Before addressing Kristina Hooper's ten issues in detail, I would like to define my position a little more clearly and also define some of the basic terminology that I shall be using. Firstly, I should clarify that I see my contribution to the conference coming from my professional role as an Instructional Developer with a big "D". The big "D" is meant to imply the concept of development as one that encompasses all the stages of a project of innovation. Development with a small "d" in a context such as the automobile industry is the work of skilled technicians putting together the first prototypes of a new model. Development with a big "D", on the other hand, would in this context be seen as the whole process of market research, identification of market needs or requirements, development of a strategy to satisfy these requirements that would be competitive, design of an appropriate model, followed by the (small "d") development of prototypes, and finally the solution of all mass production and management issues that inevitably arise in order to get the model actually to the market. The analogy to developing an educational innovation is I think clear and need not be elaborated.
The second aspect I would like to clarify is my understanding of some of the terminology which is implied by the title of this paper. Let us start with that part of the title which is also the conference title. It has been encapsulated into the acronym IMM. We need to understand the components of that complex term in order to be sure of our understanding of the term as a whole. In other words, what do we mean by the terms interactive and interactivity on the one hand, and what do we understand by multimedia on the other. Let us take these terms in reverse order.
Twenty, even thirty years ago, it was already common to come across classrooms in certain schools which were utilising multimedia materials. At that time, this would often imply that students were utilising some sort of package that involved audio tapes, slides, some text material and some guidelines as to what was in each component and how best to utilise them. It is important to realise that the concept of using several media in concert to present learning materials is by no means as short lived as the "Multimedia Revolution" that is implied by current writings on the topic (Barber and Tucker, 1990).
Another aspect worth mentioning is the confusion that seems to exist between Multimedia understood as media used in combination and Hypermedia, which is the more recent connotation of Nelson's concept of Hypertext. The essential meaning of Hypertext can be summarised in the two terms, "nodes" and "links". Theodore Nelson (1967) and before him Vannevar Bush (1945) promoted the concept of storing textual information as a network of documents linked together by meaningful "pointers," as an alternative to the linear way that books are organised at the moment The argument is that humans utilise information in a non-linear form, jumping from one aspect of a topic to another as they form their own viewpoints and, therefore, the reference and learning materials that they utilise would be more "user friendly" if organised in a similar manner. Whereas the basis of this concept would seem to be related to the structure of knowledge and therefore not necessarily of recent invention, the feasibility of producing such systems has grown as the power of computer systems has grown. With the marketing of HyperCard as a standard application tool with all Macintosh computers and the development of similar systems operating on other platforms, the concept of non-linear interlinked storage of information has left the laboratory and has entered business and educational contexts. As computer technology has also made inroads into image processing in a way similar to the earlier progress on text processing, it has been natural to incorporate the ability to store information in a variety of media within a network of information and thus the concept of Hypermedia was born.
Why is it important to distinguish Hypermedia as a concept from Multimedia? The use of a variety of media in order to better communicate a particular topic is one issue. The storage of information (in whatever medium) in a network so that it can be more easily cross referenced to other relevant information is another. Such a distinction is not always clear in the minds of practitioners in our field. For example, at the interactive multimedia conference mentioned above, almost 80% of the contributions dealt specifically with Hypermedia, rather than Multimedia. The issues discussed by most of the conference participants were related more to how to usefully link information in non-linear networks, how to analyse knowledge into knowledge structures and knowledge bases that would be educationally useful and how to develop computer based tools that would enable such Hypermedia systems to operate in a cost effective manner. A minority of the papers addressed issues concerned with which media or which combinations of media would be appropriate for a particular topic in order to maximise the clarity of communication and the effectiveness of learning. I would argue that when we are discussing the IMM field, it is useful to make this distinction so that as we listen to each other we do not enter the conversation with a built in predisposition to misunderstand.
Let us come back now to the term that precedes multimedia in our title. The concept of interactivity is also one that has many meanings in our field. It may therefore be apt to state my own understanding of this term. At a very simple level of analysis, interactivity in an educational or instructional process can be taken to imply that a series of interactions occur between teacher and learner, or between learners and other learners and, in the case specifically of machine mediated instruction, between learners and the computer based instructional system. However, to understand interactivity more deeply, we have to consider the various dimensions along which such interactions may vary.
They may vary, for instance, in purely quantitative terms, that is, in terms of the frequency with which the teaching system asks the learner for responses or responds to specific learner initiatives. These quantitative measures of interactivity are ones that are highlighted, for example, in the work of Bork (1987) as ways of measuring the degree of interactivity. A completely different dimension by which one may judge the extent of interactivity that is present would be based on qualitative criteria that look at the nature of the interactions themselves. To what extent do the actions of the teaching system promote surface level recall reactions from the learners or to what extent do they make the learners think and "deep process" their knowledge? For example, the types of interactivity that were promoted by the programmed instruction methodologies of the 1960s and 1970s are much more surface level than the types of interactivity that would be built into a contemporary cognitive science based model for an intelligent tutoring system.
I would also like to say a few things about the terms "courseware" and "networks." Let us take courseware first. This term will no doubt immediately be understood in the context of a meeting like this as "computer based educational material" including also multimedia material that is in some way interlinked and controlled by computer based software. The term courseware has indeed been born in the context of computer assisted instruction. However, it may be useful to remember that computer delivery is only one among many ways of presenting information and exercises to students. Thus, if we are considering a totally integrated approach in terms of multimedia, should we not also be considering the more traditional media such as, for example, paper based printed materials, or dramatised role plays to be enacted live in a group, or email messages received by a student in response to an assignment. Should one restrict the term "courseware" to the context of materials, media and systems that are computer controlled or delivered or should one now adopt this word as the generic term for learning materials whether a computer is involved in their delivery and control or not? My own tendency is to argue for the expansion of the use of the term in order to fight the erroneous misconception that if one is using computers as part of an instructional delivery system, then the use of other media such as group interactions or paper materials is a sign of failure on the part of the designers.
Coming now to the term networks, I would just like to point out that this term can also be construed in at least three quite different senses. The meaning that probably first comes to mind is "local area networks" or even "wide area networks" to enable the sharing of materials, or the sharing of comments on materials, between distant participants. However, another meaning is related to the structure in which a particular set of topics has been stored and interlinked in order to create a particular example of Hypermedia courseware. Yet a third meaning is the network of concepts and ideas that exist in a person's mind on a particular topic. All these connotations of network are indeed valid and in fact will be both are being addressed in this paper. However, once more, we ought to be quite clear in what sense we are using the term at any moment.
It is now time to proceed to the major issues that I wish to explore in this paper. As mentioned before, I shall organise these around the ten issues that were raised by Hooper in her summary of the 1986 Apple conference. I have reorganised the sequence of these points but I reproduce them in Figure 1 in exactly the words used by Hooper.
Bork (1987) in his discussion of interaction focuses very much on the quantitative aspects. He goes as far as to define two numerical factors which determine the degree of interaction. The first factor is the average time between interactions which he defines as ideally being in the range of 15 to 20 seconds. The second quantitative factor he suggests is the time it takes for the student to respond. In this context he favours challenges that take the student somewhat more time to respond than the typical one keystroke mechanical responses that are often all that is required to move on within a program. He argues that short times are not desirable in a learning environment "because we want to encourage thinking as part of the learning process - when a student replies to a question we want the student to think about the question, not just to react."
Whereas I agree wholeheartedly with the implied intention of involving the learner in deep reflection, I have some doubt as to whether timing the response period that a learner takes to respond is a valid measure. I feel that quantitative measures alone are insufficient for decision making in relation to when and how to incorporate interaction into a learning experience. Unfortunately, Bork offers little precise guidelines in terms of qualitative criteria that could be applied in a the design of interactive learning, beyond suggesting that this requires skill and knowledge both in instructional techniques and in the content of the lesson on the part of the designer.
Several qualitative models have been proposed. One particularly complete and complex one is suggested by Jonassen (1986) in which he identifies three dimensions along which an interactive/adaptive lesson might vary. Jonassen distinguishes rather carefully between his definitions of interactivity and adaptation. Interactive instruction is defined as a process in which the learner is involved in overtly responding to material by making selections or giving answers to questions, etc. Adaptive instruction is a process in which material presented and exercises performed by the learner are selected in some way that is sensitive to what has happened within the learning process to that point. Jonassen further distinguishes between internal adaptations within the structure and content of a given lesson and external adaptations related to selecting lessons or components in relation to differences in prerequisite skills, etc.
Using these three dimensions (interactivity, internal adaptation and external adaptation), Jonassen builds a 6x6x4 cube resulting in as many as 144 different cells that would describe materially different situations that all could be defined as variants of interactive lesson designs. A study of other authors has revealed that many other possible dimensions or factors could be included in one's analysis of what constitutes interactivity. Whether one accepts or rejects Jonassen's model, there is no doubt that the concept of interactivity can vary from one person to another and one situation to another.
In order to simplify this apparently complex picture, I would propose to compare interactivity as it may be experienced within a computer based or interactive video lesson with interactivity as it is experienced on a day to day basis in human interactions or conversations. Let us examine a series of typical scenarios. A technical instructor may restrict information presentation to practical demonstrations and explanation of a minimum of key points to be remembered in the execution of a given task. The learner would then be expected to follow the demonstration and possibly explain back the key points to the instructor. The instructor's feedback would be limited to corrective actions if the learner's performance were in any respect below criteria of expectation. This essentially is a description of the TWI (or Training Within Industry) methodology that was developed during World War II to instruct novices recruited to perform repetitive industrial operations. It is probably one of the simplest and most easily replicated models for planning instructor/student interactions. One may notice the similarity between this account of human interaction and what was common in linear programmed instruction which was perhaps the first systematic effort at machine based instruction.
As one proceeds to more complex tasks and content, the human interaction process becomes somewhat more complex. The instructor may not simply correct and demonstrate anew the poorly performed element in the original demonstration but might, as for example in the case of teaching computer programming, use alternative examples or analogies or other tactics that would simplify the message or overcome a specific learning difficulty being experienced by a given learner. We are now seeing an example of human interaction that is analogous to branching programmed instruction that is still essentially the model upon which the bulk of currently used CAI is based.
Proceeding yet further, we may observe human interaction processes that exhibit yet more complex structures. We may witness a conversation which is more student led rather than instructor led. For example, a student approaches the expert with a specific question. When this is answered, this raises yet another question in the mind of the student. The instructor in this role is a resource, supplying information that is requested by the student in the light of the student's current understanding of a topic and current needs for deeper understanding. Such a conversation is analogous to a student led search of a database on some topic or other, exemplified by research project work in the library or by browsing a Hypertext or Hypermedia document.
More commonly, however, the student may ask for help with a particular issue and the instructor in replying may take the conversation beyond the student's initial request. This may be in the form of supplying more information that the instructor realises would probably be relevant, although the student did not initially ask for it. Alternatively it may take the form of some challenge for the student to demonstrate understanding of the information by applying it to a problem situation. An instructor who uses such a technique is engaging in a conversation that is attempting. to go beyond surface interactions related to the content being discussed, to reach some sort of "interaction of minds." By posing challenges and asking questions, the instructor is probing the cognitive structure that the student has formed. By supplying extra uncalled for information, the instructor is extending this knowledge structure in ways which (from superior expertise and knowledge of the domain) the instructor realises will be beneficial to the student. This form of dialogue has often been referred to as Socratic dialogue though it can indeed take many forms, not all of them very similar to that represented as Socrates' methodology by Plato in his writings. The equivalent machine based instruction models would be those that are currently emerging from attempts to create intelligent tutoring systems.
An intelligent tutoring system should be able to handle both system and student initiated questions and should be able to learn from a given student and adapt in a manner that has not been predicted and pre-planned into some initially constructed courseware. Thus, a characteristic of a truly intelligent tutoring system would be that it could lead a conversation into areas that are validly part of the domain under discussion but have not necessarily been predicted by the designers of the system. Furthermore, once the discussion gets into such an area, the system should be able to handle the discussion in a way analogous to any intelligent human tutor. Taking this description as the ultimate goal, it is fair to say that no practical project that is currently ongoing has as yet modelled all aspects of intelligent tutoring.
To summarise, interactivity in the instructional process operates by setting tasks for the learner to deal with but its value and its nature can best be described by the "depth of processing" or the quality of thinking that is demanded from the student. In the scenarios I have outlined, we can see a continuum from simple surface level interactivity where simple responses are observed and commented, through to deep level conversational dialogues where both learner and teacher are gaining insights into each other's minds. This dimension (surface processing to deep processing) may be one of the most useful ways of evaluating the nature of the interactivity that has been built into a given instructional product. Two important questions now remain: in what way can one plan to build in an appropriate level of interactivity; how should one proceed in order to achieve effective and useful interactivity at the appropriate level?
How far have we come along the route to answering such questions since the Apple conference was held five years ago? To answer this question, it is necessary to look at the activity in the area of research and development on the design of interactive media. If we look at the research literature, it is heartening to see an ever growing tendency towards exploring aspects of cognitive science that may throw light on how to proceed in order to achieve deep cognitive processing on the part of students. If, however, we look at the practical field of development, we may be somewhat disheartened in that even the most recent products in the area of interactive multimedia are predominantly employing surface level interactivity. There is much basic research yet to be done. However, we are still a long way from transferring into practice the principles of learning and instruction that the current state of the cognitive sciences has to offer us.
As mentioned earlier, the concept of utilising media in combination has been around for many generations. What is the rationale for this? One driving force behind the use of multimedia presentations and packages in education has been the belief that due to individual differences among students in how well they learn from auditory or visual or verbal material, the provision of audio, audiovisual and textual versions of the information on a particular topic would be beneficial in that students could in some way be matched to their most effective modality of learning. A further factor that would indicate the use of a variety of media is the content to be transmitted. Certain content requires that pictures or sounds be used while other content requires structure to be shown and some sort of graphic communication is essential to this process.
Yet a third viewpoint that has stimulated the use of mediated instruction is the belief that certain media can teach more effectively than others. This has led to a very large body of research built up over a period of more than half a century that has compared the effectiveness of teaching a particular topic to a particular type of student by different media. Of late, much of this work has been attacked, notably by Richard Clark who has re-analysed much of the research and the meta-analyses of this research and has demonstrated that when other factors are fully controlled, then very little if any variability in learning due to alternative media choices can be demonstrated. This has led Clark to formulate his now notorious comment that the contribution of media to the effectiveness of learning is no more than "the truck which delivers groceries to the market contributes to the nutrition in a community" (Clark 1985).
What Clark stresses in his work is that the important factor which results in significant differences in instructional effectiveness is the instructional design of the lesson. Of course, in any comparative research of the nature that he criticises, the instructional design should be kept identical in the various modalities of lesson compared. Otherwise, this is a confounding factor. However, it is exactly in this aspect that there is the hidden promise of media in education. A given learning outcome may be achieved more effectively by a superior instructional design and occasionally, though not always, the superior instructional design may require the use of specific media that have not been used in previous instructional designs. One does not have to go to the extremes of quoting a self study course I once discovered in the literature that attempted to teach the Pitman's shorthand without illustrating any of the symbols by means of visuals or graphic diagrams, to make the point that the selection of appropriate media modalities may be a critical factor in achieving an effective instructional design in a particular topic area.
Certainly what has just been said is not new. Nor are the media elements that can be combined into an effective instructional design new. These amount to text material or graphic illustrations or more realistic still pictures or moving sequences or pure audio sounds or verbal messages. What is new with the advent of the optical laser discs and their control by computers is the ability to store all of these modalities in one form of delivery medium and to access, interweave or combine these modalities with great facility and, in some instances, in ways which were very difficult or indeed impossible with earlier technologies.
This question requires a clear definition of what we mean by instruction. In three of the four mini-scenarios of human interaction, we observed the use of the word "instructor," whereas in one Of them I used the word "resource" for the teacher element in the conversation. Whenever the word instructor was used, there was a clear implication that some fairly specific learning outcome was in the mind of both the learner and the instructor. In the first (Training Within Industry) scenario, the learning outcome was the competent execution of the task required to achieve results on the job. But even at the other extreme, in the so-called Socratic dialogue example, the instructor clearly had some idea in mind of learning outcomes that should be achieved if the interactions were to be considered successful. Only in the student driven example where the teacher is restricted to answering the questions posed by the learner could we postulate that the teacher has no clear objectives in mind for the outcome of the interactions. There may still be objectives, of course, in the mind of the learner but I am implying that the learner is not necessarily communicating these to the teacher nor asking for help in their achievement.
This is the key difference between instructional systems and information dissemination systems. In an instructional system, the teacher element not only will present information but will engage in some form of activity in order to evaluate the effect on learning of that information and use that evaluation to attempt to improve the learning process. In an information dissemination system, the emphasis is on the nature of the information, the content to be included, its organisation, structure and presentation. Naturally, learning may be expected and intended and may indeed occur. However, there is no mechanism built in to the process to evaluate whether the learning has occurred and to take remedial or other action if learning has not occurred.
It is interesting to note that the bulk of interactive multimedia products within the educational field are indeed information dissemination systems rather than instructional systems. In the area of interactive multimedia applied to training, this is not quite the same. In this area the bulk of interactive video programs that are currently being used do exhibit the elements of presentation, practice, evaluation and feedback as necessary. In the educational products, many of these functions are left for the teacher to build into the lesson plan as classroom activities, project work, etc.
But the live teacher and the classroom environment are not always available when students choose to study a topic. I am firmly of the opinion that future systems should be capable of adaptively varying between the instructional and informational modalities. A student wishing a system driven instructional presentation should find a method of easily following such a logically pre-structured sequence. On the other hand, a user with a clear aim in mind and with the requisite mastery level of database search strategies should be able to browse the same system in a very flexible and non-linear manner identifying easily and rapidly those elements within the total information database that are of most relevance to current needs. At any point, however, the learner should be able to ask to be evaluated, ask for learning guidance, and ask for constructive feedback.
The difference between how interactive multimedia are being applied in training situations and how they are being applied in education may be symptomatic of a deeper underlying problem. This problem relates to the aspect of instruction that has apparently been overlooked in the papers of the Apple conference and indeed in the progress we are making in this field to date. The training applications of interactive multimedia tend to aim at typical training objectives, such as the execution of particular tasks and procedures (relatively algorithmic content). In these areas, the relatively shallow levels of interactivity of programmed instruction and current computer assisted instruction design methodologies are usually appropriate. As a result, the chief difference between interactive video in these applications and earlier text based programmed instruction or computer based instruction is in the introduction of visual and audio messages as appropriate.
In contrast, the bulk of interactive multimedia products aimed at the education market appear to have focused on pulling together into one convenient package a diversity of materials on a particular topic area in such a way as to make them more easily available to the classroom teacher and the student in the library. The emphasis has been on the organisation and packaging of the information. Often, quite intentionally, the uses to which this information would be put and the specific learning outcomes which would be pursued have been left undefined, this task being considered more the role of each classroom teacher or each learner independently. As a result, the level of interactivity that is built into such material is also quite shallow, restricted to selecting items from a menu that describes the content stored in particular nodes of a network. This content may be organised in several alternative ways. For example, a history resource base could be indexed according to events or people's names or specific historical trends. However, these alternative classification schemes represent only alternative ways of analysing the content and presenting a structured contents list to the user. In terms of design methodology, this goes no further than typical practice in the organisation of conventional library resources.
Very little progress seems to have been made in marrying these two modalities of utilisation. I would also add that by and large, whichever of these two modalities have been attempted or even when both have been implemented in one system, the rationale by which the information has been organised and by which presentation, practice, evaluation and feedback have been designed have not always followed the available theoretical and practical principles that are already well documented in the instructional design literature.
What would be particularly interesting to pursue in IMM projects would be the last of the mini scenarios I described earlier, that is the Socratic dialogue analogy. This requires that students may use the system to find answers to questions they have in their minds but may also be challenged by the system to demonstrate understanding and, in the light of the results of their response to that challenge, be directed to further information sources. The system and the user mutually cooperate towards the achievement of agreed learning goals. Few of the commercially available interactive multimedia packages have attempted to achieve this. There are a number of experimental models, however, and one or two of these will be discussed later in the paper.
In relation to learning, Hooper states that "the interactive nature of the multimedia presentations .... as well as the emphasis on browsing modes of utilisation of such materials will encourage active involvement by users and therefore should enhance learning." She also suggests that interactions with high quality sounds and images should enhance learning "in ways we have yet to anticipate." Finally she stresses the potential of networked information databases such as offered by Hypermedia possibilities. She maintains that "though we are still naive in matters concerning how users will interact with capabilities for the interlinking of materials as well as with materials that are presented to them in interlinked contexts, it seems that this approach to the presentation of materials should make explicit to learners the importance of interrelationships of ideas and should therefore enhance learning." It would seem, therefore, that the state of the art, as represented by the Apple conference of five years ago, was more of the opinion that we have yet to develop approaches that would be adequate to the tasks in hand.
However, in the literature of instructional design, one can encounter several models that would seem to be highly applicable in the interactive multimedia area. I have already mentioned that in some training applications of interactive video, the earlier effective and efficient models derived from programmed instruction principles continue to be applied and continue to be applied with excellent results. In job related training situations, the more conventional drill and practice and branching tutorial models are adequate to achieve the learning results that are required. The use of computer based simulations has been successful in the promotion of understanding of complex conceptual information or in developing skills related to heuristic decision making.
In addition to the successes in this field, there are, of course, the failures. Some research carried out by myself a few years ago demonstrated the relative lack of effectiveness of some very spectacular interactive video situational simulations in the management development area (Romiszowski 1989). This lack of effectiveness was due to nothing else than overlooking one well proven aspect of the design of case study and simulation training, namely the need for a reflective discussion of the experience in order to synthesise the important conceptual learning that was the major objective of the whole exercise. In short, the design models for effective performance based training applications of computer based instruction and therefore of interactive multimedia training sequences are largely available in the literature but are largely ignored by practitioners.
As one proceeds into the area of learner directed search of information resources, the focus switches from models for the design of specific objectives based instruction to models for the design of knowledge bases. At a recent NATO sponsored workshop on "Hypertext and Hypermedia for Learning" in which I participated, I was somewhat surprised that the invited participants had very few principles for the design of such knowledge bases. A majority of practitioners, it seems, regard the use of Hypertext and Hypermedia environments as most applicable to what might be called "collective creativity." This implies that a group of learners or other participants interact with some existing materials and create annotations and links so that, through their collaborative activity, a Hypermedia environment is created which is meaningful to them as a group. Only a minority of the participants in the NATO workshop were at all interested in the pre-design of a Hypermedia environment that would serve as a reference resource within a particular domain of knowledge (Romiszowski 1990).
The lack of interest in design issues for domain specific knowledge bases is paradoxical given that this is the major area of production by commercial organisations involved in the supply of the interactive multimedia market. It would seem that most prominent researchers in the field are not really focused on the majority demand in the marketplace. Perhaps it is not therefore surprising that, as mentioned earlier, many of the products that have been released commercially are not exemplars of well organised structure which users might easily browse to find what would be of relevance to a particular individual need.
However, it is not the case that there is no research and development on this particular issue. Indeed a very powerful methodology that could be implemented most successfully for the design of Hypermedia environments for free browsing or guided study has been available for the last 30 years. I am referring here to the structured writing techniques, initially pioneered by the work of Robert Horn on "Information Mapping" that was developed in the late 1960s in the search for more effective technical writing and documentation methodologies (Horn, et al 1969, Horn 1976, 1989). This particular methodology has been used in practice and constantly refined and extended over the last 30 years, but unfortunately as most of the work has focused on specific business enterprises to produce materials and knowledge bases related to in company topics, the general knowledge regarding this technique has not been diffused as widely as it could have been. Consequently, many authors of educational Hypermedia systems proceed quite oblivious of the quite vast know how that has been built up over the years on how to structure information resources in such a way that they are most effective and most efficient for both initial learning and later reference purposes.
As we progress to the yet deeper levels of interactivity (as illustrated by our Socratic dialogue example of human conversation), it would appear that the area of machine mediated instructional design is less powerful and less developed than in the cases we had just discussed. The totally free dialogue exemplified by human interaction in a deep conversation would depend on significant progress in the area of artificial intelligence applied to computer based instruction.
Progress in the application of AI to education has on the whole been somewhat disappointing. When Sleeman and Brown published their classic book on intelligent tutoring systems (Sleeman and Brown 1982), the general tone of the editorial and of many of the papers collected in this book were that the millennium has arrived. The reader was left with the impression that within a decade or so progress will be such that many examples of intelligent tutoring systems will be in regular use in education and training. A few years later, the next seminal work on intelligent computer assisted instruction (Self 1988) was much more guarded in its expectations. Today, some five years later, there is a growing number of commentators in the literature who are expressing doubts as to whether intelligent tutoring systems will ever become a significant aspect in the day to day practice of education.
It is true that there are some examples of workable practical applications of intelligent tutoring in some military training contexts. This is so partly because it would appear that in military training often cost is no object, but partly also because the training content selected for these applications has been within the current capabilities of the technology. As we look from these simple applications to the complex issues of discussion of theory and philosophy between already well informed adult students and the leading experts in the field, we not only depart from the bounds of feasibility in terms of achieving workable systems for an acceptable development cost, but we are also exceeding the current boundaries of capability of artificial intelligence.
We may indeed be attempting to exceed the theoretical limits of computability. In other words, there may be whole areas of human thinking that are dependent on a complex of aspects that are not only dependent on the analysis, organisation and manipulation of knowledge but are highly dependent also on personality and emotional traits that may well be beyond the capabilities of replication within computer software. Indeed, one prominent researcher in the field of machine instruction, Gordon Pask, refuses to allow his work to be classified as within the field of artificial intelligence on the grounds that he has never seen a machine exhibit true intelligence, and believes that he never will, given that he has never seen a machine exhibit emotions.
It is worth mentioning here that Pask's contributions to the theory and practice of adaptive computer based instruction, conversation theory (Pask 1975, 1976; Pask and Boyd 1987) and the CASTE system for "conversational CAI" (Pask and Scott 1973; Pask, Scott and Kallikourdis 1973) serve as yet another example of promising instructional design models that are highly relevant to the practice of IMM, but are largely ignored by IMM practitioners.
The following personal story may help to define my position on this issue. At a recent presentation at Ithaca College, which I attended, the world famous Brazilian educator, Paulo Freire, the author of Pedagogy of the Oppressed, was asked by a young American mathematics teacher a question related to this issue. The teacher in question classified herself as a "passionate teacher," that is one who believed in the beauty and the inherent elegance of her subject and who wished to communicate her passion for mathematics to her students. However, these attempts were typically rejected by her students. She asked whether she should refrain from expressing her emotions in relation to her subject matter or continue to attempt to change the attitudes of her students both to mathematics and to her teaching style. Paulo Freire couched his response in the context of an analysis of cultural differences between North America and certain other societies on this aspect of emotion in teaching. He criticised the phenomenon of student rejection of emotional teaching that the teacher had described and suggested that we should fight hard to overcome such resistance to the emotional approach, not only at an individual but at a societal level. Unfortunately, he did not give the teacher any hints on the strategies and tactics to use in the fight.
According to Freire, teachers without passion and devotion to their subject are not teachers at all. Students who do not develop a passionate interest in the subject will not remain true students. Here we have the opinions of a great humanist, educational politician and, by his own self description, a utopian. Interestingly they do not contrast at all markedly with the pronouncements of Gordon Pask mentioned earlier, which come from the field of research on machine mediated instruction. It would seem that from quite different viewpoints, educators in general agree that an emotional element is essential in instructional systems.
If, as my story suggests, students (at least in some cultures) tend to reject emotional presentations of the content of education from the teachers they meet everyday in the classroom, then maybe this emphasises a particular and important role of media in education. After all, in the area of entertainment, media are used specifically to arouse emotions. They communicate through the affective channel much more than through cognitive or other channels. Great films are successful because they made a lot of people laugh or cry. Sometimes they are successful because they made people think deeply about a problem, but on analysis this almost inevitably boils down to having made people become emotionally engaged in the problem and therefore devote time to thinking about it. An interesting question is whether interactive multimedia are in any way more effective in this respect than well designed linear media presentations such as films, television and radio programs.
In some cultures, education has always and continues to be implemented through the emotive use of theatre, mime and dance. Recent trends in modem theatre have, in some cases, involved audience participation and interaction. Current experiments with interactive two way television for entertainment are following the same path. The proponents of these innovations are seeking to achieve more impact, greater involvement and ultimately a higher level of entertainment. However, the well written play which involves audience participation of a passive nature is still a very strong contender and indeed may not be ever replaced. Similarly, the linear pre-planned presentation of a message that has both cognitive and emotional content may remain just as important a part of the use of media in education as the new interactive multimedia.
If we look at the field of instructional design and development from the viewpoint of emotion, we probably would focus on the aspect of motivation for learning as one of the first practical indicators that an emotional set is present or that some emotional change has happened. It is interesting here to contrast two theoretical viewpoints that have both been put into practice in instructional design models.
Firstly, let us revert to the 1960s and recall what Thomas Gilbert wrote in the preface to his influential treatise that he called "Mathetics" (Gilbert 1962). This treatise explained in detail a methodology for the systematic design of instruction based rigorously on learning principles derived from the behavioural science laboratories. Whatever we may think (emotionally) concerning behaviourism and its relevance to the educational process, there is no doubt that for certain types of procedural learning, the behavioural science based methodologies launched as programmed instruction in the 1960s were effective and sufficient to achieve the limited learning goals expected. Among the methodologies for the design of instruction developed at that time, the Mathetics methodology was by far the most detailed, most rigorously science based and has stood the test of time as effective and efficient when applied in appropriate contexts. I think it is useful to re-read a whole paragraph or two from the introduction to Mathetics.
Certain assumptions about motivation need examination, for they underlie the thesis developed here. The principles and procedures I describe will assume a motivated student, one that possesses a genuine educational objective. If the animal is not motivated, or to put it differently, if the teaching agent is not a clear instrument by which he can achieve a reinforcer, the animal will not perform and cannot learn. If a student is to learn, the consequences of the mastery of a knowledge or skill must be reinforcing. We must assume that the student possesses an educational objective, because by itself frequent and apparent success in the course of learning is not intrinsically reinforcing.Gilbert's viewpoint so clearly stated here seems to place the responsibility for achieving initial interest and motivation in the learning process on elements other than the instructional materials that we may develop. Perhaps one important source of such initial motivation is the early emotional engagement of the student in being motivated to learn the subject matter for its own sake when extrinsic motivators such as keeping down a job are not present. Maybe this aspect is where the passion of the teacher or the emotional content of media that might be used by the teacher may play a key role. On the other hand, Gilbert acknowledges that an essential role of the instructional process is to maintain and not extinguish the initial motivation that the student brings to the learning task. Within the behaviourist model, this maintenance of motivation is achieved through demonstrated success towards a goal that has already been emotionally accepted by the learner. Maybe we should give greater attention to how motivation can be first sparked and then maintained.
Inherent in any well designed set of teaching materials is an attempt to maintain the motivation of the student No matter how well designed materials are, if mastery in a subject is not an objective of the student, or those who control him, the materials will fail because the student will not complete them. People are circumspect in what they choose to learn; the so- called programming principle that progress through the course of learning is inherently reinforcing is not only poor learning theory, but is, I think, an abandonment of common sense. We confuse the tendency to pay homage to education with a desire to learn. Only seven percent of those who buy a well known encyclopedia ever open its pages; this multi-volume symbol of culture is placed on the altar of the home next to the even more infrequently used Holy Bible.
One more recent model for systematically planning the development and maintenance of motivation in instructional systems is John Keller's ARCS model. This acronym stands for Attention, Relevance, Confidence and Success (Keller 1983). The first stage of the teacher's task, according to Keller, is to gain the student's attention and interest. This may well be equated to gaining emotional engagement in pursuing learning in a particular domain. The role of impactual media and passionate presentations is important here. Once a student is engaged and paying attention, e instructional process can commence. Inevitably at this stage, because of prerequisite sequencing, certain topics that are not clearly related to the ultimate goals and the reasons for interest may have to be studied. Therefore, establishing the relevance of the content of study to the reasons originally established for studying the topic becomes of importance. Then, as the learner progresses into the body of the instructional process, difficulties may reduce the initial level of motivation. Therefore one should plan the process in such a way that the student's confidence in his/her abilities to succeed is enhanced. Finally, this confidence should transform into real success in progressing towards and ultimately achieving the goals for which the student was initially motivated.
In the context of interactive multimedia, one can possibly separate the aspect of emotional involvement from the aspect of effective learning by considering the implications of the four components in the ARCS model. The trend towards what I've described as surface level interactivity, which places control over vast bodies of subject matter in the hands of students and allows them to browse in a non-linear manner, is often justified by stating that students are therefore free to follow up their own unique interests. In a way, unlike our passionate mathematics teacher who is seeking to engage the students emotionally in the study of a specific area of content, we are giving the student empowerment to select areas of interest and follow them up by accessing relevant information. However, unless the student has some goals, what will he/she choose to follow up. Surely, there is still need for the creation of motivation through engagement of the student's interest and passion for learning in the topic.
If we abandon the linear lock-step curriculum where a classroom teacher is charged with infusing motivation and interest for the standard content in all the students who come through her classroom, we must then replace that lock-step "machine for emotional involvement" by a series of smaller machines, each linked to a particular subset of content area. The encyclopaedic interactive multimedia environment should consider including "passionate front end introductions" to subtopics in order to help the student who has no clear motivation to establish one.
Once the student enters a particular domain of study, then the nature of interaction that would promote effective learning descends to the deeper level of interactions that I described earlier. Emphasis is now not only on allowing students to easily access information on a variety of topics and maybe create mental maps of how these topics are interrelated but should also be on building the student's confidence and ultimately establishing the student's success through mastery of the concepts and principles that the particular topic is composed of. This aspect of the use of interactive media is less related to the question of emotional presentations and is more dependent on the analysis of the structure of knowledge and the embedding within the material of methods by which the student can self evaluate understanding and control progress so that initially confidence is built and ultimately success is achieved.
In my opinion, this problem is and will continue to be quite serious. One recurring problem in practical projects in the area of interactive multimedia design is the establishment of a common set of understandings among team members who bring quite different perspectives and expertise to the common effort. In addition, the conceptual language of the separate disciplines is undergoing a metamorphosis as the science is developed. Gavriel Salomon has observed, for example in relation to psychology and the cognitive sciences, that we have witnessed over generations a tendency to conceptualise human thinking and learning in terms of the prevalent technologies of the time. In his article "Artificial intelligence in reverse" (Salomon 1988), he argues that maybe there has been more influence on psychologists in how they conceptualise the human learning process because of their interaction with computer scientists than there has been on computer scientists to adopt established psychology concepts in their search for models of human intelligent activity to replicate by machines. Obviously, a lot of mutual influence between these two areas of scientific endeavour is occurring. The question of who is leading whom may be somewhat of a "chicken and egg" argument. However, the mingling and fusion of concepts from the information and the human sciences is occurring, will continue to occur and is often the cause of a lot of confusion and misunderstanding.
In the context of this paper, the two quite distinct and often opposed meanings of interactivity that I described at the beginning are an example of the same word being used with, in the one case a primarily information technology based background and in the other a primarily learning technology based one. As long as we continue to use the term interactivity indiscriminately without clarifying these two facets of its potential meaning, we will be doomed to misunderstandings with our collaborators or clients. The search for a unified philosophy and scientific terminology in our field is possibly underestimated in terms of its importance.
The important contribution of Robert Gagne in the area of instructional design is possibly not so much any of the specifics of his taxonomic models or his procedural guidelines for the design of instruction, as the bringing together of concepts and principles from a number of previously conflicting camps to create a form of planned eclecticism. If we look at the work of Robert Gagne over several decades (eg. Gagne 1965, 1974; Gagne and Briggs 1974), we may observe the continual evolution of concepts, extending the net to capture ideas from an ever wider range of philosophical and theoretical positions and presenting them together as an integrated conceptual structure describing the field of instruction.
In the 1970s and early 1980s, I suggested ways in which the view of our field could be yet further integrated and extended (Romiszowski 1981, 1982). Other workers, for example David Merrill, were at the same time intent on clarifying and more precisely defining some of the guidelines that Gagne had gathered eclectically from a number of theoretical sources. Merrill's Component Display Theory (Merrill 1983) and Reigeluth's Elaboration Theory (Reigeluth and Stein 1983) are some of the resulting prescriptive instructional design models that were born in that period. However, my viewpoint was that where Gagne had contributed and where Merrill had progressed even further was in the analytic atomisation of the components of subject matter and of strategies that might be used to effectively achieve its communication and learning. What I was concerned with at the time was that, although the resultant "atomic table" of the components of learning and instruction is an important contribution, the bringing together of the "atoms" into evermore complex "molecules" and those molecules into "living structures" of knowledge that can be used productively in real life situations was the area that we should be spending more effort on understanding.
It seems that now maybe the time for this idea has come. Gagne and Merrill (1990) in their most recent writings are directly focusing on the building of models for the bringing together of the elements of knowledge and learning into powerful and useful structures to think with. Many theorists have been arguing for such a change in emphasis. However, few design models have appeared in the literature that give clear and precise guidelines on how to go about achieving these goals. Cognitive psychology has offered a number of good ideas for the enhancing of learning (West, Farmer and Wolff 1991). One example is the use of advance organisers based on the pioneering work of Ausubel (1960). However, there is still no accepted design theory for the production of effective advance organisers. Other emphases from the cognitive psychology field have been on the development of study and learning tools. such as concept mapping (Novak and Gowin 1984). This current emphasis on empowering the learner to learn, irrespective of the content of learning, is to be applauded. However, success in this area is never likely to be absolute and therefore progress in learning tools technology does not reduce the need for progress in the technologies of effective presentation and organisation of knowledge.
Firstly, I would like to mention the "questions driven environment." The user of an information system usually decides to log on because of some question that requires an answer. This question may be very general and not very well formulated. A human expert such as, for example, a skilled librarian would receive such a question and analyse it somewhat further, bouncing it back to the questioner in order to clarify it and perhaps break it down into a series of questions of greater precision. This process goes on until the questions are sufficiently precise for the librarian to direct the questioner to an appropriate book or bookshelf. This function of the librarian is critical to the usefulness of a library system and is not the same as that performed by a card catalogue or other indexing system. In order to use an indexing system, the user must have a much clearer idea of what exists or is likely to exist in the library. The key words by which the material can be accessed must already be known or accessible to the user. But often in reality, that is not the case and it is the librarian who has knowledge of these organisational features of the library and can match the specific content in the shelves to the needs of a particular inquirer through a process of clarifying the inquirer's original question.
Whereas I have been somewhat sceptical in earlier comments concerning the probability of intelligent tutoring systems having much impact on the day to day teaching in our schools and universities, I believe that the task of developing "intelligent librarian systems" apart from being more tractable is possibly a much more important task to which to devote our attention at this point in time. An example of an intelligent "guide to the literature" was well illustrated by the animated talking computer illustrated in Apple Computer's videotape The Knowledge Navigator. It may be that the first generations of such intelligent research assistants will be somewhat more formal in their conversational interactions with the user than in that hypothetical example. However, I look forward to seeing such tools widely available within not too many years. Of course, the real value of these intelligent librarians will be in the rapidly developing "electronic libraries" that can be accessed through computer networks from home or office.
The second metaphor that I think worth pursuing is the "virtual learning community." I am referring here to the increasing use of networking as a way for individuals, separated by both time and distance, to form collaborative conversational groups on a variety of topics. I look forward to the day when computer mediated communication (CMC) on networks is integrated with the use of other information resources and multimedia (IMM), so as to become part and parcel of a total information and conversation environment. As the feasibility of transmitting multimedia over distance in real time becomes a reality, the possibilities for interaction between people utilising the same information resources opens many new possibilities for creative educational applications.
My own vision is of the personal computer as an integrated communication tool that allows one to work with materials and discuss those materials with a community of others interested in the same issues. Within these communities will be experienced teachers who can serve a tutorial function and indeed can be considered as the intelligent tutor element within the system. The versatility of this approach is superior to that which is promised by the AI based intelligent tutoring system developments of the last few years. The costs may well be highly competitive with any other alternative for educational systems of the future. Current research including some that I have engaged in suggests that such distance education systems may be equal and possibly in some respects superior to "place based" small group discussion sessions. For example, we have obtained results in the using of business cases over such networks that are considered to be superior in terms of both quantity and quality of discussion as compared to using the same case material in the classical manner.
At the Open University in the United Kingdom, the CoSy computer conferencing system has been used now for several years as one of the communication media regularly available to students on certain courses. The interface for the student in this conferencing system is set up on what is referred to as a "campus metaphor." There are separate areas for messages to be stored, exchanged and commented that replicate the functions of large lecture rooms, smaller classrooms, seminar discussion rooms, coffee bars and of course libraries (Mason 1988). I see the desktop metaphor which is the currently popular one becoming part of this broader metaphor of the total educational environment available through the terminal on your desk.
How has the market moved in the last five years? The only real success story of optical disc technologies has been the music CD. Videodiscs, although growing in availability and acceptance, do not form a large part of the media that are used in training, education or even entertainment. The much promised, cheaper and more practical alternatives of CD-ROM based interactive video, whatever their format, are yet to make any impact on the real marketplace. All in all, there has been some progress since Hooper's analysis in 1987 but it is fair to say that still we are in the position of trying to create a significant market for interactive multimedia.
One reason suggested by Hooper for this slow development was the so-called "chicken and egg" syndrome. Until there is a sufficient hardware base of installed systems in the marketplace, there is no incentive for producers to launch large libraries of interactive multimedia courseware. But until there is a reasonable installed library of courseware, there is no incentive for the end users to invest in the relatively expensive workstations.
This "chicken and egg" argument was exactly the one quoted by pioneering manufacturers of video cassette systems and courseware at a conference that I helped to organise in Britain in 1970 (Leedharn and Romiszowski 1973). In Britain, I was able to witness a phenomenon that could best be explained by this argument. In the early 1970s, the National Council for Educational Technology set up a clearinghouse for the sharing of information among universities on videotapes and cassettes that could be made available on loan or sale basis to be used in other universities. The material was restricted to that produced by British universities. The number of titles in the annual catalogues published by the National Council remained in the range of a few hundred climbing very slowly from year to year for over a decade. Suddenly an exponential expansion in available materials occurred more than doubling the number of titles every year for a period of a few years.
Despite the above mentioned 20 to 30 years of research and development on interactive multimedia systems, the period of relatively easy availability of such systems for commercial production of courseware is, less than a decade at the moment. Maybe we have some time to wait before a similar exponential growth in production and dissemination of IMM courseware will take place. It is probable that this would be accelerated by significant decreases in investment costs and standardisation of delivery platforms. The promised mass marketing of videodiscs in the small CD format as a consumer item will no doubt be the route by which such standardisation and cost reduction will occur. Then and only then might we see the "chicken and egg" problem resolving itself.
Hooper suggests another reason why production and dissemination may be lagging behind research and development in this field. She argues that people in general may not be at all accustomed to be "generators rather than consumers" of media programs. It is evident that the "couch potato" syndrome is present in most industrial societies that have been reared on mass distributed radio and television programs. However, the accelerated rate at which the Nintendo revolution has been accepted by the younger generations may be a sign that "couch potato" traits may quite rapidly be unlearned if appropriate opportunities are presented. As interactive video games become distributed in a CD format incorporating video sequences that can be manipulated by the players, it is possible that a new form of literacy related to the appropriate exploitation of interactive multimedia may be instilled in our younger generation.
Just as most successful computer games and more recently arcade video games have been designed by young people hardly out of their adolescence, so future innovative formats of interactive multimedia materials, for both entertainment and more serious purposes, may have to wait for a generation of designers reared on the first generation of mass produced entertainment products. If we have to wait, therefore, for cheap versions of "entertainment" interactive video in order to influence our young people and train them in the language, grammar and syntax of IMM environments, then maybe we will have to wait more than the decade that it took for video cassettes to "take off" in the educational market.
However, there are some signs that such an IMM explosion is likely to occur. A combination of growing numbers of enthusiasts, growing interest in the commercial potential, increasing availability of development platforms and rapidly diminishing costs can almost guarantee that there will be an accelerated level of provision of accessible interactive multimedia programming for educational and training purposes. As usual the training field leads education in terms of experimentation with new technologies. One training materials supply company alone, Applied Learning International, which specialises in the computer skills development and management development area is currently offering over 200 titles of courses on interactive video or CD-ROM based delivery platforms. This represents 16% of their total course offerings. Almost all of this shift from conventional computer based instruction and linear videotape training materials has occurred within the last five years, that is since the 1986 Apple Conference was held.
There is also an increase in the number of video discs and accompanying software for computer control being offered to the education market. Among such programs is the videodisc of the Louvre which presents just about all the significant works of art in that museum (Hatala et al 1991). This is accompanied by software that enables the material to be accessed in a number of ways making the collection a versatile database on painting, sculpture, etc. This product is typical of the non-instructional interactive multimedia that are really information resources which somebody, usually the classroom teacher, must decide how best to use.
Another relatively successful product recently released in the USA is the interactive video disc on Martin Luther King (ABC News Interactive 1989). This is interesting in that it combines existing newsreel footage and photos with quite extensive textual information presenting the details of most of Martin Luther King's speeches and also a lot of other documentation referring to key events in his life and work. These documents are interrelated in the form of a Hypertext and may be browsed in a non-linear manner. Media excerpts may be accessed at points where they are relevant. These points are indicated by means of icons within the text so that the reader is always aware of what other information resources are available on the particular issue that is being studied at the time. This product comes closer to the instructional paradigm although once more, there are no self tests or other ways that a user may check whether specific learning objectives are being achieved or not.
In the area of specific instructional materials, there are less programs produced so far. Some, such as Optical Data Corporation's Life Science series (Walker, McCormick, et al 1989), are quite elaborate multimedia packages which involve texts and reference materials in addition to video discs and associated software. These products allow many alternative objectives to be pursued. Self tests are incorporated in the material allowing a student to evaluate progress and to find quick routes for review of important basic concepts, etc. Interestingly, the tests and review guides are included in the supplementary texts and not in the computer based components. The resulting material is not cheap as it includes many video discs and quite expensively printed materials to supplement them. For example, one video disc is devoted entirely to a two sided biology unit on frogs.
Experience in schools in the Central New York area where I currently work is showing that these more "instructional" materials are not being accepted by teachers as much as the more "reference" materials such as Martin Luther King or the Louvre discs. Whether this is due to their more elevated cost and complexity of use or whether it is a reaction of teachers to materials which take over the lesson planning is not absolutely clear. It is probable that both these factors play some part.
In my opinion, the real market breakthrough for interactive multimedia will come when the decisions if, when and how to use them will rest with the learners. This will be when they are available in reference libraries or on loan, or when ultimately both the video and the computer program elements will be accessible at a distance through computer networks. Then we may see the real potential of interactive multimedia as self study materials realised in the educational and training context. Such use will involve both information dissemination systems and instructional systems, but where the student has a clear goal that he or she wishes to reach, the instructional function would be essential. It is therefore a little worrying that instructional IMM materials are not being widely produced and, even when produced, are not necessarily well designed.
The factors of acceptability and effectiveness are paramount if interactive multimedia are to make a real difference to education or training. We have seen the example of television which has been accepted wholeheartedly by society as an entertainment medium and yet as a mass education medium was an unmitigated flop. This was not due to a lack of potential of broadcast television for education and training, but somehow, neither the factors necessary for ready acceptance in the marketplace nor the willingness to put into practice proven principles of effective instruction were present at the right time in the right place. As a result many millions of dollars of investment resulted in almost no impact in classrooms. How can we guard that a similar fate is not in store for our interactive multimedia revolution?
A more serious problem may be the necessary adjustment in teaching methods and organisational structure of schools in order to take advantage of interactive multimedia on any reasonable scale. Although there is always the enthusiastic group of innovators and early adopters who will try any new idea, the bulk of teachers are on the whole late adopters who can be counted on to resist innovation and change for its own sake. These teachers will have to see some very clear advantages either to the school and the students or to their own professional positions as a result of getting involved in interactive multimedia instruction.
Unfortunately, within the conventional school structure, many teachers are likely to see extra work and difficulty involved in implementing highly interactive mediated systems of instruction. Some may actually feel threatened by these innovations. It will be hard to convince them that the potential benefits outweigh the negative aspects. By and large, technology based innovations that have been introduced into educational systems over the last half century or so have been notorious for their ultimate rejection and failure. This is true not only of the much maligned programmed instruction movement but also of instructional and educational television, of individualised instruction schemes of many varieties, of resource based learning in its more traditional formats and even major curricular changes such as the introduction of the modern mathematics curriculum. Why should we expect the current interactive multimedia revolution to fare any better?
In my opinion, there is some chance for the IMM movement to survive due to a lucky juxtaposition of a number of otherwise unrelated factors. On the one hand, there has been over some time a growing dissatisfaction in political and societal circles in many countries with the apparent lack of effectiveness and efficiency of conventional educational systems. In the United States, for example, ever dwindling test score results and ever growing problems of a disciplinary and social nature within the schools, has led to the situation being described as having reached a crisis point. In 1990, the issue became one of national importance to the extent that President George Bush pronounced the intention of creating the conditions for a total turnaround in the trends that are leading to reduced educational effectiveness. This political backing has given impetus to an unprecedented amount of activity on the restructuring of schools. There must now be hundreds of school restructuring projects taking place across the United States. Funding for major innovative projects is appearing from the most unlikely sources. For example, a conglomerate of business organisations has formed a consortium called the New American Schools Development Corporation to finance the practical development of alternative designs for American schools. The slogan under which the request for proposals was launched was for "break the mould" designs that, instead of trying to patch the defects in the existing systems, propose entirely new approaches and systems for the delivery of education.
About 20 projects will be funded for two years at levels of between one and three million dollars each for the detailed development of practical plans and pilot implementation of these plans in one or more school districts. It is then expected that the most successful of these 20 or so pilot projects will be funded at a higher level for two more years in order to be disseminated and implemented across a wider range of schools. All the indications suggest that many of the proposals that will be vying for this funding are going to be focusing on high tech alternatives to existing educational delivery systems. Out of these projects may come one or several success stories that will catch the imagination of the nation and possibly get widely implemented by many school districts. In such a case the seeds for the use of interactive multimedia on a relatively wide scale may be sown.
Another factor which may support the growth of interactive multimedia utilisation is the ever increasing emphasis on non-formal educational systems, particularly distance education and open learning systems. These are fast becoming major contenders in the educational stakes and no longer the poor substitute for conventional classroom based instruction. Especially in the areas of higher education and continuing adult education, it is quite possible that by the turn of the century, distance and open learning systems may be accounting for more students per annum worldwide than conventional forms of universities and night schools. The spur for this growth is coming from two directions. One is from the less developed nations of the world, where currently available conventional educational provision is inadequate to keep pace with the demands for education of an ever growing population that is rapidly being absorbed into an industrial and commercial workplace. The other driving force comes from the industrialised nations where the need for continuing adult education and retraining is not being adequately met by existing conventional institutions.
I am of the opinion that as a country industrialises and marches into the information age, the conventional institution is no longer a model that is capable of reacting sensitively to the fast changing and varied needs for job related education and training. It is in this area that self instruction is growing at the fastest rate at the moment. It is also in this area that creative uses for networking are being discovered. Small groups of geographically scattered people that share a common need for training or development are banding together as "virtual communities," supporting each other in their quest for competence. In the not too distant future, it is probable that the bulk of job related procedural training will be delivered by computer based instruction, where the use of computer based systems is part of the job by what is becoming known as "computer supported performance systems." These encapsulate, in one system, all the materials necessary for initial training, on job reference, performance evaluation and quality control. Many routine jobs, even those involving low level intellectual tasks, will be taken over by expert systems. The more conceptual and heuristic decision making parts of industrial work will become the major aspect of many people's work. Traditionally, these are best taught by small group (seminar, case study and role play) learning activities. It is in this area that virtual groups of individuals separated by time and distance may replace place based groups to some extent. The benefits are both economics and convenience. The penalty or cost is the loss of face to face contact with one's colleagues. Some currently ongoing research seems to suggest, however, that the benefits may far outweigh these penalties. If that is the case, then such virtual groups will utilise a variety of materials as a basis for their discussions. Many of these materials could well be interactive multimedia materials. How this may appear in a future scenario will be discussed further in the next section.
An ever growing number of papers are being presented at conferences in the educational technology arena on the importance of team building and the specific peculiarities of working with such a variety of specialists as are required on a multimedia project. However, we are only just beginning to address the problem. A recent book specifically relating to the management of interactive video and multimedia projects (Bergman and Moore 1990) contains very little on these aspects. In this book, one chapter is devoted to building the project team. In an early section entitled "So What's New", the first paragraph addresses the issue that whereas in most projects team members come from a single discipline or background and share a common language, this is by no means true in interactive video disc and multimedia projects. Having said that, however, the rest of the chapter continues with a standard account of team building and organisation, never once more addressing how to handle the peculiarities of multiple languages and backgrounds.
Luckily, when bright and enthusiastic individuals are put together in a team and given appropriate resources and support, the chances are that they work out a way of cooperating to the common good. Some of the innovative multimedia projects that have come out of places such as MIT's Media Lab are more the results of such interpersonal synergy than of any pre-conceived master plan by management. One of the strengths of Xerox PARC and the reason often quoted for its prolific contributions to progress in the area of computer science and related fields is that top management has always been careful to select the best researchers available, but has then left them very much to their own ends in terms of how they cooperate to pursue project goals. However, such lucky conglomerations of bright young people are fine at early research and development stages. One cannot rely on the luck of achieving such a synergistic combination once the design and development of interactive multimedia becomes an institutionalised process in hundreds if not thousands of organisational settings. As the process becomes thus institutionalised it will be ever more important to develop methodologies of team building and project management which are specific to the peculiarities of this rather unusually multi-disciplinary industry.
I would like to conclude with some comments on another aspect of technological synergy which was not addressed by Hooper in 1987. This is the synergy of information technology, educational technology, and telecommunications technology. The third partner, telecommunications, has joined this synergistic group more recently. Computer technology and educational technology have been bedfellows since the beginning of work on computer based instruction in the early 1960s. The more recent developments in telecommunications technology, associated with digitisation of all data broad band transmission possibilities offered by fibre optics, ISDN and worldwide satellite communication networks have opened up a vast array of opportunities for the processing, storage and transmission of information.
One interesting result is that the last decade of distance education projects is substantially different in its form of delivery as compared to earlier systems such as the British Open University of the 1970s. Whereas the British Open University was based on a correspondence course model, very much conceived as a system broadcasting one message to many students, the modem crop of distance education systems are small, interactive discussion networks utilising telecommunications with print as a support medium rather than the principal medium. Initially, telecommunications was used for real time teleconferencing. The history of audio teleconferencing goes back for many decades and has both its success stories and its failures. Video teleconferencing has for some decades been foreseen as the next generation of effective teleconferencing. What has held it back is its elevated cost compared to audio conferencing or indeed (in most cases) to conventional "place based" group instruction alternatives.
What has made the big difference in this field is the networking of computers. This has created the opportunity for a completely new modality of communication. Unfortunately, due to prior traditions of teleconferencing, it was named computer teleconferencing, or computer conferencing for short. I say this is unfortunate because the name has labelled the methodology in many people's eyes as another form of exchanging chit chat as often occurs in teleconferences. However, the versatility of networks of computers as communication tools allows a lot more variety of communication and instruction modes than mere conferencing. This is now generally realised and has led to the acceptance of the more general term of computer mediated communication (CMC). This includes a multiplicity of modalities ranging from electronic mail systems that are essentially one on one message delivery systems, through bulletin boards that allow the posting of information for many to see, to conferencing systems that operate in ways not unlike an audio or video teleconference except that the participants can log on at any convenient time to catch up with the discussion in progress.
In addition to these basic forms of CMC, there are more specialist applications such as collaborative interaction on the reading, annotating and development of a Hypertext document, collaborative research and development work by a scattered team of professionals or academics, so-called "telework" which means working for an organisation from your home rather than turning up to an office, and of course "telelearning" which could be used to describe computer based distance educational systems that allow one to take single courses or in some cases whole programs without attending on campus.
Firstly, a growing proportion of educational experiences may take place by individual study supported by discussions in the context of a networked "virtual group". Let us compare this situation with our more conventional classroom based instruction model. In this model, a major part of learning beyond early elementary years also occurs through individual study outside of the classroom situation. Students read books or access other learning materials in libraries or at home in preparation for classroom activities. If they have difficulties in accessing or in understanding what they access, they can seek help from peers or from teachers through direct contact. When they come to class, they can raise issues concerning the materials they have studied. These may be related to substance and content or related to their reactions, positive or negative to the materials. Teachers, on the other hand, will have previously indicated the materials that are worthy of study and later follow up this study by appropriate classroom activities that may be practice exercises or evaluative discussions or a number of other formats. All of these activities are planned individually and integrated into a program of study, either by the teachers alone or in collaboration with the students. All of these activities are also evaluated both formatively and summatively either by the teachers alone or by a combination of teachers, students and outside evaluation bodies such as national examination committees.
Within the fully developed electronic "virtual community," all of the activities outlined in the last paragraph can occur without the necessity for students to gather in real time as one place based group. Even collaboration on the planning of an educational program between prospective students and faculty and the evaluation of the results of a program by taking examinations at a distance have been successfully implemented in computer based distance education systems. The one facet of place based education that cannot be replicated fully within the virtual community is the physical face to face contact among students and teachers. One inevitably loses some of the non-verbal communication aspects and also probably some of the camaraderie and mutual moral support that may develop during extra-curricular activities of a social nature that occur in a campus or school environment.
The economic viability of networked education and training will inevitably grow as the relative costs of place based education continue to rise. This is already happening to a great degree in the human resource development area in industrialised nations. The figures for the United States, for example, are that the annual human resource development budget is in the region of 30 billion dollars per year in direct costs. However, if one adds in the indirect costs of airline and other travel, hotel accommodation and extra time off work incurred through having to travel to a course, this annual cost rises to more than 100 billion dollars. There is, therefore, a theoretical potential of reducing human resource development costs by 213 if our place based training could be transformed into networked training, assuming that the delivery costs of networked training are not substantially different from those of campus based group instruction (evidence suggests that they may in fact be much lower).
Given these economic realities, it is not surprising that the concept of networked training is rapidly being accepted by the business HRD community in the USA and elsewhere. As this trend continues, it is clear that possibly the first major mass market for interactive multimedia courseware will be established. Once computer networks are installed, the extra costs of equipping them with the hardware necessary for incorporating multimedia is not very high. In the present state of the technology, where it is not yet fully feasible to upload and download interactive multimedia over networks, it is an added expense, but within the realms of economic viability to have CD-ROM or interactive videodisc players attached to the already existing high powered computer workstations that are essential to the productivity of the "knowledge worker." The materials may be circulated as a course package to students through the mail, thus obviating the student's travelling at much greater cost to a centrally organised course. One can balance the cost savings of reduced travel and hotel expenses against any initial extra costs of equipping workstations and commissioning high quality interactive multimedia learning materials. In the medium term, as travel and subsistence costs continue to climb and as the investment costs in both hardware and software for interactive multimedia continues to decline, the economic equation will evermore favour the use of interactive media in conjunction with computer mediated communication.
As such integrated systems of computer based interactive instruction and conversational virtual group discussions become commonplace in the business setting, the model will be followed by universities, further education institutions and, ultimately, school systems. In the long term, as the electronic transmission of multimedia over wide area networks becomes practically viable, both the educational power and the economic advantage of networked education and training will increase. The "Educational Utility" concept (Gooler 1986) may become a reality.
If, as I suggest, the major driving force for such trends is going to be economics, then the chances are very high that pressures will be put upon trainers and educators to utilise new technologies even where they may not be ideal from an educational viewpoint. Where are such methods reasonably equivalent to current methodologies of education and training? Where are they likely to be significantly inferior? Where may they even be somewhat superior to the best of what we can do at the moment? These are questions on which there is as yet a scant amount of research.
For some years now, I have been interested in this specific research agenda because it seems that in the not too distant future, we will be faced with the familiar situation of technology leading education and training in directions that may not be pedagogically ideal but happen to be economically or politically expedient. As we are unlikely to change such basic "facts of life," it seems that we should be engaged in a program of research and development that could render IMM/CMC systems more effective across a wider range of educational and training situations. How can we get the best value out of the interactive multimedia component of such systems? How can we get the best value out of virtual group discussions?
Many current implementations tend to treat these two components relatively separately as was typically the case in conventional educational systems. Students are sent a study assignment which they do on their own, or maybe in small groups, utilising some set of resource materials. Typically these are a set of texts, a CAI package, or some simple media, but in the foreseeable future they may be an integrated interactive multimedia resource base. Having carried out this prior study assignment, students then are invited to "get on line" to discuss their reactions to the materials or to respond to some more specific questions or tasks that the group instructor at a distance may feel appropriate to set them. In this sort of scenario, the designers of the interactive multimedia package have given little thought to the role and the structure of subsequent group or one on one discussions. Similarly, the instructors who lead follow up discussions have little idea of specific aspects of the design of the multimedia resources, unless they make great efforts to review and study the resources for themselves.
The model I am suggesting here is a synergistic model where the design and development of the interactive multimedia components of an instructional system, in fact, occurs along side the design and development of supporting discussion environments.
In one particular line of research and development I have been developing environments for the study of case materials at a distance. These involve the presentation of the "facts of the case," either as conventional linear presentations or as some form of interactive simulation. The discussion of the case then commences in the interactive multimedia environment by means of a rather open ended conversational tutorial form of computer based instruction. In this way, the surface level interactions related to checking that the basic facts of the case have been understood and that some form of position, interpretation, or decision has been taken by the student occur within the multimedia package. As one proceeds to deeper discussions, where the focus is on reflective analysis of the thought processes that went on in an individual student's mind during analysis and interpretation of the case, the need for deep level interactions with an intelligent tutor, or with other intelligent peers that may have opposing opinions, becomes greater. When the discussion reaches this level of deep processing, the network connections are automatically made to allow individual learners to share their ideas with other learners and also with tutors who form the geographically distributed "virtual group" for the course.
This approach is found to be both effective and efficient. It is effective in that it ensures that every participant considers not only surface level aspects of the case being presented but in fact takes positions which require a considerable amount of deep processing of the underlying concepts and principles which govern the approach to take to the particular case. This deep processing results in conversational interactions that allow intelligent constructive feedback to be received by each participant either from a distant tutor or from one or more peers. The overall effectiveness of this form of deep level interaction is reflected in the ability to transfer the newly acquired heuristic problem solving skills practiced in one case to other similar cases or indeed to real life problem solving situations. The efficiency of the system is reflected not only in the significant cost savings that can be achieved through the elimination of travel to group meetings but also through the efficient use of the time of the experienced tutors who are only called in to interact with students at the deep processing level and are spared the relatively mundane task of evaluating and correcting basic surface level misunderstandings of the facts of the case and their implications.
A more detailed discussion of this research agenda is presented in another paper included in the conference proceedings [Romiszowski 1992].
Ausubel, D. P. (1960). The use of advance organizer in the learning and retention of meaningful verbal material. Journal of Educational Psychology, 51, 267-272.
Barker, J. and Tucker, R. N. (eds) (1990). The Interactive Learning Revolution. London: Kogan Page.
Bergman, R. E. and Moore, T. V. (1990). Managing Interactive Video/Multimedia Projects. Englewood Cliffs, NJ: Educational Technology Publications.
Bork, A. (1987). Lessons from computer based learning. In D. Laurillard (ed), Interactive Media: Working methods and practical applications, 28-43. Chichester, England: Ellis Horwood Ltd.
Boyd, G. and Pask, G. (1987). Why do instructional designers need conversation theory? In D. Laurillard (ed), Interactive Media: Working methods and practical applications, 91-96. Chichester, England: Ellis Horwood Ltd.
Bush, V. (July 1945). As We May Think. Atlantic Monthly, 176, 10-108. http://www.theatlantic.com/unbound/flashbks/computer/bushf.htm
Clark, R. E. (1985). Evidence for Confounding in Computer Based Instruction Studies: Analyzing the Meta-Analyses. Educational Communications and Technology Journal, 33(4), 249262.
Gagne, R. M. (1965). The Conditions of Learning. New York: Holt, Rinehart and Winston.
Gagne, R. M. (1974). Essentials of Learning for Instruction. New York: Holt, Rinehart and Winston.
Gagne, R. M. and Briggs, L. J. (1974). Principles of Instructional Design. New York: Holt, Rinehart and Winston.
Gagne, R. M. and Merrill, M. D. (1990). Integrative Goals for Instructional Design. Educational Technology Research and Development, 38(1), 23-30.
Gilbert, T. F. (1962). Mathetics: The Technology of Education. Journal of Mathetics, 1(1), 7-73.
Gooler, D. D. (1986). The Educational Utility: The Power to Revitalize Education and Society. Englewood Cliffs, NJ: Educational Technology Publications.
Hatala, A. et al (1991). Louvre Videodisc 2: Sculpture and Objectives d'Art. ODA/Voyager Company.
Hooper, K. (1988). Multimedia in Education. In S. Ambron and K. Hooper (eds), Interactive Multimedia, 316-330. Redmond, WA: Microsoft Press.
Horn, R. E. (1976). How to Write Information Mapping. Lexington, MA: Information Resources, Inc.
Horn, R. E. (1989). Mapping Hypertext: The Analysis, Organization, and Display of Knowledge for the Next Generation of Online Text and Graphics. Lexington, MA: The Lexington Institute.
Horn, R. E. et al (1969). Information Mapping for Learning and Reference. Lexington, MA: Information Resources, Inc.
Jonassen, D. H. (1985). Interactive Lesson Designs: A Taxonomy. Educational Technology (June), 7-17.
Jonassen, D. H. (1989). Interactive Lesson Designs: A Taxonomy. Interactive Video. The Educational Technology Anthology Series, Volume 1, 19-29. Englewood Cliffs, NJ: Educational Technology Publications.
Keller, J. M. (1983). Motivational Design of Instruction. in C. M. Reigeluth (ed), Instructional Design Theories and Models: An Overview of their Current States. Hillsdale, NJ: Lawrence Erlbaum. '
Leedham, J. and Romiszowski, A. J. (1973). Video cassettes in Education and Training. London: Kogan Page.
Mason, R. (1988). The Use of Computer Mediated Communication for Distance Education at the Open University, 1988. Paper presented at the Open University International Conference on Computer Mediated Communication in Distance Education. Milton Keynes, England, October 1988.
Merrill, M. D. (1983). Component Display Theory. In C. M. Reigeluth (ed), Instructional Design Theories and Models: An Overview of their Current States. Hillsdale, NJ: Lawrence Erlbaum.
Nelson, T. H. (1967). Getting It Out of Our System. In G. Schechter (ed), Information Retrieval: A Critical Review. Washington, DC: Thompson Books.
Nelson, T. H. (198 1). Literary Machines. Swarthmore, PA: Author.
Novak, J. D. and Gowin, D. B. (1984). Learning how to learn. Cambridge: Cambridge University Press.
Pask, G. (1975). Conversation, Cognition and Learning: A Cybernetic Theory and Methodology. Amsterdam: Elsevier.
Pask, G. (1976). Conversation Theory: Applications in Education and Epistemology. Amsterdam: Elsevier.
Pask, G. and Boyd, G. (1987). Conversation theory as a basis for instructional design. In D. Laurillard (ed), Interactive Media: Working methods and practical applications, 97-115. Chichester, England: Ellis Horwood Ltd.
Pask, G. and Scott, B. C. E. (1973). CASTE: a system for exhibiting learning strategies and regulating uncertainty. Int. J. Man-Machine Studies, 5, 17-52.
Pask, G., Scott, B. C. E. and Kallikourdis, D. (1973). A theory of conversations and individuals (exemplified by the learning process in CASTE). Int. J. Man-Machine Studies, 5, 443-566.
Reigeluth, C. M. and Stein, F. S. (1983). The Elaboration Theory of Instruction. In C. M. Reigeluth (ed), Instructional Design Theories and Models: An Overview of their Current States. Hillsdale, NJ: Lawrence Erlbaum.
Romiszowski, A. J. (1981). A New Look at Instructional Design. Part 1 - Learning: Restructuring One's Concepts. British Journal of Educational Technology, 12(1).
Romiszowski, A. J. (1982). A New Look at Instructional Design. Part 2 - Instruction: Integrating One's Approach. British Journal of Educational Technology, 13(1).
Romiszowski, A. J. (1989). Structural Communication Enhancements to an Interactive video Simulation Game: In Search. of Reflective Learning. In C. Bell, J. Davies and R. Winders (eds), Aspects of Educational Technology, 22, 226-230. London: Kogan Page.
Romiszowski, A. J. (1990). The Hypertext/Hypermedia Solution - But What Exactly is the Problem? In D. Jonassen and H. Mandl (eds), Designing Hypermedia for Learning. Berlin, Germany: Springer-Verlag.
Romiszowski, A. (1992). Conversational systems for adult education and training. In Promaco Conventions (Ed.), Proceedings of the International Interactive Multimedia Symposium, 495-521. Perth, Western Australia, 27-31 January. Promaco Conventions. http://www.aset.org.au/confs/iims/1992/romiszowski2.html
Salomon, G. (1988). AI in Reverse: Computer Tools that Turn Cognitive. Journal of Educational Computing Research, 4 (2), 123 -139.
Self, J. (ed.) (1988). Artificial Intelligence and Human Learning: Intelligent computer aided instruction. London: Chapman and Hall Ltd.
Sleeman, D. H. and Brown, J. S. (eds). Intelligent Tutoring Systems. New York: Academic Press.
Walker, T., McCormick, B. et al (1989). Life Science: The Living Textbook. Warren, NJ: Optical Data Corp.
West, C. K., Farmer, J. A., and Wolff, P. M. (1991). Instructional Design: Implications from cognitive science. Englewood Cliffs, NJ: Prentice-Hall, Inc.
|Please cite as: Romiszowski, A. J. (1992). Developing interactive multimedia courseware and networks. In Promaco Conventions (Ed.), Proceedings of the International Interactive Multimedia Symposium, 17-46. Perth, Western Australia, 27-31 January. Promaco Conventions. http://www.aset.org.au/confs/iims/1992/romiszowski1.html|