Full Paper

Margaret Bearman

margaret.bearman@med.monash.edu.au

Centre of Medical Informatics (CMI)

Monash University

During an evaluation of What's the Hype?, interactive health education modules for adolescents, a year 9 student raised his hand to give feedback. '[I] liked the drugs one because it had a story,' he said. Likewise our evaluations of a large medical education hypermedia project on HIV/AIDS showed that the case studies were considered most useful out of a range of learning options.

Case-based teaching is now commonplace in medical schools. There is general acceptance of problem-based learning (PBL) as a teaching method (Barrows, 1994), where the use of 'real' cases is central. There is extensive, innovative use of technology in developing electronic PBL cases (Koschmann et al., 1992; Mahling et al., 1995; Bouchard et al., 1997). The effectiveness of computerised case-based medical education has been demonstrated for some years by the rigorous PlanAlyzer study, which concluded that learning is more efficient via the computer, and just as effective (Lyon et al., 1992; Lyon et al., 1991).

Of course, case-based educational technology is wide-spread across disciplines. In last year's ASCILITE proceedings alone, there were multiple examples of case-based teaching. Subject areas included: business (Lawrence et al., 1996), manufacturing systems engineering (Netherwood, 1996) and software engineering (Pilgrim, 1996), as well as medicine (Smith et al., 1996). These cases and simulations are examples of situated learning, which contextualises knowledge within a particular domain, an concept promoted in tertiary education by Laurillard (Laurillard, 1993).

But what of the role of narrative? The pedagogy of case-based education generally emphasises active and situated learning. It does not stress 'story', and narrative is not a necessary part of a case, although stories are often included, occasionally with highly dramatic plots. Likewise the term 'simulation' includes story-based scenarios, but the emphasis is on approximating 'real life' rather than providing a narrative framework. It's worth considering a few theoretical perspectives, to gain an understanding of why some psychologists regard narrative as important to both education and culture.

There is an increasing interest in narrative in current psychological and pedagogical literature. Schank and Abelson (Schank et al., 1995) for example, maintain that stories form the basis of almost all knowledge, memory and comprehension. For example, I know that 'Melbourne is Australia's second largest city'. But perhaps this 'fact' is actually represented in my mind by an experience: an argument with a friend about the relative merits of the state capitals. Thus narrative might be the way that the mind situates and retrieves information.

Schank and Abelson argue that we achieve comprehension by associating our stories - and thus our experiences and indirect experiences - with other peoples' stories. Through the acts of telling our own stories and listening to other people's stories, we acquire understanding. This is compatible with Schank's more educational work, 'goal-based scenarios' (Schank, 1994) , which strongly advocates the use of situated learning, in particular using specific cases - even stories - to illustrate concepts. Both works taken together suggests that understanding, stories and real experience are all closely interrelated.

Bruner (Bruner, 1996) considers the value of narrative in education from the viewpoint of a cultural psychologist. In contrast to Schank and Abelson's mind-oriented, computational approach, he believes that narratives are vital for humans to 'make meaning' of themselves and their cultures. He argues that there are both narrative and deductive-logical approaches to education, and that our system favours the latter. The sciences, in particular, overlook the value of narrative and storytelling.

Bruner stresses that narrative is central to our lives as individuals, to our culture as a whole and to our systems of education.

While there is certainly some dispute as to the more controversial claims that all knowledge and memory is framed by stories (Brewer, 1995) , there is no question that narrative is important to human beings. For example, there are empirical studies clearly showing the vital role that narrative plays in comprehension, recall and organisation of events (Thorndyke, 1977; Graesser et al., 1995) . And there is a wealth of literature discussing the ability of humans to create narratives out of seemingly unrelated events (Stratfold, 1994; Reason et al., 1997) .

So is there any theoretical consensus about defining a narrative? There are similarities and differences between all the theories. However for the purposes of this paper, and in order to address the theme of 'what works and why', I will use a working definition from Plowman (Plowman, 1996a) . She identifies elements of narrative as: 'coherence, significance of all elements, and a fixed sequence: causality and linearity.' This does not rule out multi-linear narratives, where the user's choice results in different storylines, as the user experiences only one seemingly linear narrative.

A number of people have considered the dramatic narrative within interactive multimedia and virtual reality using literary, dramatic and cinematic theory (Don, 1990; Meyer, 1995; Plowman, 1994) . Obviously the body of work on narrative in the arts is very large, and not always appropriate for educational technology, where the emphasis is on learning rather than entertainment. In many ways, studies of documentaries provide a more useful parallel to educational technology than fictional works (Plowman, 1996b) . While approaches of this nature are valuable and give insight into the role of the narrative, pedagogically-oriented studies are beginning to emerge in their own right.

It appears that narrative serves a very different purpose to interaction (Plowman, 1996b) . Interactivity may be responsible for observed learning gains(Clark et al., 1992 ), but narrative may serve to keep the learning experience directed. Our own experience evaluating What's the Hype? (Bearman et al., 1997) indicates long interactions bound by a narrative may be viewed far more positively by users than a series of short unrelated interactions. This is in keeping with current research into the role of narrative in educational technology.

Many of the studies into narrative and educational technology have been conducted at school level by the Multimedia, Education and Narrative Organisation (MENO) in the UK. This organisation's work is partly motivated by an evaluation report to the UK National Council of Educational Technology (NCET), which described on-going problems with acceptance, usage and educational quality of educational technology within schools. This study found that:

MENO is researching theoretical and empirical aspects of narrative in educational technology, in order to address this question (MENO, 1995) . Initial results, based on qualitative data from the MENO studies, suggest the need for coherence in the learning experience. The students require direction in order to satisfactorily complete their tasks. It would appear that this can be provided by a 'macro-narrative', or an overarching story line, or sub-narratives organised by a unifying interface, or even by a goal supplied by the teacher (Plowman, 1996c) .

There have also been studies into the role of narrative into adult learning. Stratfold (Stratfold, 1994) conducted a qualitative study of TerminalRISK, an interactive training videodisk at Price Waterhouse. TerminalRISK contained a complex narrative structure, which was mediated through segments of video. The users felt it was entertaining and therefore motivating. Stratfold notes:

Stratfold concluded that the narrative helped recall and understanding, as well as providing a real life context.

Although there are still large gaps in our understanding of stories in educational technology, there is some consensus about how to design the narrative encounter. For example, Jones states, from his experience in designing a story-based simulations for adult English students, that there is a need to 'reconcile the narrative sequence with the desired learning sequence' (Jones, 1991) . Likewise Plowman states in her study of four interactive packages for primary students that 'Narrative should accommodate and unify both the fictional and pedagogical elements so that tasks are integral to the narrative.' (Plowman, 1996b) .

It is evident that there is a tension between the interactive task and the overall flow of the program. The interactivity disrupts the narrative, in order that the students achieve active learning, but at the same time, they need to be able to engage with the program in an on-going fashion. Both Plowman and Jones provide a few suggestions as to how to achieve the optimal balance between these two elements.

The primary and most important step is to select a narrative that will adequately accommodate interactions, and to ensure that these interactions arise logically from the flow of the story. This equates to choosing or creating a good case study, which allows the student to develop skills appropriately within the artificial context of the electronic medium. Jones, in discussing stories which 'simulate' the real world, stresses that the development of the narrative should reflect the success or failure of the student. That is, if the student makes an ill-considered choice, the narrative should provide intrinsic feedback, by showing the consequences of the action. For example, in CMI's What's the Hype? drug and alcohol module, there are consequences from consuming too many drugs at a simulated party. If the user drinks too much alcohol, the 'party' comes to an abrupt halt, with and the user ends up 'in the garden spewing your guts out'. Likewise consequences from good decisions should result in positive outcomes.

Plowman suggests that a coherent experience for the user is achieved by keeping distractions from the technology to a minimum. Preferably, interactive tasks should be short, discrete and related to the ëaction". Her studies indicate that the use of a narrator can mediate the learning experience and provide continuity.

The initial research into narrative provides a challenge for developers of case-based educational technology. It is important to find a balance between the narrative - which is a constraining device - and the freedom afforded by educational technology. Our own observations in formative evaluation of the HIV Hypermedia Medical Education Project Version 2, again provide a good working example. Five medical students and practitioners worked through case-based narratives under recorded observation. Within the package, they took on the role of a doctor, treating an HIV positive patient. Although the users had ample opportunity to leave the tutorial and investigate other options, they did so rarely. The story became the driving force of the encounter. The engrossing nature of narrative may well limit the exploration of further materials or additional concepts.

Perhaps narratives might be appropriate for novices within a particular domain of knowledge. At earlier levels of understanding, students might require the 'scaffolding' effect of a story. Indeed, narratives fit many of the features of the scaffolding process as proposed by Wood et al. Narratives, like a good tutor, perform the following functions: they enlist the interest of the student; they reduce the degrees of freedom of the task; they keep the learning process directed; they accentuate relevant or critical features. These are four out of the six scaffolding functions of a tutor (Wood et al., 1976) . Perhaps, as students develop expertise, the role for the narrative decreases. Again, anecdotal experience illustrates the point. Experts in HIV medicine have used the case-based tutorials from the HIV Hypermedia Project at many demonstration sessions. Informal observations indicate that experts appear to be far more concerned with the 'correctness' of their response (and of the material within the tutorial) than with the story.

In general, the notion of coherence is out of fashion with post-modern approaches to literary theory (Plowman, 1996b) . Is the concept of a unifying device such as a narrative too constraining of the student, forcing them down a particular path of understanding? This may be true in some cases, however narratives are not wholly at odds with constructivism. For example educational technology based on cognitive flexibility theory often provides multiple perspectives in the form of 'mini-cases'(Stephens, 1997) . These might be regarded as mini-narratives, multiple stories dealing with the same domain of knowledge. Additionally Bruner (Bruner, 1996) argues that a story is inherently interpretive rather than prescriptive.

It is also possible that narratives are more suited to some learning styles than others, as well as different age groups. While theory suggests that stories are fundamental to learning, there is little empirical evidence in the area of adult learning.

While there are many unanswered questions and a lack of empirical evidence, narratives can be an important consideration in educational technology design. The psychologists' case for humankind's tendency towards understanding through narrative is very compelling. Moreover the relationship between a story and a 'situated' experience is a strong one. However not all successful case-studies have 'causality' or 'coherence'. The work of MENO is proving very useful in describing the necessary role of narrative in educational technology for children, that it provides coherence to their experience with the interactive media. But children and adult learners do have different needs. Within a tertiary situation, we must consider narrative as an viable and valuable option, but perhaps considering that a degree of linearity may restrict the learner's available choices. Narrative may be most valuable as scaffolding to the novice, but what is scaffolding in one situation, may be restrictive in another.

All of the above references that are available on the Internet are through a small number of sites. I will list the five sites here, as they contain numbers of relevant and interesting papers.

This research has been carried out as part of study towards a PhD thesis at the Centre of Medical Informatics, Monash University. Thanks to my supervisor, Branko Cesnik.

(c) Margaret Bearman

The author(s) assign to ASCILITE and educational and non-profit institutions a non-exclusive licence to use this document for personal use and in courses of instruction provided that the article is used in full and this copyright statement is reproduced. The author(s) also grant a non-exclusive licence to ASCILITE to publish this document in full on the World Wide Web and on CD-ROM and in printed form with the ASCILITE 97 conference papers, and for the documents to be published on mirrors on the World Wide Web. Any other usage is prohibited without the express permission of the authors.