just a couple of referees -- to judge its worth. To this argument, one may respond that there is another scarce resource, the scholar's time. No one has the time to read everything that is written in their subject area, the good, the indifferent, and the crackpot. Referees do us a service by filtering out work that is not worth others wasting their time on. They also often make useful suggestions for the improvement of those articles that are published.
Another widely advanced suggestion is that the university where research originated should publish it, rather than transferring copyright to a publisher who then makes money from it. Each university would have a WWW site containing quality-controlled work produced by its own academics and students, and WWW search engines would enable all other workers elsewhere to access relevant material. A variant on this idea, which would preserve the useful concept of journals specialized by subject, would be for each university to publish a limited number of journals electronically and for universities to agree amongst themselves as to who would cover which subjects. Gail McMillan (1994) sees this as moving from a position where each university library provides many journals for (relatively) few people -- the community of that one university -- to one where it provides a few journals for an enormous number of people -- essentially the whole world. These models are attractive, and there is no superficially obvious reason why they could not work, but anecdotal evidence suggests that internal university politics might be a major obstacle to their effective and impartial operation. One also quails at the thought of the complex multilateral, indeed multinational, negotiations between universities that would be needed to set up the McMillan model.
Despite all these alternative suggestions, it seems unlikely that existing publishers will quietly retire from the field of scholarly journal publishing. For-profit companies will continue to seek a financial return on their properties, and established journals of repute are very valuable properties. Not-for-profit publishers such as university presses and learned societies, on the other hand, see themselves as retaining a role in the control of scholarly quality through their publishing operations. Though both types of publisher will no doubt wish to remain in this field, they will only do so if they can provide a service to scholarly communities at a price those communities can afford. At present they seem intent on charging prices for electronic journals that are the same as, or higher than, those of their printed journals. There is little general difference between the for-profit and not-for-profit sectors in this respect; however, certain not-for-profit publishers, such as the Johns Hopkins University Press with its Project Muse (Johns Hopkins, 1997) and the Association for Computing Machinery (Denning, 1996), seem to be displaying a more flexible, less finance-oriented approach than some others. But the American Society for Biochemistry and Molecular Biology, which has done an excellent job on the design side with the electronic version of the Journal of Biological Chemistry, has sadly used the occasion to bring in a general price increase.
There has also been considerable debate about pricing structures for electronic publication, in contrast to the question of the absolute level of prices. The available options were usefully summarized by Tom Graham, then university librarian at the University of York in the U.K. and recently moved to the University of Newcastle upon Tyne (Graham, 1996). These ranged from national site licence agreements -- as being experimented with at present in U.K. higher education (Bekhradnia, 1995) -- through institutional site licences, to personal subscriptions with alternatives of pay-per-access or pay-per-delivery for occasional users, to grants for publication (the electronic version of page charges). The central problem for publishers is to ensure that the first-copy costs are covered from the various sources of income. The universities' ideal is that on payment of a known annual sum (preferably smaller than that now paid for the printed version) all bona fide users on their campus -- staff and students -- can have unlimited access to each journal from their own desks or from open-access student PC laboratories. Access from machines in halls of residence would also be permitted, but problems arise with off-campus students, staff working from home, and especially distance learners and part-time students studying for only a few hours per week. Where and how does one draw the line between users who are paid for by a campus licence and those who are accessing the journal from home and might reasonably be expected to pay either a personal subscription or on a pay-as-you-go basis? Efforts have been made by CD-ROM publishers and some producers of computer-assisted learning (CAL) materials to limit the number of simultaneous accesses to their products on the campus: thus the licence might allow only 10 simultaneous connections to a networked CD-ROM, say. This might be workable for journals held in electronic form locally (analogous to a networked CD-ROM), and perhaps might work for files held remotely if the publisher had suitable software to identify the origin of the search, though caching and mirroring will certainly render such an approach more difficult.
Given the shortage of funds for university libraries this is not a tenable position. New technology must be used to provide improved value for money and improved access for users. If existing publishers refuse to provide this, they can only expect the pressures for them to be bypassed will increase and, in that case, other players will undoubtedly arise to fill the need.
Conclusions
On the basis of work undertaken at Loughborough (Bell & Rowland, 1997; Meadows et al., 1997; Rowland et al., 1995; Woodward et al., 1997a, 1997b) and other studies recorded in the published literature and summarized in Peek & Newby (1996), one arrives at the conclusion that the days when the growth of usage of electronic journals was limited by technical shortcomings of hardware, software, and networks are over. This view is subject to the caveat that it is largely applicable only to the developed world. Even there, there are often problems of inadequate capacity of the information superhighway, which at times seems to approach gridlock. Users in Europe are well aware of the need to do their netsurfing in the morning, when it is the middle of the night in North America! In our studies, users on the whole were fairly tolerant of the slow delivery of information. Thus, despite these frustrations, the infrastructure is now largely adequate, and computer and networking specialists continue to make steady improvements.
The factors limiting the acceptance and use of electronic journals by academics, researchers, and graduate students now are twofold: human and economic.
Electronic journals need to be designed for the computer screen, even though at the end of the search the user may wish to print out copies of a proportion of articles. In particular, the "navigational" screens -- the ones that the user has to go through en route to the actual articles of interest -- must be designed with particular care, as must the structure of hypertext links within and between articles, and the search engines provided should operate in a reasonably intuitive manner. Given that the network is often congested and slow, publishers must try to keep to a minimum the number of different screens the user needs to go through to get from the home page to an actual full-text article of interest. Change for its own sake should be avoided. Users find it hard enough to keep up with work in their own discipline; they should not be expected to try to keep up to date in the disciplines of electronic publishing and computer networking as well. Nor should publishers assume that everyone has the latest versions of all hardware and software; in universities, at least, they certainly do not.
Commercial publishers in general believe that "page integrity" -- the preservation in the electronic version of the layout of the pages from the printed version -- is of importance, and many therefore use systems such as Adobe Acrobat and RealPage to achieve this. It is not clear that users take the same view. Pages in a portrait (vertical) format do not fit well on the computer screen in landscape (horizontal) format, especially when a two- or three-column layout is used in the print version. Furthermore the capacity of the screen is smaller if the text is going to be presented at a size that is comfortable for users of all ages to read. Free electronic journals have no equivalent printed version, and so "pages" are irrelevant and their editors have not in general tried to make their journals resemble a printed publication in appearance. There is by now a substantial number of people with expertise in WWW design and where this expertise has been used in the design of electronic journals, there have sometimes been very satisfactory results.
The preference of established publishers for page integrity, and especially their liking for Adobe Acrobat, may be due more to economic than to ergonomic factors. If the publication is prepared for printing by using PostScript, the conversion to publication distribution file (PDF) format used by Acrobat is relatively straightforward and inexpensive. Conversion to HTML, on the other hand, is more costly unless the publishers use an SGML system for internal handling of their files. Thus it may be that the attraction to publishers of page integrity systems may not be, as they claim, a matter of user preference for page integrity, but more a matter of economy for the publisher.
The question of economy for the publishers, however, is of some interest to the user. If journals -- printed or electronic -- are beyond the means of the users or of their library to buy, then it becomes a matter of little interest how well designed they might be. Publishers clearly have been investing considerably in their electronic products and reasonably expect to find a return on their investments eventually. Much of this need for investment is clearly a consequence of the sheer size of major journals, as Elsevier has clearly recognized (and some academics have not). However, as commercial undertakings they also recognize the concept of charging what the market will bear; and there is mounting evidence that they have in some cases gone beyond that point already. Those British publishers who are collaborating in the Higher Education Funding Councils' (HEFC) National Site Licence Initiative (Bekhradnia, 1995) may well have a more realistic approach than some others. If publishers do not moderate their financial demands, they may find that the universities -- which, as well as being major customers for the journals, also provide much of the content free of royalty -- may decide to refuse to assign copyright to commercial publishers.
References
Bailey, C. W. (1997). Scholarly electronic publishing bibliography (5th ed.). URL: http:// info.lib.uh.edu /sepb /sepb.html
Bekhradnia, B. (1995). Pilot national site licence initiative for academic journals. Serials, 8(4), 247-250.
Bell, I. D. (1996). Issues surrounding electronic journals and their implementation within Glaxo Wellcome Research & Development. Unpublished master's dissertation, Loughborough University, Loughborough, UK.
Bell, I., & Rowland, F. (1997). E-Journals in an industrial environment. Serials, 10(1), 58-64.
Bernal, J. D. (1939). The social function of science. London: Routledge.
Denning, P. (1996, September). Electronic publishing plan a must. Computing Research News. URL: http: //www.acm.org /pubs /JournalsEnd.html
E-journal costs. (1995, June). Thread on the hyperjournal-forum. URL: gopher: //nisp.ncl.ac.uk:70/0R168907-171823-/ lists-f-j / hyperjournal-forum /archives /1995-06andlate
Falconer, Catherine. (1995). Publishers' opinions on the future of the academic electronic journal. Unpublished master's dissertation, Loughborough University, Loughborough, UK.
Graham, T. W. (1996, March 27). Electronic journals: Pricing models. Message sent to lis-elib electronic discussion list. E-mail: lis-elib@mailbase.ac.uk
Harter, S. P., & Kim, H. J. (1996, May). Electronic journals and scholarly communication: A citation and reference study. Paper presented at the American Society for Information Science (ASIS) Midyear Meeting, San Diego, CA. URL: http: //ezinfo.ucs.indiana.edu /~harter/ harter-asis96midtxt.htm
Hitchcock, S. (1996). Open journals. Ariadne, 5, 9. URL: http: //www.ariadne.ac. uk / issue5/open /
Hitchcock, S., Carr, L., & Hall, W. (1996, January). A survey of STM online journals 1990-95: The calm before the storm. URL: http: // journals.ecs.soton.ac.uk /survey /survey.html
Johns Hopkins University Press. (1997). Project Muse online journals. URL: http: //muse.jhu.edu /pricing.html
McMillan, Gail. (1994, July 27). The VT [Virginia Tech] model. Message posted to the VPIEJ-L discussion list.
Meadows, J., Rowland, F., & Yates-Mercer, P. (1997). An online electronic journal for teaching purposes. ALT-J, Association for Learning Technology Journal, 5(1), 13-18.
Okerson, A. (Ed.). (1996). Directory of electronic journals, newsletters and academic discussion lists (6th ed.). Washington, DC: Association of Research Libraries. URL: http: //www.gold.ac.uk / history / hyperjournal /arl.htm
Peek, R. P., & Newby, G. B. (Eds.). (1996). Scholarly publishing: The electronic frontier. Cambridge, MA: MIT Press.
Rolinson, J., Meadows, J., & Smith, H. (1995). Use of information technology by biological researchers. Journal of Information Science, 21(2), 133-139.
Rowland, F. (1995). Recent and current electronic journal projects. In F. Rowland, C. McKnight, & J. Meadows (Eds.), Project ELVYN: An experiment in electronic journal delivery: Facts, figures and findings (pp. 15-36). London: Bowker Saur.
Rowland, F. (1996). The need for management of electronic journals. In R. P. Peek & G. B. Newby (Eds.), Scholarly publishing: The electronic frontier (pp. 243-250). Cambridge, MA: MIT Press.
Rowland, F., McKnight, C., & Meadows, J. (Eds.). (1995). Project ELVYN: An experiment in electronic journal delivery: Facts, figures and findings. London: Bowker Saur.
Senders, J. W. (1977). An online scientific journal. Information Scientist, 11(1), 3-9.
Shackel, B. (1991). BLEND-9: Overview and appraisal (Research Paper No. 82). London: British Library Research and Development Department.
Swinnerton-Dyer, Sir H. P. F. (1992). A system of electronic journals for the United Kingdom. Serials, 5(3), 33-35.
Tuck, B., McKnight, C., Hayet, M., & Archer, D. (1990). Project Quartet (LIR Report No. 76). London: British Library Research and Development Department.
Wood, F. E., Ford, N., Miller, D., Sobczyk, G., & Duffin, R. (1996). Information skills, searching behaviour and cognitive styles for student centred learning: A computer assisted learning approach. Journal of Information Science, 22(2), 79-96.
Woodward, H., McKnight, C., Meadows, J., Pritchett, C., & Rowland, F. (1997a). Electronic journals: Myths and realities. Library Management, 18(3), 155-162.
Woodward, H., McKnight, C., Meadows, J., Pritchett, C., & Rowland, F. (1997b). Use of electronic journals by academic staff and postgraduate students in an information-literate university. In J. Zeeman & R. A. C. Bruyns (Eds.). New Book Economy: Proceedings of the 5th International BOBCATSSS Symposium, Budapest, Hungary (pp. 274-281). Amsterdam: Hogeschool van Amsterdam.