Education & Computing 8 (1992) 229-237 Elsevier

Computer-based approach

training: an institutional

Philip Barker and Karim Manji School of Computing and Mathemaucs, Teesside Polytechnic, County Cleveland, UK

Abstract

Barker, P. and K. Manjl, Computer-based training: an institutional approach, Education & Computing 8 91992) 229-237.

Conventional approaches to the provision of learning and training opportunities often leave much to be desired in terms of their suitability, their relevance and their cost-effectiveness. Of course, in the current financial climate, educational establishments are also being called upon to achieve much higher productwlty with minimal increase in resources. In order to meet the increasing demands that are being placed upon conventional training and learning resources, many organiza- tions are actively searching for new approaches to instruction. Because of the attractwe facihties that they offer, computer-assisted learning (CAL) and computer-based training (CBT) have now become important techniques for both knowledge dissemination and skill acquisition. Five important attributes of these approaches to pedagogy have contributed to their success: efficiency, effectiveness, adaptability, automated monitoring and the high degree of student motivation that can be achieved. For the reasons outlined above, many organizations are now setting up "institutional" CAL or CBT centres. Their function is to provide courseware development services on a cross-organizational basis. This paper discusses some architectural blueprints for the creation of a successful courseware development centre. In this paper, we cover such topics as courseware engineering principles, authoring tools, media selection, workstation design, production strategies and the marketing of courseware products.

Keywords: electronic learning; computer-assisted learning; computer-based training; courseware development; authoring tools.

Introduction: why CAL and CBT?

Over the last decade, there has b e e n a t r emen- dous growth in the use of computers for the imp lemen ta t i on of l ea rn ing and t ra in ing pro- cesses at all levels wi thin the educa t iona l system [1,2,12]. Undoub ted ly , this growth was ini t ia ted by the commercia l availability of relatively low- cost (but very powerful) mic rocompute r systems. However, in recent years the growth of instruc- t ional compute r usage has b e e n amplif ied by the availability of easy-to-use courseware develop- men t tools and high capacity (and robust) storage media to facili tate courseware dis tr ibut ion.

Historically, pedagogic appl icat ions of comput- ers have always b e e n classified into one or o ther of two b road categories. The first of these (called computer-ass is ted learning, or CAL) emphasizes the use of computers for the in i t ia t ion and pro-

mot ion of l ea rn ing processes. The second cate- gory (called compute r -based t ra ining, or CBT) is more concerned with providing compute r envi- ronmen t s that will facilitate skill deve lopment and e n h a n c e m e n t within par t icular task domains.

A useful acronym that is somet imes used as an umbre l l a te rm to describe both of these ap- proaches is CBI; this is an abbrevia t ion for com- pu te r -based ins t ruc t ion [13]. Al though there are subtle differences be tween CAU and CBT, in this paper the terms will be regarded as synonymous and they will be used in terchangeably . Thus, any reference to one will also be t aken to imply a reference to the other.

Both CAL and CBT have a n u m b e r of impor tan t and appeal ing a t t r ibutes which make them very attractive techniques for the imp lemen ta t i on of many different types of l ea rn ing and t ra in ing activity. Therefore , in order to provide a basis for

0167-9287/92/$05.00 © 1992 - Elsevier Science Publishers B.V, All rights reserved

230 P. Barker and K. Manji / CBT: an mstitutional approach

Table 1 Attractive features of CAL and CBT

1. They are based upon electronic information; pedagogic resources are therefore easy to disseminate, deliver and update.

2. They are flexible in both approach and application and so are ideal for use with open and distance learning.

3. They are able to facilitate both individualized and group- study approaches.

4. They can be used to teach subjects that could not easily be taught by other means.

5. Automatic record keeping and assessment can be used. 6. They can be used to facilitate a shift from teacher-centred

to learner-centred pedagogy. 7. It is easy to implement learner control. 8. Students like using interactive computer systems. 9. Computer-based multimedia learning is effective.

10. More pedagogic resources are becoming available to sup- port these modes of learning and training.

the material that is to be discussed in subsequent sections of this paper, we present in Table 1 a list of reasons why we feel CAL and CBT are useful pedagogic candidates for the implementation of learning/ training processes.

Although they represent very powerful peda- gogic methods, CAL and CBT require substantial planning and administrative control. This is par- ticularly true in situations where large-scale in- house courseware development is undertaken. This paper therefore discusses some of the im- portant issues associated with the institutional use of CAL and cBx methods. In the following sections of this paper, we discuss three major points. First, we discuss some different ap- proaches to electronic learning. We then outline the basic principles underlying courseware devel- opment to support these approaches. Finally, we proceed to a description of a blueprint for the creation of a corporate CAL/Cm" development centre.

Approaches to electronic learning

Th¢ term "electronic learning" is one which is often used to describe the various types of peda- gogic activity that can be brought about by infor- mation or knowledge that is held in electronic form. Johnston gives a useful description of this term [18]. He also offers a useful taxonomy of media and technologies to support this approach

'4 Networked Computers

(WAN)

3 Networked Computers (LAN)

2 Multiple Microcomputers

I' I Indivtdual Microcomputers

Groups of Users

Distributed Learnmg

Nattonal and International Dtstance Learning

Fig. 1. Levels of CAL/CBT involvement.

to learning. This taxonomy is based upon the "reactivity" and the degree of end-user control that a communication medium allows. Three broad classes of media are introduced: pgssive linear media; interactive communication tech- nologies; and those that support interactive user command and control. In this paper, we are con- cerned only with the last of these categories. That is, systems that are dependent upon the interac- tive use of computers.

Computers can be used to implement a very diverse range of approaches to electronic learn- ing. The complexity of the systems that can be employed for pedagogic purposes varies enor- mously from one application to another. Some of the most important types of pedagogic computer configuration that are in current use are illus- trated schematically in Fig. 1.

Undoubtedly, one of the simplest approaches to electronic learning is through the use of indi- vidual stand-alone microcomputers. These are of- ten referred to as "student workstations." Work- stations may take the form of portable microcom- puter systems that are located in a home environ-

P. Barker and K. Manji / cBr: an mstttuttonal approach 231

ment, in an office or in a study bedroom within a college campus. Of course, students may also be required to use various types of non-portable machine for some subjects and courses. Again, these can be located in many diverse locations, such as a library, a common room or purpose-built study areas. The detailed specification and type of microcomputer that is used to implement learning and training will depend critically upon the nature of the subject being studied and the kind of work that students are to undertake.

In order to facilitate the teaching of larger groups of students, collections of microcomputers may be brought together into one physical loca- tion to form either an open learning centre or an electronic classroom. These types of learning con- figuration have similar intent and functional r61es, but differ in the ways in which they are used and administered. Sometimes it is convenient to link a collection of microcomputers together, using a local area network (LAY). Such an arrangement, of course, allows the creation of "logical" elec- tronic classrooms, as well as physical ones. An- other way of achieving the creation of logical electronic c l a s s rooms- -bu t which are distributed over wider geographic a r e a s - - i s through the use of a national or an international wide area net- work (WAN). As well as facilitating the creation of electronic classrooms, networks of this kind also allow various types of distance learning to be implemented.

The availability of computer-based communi- cation networks means that the delivery of peda- gogic resources can be accomplished fairly easily. Two approaches are possible. The network can be used either for providing access to courseware that is resident at some particular remote host location on a WAN, or for distributing courseware to various remote network locations, using an electronic mail file transfer facility. Another use- ful way of distributing courseware is through the use of digital optical storage media (compact disc read-only-memory, or CD-ROM). Of course, the use of this medium requires that student worksta- tions are fitted with a suitable CD-ROM unit that will enable materials to be accessed. Currently, most courseware is distributed on discs that con- form to the High Sierra standard. However, new ways of using CD-ROM are emerging. Two impor- tant developments are ci>i (compact disc-inter- active) and DvI (digital v ideo - in t e r ac t i ve )

[9,19,20]. These standards define ways of using compact disc that will facilitate the development of many new and novel types of interactive multi- media pedagogic products. One important type that we are researching is the production of elec- tronic books for use on portable microcomputers. Details of this work are presented elsewhere [9,11].

The importance of ci>~ and Dw lies in the fact that they are "all digital" methods of storing multimedia pedagogic resources on an optical storage medium. That is, they offer facilities for storing text, static pictures, sound and (using compression techniques) Tv-quality video pic- tures, on one common medium. CD-I is intended to form the basis of a bespoke range of delivery stations that consumers would purchase outright. On the other hand, DW is intended to be a system builder's approach to desktop digital multimedia. That is, users of DVI simply purchase add-on boards that they use within conventional micro- computer equipment.

Courseware engineering for CAL and CBT

The widespread use of CAL and CBT, both in academic and non-academic organizations, re- quires the availability of a substantial repertoire of high-quality instructional software. There are various ways in which an organization can build up its stock of courseware. Three of the most commonly used approaches are: purchasing of bespoke packages; using a "shareware" facility; and in-house production. In this paper, we are primarily concerned with the third of these ap- proaches.

The in-house design and fabrication of instruc- tional software requires the initiation, control and coordination of a range of developmental pro- cesses, all of which contribute to the ultimate goal of producing operational courseware prod- ucts. The various processes that contribute to this overall product development activity are often referred to collectively by the term "courseware engineering." The major processes involved in a typical courseware engineering exercise include: needs analysis; requirements specification; in- structional design; prototyping; courseware im- plementation; evaluation and testing; revision; bulk production; distribution; and marketing. De-

232 P. Barker and 1<2. Manji / cBr: an instUutional approach

scriptions of each of these constituent processes are presented in most of the standard text books on courseware development [13,14].

Courseware engineering activity can be con- ducted at two basic levels of involvement. The first involves encouraging groups of individuals to produce computer-based learning resources for the particular courses in which they are involved. The second level of involvement requires the creation of a multidisciplinary courseware devel- opment team that has an institutional responsibil- ity for producing instructional software on de- mand, to precisely specified contractual specifica- tions. This latter approach is further discussed in the following section. When individuals are en- gaged in preparing courseware two basic ap- proaches can be employed, depending upon the type of personnel involved: staff or students. Staff can produce instructional software either for par- ticular courses that they run or in order to meet particular market demands. Staff involvement forms the basis of the electronic learning-package factory (ELF) methodology [3,16,21]. Alternatively, because courseware production is itself an educa- tional process, students can be encouraged (as part of their learning process) to develop course-

ware for use in other courses. When these meth- ods of producing courseware are employed, it is imperative that suitable courseware development tools and strategies are provided. Some course- ware development strategies (and the types of tool that their use implies) are illustrated schematically in Fig. 2.

Depending upon the extent of automation that is involved, this diagram allows three basic types of courseware implementation strategy to be de- rived. Conventional programming approaches (using such tools as BASIC, PC/PILOT, Microtext, etc.) do not allow any substantial automation of the courseware implementation process. These tools are used primarily by professional course- ware engineers. CASE tools, which allow partial automation of software production, also fall into this category. For staff and students who do not have any programming ability, we recommend the use of courseware generators that totally auto- mate the courseware implementation process. Some popular examples of such generators are PROPI, LinkWay, HyperCard and Course Builder. Further details of these packages are presented elsewhere [4].

Historically, of course, one of the major prob-

Computers

Partml Automat=on

Productivity Tools I (CASE Packages)

~[ PEDAGOGIC COMPUTER APPLICATIONS

\T! L, ) l OOURSEWARE j.

Programming t Total Automat=on

Engineers

Basic 1 ] Courseware Development Courseware Tools Generators

M I Techn,ques~ .... T End-Users I

Instructors I Students

Fig. 2. Approaches to courseware engineering.

P. Barker and K. Manfl ./CBT: an mstttuttonal approach 233

Organisations j I and Open Learning Centres Marketing Unit institutions

I The Teaching Departments ~ CAL/CBT Centre I Courseware Development J

]Research and I ~ ~ ' ' ~ ~ P : : s ~ ~ ~ Staff Tr ..... g ] Oevo,oprnoo, l Site ,,°°noes [

Special I ~ Services

Fig. 3. Blueprint for an institutional CAL/CBT centre.

lems that courseware engineers have always faced has been the substantial amount of development effort that has had to be expended in order to produce a relatively small amount of CAL or cBx resource. Typically, cm requires as much as 100- 300 hours of development t ime per unit of deliv- ered instruction. Ellis [15] has referred to this as the "entropy p r o b l e m " - - t h a t is, having to put a lot of work and effort into producing relatively little output. Although automation tools, of the type described earlier in this section, can help solve some of the productivity problems associ- ated with cm, it is obvious that a more fundamen- tal solution is needed.

One potential solution has been described by Ellis [15]. He claims that the successful use of computer -based education at his institution (Queens land Univers i ty of Technology) is attributable to the use of tutorial CAL that in- volves only Socratic dialogue based upon ques- tions-and-answers. He claims that producing cm resources of this type, which can be used as an

adjunct to conventional lectures, is probably the most cost-effective way to proceed towards insti- tutional use of CAL. Indeed, the figures that he quotes, both for student usage and for develop- ment costs, would seem to support this claim.

In the past, courseware engineering activity has most often been associated with the develop- ment of instructional software, which, to a large extent, was intended to act as a substitute for conventional teaching. In many situations (prim- arily, self-study in open and distance learning situations), this requirement still needs to be met. However, in institutions that use conventional "chalk and talk" approaches to teaching, the adjunct approach used by Ellis (amongst others) may offer an optimal way to progress.

The CAL / CBT centre

CaL and CBT can be used in a variety of differ- ent ways to augment the ongoing educational and

234 P. Barker and K. Manji / CBT: an institutional approach

training activities that take place within both aca- demic and non-academic organizations. Indeed, computer-based techniques can be used both as an adjunct to existing training methods and as a replacement for conventional approaches to in- struction. The extent to which CAL and CBT are used will obviously dictate the institutional strat- egy for the provision of access to centralized resources. Once activity grows above a certain threshold level, it is worth considering setting up a central CAL//CBT centre to provide corporate services in this area. A possible organizational structure for such a centre is illustrated schemati- cally in Fig. 3. In the remainder of this section, we discuss this blueprint for a CAL centre in terms of: its functional r61e; the provision of campus services; staffing and equipment requirements; and the marketing of courseware resources and CAL//CBT services.

Functional r6le

We envisage that the functional activities of a CAL//CBT centre would fall into four broad areas: research, development, marketing and education. The research function is important from the point of view of designing and exploring new ap- proaches to computer-based pedagogy. The de- velopment function is intended to encompass a variety of basic r61es, such as the creation of new educational products, the production of new tools and the design of better techniques for using computers for learning and training. The educa- tional function of the centre is intended to pro- vide advice, awareness and training facilities for staff. Finally, the marketing function is intended to generate revenue through the marketing of the courseware products that the centre generates, the provision of specialist services and the run- ning of external courses. As we discuss in the following section, much of the functionality of the ¢AL centre would be realized, at a pragmatic level, through the campus services that the centre provides.

Campus seruices

The range and quality of campus services that a ~AL//CBT centre is able to provide will depend upon a variety of financial, organizational and political factors. In general, we believe that the

major campus services for which a CAL centre should be responsible should be categorized un- der four broad headings: administration; staff training; courseware development; and the provi- sion of specialist services.

The administrative r61es of the centre include: the organization and running of the various open learning centres and electronic classrooms that are available; the implementation of courseware procurement procedures; and the acquisition and control of site licences for courseware packages and the various authoring tools needed to sup- port the in-house production of instructional soft- ware.

The staff training responsibilities of the centre would cover two basic areas. These would provide general courses of instruction on CAL and CBT methods; and various courses on particular in- structional packages, courseware development tools and other necessary equipment. The latter aspect of training would include instruction in the use of author workstations, student delivery sta- tions and special equipment, such as image scan- ners, video discs, CD-ROM equipment, etc. De- tailed descriptions of our own in-house ap- proaches to training are given elsewhere [5].

One of the most important facilities that the CAL centre should offer is a courseware develop- ment service for those members of staff who do not wish to undertake this activity themselves. Of course, the successful in-house production of courseware requires a number of basic prerequi- sites to be satisfied. Amongst the more important of these we include: the use of a multidisciplinary team approach (this is further discussed in the following section); using well-founded and sound design guidelines and models; the use of appro- priate courseware development tools; adopting well-defined production strategies; and making available a range of specialist services that deal with the use of different media and technologies.

The provision of specialist services is an im- portant function of the CaL/CBT centre. As well as providing advice and assistance with respect to the appropriate use of media (sound recording, image production, animation, use of video disc and CD-ROM, etc.), a range of other specialist services need to be made available. Some of the more important of these are: advice and guidance with respect to pedagogic matters; the provision of design studios and prototyping facilities; evalu-

P. Barker and K. Manjt ./cBr: an mstttuttonal approach 235

ation and assessment of courseware; assistance with documentation; and bulk duplication and distribution facilities.

Staffing and equipment requirements

The successful operation of a CAL/CBT centre depends critically upon the way in which it is staffed and equipped. As we have suggested pre- viously, we believe in using a multidisciplinary team approach. The composition of development teams should be chosen from a pool of expertise in which staff skills complement (and in some areas duplicate) each other. Of course, it is ex- tremely important to provide development teams with an appropriate infrastructure and suitable equipment with which to undertake their individ- ual development tasks. Providing these require- ments are met, it is possible to create an effective and efficient institutional CAL/CBT courseware development centre.

Marketing courseware and CAL/CBT services

The major r61e of an academic CAL/CBT cen- tre will undoubtedly be the provision of the types of service and facilities that have been described in the earlier parts of this section. However, it is important to realize that academic CAL units can act as invaluable sources of substantial amounts of courseware development experience, expertise and equipment. In principle, these can often be used to augment the CBT activities of commercial and industrial organizations. Indeed, for small companies wishing to utilize CAL/CBT methods for in-house training (and which do not have any CBT experience themselves), academic CAL cen- tres can provide invaluable advice and assistance. At the same time, the provision of these services, along with contract courseware development, can act as a source of revenue to support the unit's own ongoing research and development activity. Obviously, the creation of new learning and train- ing products - - such as Texas Instrument's Chat- ter-Book, Questron's Electronic Books, Comp- ton's Multimedia Encyclopedia, and so on [6]-- can also act as a source of income to support research work. However, if these avenues are to lead to fruitful results, there is a need to develop aggressive promotional and marketing strategies. This is the function of the marketing unit de-

picted in the upper left-hand side of Fig. 3. It is unlikely that this unit will become involved in the large scale production of CAL/CBT resources. Its major r61es will therefore be in promoting the services available within an institution and in contracting out the actual production and distri- bution processes to other suitable organizations.

Future directions

For some years, we have been involved in developing courseware for a variety of different kinds of teaching application. We have found computer-based methods extremely useful in the context of designing and implementing open and flexible learning/ t raining systems. Based upon our own success in this area, we should strongly recommend the use of computer-based methods to those other organizations that are involved in any form of learning and training activity. Indeed, we should urge other potential users of technol- ogy-based training to consider the use of CAL and CBT for the realization of some, if not all, of their pedagogic activities. Obviously, as the demand for these techniques increases, so their implementa- tion will become easier and less cost ly-- the costs of delivery stations will fall and the range of quality courseware will increase. Of course, within academic organizations, there will be greater scope for sharing courseware and, hopefully, movements towards the realization of collabora- tive development ventures [3]. Naturally, progress towards the widespread use of CAL and CBT will not proceed in an entirely problem-free fashion. One major problem that we foresee is that of global, large-scale courseware portability. A num- ber of research groups are currently investigating solutions to this problem. Our own solution in- volves introducing the concept of meta-level au- thoring. The way in which this technique is being used is described in more detail elsewhere [6].

Obviously, the widespread availability of digi- tal optical storage discs will have a considerable impact on courseware por tab i l i ty - -no tab ly through the CD-~ and DvI standards mentioned earlier in this paper. Optical disc technology will also strongly influence the way in which comput- ers are used for pedagogy and the types of metaphor that CAL and CBT systems are able to embed [23]. In our own work, we strongly advo-

236 P. Barker and K. Manjt / CBT: an institutional approach

cate the use of electronic book metaphors [7,8,9,11]. Indeed, the advent of rewritable optical CD-ROM disc storage and low-cost production fa- cilities will make the prototyping and distribution of such books a relatively easy matter [10]. How- ever, it is unlikely that metaphor will become reality until we achieve the widespread availabil- ity of inexpensive and highly portable delivery systems, similar to the Dynabook [17,22]. Of course, these will then need to be available on a one-per-student basis.

The ability to deliver educational resources via compact disc, using portable computer systems, could have a significant impact on both the way in which an institution organizes its courses and the way in which these courses are managed. In the long term, this could have a marked influence on the r61e that an institutional CAL centre might have. Instead of just being a producer of course- ware resources and a centre of expertise, the CAL/CBT unit might also have to adopt the r61e of an electronic pedagogic publisher.

Conclusion

Over the next decade, computer-based meth- ods are likely to play an important part in the growing movement towards the realization of quality-driven educational systems. However, if computers are to be widely used as an educa- tional resource, it is essential that four basic requirements are fulfilled. First, appropriate hardware resources must be provided. Second, a wide range of quality courseware must be made available. Third, staff involvement must be en- couraged. And fourth, an appropriate infrastruc- ture must be created in order to allow electronic learning to take place.

Within many large organizations, we believe that the use of an institutional CAL/CBT centre is a necessary part of the infrastructure needed to support the use of computers for education and training. In this paper, we have sketched out a blueprint for the creation of such a centre. We have also outlin¢d some of its functional r61es and the services that it should provide.

Acknowledgements

An earlier version of this paper was presented at a workshop on "Computer-Assisted Open Ac-

cess Education," held at Napier Polytechnic of Edinburgh, Scotland, on 2-3 July 1990. Our work on electronic books was funded by the United Kingdom's Training Agency. Fig. 2 has previously appeared in an article on automating the produc- tion of courseware, in Vol. XXII I of Aspects o f Educational Technology (Kogan Page, London, 1990, pp. 203-208), and is used with permission.

References

[1] P.G. Barker, Author Languages for CAL (Macmillan, London, 1987).

[2] P.G. Barker, Multi-media CAL (Kogan Page, London, 1989).

[3] P.G. Barker, Adoption of CAL in higher education: a cooperative approach to research, development and im- plementation. In: Proceedings of the 7th International Conference on Technology and Education, Brussels, Vol. 2 (CEP Consultants, Edinburgh, 1990) 92-94.

[4] P.G. Barker, Tools for hyper-media authoring. Paper presented at ETIC 90, City University, London (1990).

[5] P.G. Barker, Courseware engineering. Short course de- scription (Teesside Polytechnic, Cleveland, 1990).

[6] P.G. Barker, The concept of meta-level authoring. Work- ing paper (Interactive Systems Research Group, Teesside Polytechnic, Cleveland, 1990).

[7] P.G. Barker, Electronic books, Learning Resources Jour- nal 6(3) (1990) 62-68.

[8] P.G. Barker, Interactive electronic books, interactive multimedia. In: R. Wintersburn, ed. Aspects of Educa- tional and Training Technology, Vol. XXIV, REalising Human Potential (Kogan Page, London, 1991) 208-212.

[9] P.G. Barker and S. Giller, An electronic book for early learners: a CD-ROM design exercise. Paper presented at the workshop on Computer-aided Open Access Educa- tion Napier Polytechnic, Edinburgh, 1990.

[10] P.G. Barker, S. Giller and S. Richards, Screen design for CBT: outline design documentation. Working paper (Centre for Electronic Pedagogy, School of Computing and Mathematics, Teesside Polytechnic, Cleveland, 1990).

[11] P.G. Barker and K.A. Manji, Designing electronic books, Journal of Artificial Intelligence in Education 1(2) (1990) 31-42.

[12] P.G. Barker and H. Yeates, Introducing Computer As- sisted Learning (Prentice-Hall, London, 1985).

[13] E.L. Criswell, The Design of Computer-Based Instruction (Macmillan, New York, 1989).

[14] C. Dean and Q.A. Whitlock, A Handbook of Computer- Based Training (Kogan Page, London, 1988).

[15] H.D. Ellis, Using computers in university education: ex- periences from computer-based education at Queensland University of Technology (Queensland University of Technology, Brisbane, 1989).

[16] S.J. Fallows and T. King, ELFsoft: a simple route to CAL authoring in higher education, Educational and Training Technology International 26(4) (1989) 342-347.

P. Barker and K. Manji / cBr: an mstitutional approach 237

[17] A. Goldberg, Educational uses of a Dynabook, Comput- ers and Education 3(4) (1979) 247-266.

[18] J. Johnston, Electronic Learning: From Audtotape to Videodisc (Lawrence Erlbaum, Hillside, NJ, 1987).

[19] A.C. Luther, Dtgital Video in the PC Environment (Mc- Graw-Hill, New York, 1989).

[20] J.M. Preston, Compact Disc-lnteractive: a Designer's Overvtew (Kluwer, Deventer, 1988).

[21] T. Stonier, S.J. Fallows and A. Radtke, Computerisation of a UK university: the Bradford experience. In: Proceed- mgs of the 7th Internattonal Conference on Technology and Education, Brussels, Vol. 2 (CEP Consultants, Edin- burgh, 1990) 441-443.

[22] G. Swarbrick, Dynabook, Personal Computer World (May 1989).

[23] D.D. Thornburg, From metaphors to microworlds: the challenge of creating educational software, Computers and Education 12(1) (1988) 11-15.

Philip Barker is Professor of Applied Computing within the School of Com- puting and Mathematics at Teesside Polytechnic, Cleveland, UK. His re- search group is actively involved in research in three major areas of com- puter science: human-computer in- teraction, knowledge engineering and the design of interactive learning sys- tems. He has published several books and many papers within each of these areas.

Karim Manji is a lecturer working with the Interactive Systems Research Group at Teesside Polytechnic. His research deals with pictorial commu- nication methods and their applica- tion to the design of human-com- puter interfaces. He is particularly in- terested in designing interfaces to new types of electronic book for the sup- port of interactive learning.