The libraries of institutions and organizations account for a majority of scientific journal subscriptions. The Journal of Biological Chemistry is no exception. There are approximately 1,200 institutional subscriptions compared to 200 individual subscriptions. It is reasonable to assume, therefore, that the perceptions of professional librarians about electronic journals will be an important determinant of whether the organization subscribes to the JBC or not.

The questionnaire whose results are shown below was sent to the e-mail addresses of approximately 300 institutional subscribers. Because there was no indication on the addresses of the status of the individual originating the subscription, recipients were asked to forward the questionnaire to the appropriate person responsible for making subscription decision.

Approximately 60 recipients responded for a 20 percent response rate. This is considerably lower than the response rate of scientists in the Editorial Board and individual subscriber surveys. The lower rate is due in part to the large number of foreign institutions that subscribe to JBC. They are substantially under represented in the results. It may also be due to the fact that the questionnaire was sent to the address of the person who initiated the subscription and not the individual(s) responsible for decisions regarding subscriptions.

The results of each question are presented in three parts: (1) a table showing the tabulated results, (2) interpretive comments that discuss the implications of the tabulation, and (3) verbatim comments of respondents that elaborated on their tabulated answers. Although these comments are extensive and somewhat time consuming to read, taken together, they offer insights that go well beyond the tabulated tables. Especially interesting are the comments made on the last question when subscribers were asked how, in the presence of stagnant library budgets, they would trade-off their print titles against the emergence of electronic versions of many of these same titles. ("Don't know" was a popular answer!)

Table 1 reflects two things: (1) to date, most libraries have had very little contact with on-line journals. That is like to change dramatically in the next year or so as important society journals continue to come on-line and as publishers with a large number of titles, e.g., Elsievier Science and Academic Press, provide electronic versions of their offerings.

The table also reflects the fact that the librarians who received the questionnaire were located at different points in their institution's hierarchy. For example, the majority who have purchased only a limited number of journals appear to be located in academic departments and companies. Those purchasing more than 20 journals appear to be responsible for serials acquisition in large research libraries.

Several journals are providing free on-line access to print subscribers at present and we are making use of this facility. While the choice remains I would not feel happy about going only for the on-line version until we have more experience of ease of access, particularly for titles from the USA.

JBC is the only e-journal for which we directly pay. We currently have access to more than 20 other titles that are presently "free" with our print subscriptions.

Table 2 shows a wide range of institutional approaches to decisions about electronic journals. The titles of the responding librarians suggest that there is again a marked difference between the way in which smaller (often corporate) groups decide upon subscriptions as opposed to large research libraries. In the former, journal acquisition is more often done by a joint scientist-librarian committee whereas in the latter, it was most frequently done by the cognizant librarian responsible for a particular discipline or field of study.

Individual scientists have taken the initiative to sign up for subscriptions. University-wide groups may also get involved.

Serials Team and/or other members of staff who identify titles; previously done by a medical student.

Recommendation from faculty to collection librarians for journals. Collection librarian selects monograph material in all formats.

The answer to this question obviously reflects the librarians' views of their role and not necessarily the reality of their situation. However, even taken as a test of perception, this answer, in conjunction with the previous answer regarding subscription decisions, supports the argument that librarians are important players in determining how rapidly electronic journals will be adopted. This is particularly true in large organizations where the cognizant librarian may be somewhat removed from the individual users.

Both things have happened in our library. Sometimes the librarians identify appropriate journals, sometimes faculty request that we subscribe.

Number of people with interest, and availability at the nearby University library.

Annual User Survey followed by recommendation of Library Committee (mainly Management)

The main reason [for having the on-line version] is to have a more up-to-date copy. However, getting used to on-line journals will require a cultural evolution.

Sample issues are displayed June/July in Library reading rooms, and a weighted vote system is applied after customers have signed survey sheets. Interest in new journals is then balanced with cancellations.

Selection is done by bibliographers and departmental librarians (science and non-science). The use the same criteria as for print journals, along with access issues (e.g., do we have the appropriate equipment or connections). Also whether the license agreement can be negotiated to meet our needs/constraints.

No. of requests (verbal and written) + no. of times journal is cited in print requests from Paper Chase searches + availability nearby + cost to library (some departments currently fund half the cost of library subscriptions in that discipline).

We usually obtain Sample issues to determine opinions on entering new subscriptions. We also have a Recommendation for Library Purchase that can be submitted by faculty and students.

a. if indexed in Index Medicus

b. cost

c. number of requests

d. relative number of titles in that particular field that we already own e. ILL statistics; if cheaper to own rather than pay copyright, for instance

Subject matter must reflect "core business" of site. Consultation with scientists to ensure current subscriptions are still valid and to enlist suggestions for new titles as necessary.

I utilize the statistics from 2 separate 6-month surveys. One is a survey of the use of newly-arrived journals (a small survey form is attached to each issue, users initial). The second is a reshelving survey to assess the use of older issues. A simple formula results in a "usage percentile." Our Library Committee, made up of a rep from each of the research programs at the Institute, meets and reviews the information prepared and presented on a spread sheet by the Librarian. Spread sheet includes subscription cost for each journal, the usage data, availability elsewhere on campus, etc. A "hit list" is defined, it is e-mailed to all staff who have opportunity to comment or argue for renewing a journal important to them. Final "drop list" is again e-mailed to all for their information.

Regrettably, we are not very organized about this yet. In general, easy access, such as by domain name, is essential; pricing is also an issue. Links to MEDLINE will become essential in the next 2 years!

Many criteria must be examined: subject area; publisher; editorial board; cost; faculty opinion. Addition of a new journal subscription requires cancellation of (an) existing subscription(s) in order to generate the necessary funds. This policy has made librarians very cautious about the purchase of new subscriptions.

We have a list of criteria that we use, including where is a journal indexed, have we had any ILL requests, has any patron requested us to purchase a subscription, does it match the research areas of the institution.

Presently we use the popularity of the print journal, and user requests as a guide to subscribing to on-line journals. We also have statistics on the use of our online journal collection from Ovid.

The decision is based on requests from users and use as determined by document delivery records.

We assess whether or not the journal has value as a print version/ and then whether it has added value in on-line mode. This decision is made one by one now based on price and need.

Based on the following criteria: 1. within the scope of our library's collection; 2. recommendations by scientists and reviewed by the library committee; 3. cost of the journal; 4. available space for storage; 5. copy right compliance when requesting photocopies by interlibrary loans.

Depends on use of the title, its cost, and its method of password protection or distribution.

By its usefulness in the focus of our manufacturing area (pharmaceutical), by its availability/lack of availability from other libraries, and by its cost.

A serials review process, which includes faculty interest, that looks at relevance to the curriculum, cost and accessibility.

Retention of a subscription is based on usage statistics and price. Purchase of new subscriptions are decided by the librarian and the library committee based on anticipated need, reputation of the journal editors, price, no. of past ill requests, etc.

We subscribe to a number of printed copies of journals in our research group, and at the institute. The assessment is determined on the basis of research interests and funds available.

It is a decision of the library committee

By the cost of the journal, who is requesting, who the prospective users might be, i.e., what programs within our institution would the title relate to -- the broader the application the more likely the title will be purchased.

It is a cost/benefit analysis of those in-scope and index in MEDLINE or BIOSIS.

Decision was taken at the Department level and based on number of laboratories interested

Reputation of corresponding print journal and whether we are the archival location of the print.

We consider scientists' requests and results of questionnaire on journal.

Requests from patrons and faculty. ILL stats. Adjustments to our changing curriculum. Space required for the run in our stacks (some prolific titles we acquire on film or electronically). Price of print vs. other formats. Electronic titles replace print in instances where they are readable online (few very long articles) and they quickly fill up stack space.

To be brief, we look at price, faculty support, curriculum support, research program support, and to a lesser extent indexing coverage and journal impact factor (ISI rankings).

We have a very limited budget for journals so we ask our "subject experts" for help in decided what to keep and what to discontinue.

Feedback regarding the print title is solicited from information/access services staff and is combined with collection development knowledge and expertise, as well as a review of issues, to reach a decision. Collection development policies and objective criteria (which include regional holdings, indexing, subject scope, faculty involvement, full-text access) are evaluated for each title under consideration.

Given the cutbacks we have been experiencing for the past few years, the primary concern is whether the journal will support our research programs in a cost effective manner. Titles peripheral to our core programs will no longer be purchased. Titles must support the department's research programs to be considered. As a result, scientist recommendations are a key factor in considering new titles. If not recommended, few titles are actively considered by the librarians.

Determine level of previous use through interlibrary loan and/or assess appropriateness to work being done in the institution. Factor in cost and local availability.

Strictly according to co-financing by scientists from their research funds.

Librarian committee evaluates journals usually from faculty requests. We consider cost, potential audience, indexing, holdings at other libraries and other aspects. We consider the purchase of a new subscription a long term investment in funds and space (which is a major concern for our library).

Normally when any Library customer's request is authorized by their Head of Department. If the library budget is stretched, the department may be asked to pay. If the librarian feels there are good reasons why a specific item may not be worth the cost, this will be discussed with the requester.

We have a joint library/scientific staff committee who evaluate. Impact factors play a roll and experience of the scientific members.

Subject analysis; indexing; scope; document delivery requests; cost; benchmarks; Brandon-Hill list

Faculty support for title (there is a need), scope appropriate, indexed in a major index, affordability, reputation of publisher, availability (from another Library)

We usually conduct a survey and consider subscription when more than four labs (out of 28) are interested in a particular journal. We also take into account the price of the journal and its availability at close-by university libraries.

The answers to questions 4 and 5 underscore the care with which librarians are scrutinizing their serial acquisitions. As might be expected, costs and budgets are mentioned frequently and, to this extent, one might conclude that there is a difference in perspective on serials acquisition (including e-journals) between librarians and researchers. However, there are a surprising number of references to departmental or laboratory cost-sharing. ("Strictly according to co-financing by scientists from their research funds.") The comments also indicate a substantial amount of consultation, perhaps more so than the tabulations suggest.

Even if there is consultation regarding the ranking of titles, however, the day-to-day decisions about what to add or subtract--and hence whether budget constraints are met--are the responsibility of the cognizant library unit. Looked at from a model of bureaucratic risk taking in which the penalties for budget shortfalls are may be significant, the result is likely to be a conservative approach to change. The prediction of such a model would be that librarians will move slowly and cautiously to substitute e-journals for print unless significant cost advantages can be demonstrated.

It would be difficult to judge the cost effectiveness of e-journals from general user statistics. The best judge would be if one had user statistics of your own users, but how do you get these without subscribing? J. Biol. Chem., our main journal will probably be kept in any case.

I should like to compare that number to the number of people reading the printed version.

By the number of users and the seniority of the scientists involved.

The library budget is the most significant determining factor in whether we would be subscribe to the e-version of a journal. The number of people in a particular laboratory and the number of labs making use of a title is also a factor.

Same as above (#4) with the exception that we would have to decide between an electronic and a print version if there is an additional cost to get both. We would be unlikely to replace a print journal with a more expensive electronic version and there is a great concern about archive copies.

Whether on-line usage equaled or exceeded current print usage. Whether very current information was being accesses - i.e. because on-line issues are available before we receive the printed version. For on-line only journals - what would we be spending on interlibrary loans if we did not hold the electronic subscription.

I don't know. I'm not sure what you mean by user stats. Only number of users? What they used? (i.e. abstracts, table of contents, full text, browsing only, down-loading???) In the few cases where the publisher has offered a low-cost or free trial of their e-journal, I have notified Institute staff of availability, then polled them later to find out their thoughts as to value. Trial use by my own users would be the most useful to me.

I would want to know the following by month and calendar year-to-date:

a. Total access to the journal for any reason

b. Total access to the table of contents

c. Total access to an abstract

d. Total access to the body of an article

e. Total access to references

f. Total access to illustrative material (tables, images, etc.)

Global user statistics would be of little use. Statistics of use by users at this institution would be of great help. One could then compare various e-journals with respect to use/cost. I suspect that the use numbers will be low at first and take several years for use to reach a plateau. So use results will become increasingly valuable with time.

I'd like to know how many repeat visits were made to a site, not just one-time visitors.

I'd like to see the number of connections by IP numbers. I realize that with web servers, statistics usually count every page or graphic downloaded as opposed to a count of sessions as on older systems, so I'll need a detailed description of how the statistics are created.

For me it is important how many users are from Europe, because this reflects the speed and stability of the connection

I think my users would utilize on online version of a popular, well respected journal in their field - like JBC. New e-journals would be a harder sell. Outside statistics might not be that influential. A publicized trial would be more effective.

We would use those statistics much the same way we use circulation statistics. Right now we have a brief survey on the web page asking faculty opinion of the online titles.

Most of the cost is minimal when compared to the printed copy. Our users are largely sophisticated computer users and research is "the" main duty of their work. Publications are their product of performance. In this respect, journals are essential channels and means for our center. Journal cost has not been big issue to us.

By the number of my library users accessing the e-journal, by the level of satisfaction my users have with e-journal, by the extent of archived issues available for print-offs (is a long run available?), and by the costs (time and money) involved in obtaining hard copy prints from e-journal.

With very few exceptions our electronic subscriptions come at no extra cost; paying for the print subscription entitles us to the electronic version. A high degree of use would have to be in evidence to make an e-version worth its cost.

I am not interested in user statistics, I know what I have to read to support our own research.

User statistics can tell how many people have looked at a journal, but it is not so easy to decide if they found it useful. We tend to try to subscribe to the e-version of as many of the hard copy titles we take largely because our user group is dispersed over a wide geographical area and it is one way of ensuring equality of access.

User stats helpful, but quality is still paramount.

We do assess the retention of journals by usage statistics both for in-house use and for Interlibrary Loan services to other libraries. With electronic journals, it appears that ILL would no longer be possible. A subscribing library could not download or copy an article and forward it to another non-subscribing library without violating the agreement with the publisher. This could have a major impact on the distribution of research literature as it is done now. If a title has low use and is moderately to very expensive, it is likely to be cancelled.

Determining who uses the online pub, how often it's accessed and whether valuable information is obtained vs. the cost of the online version and whether there are conditions and restrictions governing the use of that online publication.

For major journals, such as JBC, we are moving to electronic and the decision is really how long to keep the print. Our major efforts at present are to get prepared for electronic journals by upgrading online catalog stations to microcomputers with good printing options, including networks printing when we can, working with Faculty & staff on training, and --very important--getting the electronic content linked to the Web interface for MELVYL's databases. Usage statistics may be a factor for lesser-used journals than JBC in determining the cost/benefit of electronic access, but usage statistics will have more validity when there is a critical mass of online content and users have the equipment to access them & do so regularly. During this transition phase, our in-house usage data are also good indicators of candidates for electronic journals.

Given our experience with other "cutting-edge" technology, I doubt this would be a valid measure for the first 1-2 years of subscription. For the time being, we will compare a price of the e-version to that of the print version.

Statistics would influence our decision, but overall cost is much more important to us. Once the price gets into a certain range (maybe over $1,000 for science journals), then we are not interested.

It will also depend on whether we get print as well, and what the combined costs are. If we have the print version, a supplementary online version is only worth it to us if the cost is minimal. If it is available only in electronic format, then we will evaluate it like any other journal subscription

I suppose the number of times the online version is accessed versus the costs for interlibrary loan of articles. Perhaps the cost for online divided by the number of access sessions to come up with a figure of cost per use.

We would compare the "per use cost" vs. that of the print version. This would, of course, be only one factor in judging the worth of the electronic version.

I would have to weigh use statistics with other factors such as convenience of having the journal available on a network, bound volume storage considerations, and availability of backfiles in the future. I'm not sure how I would determine cost-benefit exactly.

We are logging connections to on-line journals via our web pages. In my opinion, for the next year electronic journals should be offered regardless of use statistics, if the price is not more than a 10% extra on the price of the print version (I prefer the approach of IOP and Springer-Verlag). Then we shall see. I think the deciding factors will be ease of use (no passwords for registered sites, fast searching, print-quality articles on demand) and network speed. As always, we shall follow the wishes of our scientists. I must tell you we will not continue our subscription to the electronic version of JBC, due to its excessive cost.

We have established a cost per use maximum of $40.00. Journals subscriptions which exceed that maximum over time are considered (along with other factors) for cancellation.

Number of times accessed (not necessarily printed/downloaded) combined with user's views on the system user-friendliness and journal content. If our users preferred the on-line version then they would decide.

If it is cheaper than getting an ILL (staff costs included) then it is worth its cost.

For the time being, we keep our subscriptions to the print versions because not everybody is equipped with the hardware (that is computer and printer) necessary to obtain a good copy of an article. No doubt in my mind that we will see more and more well equipped researchers who will want everything on-line and that will surely help us in deciding to subscribe to the e-versions. That said, we suffer from serious budget cuts and it is very hard to know in advance what amount of money will be available for subscriptions...

Table 6 shows that a major concern of librarians about e-journals, again not mirrored in the interests of researchers, has to do with the problem of archives. Unfortunately, based on the dozens of responses to the topic in the ARL e-journal list, there are widely differing ideas on how the question of long-term availability of electronic materials can be assured. The inability to provide a cost-effective solution to the problem that everyone: users, libraries, and publishers, can live with is likely to create a significant incentive to retain print versions, at least in the short-run. Again the model of bureaucratic risk aversion is helpful. Reliance on a system of archiving that is based on relatively unproven technical and institutional arrangements can lead to serious criticism. Librarians need purchasing options that allow them to gain experience with online formats without putting their archiving functions at risk. There are several models which may provide safe alternatives: bundling print with online; offering a very inexpensive microfilm archive as an "extra" with the online version; offering a set price for the content with additional fees for print/online/microfilm.

Remote access is sometimes very slow, communication lines are not as reliable as they should be, and of course there is the issue of cost. 7. Have there been discussions in your institution about increasing the serials budget to incorporate the costs of electronic versions?

Archive of back volumes - how long will they be available, with hard copy you have control of that. What happens to the Archive when you stop subscribing - you lose access, with hard copy you still have the hard copies for the years you subscribed. The subscription cost SHOULD be cheaper as the publisher has lower distribution, printing and administration cost, and the subscriber library has less control. Need real guarantees that the material will remain in some electronic form in the years to come. How about publishers setting up an independent organization funded by a small percentage of the subscription to ensure long term availability of older material - e.g. a national library.

There is no reservation for me. Probably just a matter of becoming used to it.

Is access available/supported via multi platforms? (e.g. PC, Macintosh, Unix workstation....). Is site-wide access available? (i.e. IP registration or domain registration available). Is the subscription for a site-wide license? (i.e. not charged by usage)

So far access to the Internet is not so common and some people have told that they prefer the printed version.0

I worry about future access to past issues and am not happy about being solely in the hands of the publishers for this. Online backed by CDRoms would be better.

There are selectors who feel that not enough people on campus have/use web access and that we should keep the print versions for them. I don't necessarily agree, but I'm not a selector.

Archiving! How are we to be provided with archives of years subscribed if/when we cease subscribing to journal?

We would subscribe to both versions for a minimum of one year before we would consider canceling the print version and only after determining the usage of the print version.

a. archival ability, b. ability of ALL our clientele to make use of on-line journals c. ability to loan on-line articles through ILL; currently, many licences to not allow that

Continued access to old issues that have been paid for. Ease of access. Limits on numbers of concurrent users. Costs.

Users need re-training to access information regularly, as onus is on them to remember to look for it - not as easy as seeing all the journals actually in the reading room.

They are many! Mostly just the often-stated difference between being able to hold a journal in your hand, move around with it, browse through it easily, photocopy what you want, and VERY important, be assured that YOU will be responsible for archiving it and will know that it will be available in the future. What guarantee is there that the publisher will archive appropriately for your needs? What is to keep a publisher from charging whatever they wish for your future access to archived issues? So to sum up, lack of portability, lack of control, lack of aesthetics. I don't think it's just me... I think reading long articles at the computer is DEADLY. It's a medium of speed, not contemplation! What can replace coming to a library, sitting in a comfortable chair, browsing through all the journals there, if you wish, talking quietly with a colleague about something you've just read? Somehow, the scenario of each isolated scientist sitting in his/her office in front of a flickering screen, printing out his/her very own copy of every article of interest, is not a pleasant one!

1. lack of equipment for accessing e-journals (both in the number and power of workstations)2. unreliability of the Internet - it is difficult to know why it is slow or does not respond 3. lack of campus network connections - this is improving but they are not now, nor do we anticipate them to be ubiquitous in the foreseeable future. However, the Library is well connected. 4. user resistance - many believe e-journals are the solution for someone else as long as they get to keep their print journals. Also, some of the more mature users (even some in their 30-40's) are reluctant (or resistant) to learn the skills needed to make use of e-journals. These are very intelligent people and it may be hard for them to admit they don't have skills needed to do basic research. 5. printing costs - in most cases the final result of the use of an e-journal is still a printed copy of an article or part of an article. The best printers are laser printers or laserjets. We have no way of collecting for the cost of printing right now, so we're giving it away. We cannot afford to do this, especially when the most important journals move to e-format. 6. printing equipment - the variety of printers and their compatibility with various workstations and products is becoming more complex. Anticipating the best equipment is becoming more difficult. Figuring out if the printing problem is in the product, the interface software, the printer, the network, etc. is very complex and time consuming. 7. training users - learning how to use a journal used to be a by-product of the teacher/student relationship, especially at the graduate level. Once you learned to use a journal you could figure out how to use other journals on your own because they're generally alike. Libraries used to only provide the organization of the journal collection that facilitated its use. It seems that most e-journals are different now, with different access methods and different organization, arrangement and capabilities. The library role in teaching students and faculty how to use specific journals is increasing, while the number of staff to do this is not. On the positive side, we are seeing interest among graduate students in taking on some instructional role in their discipline.

A big problem is the archive. Will publishers maintain the archive, and if so, in what format (print; fiche; CD-ROM; magnetic) and at what additional cost? Control of access must be convenient (i.e., by IP number rather that requiring each patron to obtain a library subscription number from library staff. So far there has been no rush by faculty or students to electronic journals. There is the problem of additional workstations in the library; and for external users there is the problem of adequate hardware and software. Access over the internet can be slow at particular times. Many computer monitors are still too small with too low a resolution to make reading papers easy. Browsing the contents of paper are easier in print format than in electronic format.

Depriving those users who don't have an adequate computer of access to a journal.

Archive availability when subscriptions cease, licensing arrangements, ease of access.

Many users still prefer photocopies.

e.g.slow connections from europe to the states.

I am afraid that back volumes will not be archived. Reading lengthy journal articles from the screen is not comfortable and will usually be printed anyway.

I am afraid that copyright costs will become exorbitant and that we will end up paying for the same material many times. When networks are down (more often than we like) material is not accessible.

Our concerns are based escalating costs for online and print version. Right now we are not dropping the print version of the title, we are adding them and evaluating their usefulness. This means added cost. We are also concerned about reliability of the online version and archiving of the online versions.

We still have to keep the printed version of journals because of needs to produce photocopies, slides, and graphs of biomedical images. It is difficult to "read" in-depth articles on the computer screen for a long time. There are still people who prefer the traditional paper for reading. Total dependence on internet is unpredictable, not 100% available all the time; Access to Internet is becoming slower, connection can be interrupted any time. There will incur more cost in hardware, for computers, printers, scanners;, software, for browser, data base interface; paper consumption, and escalating cost of service maintenance agreements for the equipments. Worse yet, a library still needs to maintain numerous copiers regardless how many computer printers are available.

The longevity of the service and access to older copies. Gaps in the collection if we cancelled the print version and later had to reinstate it. Access from all locations.

My concerns are accessibility to many years of back-issues for hard copy prints, ease of access to e-journal, licensing issues (e.g. e-journal reasonably available from LAN desktop PCs), reliability of on-line connection, simplicity of issue-searching for specific references and for articles by subject or authors (multi-year index would be crucial on-line), cost of e-journal subscription and how it compares with actual publisher costs--are publishers going to charge more and more when printing and shipping costs are reduced by going electronic?

Access, archivability, cost, inability to ILL.

Graphics quality, ease of use of the site, price, accessibility and speed of the connection, ability to print easily, and will this online only version be at least as functional and easy to use as print.

I like to browse through the paper version, also to find articles that I did not search for. Besides, figures and photos, in particular electron micrographs, are usually of higher quality than on the screen.

Who guarantees the availability of older material and who guarantees that the material is not tempered with after the first "publication"?

While we don't really have major reservations about subscribing to as many e-versions as possible - for the reason as above - it is our experience that our users still require good access to the last 10 - 15 years of the core titles for our subject areas, and we need to have the hard copy to minimize ILL costs.

I am a librarian and my primary concern about journals, in any format, is increasing costs. If we buy the on-line version and cancel the print, then the publisher is our sole access to, not only the current issues, but also backfiles. I worry that this might set up a scenario where we have to purchase the information twice!

Currently, there are many limitations on the hardware needed to download and produce quality print copies. Many users do not have the skills necessary to use plug-in software and other requirements with the Internet. The quality of photographs, illustrations online is a concern. The means of archiving a journal especially if a subscription has been stopped is a major concern. If a library no longer has a current subscription, does that mean that it also no longer has access to the years of the journal that it had subscribed to. Print copies you have on the shelf once they are paid for and canceling the current subscription does not delete the previous holdings.

Primarily accessibility, downtime and archiving.

Faculty state poor printing capabilities as a drawback for their own sites, and the library here has yet to convert from its dumb-terminal OPACs, although we clearly must do so soon. When we attain critical mass for electronic content, faculty workstations will be upgraded, though.

Printing quality of figures: printing is time consuming; bandwidth; continued access to title if there is ever a need to cancel

We have no reservations now. To purchase the electronic version of JBC is a test case in our library.

Unpredictably of the network. Archiving of back issues by publisher. Quality/Accessibility of charts, graphs etc. over various browsers (here I assume web access). Demand of paper copy by patrons (if print/e-mail functions were good this may decrease).

Will limit access to undergraduates who don't have easy use of a computer. Worry how we'll have access to archives and backfiles, especially in the future if we no longer subscribe and/or the journal publisher changes.

The quality of the micrographs and photographs in the online versus the print version

Also archiving - will past volumes be available on CDRoms, how far back in time will online be available.

Lack of appropriate infrastructure on campus has been the main obstacle. Second to that would be determining whether our users actually do prefer the online version to the print. Archival issues are also important, especially in the biomedical literature.

Accessibility for our staff. We are a very large federal department with employees located across the country. Not all sites have equivalent computer capabilities. Some sites still have no Internet access whatsoever. Accessing "archival" issues. Will we be obliged to maintain a print subscription merely to have access to older issues of the title? Copyright concerns.

Uncertainty about continuing availability and additional training necessary for users.

I think we could move if the journal data files were downloaded to our local computers as soon as they became available on the net, and if an archive CD-ROM was included in the subscription.

We worry about archiving of back issues, equipment compatibility of our users, future cost considerations and more.

Some library customers are still very uncomfortable with browsing online. Speed of access/downloading. Current electronic versions often don't offer any advantages i.e. they simply mirror the paper copy but are more cumbersome to read. 4. Future availability of archive copies - particularly chemistry. We will have no control over what back numbers cost in the future.

Stability of the product, archival issues, and accessibility for all of our users.

Not all users have online access or are computer literate enough to gain access. Archiving of back issues of online versions has not been resolved. Systems crash and access is cut to everyone.

For the time being, we keep our subscriptions to the print versions because not everybody is equipped with the hardware (that is computer and printer) necessary to obtain a good copy of an article. No doubt in my mind that we will see more and more well equipped researchers who will want everything on-line and that will surely help us in deciding to subscribe to the e-versions. That said, we suffer from serious budget cuts and it is very hard to know in advance what amount of money will be available for subscriptions.

The results of Table 7 mirror those of Table 1 in that many groups have not yet come to grips with the movement toward on-line journals. However, regardless of whether there has been a discussion about serials budgets, the bottom line from the comments reported below is that no one is anticipating that there will be more money available to spend on e-journals. If budgets remain stagnant and if publishers follow a practice of bundling that includes a surcharge of 15-20 percent, there will have to be a reduction in print titles to make up the difference. (In a few cases, librarians reported that they would not pay twice for the same content implying that they would be willing to choose between print and online for a particular journal rather than reducing the number of titles.)

I don't see why electronic versions should cost more than printed versions. So far the added cost of obtaining both forms has not raised any problems. For instance, the cost of binding journals is very high and we do not bind if we have CDRoms. The cost of networking CD Roms effectively is quite high.

We have talked about it but our budgets are shrinking not increasing. It won't happen. Our library serials budgets are unlikely to increase.

I'm not sure how to answer this one. Certainly, we have initiated discussion of the need to have additional funding, especially for this transition period, but every campus unit is clamoring for more funds and the library is everyone's second favorite thing. All budgets are shrinking, including serials budget!

There is no question of increasing the serials budget. If we move to on-line journals we will be forced to cancel the print. Due the rise in the cost of journals in general, we have had to cut more and more each year to accommodate a (basically) flat budget.

It won't happen if it means buying duplicate versions in different formats. That is seen as a waste of money.

With a few exceptions, librarians clearly intend to treat the on-line version of print journals as an another title to be ranked alongside the entire list of journals that compete for scarce resources. There may be several consequences of this move. It surely will spell difficulties for second-tier print journals that are regarded as marginal to particular disciplines, journals whose print runs are already quite small. If these journals are to survive at all, their costs will need to be reduced significantly, a development that may push them to move to on-line only. By the same token, it may mean that more start-up journals will be on-line only. There is simply no room, given the budgetary pressures and the increased costs of the established core journals, for additional print journals to find a niche in the market place.

Unfortunately, we are taking serial cuts at a time when we are also trying to get titles on-line. We have dropped $45,000 in serials for 1997, but we have added a half dozen titles in on-line format.

We currently view electronic journals as highly valuable but still experimental and are encouraging users to try them out as widely as possible. We anticipate that the way we subscribe to journals will go through a period of turbulence for the next 2-3 years as systems and suppliers fight for prominence, then we will adopt new purchasing policies.

Cost of space

There is no need for both print and on-line. The department will probably concentrate on-line journals, except for science 'newspapers' (Nature/Science). The outside departmental libraries I do not know much about.

Each title/version will be surveyed on its own merits/popularity. From our experience so far, electronic versions are acceptable to our community, so they may replace subscriptions to printed versions. We can replace some of the copies of a journal by the electronic version.

By continuing to cancel less needed journals, probably including print versions of those we want to acquire in electronic format. Availability of document delivery is a factor.

If budget remains constant, and cost of electronic journals remains this high, there's no question - print journals would have to be given priority. (more titles can be purchased per dollar)

Any trade off would occur after the following steps have been taken:

1. Our library participates in a cooperative collection development program with the libraries of Cornell University Medical College, Memorial Sloan-Kettering Cancer Center and Hospital for Special Surgery where we enter and cancel subscriptions to avoid unnecessary duplication. 2. The same institutional libraries share bibliographic electronic databases which no individual library would be able to afford. 3. Where possible, we would look at one of the libraries as an appropriate place to maintain a paper copy and make the electronic version available to all four institutions by sharing the cost as a consortium.

If we only knew!!! That is truly the 64 million dollar question. AT THIS TIME (it may change), we would maintain the status quo until issues of copyright, ILL, and archiving are answered. Currently, we purchase both formats but plan to experiment by pulling one or two paper copies off the shelf and "hiding" them to see what the reaction is to the electronic-only version. Additionally, we will not purchase a title electronically unless there is some sort of discount to maintain the paper copy. JBC is an example of a title that costs the same for each format; if we choose to purchase both versions, there should be a discount to do so. We will not continue to purchase duplicate electronic titles for which there are no discounts offered for the paper copy (or vice versa). Finally, we know we cannot continue to purchase both versions of titles. Most likely, as stated above, we will opt for the paper version unless issues of archiving and ILL are adequately addressed.

Currently we are making journal by journal decisions. If an electronic version is cheaper, doesn't have excessive restrictions on use, is easily accessed and used, has provisions for archival copies, and is aimed at a user group that is comfortable with electronic access, we will lean toward the electronic version. If the electronic version is available at no additional charge with the print version we will provide it. Publishers have to realize that our materials budgets are not growing.

Following consultations with scientists, only journals which are heavily used - whether in print or electronic will be subscribed to. Others, or those which are expensive and readily available via interlibrary loan would be cancelled.

I don't know yet. I guess I will base those decisions on input from Institute users and from the Library Committee.

This is a complex question. If the real budget for serials remained constant it would mean that we could keep the print journal we have. However, if we elected to convert to e-journal, our budget for equipment to use them and the campus budget for networking resources would also need to grow, and I'm not sure this will happen. Nevertheless, I believe that HighWire in on the right track by pushing the situation to a decision and by providing a quality product that makes the transition to e-journals actually feasible.

Print journals keep increasing in price so that, even without electronic journals, it is necessary to undergo extensive review and cancellation of journals every 2 or 3 years to cut the expanded serials budget to the base level. Under these conditions, some subscriptions to specialized print journals will be cancelled in order to purchase electronic versions of print journals. I expect that only a few very well established and high use journals will be purchased for electronic access. There is no stampede here to electronic journals.

By reducing the number of print subscriptions we carry - based on use of the print version (regardless of whether or not an online version is available)

Same way we currently trade off one print journal against another: significance of the research, anticipated (and observed) use.

We will continue to get dual versions of select journals, probably by using special money and judiciously cutting print serials.

We would have to delete less-used titles from subscriptions; encourage publishers to give discount on dual subscriptions to online and printed versions of the same journals -- after all the contents are exactly the same. Or else, we just have to go without subscribing to both. Publishers should not be too greedy busying making money.

We will have to be very selective about online subscriptions and only purchase the most heavily used, i.e. very few.

It will be very much an economic decision. Publishers simply must keep electronic version subscription costs reasonable, that is, comparable to print, and perhaps even lower, as technology improves. A consideration for me is, will e-journal subscriptions be made available via subscription agencies? The tremendous potential of e-journals are their space savings, their searchability, their accessibility from remote stations, and their immediacy.

As we always have done, we will have to cancel less-used titles in order to afford new subscriptions, be they print or electronic. One other possible way to afford a new e-journal is to charge each interested user a portion of the subscription cost. That's what we're doing for JBC Online for 1997, but I'm not sure this will be feasible for a larger number of titles. Individual scientists' budgets have their limits also.

When I find an article of importance for me in the electronic version, I print it out to store a paper copy for a period of time. I can not store many print journals in my office or the lab and I am usually only interested in a few per cent of the articles in a given volume. Electronic transfer is faster, 1-2 days rather than 3-4 weeks for a volume of a journal in traffic across the Atlantic. Therefore I prefer electronic transfer, if the quality of text and illustrations in the printed copy is sufficient.

We wait to see, how the demands of the scientists develop and how our reservations are handled.

This is a very good question and if you have a good answer which will convince the accountants then please let me know. However, as the research focus for our users changes on a regular basis we have to have a certain flexibility in the way we approach our journals subscriptions. We certainly cannot ever continue to take a title just because we always have. In the same way the quality of titles changes over time, and they may become less useful or relevant. This ability to respond quickly to changed needs within our organization means that there is more readiness to accept new ways of accessing information, and also an appreciation that better access to current information will cost more.

Also, we have been exploring the price reductions which can be negotiated with publishers by consortia of similar libraries, and have also been asking publishers if they will give serious consideration to offering long-term subscribers (to hard copy) a "loyalty bonus" discount.

We will continue to increase our electronic subscription at the expense of ur print collection.

This is still under debate. Duplicate subscriptions for print serials and electronic serials (primarily indexes on CD-ROM) were eliminated for 1996; this meant the paper versions of titles held in both formats were cancelled.

It is already a very slow, cautious trade off. We continually are assessing the use of our journal collection and do delete low use titles. Historically, we used the savings from cancellations to subscribe to new print titles. Now, we will probably use the cost savings from cancelled print titles to accommodate new titles in both print and online formats or to add the online format accessibility for existing print titles in the collection.

Continual feedback from users is crucial as well as information provided by the publisher/provider of the online publications. Print versions are fairly predictable in; their annual cost increases, quality of publication not only of information, but quality standards for reproductions of photos, graphs, charts, etc. If electronic versions cannot provide a quality product at a reasonable price, there will be no trade offs.

We are using usage data to determine which print titles to cancel. We will have to make deeper cuts of the print collection to purchase the electronic content.

We had to cancel departmental subscriptions to all printed journals. No decision yet for electronic versions.

Cancel more print. We will convert print journals to e-version at the scientists' requests.

We are not quite sure yet.

Our discussions will be in terms of getting the greatest *number* of journals and not duplicating titles in both print and electronic formats to the exclusion of additional titles.

I have no idea at this time.

Realistically speaking, until technological and infrastructure issues are settled, it will be the books budget which will be affected if the serials budget remains constant. We would continue to evaluate and perhaps purchase electronic versions (as long as they added value over the print and/or our users requested them) either on our own, or through cooperative efforts on campus or locally/regionally.

We would only consider replacing print versions when our concerns from question 6 have been resolved. We are not, at this time, prepared to commit exclusively to any on-line versions of journals.

We will purchase the ones we have to have, canceling print editions if that option is available, or choose the format which is less expensive. Other marginal print titles may have to be cut.

Not my decision, we will follow the wishes of researchers, who have to participate up to 30% of journal costs directly from their research grants.

Your guess is as good as mine (see above response).

The medium (electronic or paper) will often be irrelevant. We will be more concerned about whether we need to purchase full copies or whether it is adequate to obtain individual articles on demand. We will establish which journals fall into each category (full or article) by (1) assessing the demand for articles per title from our electronic current awareness service and (2) asking users which titles are essential for full browsing and their preferred medium. (This is already underway)

For now, we have a separate budget for electronic formats, so our print journals are not affected. When we decided to add OVID's Core Biomedical Collection, we knew that we were duplicating the same information in a different format but we were giving our patrons an additional, value added service by providing access from the citation directly to the article(s). Most of the electronic version duplicate the printed version. We do realize that at some point we will have to consider the budgetary implications and move to an either/or situation. Deciding how we will trade off the electronic version against the number of print journals is something that we will have to discuss and evaluate. For now, it is a case-by-case decision and we are very reluctant to relinquish print in favor of electronic.

If we are forced to spend money on electronic versions, print journals will be cancelled to finance the electronic versions.

As our library budget remains constant we cannot sacrifice other printed journal in order to subscribe the electronic version of J.B.C.

It is hard to know. I guess it will come gradually and a decision will have to be made for each one of them, considering the price and the need for it.

Don't know.