IIMS 92 contents
[ IIMS 92 contents ]

Multimedia: Some promises, some problems and some issues in human-systems interaction

Jenny Preece and Gordon Davies
Computing Department
Open University, UK
From ancient times written works were intimately connected with other forms of artistic endeavour. Today with the aid of graphic displays, computer aided technology and windowing technology we have returned to this desirable situation. We have corrected the technology driven separation forced upon us for centuries by conventional practices of mechanical printing and data encoding'. (Fox, 1989, p 195)
The quote by Fox (1989) paints an optimistic view. Certainly the technology is becoming available which will enable designers to integrate not only graphic displays and text but also moving images from film or animation and interesting uses of sound. The key question is 'how can this technology be harnessed to facilitate learning and human endeavour?' Human-computer interaction and human factors specialists have much of value to say about the human issues which need to be considered in the design of interactive systems. However, until recently the way that different media can be integrated and harnessed in multimedia systems has been of minor interest because the technology has not be available to support such activities. In this paper we consider some of the promises of multimedia, some obvious problems and a number of issues needing solutions.

1. Introduction

The development of multimedia systems promises to enhance learning and to enable users to interact with information in new ways. Despite these promises, if such sophisticated systems are not well designed they can also create problems for humans in terms of cognitive load and by inadequately supporting changes in individual and group working practices.

Sound, direct manipulation of interface objects, visualisation of processes and dynamic video images are all features of today's state of the art multimedia systems. Increasingly powerful computer graphics enable designers to create sophisticated three dimensional images. Using specialist devices, users have the illusion of manipulating these images in virtual space or even of walking around a virtual world. What are the promises and problems of these technologies?

In this paper we shall start by describing three different categories of multimedia systems which currently exist. We shall then briefly examine the components that contribute to the way people interact with computer systems and the nature of that interaction. Then we will examine some of the promises that are made for multimedia systems, some obvious problems and raise some issues that need to be resolved.

2. Multimedia and hypermedia

The term multimedia can be interpreted in different ways and one of these ways is based upon the type of media involved. Four types of multimedia systems can be identified, namely: 'non- computerised multimedia systems', 'multimedia', 'hypertext' and 'hypermedia'. Although these systems are different in a number of ways they share some promises, problems and issues.

'Non-computerised multimedia systems' are systems in which a number of different media are integrated into a single learning package or module. This a now a common approach for distance learning or self instruction. For example, The Open University in the UK provides degree level courses for students to study at home in their own time and without the constant presence of a tutor (Davies and Preece, 199?). Typically these courses comprise specially prepared self instructional texts, television programs (or a video), a home experiment kit (for science course), audio cassettes, and often a computer with software. These packages have been called 'multimedia' for several years as they are based on carefully integrated collections of material delivered via different media.

Recently the term 'multimedia' has been used to describe systems in which a variety of media are presented and integrated through a computer system and this is now the accepted use of the term. These packages are often prepared with specially developed authoring packages and they make use of voice over still transparencies which are digitally scanned into the system, video, animations, speech, music and a variety of sounds.

'Hypertext' as the name suggests is a system for presenting active text. Vannevar Bush is usually credited with the notion of a medium which permits linking by association rather than by indexing (Bush, 1945). The key feature, from the learners point of view, is that the text has many nodes and links allowing learners to determine their own routes through the material. Hypertext may contain text and graphics and, if the latter are present, it may be referred to as multimedia. Definitions of hypertext can be confusing. Gygi (1990) defines two broad types of hypertext; broad spectrum and a more clinical variety. The broad spectrum definition is characterised primarily by the recognition that hypertext is a format for non-sequential representation of ideas, it is dynamic and non-linear. The second type is more rigorous and specific but at its crux is the notion of computer supported links (p. 283 Laurel, 1990) Hypertext has many uses including uses as a presentation medium for information management and browsing, in public information systems and by educators for learning. When used in collaborative work applications the system needs to support the ability to represent changes in the structure of the information as group work progresses over time.

'Hypermedia', as the name suggests combines aspects of hypertext and multimedia. Typically, the branching structure of hypertext is used with multimedia. This produces a system in which learners can determine their own paths through well integrated combinations of different media. There is considerable talk about the use of hypermedia for different tasks but at present attention seems to be concentrated upon its use in education and training (eg. Preece et al., 1991) and for public information systems, such as tourist information (eg. Davies, Maurer & Preece, 1990).

3. Components of human-system interaction

In order to understand how people interact with any kind of system (even the non-computerised multimedia) the way that four key components interact must be considered. These four components are: the learner(s), the learner's tasks, the environment in which the task(s) are performed, and the design of the system or material being used. Each of these components has, in turn, many characteristics that need to be taken into account and all four components interact in complex ways.

Learners have physiological, psychological and social needs. For example, physiological requirements determine such things as how long the learner can sit in one position or how easy it is to reach certain buttons etc. Psychological requirements need to be taken into account in determining aspects of design concerned with memory load, perception of colour, brightness and more complex issues such as case of learning, motivation etc. Social needs include the need for support in difficult learning situations and contact with others

The kinds of tasks that learners want to do can also vary. For example some are done frequently, others infrequently, some are repetitive whilst others are varied, some are done best by individuals and others in collaboration, and so on. Furthermore the complex task of learning can be broken down into subtasks such as reading which is done for different purposes so there are many and different intermediate goals for learners.

The environment of the learner can also affect learning in a different ways. Physical conditions, working regulations and ethos, social conditions are all important considerations when developing learning and training materials.

The characteristics of the hardware and software that are used in the development and delivery of the system can be important too. In computer based learning systems for example, aspects such as response time and screen size can be important.

A well designed system takes account of all these aspects. In the case of a hypertext or hypermedia system the major problem is how to avoid overwhelming users with complexity by managing that complexity in such a way that users do not get lost in vas amounts of information.

4. Interaction and learning

Prior to designing a multimedia system for learning, designers need to carry out a thorough task analysis so that they know who will use the system, what they will do with the system, where they will do it and with what equipment, just as they should in the design of any system. However, this can be problematic as we often do not know the exact tasks of learners and they are difficult to define. Furthermore, despite many years of study, there are no adequate learning theories upon which to base the design of such systems (Wright, 1989), although when designing hypermedia for learning, three learning processes seem to be best supported by hypermedia - information seeking, knowledge acquisition and problem solving (Jonassen and Grabinger, 1989).

The distinct advantage of hypermedia for information seeking is that very large volumes of information can be presented and learners can examine them in different ways. This is facilitated even more if the system designers have provided ways of allowing learners to examine different representations of information. The role of hypermedia in supporting knowledge acquisition is more complicated to make but broadly follows the line that exploring and active learning by interacting with the contents of a hypermedia system helps students to develop their own schema. Hypermedia may support problem solving by providing alternative representations. Much work is being done and remains to be done on these complex issues (see, for example, the collection of papers in Jonassen and Mandl, 1989).

Another way of looking at the how well the design of hypermedia supports user activity is by considering the nature of the interactions that occur. Norman (1988) and Hutchins et al. (1988) have proposed a high level theory of interaction which helps to explain some of the reasons why people have problems learning to use computer systems.

The user of a system starts off with goals which are expressed in psychological terms. The computer system, however, presents its current state in physical terms. Consequently, the user's goals and system state differ significantly in their form and content, which creates two 'gulfs' which must be bridged if the system is to be used successfully by users to achieve their goals. One gulf arises from the need to map the user's psychological goals into executable physical actions, the gulf of execution, and the other arises from the user's need to evaluate the physical changes of the system, the gulf of execution. Successful interaction with the system relies on these gulfs being bridged, which can occur in two ways. The system designer can design the system so that the input and output characteristics of the system match the psychological needs of the user well. Alternatively, the user can bridge the gulfs by creating plans, action sequences and interpretations that move the normal description of the user's goals closer to the description required by the physical system. In practice, both options need to be used. In particular, well designed systems are those in which designers have been able to move the system closer to the user. There are many ways of achieving this. A well selected metaphor, for example, can be used to help narrow the gulfs. Other aspects, such as consistency and transparency in the behaviour of the system, and meaningful feedback when things go wrong, can also help to bridge the gulfs and make the system straightforward to use.

When considering systems for learning and training, such as hypertext and hypermedia, we need to be particularly aware that learners differ in knowledge, skills, experience and learning style as well as there being the other difference mentioned in the previous section. Similarly, their needs for help, support, feedback, to explore and to be motivated also differ. Furthermore, learners change their requirements as they learn and for different parts of the task, so designers cannot assume that characteristics will be static. Long explanations, prompts, help and detailed menu lists may be useful to inexperienced users and extremely annoying for experienced users.

There is now a rapidly expanding body of knowledge about human-system interaction available from texts, handbooks (eg. Helander, 1988) and guidelines (eg. Smith and Mosier, 198?), but few studies have focused directly on multimedia systems. Why is this? There are two likely explanations. One is that people believe that the same general principles may underpin successful interaction with multimedia systems as with any other kind of systems; designers just need to be careful to take account of the increased complexity of integrating different kinds of media. Another is that multimedia systems are too new and there has not been enough time to carry out detailed studies, partly because attention has been focused on technical problems such as image compression and partly because there have been few multimedia systems in existence with which to do user testing.

The design of complex hypermedia systems in which complex colour graphics, film and animation may be displayed on several screens or windows and interrelated with sound could put even more burden on designers and have even more potential for confusing users, who have to make sense of a variety of different stimuli associated with different kinds of information. Alternatively, is it possible that such systems might have some characteristics that make them easier to use than, for example, hypertext?

Similar promises are heard about hypermedia as hypertext, and they even some with the non-computerised multimedia described in section two. Promises that are heard frequently are that learners will:

Will hypermedia live up to these claims? What kind of clues might studies of hypertext and experience with non-computerised multimedia provide which could be used to examine the promises and problems of hypermedia, or at least to raise the pertinent issues?

5. Some promises, problems and issues

Any well designed hypermedia, hypertext or non-computerised multimedia learning should have the following four characteristics: Non-computerised multimedia has been around for some years now (the UK Open University, for example, has provided these kinds of learning materials for twenty one years), so it is interesting to consider how it fulfils the criteria listed above. Hypertext has been actively researched for over five years. The common features of these two media with hypermedia is that all three have a branching structure and in all three students are encouraged not just to learn by reading linear text. Designers and advocates of these methods acknowledge that the effort and concentration required to retain several tasks or trails at once can be overwhelming for students. Consequently, according to Conklin (1987) both designer and learner must be more aware of process than they are when using traditional linear documents.

Table 1 contains a comparison of some of the ways that are used to support learners in each type of learning material.

Table 1: A comparison of non-computerised multimedia, hypertext and hypermedia

System Navigation Motivation Evidence of learning Distracting features
Non-computer mm signposts - clear instructions, sometimes a symbol, study chart shows when different media are used materials designed to encourage active learning - lots of in text exercises and practical work testing with students, number of registrations, performance on assignments and exams all suggest successful learning poor signposting, superfluous detail, or inadequate explanation known to cause problems
Hypertext known to be a big problem. Facilitated by maps, trails, footprints, overviews, ability to retrace path or return to home position easily. Few obvious natural landmarks claim that exploration and determining own routes motivates not much but difficult to prove and media too new to have reliable practitioner experience as above few obvious ones - if anything the featureless nature of text is a problem. Lots of text on screen is to be avoided - poor readability and not liked by users (ref)
Hypermedia same problem as with hypertext but could be less acute as the graphics could provide landmarks. claim that graphics, film etc. will be more motivating than hypertext. Scope for realism. Scope for creative graphics designers to use design 'tricks'. much too new to say at present essential to be aware of good human factors practice. eg. arbitrary use of colour and highlighting to be avoided. Calls for knowledge and restraint from designers

6. General conclusions

Hypermedia is expensive to produce and although the technology will become cheaper some development costs are likely to remain high. Labour is expensive and so is good quality film; copyright costs for existing materials can be high. However, hypermedia offers exciting prospects compared with hypertext and conventional learning materials. Furthermore, the visual nature of hypermedia could well reduce the navigational problems that learners often experience when using hypertext. However, it will be important for designers to have a good knowledge of human factors so that distracting stimuli do not distract, annoy or disorient learners. This could become even more imperative with the continuing increase in computing power and the development of novel devices. Although virtual reality holds promises of making interaction even more direct, there are obstacles to overcome. At present the lack of tactile feedback and perspective cause users problems.

Nor is it any longer appropriate to think of user system interaction in terms of one user and one machine. Multi-user multimedia systems in which users work collaboratively, sometimes separated by large distances will be available in the future.

Even with all this technology it will not be easy to establish how much students benefit from learning with it, except from anecdotal evidence and careful observation. Studies in education have shown that it is probably impossible to prove quantitatively that one form of educational media helps students to learn more effectively than another, there are just too many variables to make such studies viable. Even if definitive results could be obtained, it is more valuable to answer questions such as: 'what are the characteristics of a system which facilitates learning in a particular situation?' What kinds of learning tasks are best achieved through which kind of media by which students? Multimedia will not be a universal learning panacea; there will be problems and it will be expensive. However, well designed multimedia, used appropriately, will be exciting and motivating for learners.

7. References

Bush, V. (1945). As we may think. Atlantic Monthly, pp 101-108. http://www.theatlantic.com/unbound/flashbks/computer/bushf.htm

Conklin, J. (1987) Hypertext: An introduction and survey. IEEE Computer, September, 17-41.

Davies, G. & Preece, J. (1989). Home computing as an integral part of distance. Proceedings of the IFIP TC 3/WG 3.4 Working Conference on Methodologies of Training Data Processing Professionals and Advanced End-Users. Education and Computing, 6, Nos. 1, 2 (1990).

Fox, E. A. (1989). The coming revolution in interactive digital video. Communications of the ACM, 32(7), 794-801.

Gygi, K. (1990). Recognising the symptoms of hypertext and what to do about it. In Laurel, B. (Ed), The Art of Human-Computer Interface Design. Addison Wesley Publishing Co., Reading, pp. 279-287.

Helander, M. (1988). Handbook of Human-Computer Interaction. North-Holland, Amsterdam.

Hutchins, E. L., Hollan, J. D. and Norman, D. A. (1986). Direct Manipulation Interfaces. In Norman, D. A. and Draper, S. W. (Eds.), User-Centred System Design: New Perspectives in Human-Computer Interaction. Lawrence Erlbaum Associates, Hillsdale, New Jersey, pp. 87-124.

Jonassen, D. H. & Grabinger, R. S. (1989). Problems and issues in designing hypertext/hypermedia for learning. In Jonassen, D. H. and Mandl, H. (Eds.), Designing Hypermedia for Learning. NATO ASI Series, Vol. 67, Springer-Verlag, Berlin, New York, pp 3-26.

Jonassen, D. H. and Mandl, H. (1989). Designing Hypermedia for Learning, NATO ASI Series, Vol. 67, Springer-Verlag, Berlin, New York.

Norman, D. A. (1986). Cognitive engineering, In Norman, D. A. & Draper, S. W. (Eds.), User-Centred System Design: New Perspectives in Human-Computer Interaction. Lawrence Erlbaum Associates, Hillsdale, New Jersey, pp. 31-62.

Preece, L., McKerlie, D., Griffiths, R. & Davies, G. (1991). The Human-Computer Interaction Multimedia Project, Dept. of Computing, Open University, Technical Report Series (in press).

Preece, J. L, Maurer, H. and Davies, G. (1991). Presentation metaphors for a very large hypermedia system. Journal of Microcomputer Applications, 14, 105-116.

Smith, S. J. & Mosier, J. N. (1986). Guidelines for Designing User Interface Software. Mitre Corporation, Bedford, MA.

Wright, P. (1989). Hypertexts as an interface for learners: Some human factors issues. In Jonassen, D. H. and Mandl, H. (Eds.), Designing Hypermedia for Learning, NATO ASI Series, Vol. 67, Springer-Verlag, Berlin, New York pp. 169-184.

Authors: Jenny Preece and Gordon Davies, Computing Department, Open University, Milton Keynes, MK7 6AA, UK

Please cite as: Preece, J. and Davies, G. (1992). Multimedia: Some promises, some problems and some issues in human-systems interaction. In Promaco Conventions (Ed.), Proceedings of the International Interactive Multimedia Symposium, 259-266. Perth, Western Australia, 27-31 January. Promaco Conventions. http://www.aset.org.au/confs/iims/1992/preece.html

[ IIMS 92 contents ] [ IIMS Main ] [ ASET home ]
This URL: http://www.aset.org.au/confs/iims/1992/preece.html
© 1992 Promaco Conventions. Reproduced by permission. Last revision: 4 Apr 2004. Editor: Roger Atkinson
Previous URL 8 Apr 2000 to 30 Sep 2002: http://cleo.murdoch.edu.au/gen/aset/confs/iims/92/preece.html