COMPUTERS & EDUCATION PERGAMON Computers & Education 31 (1998) 41±53 ISDN-based distance learning: tangible costs, intangible bene®ts Gabriel Jacobs a, *, Catherine Rodgers b a European Business Management School, University of Wales Swansea, Swansea SA2 8PP, UK b Department of French, University of Wales Swansea, Swansea SA2 8PP, UK Received 29 July 1997; accepted 13 February 1998 Abstract The issue of the ®nancial costs associated with distance learning using live video is most often discussed only tangentially in the relevant research literature. This paper addresses the issue directly by analysing the case of a trans-European video teaching link principally from the standpoint of the expenditure required to implement and run it, set against the bene®ts it o€ers. Both ®xed and recurrent costs are considered, and an attempt is made to assess the value of the bene®ts. A generally pessimistic conclusion is reached, namely that despite enthusiasm in the literature and in ocial reports for interactive tele-teaching, the ®nancial burden it imposes may in most cases outweigh the bene®ts. # 1998 Elsevier Science Ltd. All rights reserved. 1. Introduction The vast majority of published research in the area of distance learning using live video is concerned with assessing its pedagogical e€ectiveness as compared with physical face-to-face teaching, and determining its level of acceptance by the teachers and students involved, taking into account factors such as student attrition rates, and the appropriateness of certain technologies. Given the generally favourable results reportedÐand even taking into account a certain bias which will naturally be found in reports of telematic links written by those who have conducted themÐthere can be little doubt that in these respects the virtual classroom compares well with the physical classroom in terms both of learning outcomes and of student and teacher satisfaction. Rarely tackled directly in the relevant literature, however, is the * To whom all correspondence should be addressed. E-mail: g.c.jacobs@swansea.ac.uk. Tel.: 01792 295577, Fax: 01792 295626. 0360-1315/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved. PII: S 0 3 6 0 - 1 3 1 5 ( 9 8 ) 0 0 0 1 6 - 5 42 G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 question of the ®nancial costs associated with such links: for the most part, in both case studies and broad discussions, questions of ®nance, if they are discussed at all, are restricted to generalizations. This is no doubt partly because of the perception that any paper dealing directly with costs in the domain of information technology has an inevitably shortened shelf-life. This paper nevertheless deals directly with the issue of ®nances (involved in a trans-national video teaching link), and while quantifying costs inescapably means a certain built-in obsolescence, we feel justi®ed in this instance in citing actual ®gures on two grounds. In the ®rst place, if ®nances are dealt with somewhat in the abstract (such as by citing only percentages of institutional turnover or portions of average earned income), they can provide only the most general of guidelines and thus do not give either decision-makers or intending practitioners an authentic feel for what has to be paid. Secondly, while the cost of information technology continues to fall, it does so at a reasonably predictable rate, following Moore's Law formulated 30 years ago and so far proved accurate, that microprocessing performance would double every 18 months but that actual outlay would fall much more slowly and perhaps not even signi®cantly. A corollary to the arguments presented in this paper is therefore that it is very unlikely that costs will fall so low in the foreseeable future as to render those arguments completely out of date. Indeed, if Moore's Law continues to apply, the capability of videoconferencing equipment will change far more rapidly than what has to be paid for it. The video teaching link described here consisted of two-way videoconference sessions between, on one side, the European Business Management School (EBMS) and the Department of French at Swansea, and on the other HEC (Hautes EÂtudes Commerciales) LieÁge, a leading Belgian business school. The phase lasted eight weeks during the 1996/1997 academic year. For 2 hours each week, teaching sta€ of EBMS provided interactive tutorials in English for between 15 and 18 LieÁge students, divided into two groups, in a variety of business-related subject areas, each group being allocated 1 hour of teaching time. LieÁge teaching sta€ provided similar tutorials, in French, for 14 Swansea students of French and Business, again divided into two groups each allocated 1 hour. Our experience brought with it a number of unexpected pedagogical and psychological ®ndings which were not predicted by our review of the appropriate research literature. We have reported in detail elsewhere on these aspects of the link (Jacobs & Rodgers, 1997), but it is worth brie¯y mentioning hereÐsince they impinge on the question of costÐthat many of the problems we encountered resulted from the limitations of the ISDN (Integrated Services Digital Network) technology which we used. ISDN provides a digital end-to-end connection, via ordinary telephone lines, with enough bandwidth (data-throughput) even at its lowest level of implementation for real-time, full-screen colour digital video with sound. Unfortunately, this does not meet average user-expectations of analogue broadcast quality video, and especially so in our trial since we were forced, for reasons of available equipment, into using only twochannel Basic Rate ISDN (ISDN-2). This makes the connection twice as fast as can be achieved with the fastest available modem. However, this is still insucient not only to provide either a genuinely seamless motion video or perfect picture quality, but also accurate synchronisation between the audio element and the movements of a speaker's lips, this lack of lip synchronisation being especially noticeable in close-up shots. Problems of this kind have been described by some distance-learning practitioners as being major barriers to teaching and G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 43 learning (for example, Sebrechts, Silverman, Boehmdavis & Norman, 1995), and some distance programmes delivered via Basic Rate ISDN have therefore relied on audiographics (still images with sound) rather than motion video (Davis & Smith, 1994; Kaye, 1994). This results in pictures of better quality but also entails some loss of the sense of a live presence which we were keen to avoid. We found, in any case, that Basic Rate ISDN video was sucient for our teaching purposes. Indeed, its limitations tended to fade into the background once one became accustomed to them, an e€ect which Colbert, Voglimacci and Finkelstein (1995) rightly report. It is possible to multiplex ISDN channels in order to increase bandwidth, but this involves greater expenditure, both as capital investment and in running costs. It is also possible to improve the quality of the video and to achieve better lip synchronisation by employing realtime data-compression techniques beyond those already incorporated into ISDN codecs (which compress and decompress the video data), for example by using proprietary systems such as PictureTel at both ends of the link. However, for the purpose of calculating costs for this paper, we have assumed Basic Rate ISDN with no extras, and any increases in bandwidth or add-on facilities would have to be taken into account accordingly by those considering similar links if attempting to assess their true ®nancial costs. 2. Breakdown of costs 2.1. ISDN installation and call charges If ISDN, at least in its most basic implementation, cannot meet normal user-expectations, it is nevertheless the only available system for which one can currently guarantee acceptable trans-national video links without a comparatively large ®nancial outlay for equipment, dedicated point-to-point lines and specialized software. The InternetÐsurrounded as it is at present by all manner of hyperboleÐcan in theory provide educational institutions with fast desktop links costing little or even nothing at all. In practice, the perhaps inaptly named information superhighway, which admittedly works very well for electronic mail and adequately for World Wide Web pages, is unreliable and can be frustratingly slow for interactive applications involving the enormous amounts of data required for motion video and sound. In contrast, ISDN does not inherently su€er from breaks in transmission, is not a€ected by any build-up of telematic trac because it uses a direct connection via ordinary telephone lines, requires no ``plug-ins'' or high-speci®cation computers, and is reasonably inexpensive to install and run. Costs across Europe for line installation and connection to Basic Rate ISDN, which provides a usable throughput of 128 kilobits per second, vary to some degree, charges being higher in the UK than in, say, Germany, although the UK telecommunications watchdog Oftel has for some time been pressing BT to bring its ISDN charges in line with those of other countries (Oftel, 1996). Depending on the particular type of contract taken out, at the time of writing (January 1998) in the UK there is an initial one-o€ connection charge of between £135 and £400, while line rental lies between £110 and £130 per quarter, certain contracts also o€ering some free calls and other free extras. As far as many UK higher-education institutions are concerned, some further-education institutions and 44 G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 even some secondary schools, the initial outlay and even the line rental can be discounted because the ISDN connection is already in place and used, as is the case for both Swansea and LieÁge. ISDN call charges between the UK and Belgium amounted to approximately £50 per h for the Basic Rate ISDN link (call charges are broadly the same for all trans-European ISDN links). 2.2. Studio and equipment costs While it can be seen as legitimate to discount ISDN one-o€ and line-rental charges, the same cannot be said of some other ®nancial investment requirements, namely (outside the area of human resources) those related to the teaching studio, including the necessary ISDN equipment. An ordinary seminar room or small lecture theatre can be used for remote interactive tutorials, but neither is ideal. A number of research ®ndings (Coiro, Davoli, Maryni, Pulia®to, Pavan et al., 1993; Comeaux, 1995; Harris, 1996) indicateÐand our experience fully bears this outÐthat a properly equipped telematic teaching studio, at each end of the link for two-way teaching, is a major factor in ensuring success. Such a studio need not be ®tted with very sophisticated equipment such as robotically controlled cameras or special-e€ects generators, but it should consist of desks or a suitably shaped table and comfortable seats placed in a semi-circle around a large TV monitor, on top of which, or immediately to the side of which, is placed the main camera. We have found that, at the student end, only in this con®guration does the monitor showing the teacher become an integral part of the group, because the con®guration best simulates a normal physical face-to-face teaching environmentÐin any other con®guration, the technology tends to protrude. And at the teaching end, the camera should be placed very close to the monitor, otherwise a seemingly insigni®cant but in fact very serious problem of misalignment of eye contact between teacher and students can arise (Acker & Levitt, 1987; Rose & Clarke, 1995; Jacobs & Rodgers, 1997). A properly equipped studio takes into account all such matters. It is also important, in most circumstances, that the main camera is supplemented by one or more subsidiary cameras such as one focused on a blackboard, whiteboard or ¯ip-chart, orÐ more signi®cantlyÐa dedicated document camera, mounted vertically and preferably with an easily controlled zoom lens, for the teacher to switch to as a replacement for an overhead projector. Alternatively, the main camera can be tilted downwards to act as a document camera, although if more than one document is to be used during a session, and/or if the teacher needs to point to parts of a document or to write on it, such a set-up will not be perfect because the teacher may have to move more than is ideally necessary. The quality of a video signal using Basic Rate ISDN technology, with its limited bandwidth, deteriorates with the amount of movement in the pictureÐthe more movement, the more diculty ISDN has with keeping up an acceptable frame rate (the number of images per second). Precedence is always given to the audio element, since any break in sound transmission may result in incomprehensibility, whereas a small number of frames can be dropped with little perceptible di€erence in video quality. Where a great deal of movement is involved, the compression technique of frame di€erenceÐthat is, that only the di€erences between a frame and its preceding frame need be transmitted instead of the complete frame, thus saving on data G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 45 spaceÐis not so e€ective because of the large di€erences between contiguous frames. The result, in extreme cases, is an almost unwatchable jerky, stripy and/or ghosty video, and even a complete breakdown of the video signal. With Basic Rate ISDN, therefore, a camera focused on a teacher standing in front of, say, a whiteboard will not prove a successful method of presenting that teacher unless he or she is very careful not to make too many movements, the result of which is not only ghosting of the image of the teacher but also a distortion of what is written on the whiteboard. With most of the teachers involved in our project, limiting body movement proved a very dicult skill to acquire: any way in which the temptation to move can be avoided is desirable, hence the need for a dedicated document camera. The cost of installing such a document camera is a little higher than that of the primary camera because it needs to be mounted on a stand or tripod. It should be noted that the cost of good cameras can be very high, in some cases amounting to tens of thousands of pounds. However, since the ®nal quality of Basic Rate ISDN video is comparatively low, it is not worth making an inordinate investment in high-quality cameras unless they are also required for other purposes. Zoom cameras of adequate quality with a pan-and-tilt head can be had in the UK for under £500, though those with infra-red remote zoom control are more expensive. High-quality microphones are, however, essential. It is crucial that students should not have to shout to be heard, nor to have to lean forward each time they wish to speak: the principal aim of a video link should be to simulate as far as practicable a physical teaching situation, and the technology should therefore impinge as little as possible. Flat table microphones of the pressure-zone variety are ideal because they are unobtrusive, and there should be sucient of them (three for a group of six or seven students). Such microphones, of a quality sucient for the purpose described here, cost in the region of £30 each. The teacher should be equipped with a small lapel-mounted microphone, the cost of which is approximately £20 or a little more if it is cable-less. Studios at each end of the link should also be equipped with a normal voice telephone so that contact with the remote site can be made in the event of a technical failure (see also Section 2.3). Without such a telephone link, it can be impossible to know at which end the failure has been caused, and which end should try to re-establish the connection. The cost of installing this normal telephone link is quanti®able if, of course, almost negligible. Not so, other studio equipment (TV monitor, lighting, etc.) and furniture, for which it is dicult accurately to estimate costs for a particular institution if one's aim is, as here, to arrive at a ®gure per student/hour, because much depends on whether the studio will be dedicated to telematic teaching, and how much such teaching will take place, or used also for other purposes. Ideally, it should be used only for telematic teaching because the seats, desks and equipment can then be left permanently in place. But whatever arrangements are made, the cost of studio equipment and furniture cannot be ignored, and if true expenditure is to be established, a ®gure to cover it should be added into the total ®xed costs of a link. The cost of studio space will, too, depend heavily on many variables peculiar to institutions but may well have to be taken into account (see below for an estimate of general studio costs). It is not normally necessary to buy a professional TV monitor (costing £1500 or more) since the di€erence in perceived picture quality between such a monitor and a domestic TV set is 46 G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 hardly noticeable. The only situation in which a professional monitor becomes essential is if one is considering a transportable studio: professional monitors are far more rugged than domestic TV sets. Otherwise, approximately £400 will buy an adequate monitor. The cost of the ISDN communications equipment required for the studio can vary from between approximately £2000 and £40 000 or even more in the professional range. At the lowest level is a Pentium PC equipped with an ISDN card, a simple camera, a microphone and speaker. This is adequate for one-to-one communication (in other words, acting as a videophone) but inadequate for an acceptable ISDN teaching link. With such a low-level setup, the frame rate will be in the region of 7 or 8 per second, whereas at least 15 frames per sec are required to satisfy (or, at least, nearly satisfy) normal user-expectations, nor will the picture resolution (the number of pixels) or the sound quality be satisfactory. What is required is a dedicated codec, as well as a reasonably powerful sound ampli®er (40 watts) and two goodquality speakers. Such items can be bought separately (a few hundred pounds for ampli®er and speakers), but several available proprietary systems incorporate them all, and a camera, and cost in the region of £5000 to £12 000 depending on quality, with the one-o€ connection charge to ISDN normally included in the price. One can expect to ®nd a system which will be quite satisfactory for interactive tutorials at around the £7000 mark. The studio must also be ®tted with good lighting. Domestic ¯uorescent lighting is ®ne, since modern cameras are very sensitive and able to cope with a wide range of lighting conditions. Only 10 yr ago, lighting represented a major item of studio expenditure because cameras were less sophisticated. Even today, a balance has to be struck between lights which are too bright for comfort but which allow a camera excellent depth of ®eld, and lights which put a strain on its focusing mechanism, but this has now little bearing on cost. Chairs and desks or a table also have to be found, together with a heavy plain-coloured curtain to be used as a backdrop, and a good-quality carpet which will minimise echo and (more or less) eliminate the sound of chairs being moved and feet being shu‚ed. See Table 1 for a breakdown of estimates for the above investment costs for a teaching studio suitable for the purpose of remote interactive tutorials. The estimates do not take into account certain labour costs associated with installation (such as those of ®tting ¯uorescent lights), and we stress once again that they will of course vary from institution to institution. Table 1 Guideline of expenditure for equipment and furniture required for a teleteaching studio Codec, primary camera (ampli®er, speakers, ISDN connection) Document camera Tripod/stand Microphones Lighting Backdrop Carpet (30 m2) Seating furniture Total Estimated studio costs (£) 7000 500 100 100 200 100 500 500 9000 G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 47 They are therefore intended as broad guidelines for a minimum investment in the UK if starting from scratch. 2.3. Administrative and personnel costs As with studio costs, it is impossible accurately to determine certain ongoing peripheral costs, and recurrent human-resource costs, for a given institution, but it should be remembered that they go beyond those of merely paying the teaching sta€. In the ®rst place, there are recurrent costs which can be loosely categorized under administration. These can include the international telephone calls which are necessary in order to keep contact, and therefore good relations, with remote partners, as well as to provide immediate post-session feedback to remote teachers. It is cheaper to use only a voice-line for this latter purpose than to keep the ISDN connection open. E-mail, in our experience, proved insuciently interactiveÐmost teachers wanted to talk to their opposite numbers about the success or otherwise of a session. Administrative costs can also include those associated with the time required to check, photocopy and distribute, and perhaps answer questions about, materials sent in advance of a session by a remote teacher, or with the time required to ensure that local teachers are fully briefed before a session, that all required materials have been sent and received by the remote institutionÐand simply that teachers and students turn up on the correct day at the correct time. FurthermoreÐand this is much more costly than administrationÐa telematic teaching link ideally requires two people at each end not normally present in a physical teaching environment: a technician and an anchor person. The technician need not be present for all sessions, but should be at hand at all times while they are taking place: ISDN is reliable, but no technology is wholly so. It is also advisable, in a trans-national link, that at least one of the technicians involved should speak the other's language with reasonable ¯uency. Failing this, an interpreter with some relevant technical competence should be available (the students are unlikely to have the technical vocabulary which may be required), and this too may have to be costed, although in the case of our experiment the anchor person, to whom we now turn, acted also as interpreter. The role of the anchor person (sometimes referred to in the literature as a monitor or chairperson) is crucial, at least in the early stages of a link and preferably throughout. It is in essence a managerial role (Yi & Majima, 1993; Berge, 1995), though our experience is that it is best ®lled by a teacher involved in the relevant discipline. Student reticence is ampli®ed in a virtual classroom, and we have learned that classes tend to run more smoothly if a teacher known to the students is present in the studio or teaching room, provided that this teacher does not participate unless such participation is genuinely essential. In other words, the anchor person should not attempt to take the place of the remote teacher, or attempt to become the star student (this happened once in our link, with pedagogically disastrous results). The anchor person's duty is to observe that all is running well technically, and to unburden the remote teacher, especially in the case of an inexperienced one, of having constantly to think about mattersÐsome quite trivial, others less soÐnot normally associated with a conventional 48 G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 teaching situation. Such matters may include, at the teaching end, reminding the teacher to switch back from the document camera to the main camera when he or she has ®nished using it, or, at the student end, pointing out to him or herÐat the feedback stage and even, though only if absolutely necessary, during a sessionÐthat the students appeared or appear to be having diculty in understanding a point. The anchor person can see subtle facial expressions and body language more clearly than the remote teacher who, without zooming in on a particular student, sees only the equivalent of a thumbnail picture of each student in the normal wide-angle view encompassing them all. The question of wide-angle and zoomed views brings us to yet another cost-related question: that of the number of students who can participate in any one tutorial, and thus to the recurrent costs per student per tutorial. 3. Cost per student First, it is necessary to make the almost self-evident point that there is little value in setting up a live remote teaching link merely for delivering a traditional lecture. The aim of a live link is to provide real-time interactivity, and the number of students who can participate in any one tutorial is therefore limited, as it would be in any interactive teaching environment. However, a further constraint in this respect, which arises out of the limitations of videoconferencing technology of any kind, is the number of students whose faces can be clearly distinguished even on a large TV monitor. We found by experiment that the maximum number of students for a successful session is only about six or seven, something which clearly has serious cost implications. A cost of roughly £7 per student/hour covering only call charges (£50/7 students) may already appear disconcerting to some. If one adds in the cost of the teacher, the technicians and the anchor person, together with an amount for the use of equipment, space and administration, the cost per student/hour rises dramatically. Presented here as a rough guide is a nominal breakdown of that cost, assuming seven students per trans-national 1-hour session. The ®gure has been estimated by taking half the total cost of two tutorial hours in a bi-lateral linkÐeach side providing tuition for 1 hour and receiving tuition for the otherÐand by assuming the same conditions at each end (for example, the use of a full studio for teaching, though it would be possible to teach from a portable unit almost anywhere). The arrangement between Swansea and LieÁge was that the institution receiving the teaching covered the cost of the ISDN calls by dialling into the remote institution instead of setting its ISDN equipment to Receive. Thus, in our model calculation, each side either pays for call charges or for the teaching (which, usefully for the purposes of the calculation, amount to the same ®gureÐsee Table 2 for an explanation of the teacher/hour cost). The total (approximate) cost for one session is then divided by 7 to give a student/hour ®gure. The ®gures of £50 per teacher/hour and £30 per technician/hour are based on estimates supplied to us by the Language/Export Centre at Swansea, where they are used as a guide to the cost of employment of teachers and technicians (not taking into account any pro®t margin) for the delivery of its courses. Administrative costs are based on information gathered from various departments at Swansea. The breakdown therefore assumes costs in the UK, though we are G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 49 Table 2 Breakdown of approximate costs per student/hour ISDN call charges or Teacher Anchor person Technician Administration Studio space, equipment Total Approximate cost per student/hour (150/7) Approximate average hourly running costs (£) 50 50 30 5 15 150 21 assured by HEC LieÁge that they also largely apply in Belgium, and by our French partners about to take part in the next phase of the project that any di€erences in France will be insigni®cant. It is worth mentioning that the estimate of £15 per hour for the studio and its equipment, an estimate provided by the Media Resources Centre at Swansea, contains more than a little guesswork since the dedicated studio is used for a number of other purposes within the institution. In any event, the commercial rate for the use of a fully equipped studio is much higher, especially since it has to take in all manner of overheads, depreciation, etc., as well as pro®t. No ®gure has been included in our calculation for line rental since this cost is minimal when spread over several sessions. 4. Bene®ts The actual expenditure in both Swansea and LieÁge was in fact restricted to call charges only, since the teachers, technicians and anchor people were all unpaid volunteers taking part in a trial from which they hoped to gain expertise in interactive distance teaching, while the space and equipment at both ends of the link were also provided free of direct charge by the two institutions. However, if trans-national tele-education of this kind is to be seriously considered as an ongoing teaching and learning medium fully integrated into curricula, one has to justify an expenditure of over £20 per student/hour. Can the potential bene®ts justify the cost? The only bene®t which can be estimated in purely ®nancial terms lies in the potential savings on travel, and to some extent subsistence, incurred in student (or teacher) mobility, a bene®t much stressed in the literature (Lange, 1994; Donahue, 1996; Faulhaber, 1996). In the case of a language learning link, students who cannot for one reason or another go abroad to a country where their target language is spoken, can nevertheless participate in live tutorials in that language, and also experience something of the ethos and general teaching environment of an educational institution in the target country. But this ®nancial bene®t does not outweigh the costs involved. A 20-hour interactive course taught telematically, assuming seven student participants per tutorial, will amount to £500 or so per student. And tele-students do not receive any peripheral bene®ts such as that of imbibing atmosphere 50 G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 and culture which can come only from actually living for a period of time in a foreign country. A longer course compares even less favourably, since the cost of data-transmission remains always in direct relation to the time spent online, whereas the cost of basic travel (though not subsistence) is ®xed, whatever the length of a course. From the point of view of an institution which receives foreign students, there are perhaps quanti®able bene®ts with respect to savings on administrative costs, but such costs may not amount to a particularly signi®cant percentage of the total cost of accommodating and providing courses for visiting students. True, the cost of ISDN video may well be justi®able in certain speci®c educational or training situations, for example those of part-time courses delivered to under-developed or inaccessible geographical regions. ISDN video may also be cost-e€ective in certain noneducational situations where many people whose time is very expensive would otherwise have to be involved in travelling to meetings, as in the case described by Merwyn (1993) where substantial economies were e€ected using videoconferencing for both ad hoc and scheduled interaction between German ocials planning the move of administrative units from Bonn to Berlin. It may be economical to use ISDN video in order to assess the quality of teaching in franchised or otherwise validated degree schemes: it is clearly less costly to monitor occasional classes remotely than to send out teams of academics for short periods of time to do so in situ. It may be desirable, and therefore cost-justi®able, to set up ISDN video links to provide students with the opportunity of making virtual visits to foreign campuses in order to see a range of teaching cultures and methods, levels of academic standard, etc., and to base their choice of destination on that virtual experience (though, to our knowledge, videoconferencing has not yet been used in this way in Europe). Such examples must nevertheless be seen, at the present time, as either exceptional or, if they seem to o€er an interesting potential for general use, too dicult to set up for whatever reasons. Moreover, they represent something di€erent in kind from live remote interactive tutorials given to full-time students in higher-education institutions of major European cities. Indeed, it is all but impossible, save in exceptional circumstances, to justify the cost of such video teaching links purely in terms of savings made by avoiding the physical displacement of students. Even if one argues that distance learning brings with it personal skills worth cultivating, more than £20 per student hour seems a high price to pay for what would essentially be, for most students, a means of gaining expertise of marginal value. And while our experience has taught us that education by video link challenges the teacher far more than the student, and that the necessary teaching skills are therefore perhaps all the more worth cultivating, one might nevertheless legitimately raise a similar objection as far as teachers are concerned. In the eyes of some, however, certain unquanti®able factors weigh heavily on the bene®t side. For example, from our analysis of the feedback questionnaires regularly completed by the Swansea and LieÁge students taking part in our trial, it is clear that the experience of being taught business topics in a foreign language by genuine foreign teachers of Business Studies was found to be rewarding and useful (Jacobs & Rodgers, 1997), especially since both the Swansea and LieÁge students were due to spend the coming academic session in either the very institution at the other end of the link or in a similar institution. When one takes account of such factors, and adds in some of the bene®ts mentioned above, it can be argued that transnational teaching by ISDN video is a worthwhile activity. This is particularly so if one also G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 51 counts certain unquanti®able bene®ts such as that of being able to use video recordings of sessions as, essentially, customized future teaching resources, or the bene®ts gained by raising the pro®le of an institution as pioneering innovative teaching methodsÐboth Swansea and LieÁge bene®ted from coverage in the press and, in the case of Swansea, on local radio and national television. 5. Conclusions Despite these latter points, the enthusiasm shown by convinced advocates of interactive teleteaching must be set against both the diculty of quantifying certain bene®ts, and the relatively light weight of others when balanced against the ®nancial burden. Yet within the ever-expanding research literature associated with this form of teaching, most of those who write about it start from the premiss that it is a Good Thing which, at worst, requires some tuning in order to determine the best technical and/or pedagogical method of delivery. Thus one does not often read a report of an interactive video distance-learning link which failed because it was not deemed cost-e€ective: Latchem, Mitchell and Atkinson (1994) cite an Australian experiment in language teaching using videoconferencing equipment where the costs per student were eventually considered too high to justify continuing, but such reports are rare indeed. As ISDN tele-communications become more widespread, the costs involved, both recurrent and ®xed, are likely to fall even if not signi®cantly so. One can also for certain look forward to a time when broadband technologies will allow fully interactive, broadcast-quality video between sites at signi®cantly lower cost than at present. Indeed, there are examples of distanceteaching programmes using ®bre-optic networks for broadcast-quality interactive video, such as that of a group of medical schools in Baltimore (Itzel, 1996), but these dedicated networks are intended only to link various sites in geographical proximity to each other, and the cost of using them is often subsumed into other costs such that they appear to be free of charge for legitimate users. Finally, one can perhaps envisage a time when remote interactive teaching has become so well integrated into educational environments that tele-teachers and tele-students will no longer be pioneers, when the technology has become as commonplace as that of the telephone, when the need to justify capital-investment costs has all but disappeared, and when the now necessary presence of technicians or anchor people will be a thing of the past. In the meantime, the assumptionsÐin many cases the claimsÐof some who appear to consider only the bene®ts of fully interactive tele-education by video link should be treated with a measure of circumspection. More than two years ago, Walisko (1995) saw the international digital video market at a turning point, arguing that new methods of digital compression would in the very near future allow the cost of satellite video channels to fall to a level at which they could be used on a global scale for interactive educational purposes. There is, as yet, no realistic sign of this happening. Even in the supremely successful example of a distance-learning organization, the one whose model has been adopted world-wideÐthat of the Open UniversityÐthe use of live interactive video is minimal. Should we be equally circumspect concerning attitudes expressed about the desirable future of distance learning in ocial documents, such as those to be found in the important White Paper on Education and Training published in 1995 by the European Commission? (DG XXII, 52 G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 1995). Here, there is a clear bias in favour not only of physical student mobility within Europe but also of the bene®ts of trans-national distance learning where substantial growth is forecast due to advances in information technology. It is easy enough, as the White Paper does, to generalize about these bene®ts with statements such as: Education and training must draw on the new communication technologies and harness their full potential [ . . . ] This means, in particular, that Europe must adopt new high-quality teaching instruments adapted to its educational and cultural needs. (p. 35) It is harder to justify their cost. Certainly, some of the distance-learning projects funded by the European Commission can barely be faulted in the bene®ts they bring, and cost should hardly be a factor when considering their value: projects aimed, say, at helping socially disadvantaged people or the children of itinerant workers. But it has to be added, as the White Paper does at one point, though only very obliquely (p. 27), that costs will in the end be a major determining factor in most cases. Equally, Sir Ron (now Lord) Dearing's National Committee of Inquiry into Higher Education has attached great importance in its report (Dearing et al., 1997) to the development of distance learning using communications technology (Chapter 12, Item 26), and indeed has perceived increasing future competition for UK higher education coming from, among other sources, ``institutions overseas making extensive use of distance learning through modern technology'' (Chapter 1, Item 20). Nowhere in the Dearing Report, however, is there a direct consideration of the comparatively high costs associated with using live video for distance teaching. The Report refers, for example (Chapter 13, Item 27), to a survey of academic sta€ in the UK indicating that the less than widespread use of videoconferencing is probably due to the lack of availability of the technology, in this case on the Joint Academic Network (JANET), rather than to a lack of the desire among such sta€ to use it. No mention is made of the inhibiting cost factors in terms of human and material resources when combined with the relatively small number of students able to participate in any one session. There are two ways in which costs can be reduced. The ®rst is to combine face-to-face teaching with tele-teaching, in other words for the teacher to conduct a conventional seminar with local students while at the same time being joined telematically by the remote students (this is currently being done in a number of US institutions, for example at the University of Arizona). The second is for the teacher to teach groups of students in di€erent institutions (or even in the same institution) at the same time, using multiple ISDN connections. Both approaches reduce the cost of a student/hour. Yet whatever cost-savings can be made, the overall conclusion which imposes itselfÐnamely that the enthusiasm displayed by exponents of telematic distance learning should always be tempered by a cold look at the ®nancial realityÐ must remain a prime consideration. References Acker, S. R., & Levitt, S. R. (1987). Designing videoconference facilities for improved eye contact. Journal of Broadcasting and Electronic Media, 31(2), 181±191. Berge, Z. L. (1995). Facilitating computer conferencing: recommendations from the ®eld. Educational Technology, 35(1). G. Jacobs, C. Rodgers / Computers & Education 31 (1998) 41±53 53 Coiro, S., Davoli, F., Maryni, P., Pulia®to, P. P., Pavan, M. A., & Talone, P. (1993). Design and experimental monitoring of an ISDNbased multimedia distance learning service. IFIP Transactions AÐComputer Science and Technology, 29, 157±166. Colbert, M., Voglimacci, C., & Finkelstein, A. (1995). Live, audiovisual communication-systems for distance learning: experience, heuristics, and ISDN. Behaviour and Information Technology, 14(5), 267±288. Comeaux, P. (1995). The impact of an interactive distance learning network on classroom communication. Communication Education, 44(4), 353±361. Davis, J. L., & Smith, T. W. (1994). Computer-assisted distance learning: audiographic teleconferencing, interactive satellite broadcasts, and technical Japanese instruction from the University of Wisconsin-Madison. IEEE Transactions on Education, 37(2), 229± 233. Dearing R., et al. (1997). Higher Education in the Learning Society, Report of the National Committee of Inquiry into Higher Education. London: NCIHE Publications, HMSO (July). DG XXII (1995). Teaching and Learning: Towards the Learning Society, White Paper on Education and Training. Brussels: Commission of the European Communities DG XXII (November). Donahue, S. (1996). Programs and resources in distance education. Journal of the American Society for Information Science, 47(11), 870±874. Faulhaber, C. B. (1996). Distance learning and digital librariesÐtwo sides of a single coin. Journal of the American Society for Information Science, 47(11), 854±856. Harris, A. (1996). The evolution of video technologies for distance learning. SMPTE Journal, 105(12), 742±747. Itzel, W. J. (1996). Distance learning through wide area networking. Media and Methods, 32(4), 6. Jacobs, G., & Rodgers, C. (1997). Remote teaching with digital video: a trans-national experience. British Journal of Educational Technology, 28(4), 292±304. Kaye, T. (1994). Co-Learn: an ISDN-based multimedia environment. In ISDN Applications in Education and Training (pp. 179±199). R. Mason and P. Bacsich (Eds). London: IEEE. Lange, J. (1994). ISDN videoconferencing for education and training. In ISDN Applications in Education and Training (pp. 127±143). R. Mason and P. Bacsich (Eds). London: IEEE. Latchem, C., Mitchell, J., & Atkinson, R. (1994). ISDN-based videoconferencing in Australian tertiary education. In ISDN Applications in Education and Training (pp. 99±113). R. Mason and P. Bacsich (Eds). London: IEEE. Merwyn, A. (1993). Videoconferencing goes to work. New Media, 3(11), 60±64. Oftel (1996). Oftel Considering BT's Revised ISDN Prices, Report of 21 August, 1996. London: Oce of Telecommunications. Rose, D. A. D., & Clarke, P. M. (1995). A review of eye-to-eye videoconferencing techniques. BT Technology Journal, 13(4), 127±131. Sebrechts, M. M., Silverman, B. G., Boehmdavis, D. A., & Norman, K. L. (1995). Establishing an electronic collaborative learningenvironment in a university consortium: the CIRCLE project. Computers and Education, 25(4), 215±225. Walisko, W. V. (1995). Digital compression and the international TV marketplace. Proceedings of IBC 95 (International Broadcasting Convention). (pp. 372±376). London: IEEE. Yi, H., & Majima, J. (1993). The teacher-learner relationship and classroom interaction in distance learningÐa case-study of the Japanese-language classes at an American high-school. Foreign Language Annals, 26(1), 21±30.