Interactive multimedia: Educational desert or educational oasis?
David M. Kennedy

Multimedia Education Unit

The University of Melbourne


Carmel McNaught

Academic Development Unit

La Trobe University




The theoretical framework developed by Ramsden and used in the Course Experience Questionnaire offers a wealth of empirical data about teaching and learning from both studentsí and teachersí perspectives. The problem for an educator remains about how to transform what is already known about what constitutes good practice in the classroom and lecture halls into computer-aided learning and the world of interactive multimedia (IMM). One aspect of IMM design is the theoretical perspective of learning held by teachers and lecturers. Their perspective influences everything they do as educators, from what they do in a lecture theatre or classroom to how they go about designing IMM. Laurillard argues that learning is an iterative process in which students discuss, interact, adapt and reflect on the knowledge they are striving to assimilate. This paper discusses how the theoretical framework of Ramsden, and Laurillardís perspective on teaching and learning may be incorporated into the IMM design process.

1 Introduction

Implementation of a new curriculum or an innovative approach to teaching an existing curriculum in the past involved examination of the literature in the content domain to evaluate what were the difficulties experienced by students, peer review of the sequencing of content, and examination of other similar curricula. It is significant to observe that the design of interactive multimedia in higher education has developed with often minimal reference to the educational research available, both within a particular discipline and about student learning in general. This is in sharp contrast to the manner in which most researchers in higher education normally undertake a research activity. Instead, development of IMM has tended to focus on the hardware (e.g. which platform, CPU clock speeds, delivery platform verses development platform) and software (e.g. screen design, development tools, the use of colour, navigation, and budget and time restraints) issues rather than the educational perspective. The focus is now changing, however.

The educational design of IMM has until recently ignored much of the recent research into the synergistic and interactive processes of teaching and learning. A number of researchers argue that one of the major difficulties in the design of IMM yet remaining to be addressed is the gulf between the instructional or educational design of IMM and what the research literature indicates is good teaching practice. Ramsden [1] lists a number of features that have been shown to be appropriate strategies for effective learning (p. 89). The five major categories are:

ï Good teaching practice

ï Emphasis on independence

ï Clear goals

ï Appropriate assessment

ï Appropriate workload

(after Ramsden [1])

The five categories have been extensively trialed through the use of the Course Experience Questionnaire and have been shown to be reliable and valid performance indicators for teaching quality in higher education. In section three of this paper each of the five categories are examined and the relationship between good teaching and IMM design is explored. Examples of IMM design that illustrate the practice are given along with examples of existing software that illustrate the design principles discussed.

2 Teaching and learning

Learning is the way in which an individual changes the way s/he conceptualises the world. A studentís approach to learning is a qualitative aspect of learningóthat is, how the subject matter is experienced and how s/he subsequently organises it.

Teaching involves a lecturer constructing learning opportunities for students. Traditionally the settings for academic learning have been in the form of lectures, tutorials, practical classes, problem solving exercises and assignments. The development of IMM for learning has generally been seen primarily as an extra learning opportunity for students, either in the form of a self-paced tutorial done in the studentís own time or as a aid to revision. Only limited examples exist where IMM has replaced part of the traditional approach to academic learning in higher education. However, the teaching methods chosen as the focus of the learning opportunities, teacherñcentred or studentñcentred or a mixture of both, are strongly influenced by the educational beliefs held by the lecturer. This paper is concerned with how these perspectives are applied to the design of IMM.

In order to build a more complete theoretical framework it must be recognised that the value of any piece of IMM can only be considered within the total context in which it is used. What we may consider to be a good IMM design may not be used to its full potential, while software perceived as merely adequate may be used with great effect. Laurillardís [2] four processes are valuable here in deciding how active students are in their own learning. These four processes are:

ï discussive;

the learner and the teacher negotiate the goals of the task, in that the teacher provides descriptions of the task which are meaningful to the learner, while the student articulates an understanding of the task;

ï interactive;

the learner acts to achieve the task goal while the teacher provides intrinsic feedback relevant to the task;

ï adaptive;

the teacher evaluates the ëdistanceí between student and the intended goal and suggests tasks to achieve said goal; and

ï reflective;

the teacher provides support to facilitate the studentís reflections on achievement levels.

In the literature on student learning, one contrast which emerges is between academics who think of learning as reproducing knowledge (and of teaching as organising and presenting the knowledge to be reproduced), and others who think of learning as a process in which understanding is constructed by the student with the assistance of the teacher. Available evidence suggests that these two different conceptions of education result in different learning experiences and outcomes for students. One, the reproducing/transmitting/expository conception, tends to encourage surface or reproductive learning in which understanding is limited in scope and is not integrated into studentsí ways of seeing the world. The other conception favours deep or transformative learning through which understanding is refined and assimilated (Bain and McNaught [3]).

Although some forms of computerñfacilitated learning (CFL) offer students accessible ways to represent and manipulate knowledge, most programs do not provide an opportunity for conversational convergence of ideas between student and tutor. Many academics who design IMM systems may be sensitive to past studentsí learning difficulties; however, instead of designing conversational experiences for their students, they adopt the view that their role is to provide better explanations of problematic concepts. This preñemptive approach does not necessarily result in students challenging and changing their own unhelpful conceptions.

Laurillard [2] is insistent that, although some forms of CFL offer students accessible ways to represent and manipulate knowledge, most programs operate at a pre-emptive rather than conversational level. However, as Laurillard [2] and others (Wills and McNaught [4]) have observed, CFL is only part of the context in which teaching and learning occur in a university subject. Many teachers have conceptions of teaching and learning which are student-centred and transformative. They can incorporate pre-emptive CFL into their teaching strategies in ways that assist conceptual change in students (White and Horwitz [5]).

The theoretical framework presented here will attempt to show how IMM can address good teaching and learning practices which aim to engage students in active rather than passive learning, through a transformative rather than a preñemptive or expository model of design.

3 Pedagogy and design: Some examples

Ramsden [1, p. 89] states that studentsí perceptions of a good lecturer include: organisation, stimulating interest, providing understandable explanations, empathy with students needs, giving feedback on work, setting clear goals, and encouraging independent thought. It is salutary to note that they do not necessarily include the lecturerís sense of humour or personality. He argues that good IMM programs should be designed to respond to the learning styles and needs of students as part of the process in which the program operates. Poor IMM programs have little learner control either with the manipulation of the content, the sequencing of the instructional material and learning activities, or allowing the student to create her or his own knowledge construction of the subject matter.

Each of the five factors developed by the Course Experience Questionnaire are examined below with their associated aims. Although good teaching is undoubtedly a complicated matter, there is a substantial measure of agreement among these empirical studies about its essential characteristics (Ramsden [6]). The relationship between good teaching and IMM design is explored and, where possible, examples of IMM design that illustrate the concept are given. In order to facilitate the relationship between what we know about teaching quality and the interactions students experience when using IMM, the relationships between Ramsdenís criteria for good teaching and IMM developed from expository, preñemptive or transformative models of design are summarised in Table 1.

Table 1 Relationships between criteria for effective teaching and models of IMM design
Ramsdenís [1] criteria for effective teaching
Didactic or transmission modes of IMM design
Pre-emptive modes of IMM design
Transformative or conversational modes of IMM design

(Laurillard, [2])
Good teaching practice

ï showing respect and concern for students,

ï sharing the love of your subject with students,

ï being able to make the material to be taught both interesting and stimulating,

ï engaging students at their level of comprehension,

ï explaining the content using clear and appropriate language,

ï improvising and adapting to new demands,

ï learning from students and other sources (e.g. journals, colleagues) about the effects of teaching and how it can be improved.

The traditional lecture is often characterised by poor teaching practice. While individual lecturers are passionate about their subject, student learning may be passive rather than active, adaptation to new ideas can be generally slow and student engagement can be minimal. IMM designed within a didactic model may have:

ï the lecturer as expert, transmitting knowledge to students,

ï the content and sequencing prescribed by the lecturer,

ï a passive, rather than an active model of student learning,

ï minimal credence given to alternative models of representing knowledge.

Pre-emptive models of IMM design acknowledge the student as fundamental to the design of the program. The use of the ëbetter explanationí is a defining feature.

ï use of appropriate language,

ï on-line help and glossary,

ï formative design process,

ï multiple perspectives of concepts,

ï multiple paths through the software or a greater degree of student control,

ï models of IMM design based on misconceptions,

ï adaptive hypermedia (navigation, presentation),

ï use of life-world experiences of student,

ï attempts to actively engage students in modes of problem solving,

ï good use of visual and audio material.

Transformative indicates an iterative approach to learning through the processes of discussion, adaptation, iteration and reflection. The challenge is to design IMM to respond to the different learning styles and needs of the students. A suitable context in which the IMM can be utilised is essential to the design.

ï use of appropriate language,

ï on-line help and glossary

ï formative design process,

ï encouraging development of a personal perspective,

ï multiple perspectives on concepts,

ï models of IMM design for conceptual change (conflict),

ï multiple paths through the software, with a high degree of learner control,

ï tasks which allow studentsí to build their own representations,

ï students negotiating tasks,

ï context which IMM is embedded is especially crucial,

ï good use of visual and audio material, often associated with multiple representations of concepts.

Emphasis on independence

ï providing opportunities for students to become more independent,

ï implementing teaching techniques that require students to learn actively, act responsibly and operate cooperatively.

There is little or no opportunity for students to be independent in this model of IMM design.

ï only one navigational path,

ï passive learning, characterised by click-and-point interfaces,

ï minimal learner control.

There is considerable effort made to engage the learner in active learning.

ï IMM provides alternative paths for navigation which attempt to address different learning styles in students,

ï activities involving problem solving may be present,

ï software may be designed to be used in groups rather than with individuals.

The focus is to make students metacognitive about their own learning processes.

ï the content may be sequenced by the student,

ï the software is integrated with a specific context which promotes an iterative discussion process,

ï cooperative problem solving.

Clear goals

ï being committed to explicating what must be understood, the level of understanding and why this level is appropriate,

ï valuing understanding rather than rote learning.

In traditional lectures the syllabus outline was generally provided however the level of understanding and why it is required were not made clear.

ï IMM focuses on browsing rather than engagement in relevant tasks,

ï understanding is valued and desired; however, the model of learning doesnít foster such outcomes.

Active efforts to integrate the goals with the content of the IMM.

ï hypertext links from text, exercises and interactions to the syllabus outline,

ï IMM is intended to provide a ëgood explanation of goalsí,

ï a clearly articulated desire for more than surface learning,

ï on-line frequently asked questions.

The student is engaged in the shared determination of the goals of the academic content.

ï engagement in construction of relationships linking goals to academic content,

ï hypertext links from text, exercises and interactions to the syllabus outline,

ï relationships between prior, current and future course directions are explicated.

Appropriate assessment

ï applying appropriate assessment methods, the purpose of which are clearly understood,

ï giving feedback of the highest quality on student work.

Early CFL was characterised by either multiple choice or one word answers.

ï predefined (algorithmic) relationships between student responses and feedback, or

ï feedback is limited to yes/no or right/wrong answers, or

ï limited feedback (e.g. a statement of the algorithmic answer to the problem).

This model focuses on multiple modes of assessment but is still characterised by multiple choice or one word answers.

ï the focus is on the ëbetter explanationí for elucidating conceptual problems to students,

ï immediate feedback.

Assessment is focused upon determining the level of understanding and explicating personal perspectives representing the academic content.

ï extended answers which may be self-assessed from a number of models of expert answers,

ï multiple modes of assessment,

ï the student may negotiate the modes of assessment with the academic.

Appropriate workload

ï focusing on key concepts, and studentsí alternative frameworks, rather than on just covering the ground.

The general process is on delivery of the course and covering the material. In IMM this results in

ï only including the prescribed academic content,

ï ignoring studentsí alternative frameworks.

The focus is on addressing studentsí prior knowledge and misconceptions rather than just covering the ground.

ï the workload is adjusted. IMM is not merely an adjunct to conventional lectures, tutorials or practicals.

The focus is on key academic concepts in consultation with students.

ï outcomes and timelines are negotiated and students are focused on developing relationships between key concepts within an appropriate time period.

3.1 Good teaching practice

3.1.1 Showing respect and concern for students

Students appreciate the effort made by academic staff in the preparation of IMM courseware (and conventional multimedia materials, course outlines, solutions to problems, etc.). They perceive that IMM is being designed to enhance their learning opportunities. IMM that contains an onñline help and/or glossary recognises individual student needs. The use of language that recognises student prior knowledge and experience is also perceived to be an appropriate IMM design principle.

3.1.2 Sharing the love of your subject with students

At first glance, these social factors may be considered problematic when using a computer as a cognitive tool. However, research indicates that students perceive that teachers and lecturers who make an effort to make their subject more interesting, more accessible and more enjoyable by using IMM are respected and appreciated for their efforts (McTigue et al. [7]). A formative design process that involves teachers, students and instructional designers will enhance the design of IMM.

3.1.3 Being able to make the material to be taught both interesting and stimulating

Initially, computer software consisted of drill and practice derived from a textñbased format. As the hardware and software tools became more powerful, IMM utilised text, animation, graphics, sound, and video. Early instructional design in IMM limited the opportunity for students to interact with, or sequence the content. Arguably, IMM is enhanced by creative uses of multimedia (avoiding electronic page turning, incorporating high degrees of interactivity). However, the content, and the sequence in which it was organised was often prescribed by either a content expert or the programmer. Limited thought was given to differences between learners (e.g. prior knowledge or individual learning styles). Learners had little control over the sequencing or the style of their learning. The commonly used clickñandñpoint interface also diminished the ability of students to gain their own perspective on the content. Ramsden [1] and Laurillard [2] have suggested that increasing learner control will improve student motivation and interest in the content. Also, the use of lifeñworld experiences of the learners (where possible) in IMM design stimulates the learner to develop knowledge from a more personal perspective. Decontextualising content does not encourage a deep approach to learning, which has already been discussed as a more desirable and satisfying approach to academic study.

3.1.4 Engaging students at their level of comprehension

Ramsden [1] argues that a single prescribed path through the program, imposed by the ëcontent expertí or ëinstructional designerí would seriously inhibit studentsí access to the content and the potential for higher levels of cognition. Therefore, IMM should provide opportunities for students to access the content in a highly individualised manner (Reeves [8]). IMM design needs to address issues of how the learner may want to think about or study the content. To enhance student interest and engage students at their level of comprehension, studentsí prior knowledge should be included as part of the content of the IMM (Kennedy [9], Ausubel [10]). Studentsí prior knowledge includes lifeñworld experiences appropriate to the content, previous studies in the content area, and alternative frameworks already developed.

The work of Alexander and Cosgrove [11] looks at studentsí prior knowledge, and then confronts them with their strongly held prior knowledge constructions, ultimately challenging them to defend their beliefs over the scientific view of electricity. This IMM program adopts a conversational, transformative model of learning. Alexander and Cosgrove [11] reject the transmission model traditionally used by lecturers and teachers: the standñandñdeliver approach so often seen in university lectures. They argue that transmissive models of teaching often force students to operate algorithmically because the student is denied the opportunity to generate their own understanding of a concept through the normally iterative process of scientific discourse. Instead, their approach is one that attempts to align studentsí personal theories with scientific theory. This IMM focuses on the three realms of knowledge:

ï the (prior) knowledge of the student;

ï the domain knowledge (the scientific view of the concepts to be studied); and

ï the tutorial knowledge (what we know about how learning occurs).

Studentsí prior knowledge constructions are very resilient to change, after all, their life experiences tell them that things are consumed as they are used; the wood for a fire, the gas in a stove, and the energy they have before exercise. To change their beliefs, students must unravel the reasoning that led them to their current understanding and construct a new personal view of the concept (Alexander and Cosgrove [11])

3.1.5 Explaining the content using clear and appropriate language

Scientific knowledge has often been stated as being ëcounter intuitiveí (Wolpert [12]). This view is supported by Resnick ([13], p. 5) who argues that ìbasic scientific concepts are in fundamental epistemological conflict with many commonplace everyday conceptionsî. These two views reflect our experiences of teaching and learning. Studentsí prior knowledge of a particular content domain often contains many alternative frameworks and uses language in a nonñprecise form. Kennedy [14] indicates that students use everyday language and expressions to describe scientific concepts. These expressions are often imprecise or exhibit alternative frameworks. For example, in a study of high school students studying a preñuniversity chemistry course, students often used the word ëheatí when ëtemperatureí would have been more appropriate, and in some instances they described physical processes in terms which indicate they believed a chemical processes had taken place (Kennedy [14]).

In StatPlay, Thomason, Cumming and Zangari [15]) use multiple representations of statistical concepts to encourage students to immerse themselves in microworlds. In StatPlay students may approach statistical concepts from multiple perspectives in order to develop congruent understandings.

IMM developed only from the perspective of a content expert may not address studentsí alternative frameworks, prior knowledge or present multiple frameworks. The IMM designerís tasks are to determine the appropriate language to be used, provide an onñline glossary, help files with examples of procedural approaches to problem solving, and multiple perspectives of concepts.

3.1.6 Improvising and adapting to new demands

This is one of the most difficult issues facing developers of IMM today. An experienced teacher is able to monitor the understanding of the learner very closely (appropriate questioning techniques, direct observation of student practice and responding to student questions) and adapt his or her instructional strategy as appropriate. With careful design and a sufficiently large database, IMM can be produced which adapts to the userís individual needs and interests. Beaumont and Brusilovsky [16] have designed and implemented Adaptive Hypermedia Systems (AHS). These systems either adapt their presentation or adapt their navigational support depending on the individual student differences, prior knowledge or navigation constructions that develop as the student uses the software. AHS build a model of the goals, preferences and knowledge of the individual user and then uses this information to adapt the hypermedia to suit user needs. It is designed to support studentñdriven exploration of educational content.

With adaptive presentation, an initial questionnaire is provided for students and the information gathered is used to alter the onñscreen material. The information gathered includes the semester the student is currently in, and the purpose for using the hypermedia (exam preparation, introduction to the topic, and the bias the student wishes to apply to the content).

Adaptive navigation involves modification of some of the links. They may be hidden (thus reducing the cognitive load on novice students), the ordering of visible links may change, and the visible links may be adaptively annotated. Adaptive annotation at the simplest level is exemplified by the World Wide Web Netscape Navigator® browser in which sites already examined change colour to indicate that the user has already been to that site.

An example of this type of software is ANATOM-TUTOR (Beaumont and Brusilovsky [16]) which uses an adaptive presentation model. The text level is determined by determining the userís goals (as above). At the text level, the level of the studentís prior experience (what lessons s/he has already worked through, lectures already attended, etc.) then determines the expository style presented to the student. Greater levels of experience result in a more functional style being presented. With this example however, the student models are based upon the premise that students have already mastered the previous content. Much of this research is driven by studies into Artificial Intelligence (AI) and neural networks. In these areas of study the aim is to develop software that can mimic or extend the functionality of an expert teacher. Thus far, the systems described above rely on algorithmic design parameters in order to either modify the navigational paths or the content of the software viewed by different students.

3.1.7 Learning from students and other sources (e.g. colleagues, journals) about the effects of teaching and how it can be improved

A number of researchers have experienced surprise (Dickinson [17]) when confronted with the interpretations and understanding expressed by students asked to provide feedback for IMM (McNaught et al. [18]). Research has indicated that a formative, iterative design process which involves students produces more usable and effective IMM. Lecturers and teachers need to be reflective about their teaching practice, and beliefs about teaching and learning. In IMM design this involves providing a mechanism for the students to provide feedback (preferably screen by screen) regarding all aspects of the interface designócontent, screen display, navigational options, animations and the response of the software to actions by the learner.

An existing IMM resource at The University of Melbourne is the set of ChemCAL modules (McTigue et al. [7]). ChemCAL uses on-screen video and animations, a range of question formats and three levels of direct response to students; it also has built-in logging that provides two-way feedback to both students and course supervisors. Every screen is designed to allow students to comment on any aspect of that particular screen. In addition there is a more extensive form at the conclusion of the chemistry content where students may make more detailed comments about the package. The formative and summative evaluation done on these materials indicates that the students like using the software and achieve similar or better academic results in examinations (McNaught et al. [18]). The formative evaluation of the software from the studentsí perspective encouraged many alterations to the design of the final product.

3.2 Emphasis on independence
3.2.1 Providing opportunities for students to become more independent

As stated above when students are given control of the learning materials they exhibit a wide range of navigational routes. Some begin by looking at what they already know while others start with unfamiliar concepts and principles. Some work though the materials in a linear fashion, while others leave an exercise half done to explore another section before returning to complete the initial exercise (Laurillard [2]). Clearly learners require a range of navigational opportunities in order to facilitate their own style of learning. The work of Pérez et al. [19] is an example of adaptive IMM which addresses student learning styles and prior or current knowledge structures. They have designed a system that has adaptive navigation and content, based upon a profile of studentís learning with the software.

3.2.2 Implementing teaching techniques that require students to learn actively, act responsibly and operate cooperatively

The goal of good IMM should be to involve the student actively in the construction of knowledge. Most current researchers would regard the notion of ëclick and pointí in software as being only marginally interactive. Sims [20] has suggested that there are seven levels of interactivity, consisting of passive or electronic page turning, using hierarchies, updating, constructing, using simulation, using free interactivity, and being actively situated. The levels have implications for:

ï the way in which learners interact with an application;

ï multimedia design and development; and

ï the link between learner control, interaction and navigation.

Knowledge construction by students requires software that allows students to actively construct knowledge. In Simsí seven levels the degree of learner control and software complexity increases with each succeeding level. The model encompasses the range of passive interactivity (electronic page turning) to highly interactive IMM environments situated in a particular context. Interactivity may be enhanced by using problem solving exercises, case study scenarios, or interactive experiments. The software Investigating Lake Iluka (Hedberg et al. [21]) illustrates the nature of situated learning in which the students are actively involved in constructing their own knowledge (Hedberg and Harper [22]). Interactivity in the Lake Iluka software has been designed with situated learning as a major premise. It is based on the premise that ìthe activity in which knowledge is developed and deployed is not separable from or ancillary to leaning and cognition, but an integral part of what is learnedî (Brown et al. [23], p. 32). However, it remains primarily preñemptive in design. The students do not have the opportunity to negotiate their own problems. The context in which the software is used determines the degree to which it can be transformative.

3.3 Clear goals
3.3.1 Being committed to explicating what must be understood, the level of understanding and why this level is appropriate

Providing clear educational goals within an IMM learning environment is a straight forward process involving the inclusion of suitable textñbased materials. Surprisingly, it is not often done. However, IMM designs allow for not only the provision of program outlines but hypertextually linked content within the program itself. The opportunity exists for good IMM to have interactive linkages that relate the academic content of the software to the goals the students are expected to achieve. Good IMM should therefore contain a syllabus outline which is linked by hypertext to the content, questions and exercises in the software. Sound instructional design of IMM also has the potential to display to students the relationship of the current academic content to prior and future work in the subject domain.

3.3.2 Valuing understanding rather than rote learning

Much of the interactive multimedia currently produced has low levels of interactivity. The click-and-point model of IMM

ï encourages surface learning rather than deep learning, and

ï is more likely to favour rote learning rather than understanding.

Models of IMM design which facilitate the construction of studentsí knowledge are more likely to encourage deep learning with associated understanding. Mayes et al. [24] have developed learner support environments which involve dynamic hypertext linking called StrathTutor. Students have the opportunity to examine hypotheses to develop relationships between the attributes of a system. Students may interrogate the software by developing a hypothesis with particular attributes. The system responds according to preñdefined attribute coding, offering the student a ëguided tourí of all screens that are coded with that particular set of attributes. The student is able to dynamically form hypertext links appropriate to their learning needs. This model of software design is potentially transformative as it allows students to actively construct their own knowledge and develop hypotheses.

3.4 Appropriate assessment
3.4.1 Applying appropriate assessment methods, the purpose of which are clearly understood

IMM should explicate the model of assessment presented in the software, the purpose for which it is designed (formal assessment, mastery learning, or informal feedback for the student), and the number and type of questions the student will experience in using the software. The model of assessment often drives the type of learning required by the student. Short answer, and oneñword answer assessment items are more likely to foster a shallow-rote learning approach in the students since such questions tend to address knowledgeñbased questions rather than involve problemñsolving, analysis or synthesis. There are many methods available to IMM designers for assessment. They include:

ï providing opportunities for extended answers within the software to be entered by students, which may be then downñloaded to the tutor or lecturer;

ï onñline facilities to email the tutor or lecturer with problems or questions;

ï paperñless submission of assignments.

Studies in Artificial Intelligence by Dowling and Kaluscha [25] and Petrushin et al. [26] has suggested methods of providing computerñbased adaptive assessment of student knowledge. This work is predicated on predefined relationships within a knowledge hierarchy as defined by content experts. Students are provided with questions from a large database of items. The student response guides the response of a software algorithm in selecting the next test item. This area of research is focused specifically on designing alternative procedures for elucidating and grading student knowledge: the number of possible relationships in a given hierarchy increases exponentially with the number of items in the database of questions (Dowling and Kaluscha [25]). This method of assessment largely follows a model of preñemptive design. The students do not have the opportunity to negotiate goals or the nature of the assessment model.

3.4.2 Giving feedback of the highest quality on student work

Students of higher education indicate that they value immediate and comprehensive feedback. IMM can provide timely (when the student wants it) and iterative feedback for the user (Marchionini [27]). IMM can also provide the basis for individualised student feedback by the use of an iterative approach to the design which allows the on-line help and the basis for the feedback to be constructed from typical student questions and problems. There exists opportunities for IMM software to provide model answers for students. For example, a student may be required to provide an extended answer to a particular question. Once the student enters an answer, s/he could be provided with a number of model answers to the same problem. This approach provides both immediate feedback and a form of self assessment that is difficult for a lecturer to provide in all but the smallest class groups.

Another example, the software mark and its successor, xmark, developed by Ho and Whale [28] is designed to both assess student work and facilitate student feedback onñline. The software xmark is able to accept documents from students which contain text, diagrams, sound, and movie objects. The tutor or lecturer has the opportunity to respond in kind, with text, sound and movie filesówhichever may be the most appropriate form of feedback for the student. The tutor may also assign grades, and save her or his responses for further use as standard or user comments. While this software operates on a Unix system and is still in the development phase, it shows the potential for lecturers and teachers to provide appropriate feedback to students in a wide variety of formats. The authors suggest that xmark has the potential to assist university staff faced with increasing class sizes and time constraints to provide more effective comments and feedback to students.

The use of the internet is also providing a mechanism to deliver appropriate and ëinñtimeí feedback to students for IMM being developed for, and used on the World Wide Web. Freeman [29] has used the model of a managed internet bulletin board in order to provide feedback for a very large group of business finance students. A statistical evaluation of the use of this mechanism indicated that students who rated the Frequently Asked Questions (FAQs) on the bulletin board as the most useful resource, actually performed better in their assigned case studies. The students also indicated that the speed at which their questions were answered using this system was very important for their overall understanding of the content matter. The examples illustrate the range of approaches modern software tools offer IMM designers to provide appropriate feedback for students using software.

3.5 Appropriate workload
3.5.1 Focusing on key concepts, and studentsí alternative frameworks, rather than on just covering the ground

It is important for IMM to address the misconceptions or alternative frameworks or alternative conceptions (Driver and Easley [30]). There is evidence that students may hold two conflicting views of science at the same time. The first based upon their everyday ëcommonsenseí views of the world around them, and the second based upon the scientifically accepted view used at school to pass exams (Pines and West [31]). The alternative frameworks they develop as a result of everyday experiences, such as exposure to the media, (radio, newspapers, and TV), other educational experiences and adult comments which may have been interpreted in a totally unexpected manner, are often more resilient and more integrated into the learnerís conceptual knowledge (Sunal and Sunal [32]). Solomon ([33], p. 78) argues a similar position in that ìsocially acquired life-world knowledge is stored in the memory separately from symbolic school knowledgeî, and the socially relevant knowledge is easier for students to retrieve.

An expository model of IMM, by its nature, focuses on covering the ground. A preñemptive model, however, will endeavour to include alternative frameworks commonly held by students, by offering alternative pathways through the software, but still fails to address the time taken by individual students to work through the software. A transformative model involves discussion between the teacher and the student, or interaction between the software and the student, to determine the appropriate workload to satisfy the academic requirement of the course. This implies radical changes to curricula with consequent impacts on timetables, teaching spaces, and relationships between subjects. Both staff and students need to culturally adjust to new patterns of teaching in higher education (Wills and McNaught [4]) and this adjustment needs to be recognised in IMM evaluation.

4 Conclusions

The purpose of this paper has been to focus on relationships between:

The difficulty still facing instructional and educational designers is how to implement software designs which are truly transformativeóthus enabling students who use such software to change their own knowledge constructions at a fundamental level.

Arguably, the educational perspective of IMM design has often been lost in budgetary constraints, pedagogical disagreements, disagreements of content experts, restrictive timelines and delivery dates, inadequate evaluation, graphical design problems, developing appropriate navigation structures, IMM team conflicts, academicsí with increasingly busy workloads, and the other myriad difficulties of developing IMM, not the least being the 500:1 ruleó500 hours of development time for the first hour of IMM.

IMM is argued by many to still be an educational desert, however there are sufficient examples given to indicate that the oasis may be in sight.

5 References

1. Ramsden, P. Learning to Teach in Higher Education. London, Routledge. (1992).2. Laurillard, D. Rethinking University Teaching: A Framework for the Effective Use of Educational Technology. London, Routledge. (1993).3. Bain, J. and McNaught, C. Academicsí educational conceptions and the design and impact of computer software in higher education. pp. 56-59 in C. McBeath and R. Atkinson (Eds.). The Learning Superhighway. New World? New Worries? Proceedings of the Third International Interactive Multimedia Symposium. Perth, Western Australia. 21-26 January. (1996).4. Wills, S. and McNaught, C. Evaluation of computer-based learning in higher education. Journal of Computing in Higher Education, 7(2), 106-128 (1996).5. White, B. and Horwitz, P. Computer microworlds and conceptual change: A new approach to science education. In P. Ramsden (Ed.). Improving Learning: New Perspectives. London, Kogan Page. (1988).6. Ramsden, P. A performance indicator of teaching quality in higher education: The Course Experience Questionnaire. Studies in Higher Education, 16(2), 129-150 (1991).7. McTigue, P. T., Tregloan, P. A., Fritze, P. A., McNaught, C., Hassett, D. and Porter, Q. Interactive teaching and testing tutorials for first year tertiary chemistry. pp. 466-471 in H. Maurer (Ed.). Educational Multimedia and Hypermedia, 1995. Association for the Advancement of Computing in Education. Proceedings of Ed-Media 95óWorld conference on Educational Multimedia and Hypermedia, Graz, Austria, 17-21 June. (1995).8. Reeves, T. C. Research foundations for interactive multimedia. In C. McBeath and R. Dickinson (Eds.). Proceedings of the first International Interactive Multimedia Symposium. Perth, Western Australia. (1992).9. Kennedy, D. M. Studentsí prior knowledge: Implications for instructional design of interactive multimedia. pp. 288-296 in J. Pearce, A. Ellis, C. McNaught and G. Hart (Eds.). Learning with Technology. Proceedings of the Australian Society for Computers in Learning in Tertiary Education ë95 conference, The University of Melbourne, 3-7 December. (1995).10. Ausubel, D. P. Educational Psychology: A Cognitive View. New York; Holt, Rinehart and Winston. (1968).11. Alexander, S. and Cosgrove, M. The design of an interactive multimedia program to facilitate understanding of basic electrical concepts. In proceedings of the European Association for Research in Learning and Instruction Conference. University of Nijmegen, The Netherlands. (1995).12. Wolpert, L. The Unatural Nature of Science. Cambridge, Massachusetts; Harvard University Press. (1993).13. Resnick, L. B. Introduction. In L. B. Resnick (Ed.). Knowing, Learning and Instruction: Essays in Honour of Robert Glasser. Hillsdale, NJ; Laurence Erlbaum. (1989).14. Kennedy, D. M. Meaningful learning in chemistry: A pragmatic constructivist approach to instructional design of hypermedia incorporating studentsí prior knowledge. Unpublished Master of Science Thesis. Curtin University of Technology. (1995).15. Thomason, N., Cumming, G. and Zangari, M. Understanding central concepts of statistics and experimental design in the social sciences. pp. 59-81 in K. Beattie, C. McNaught and S. Wills (Eds.). Interactive Multimedia in University Education: Designing for Change in Teaching and Learning. Elsevier, Amsterdam. (1994).16. Beaumont, I. and Brusilovsky, P. Adaptive educational hypermedia: From ideas to real systems, pp. 93-98 in H. Maurer (Ed.). Educational Multimedia and Hypermedia, 1995. Association for the Advancement of Computing in Education. Proceedings of Ed-Media 95óWorld conference on Educational Multimedia and Hypermedia, Graz, Austria, 17-21 June. (1995).17. Dickinson, R. Diverse functions: The creative design of a hypermedia authoring system. pp. 118-120 in C. McBeath and R. Dickinson (Eds.). Proceedings of the Second International Interactive Multimedia Symposium. Curtin University of Technology, Perth, Western Australia. February. (1994). 17. McNaught, C., McTigue, P. T., Fritze, P. A.,Tregloan, P. A., Hassett, D., Whithear, K. and Browning, G. Evaluating the impact of technology on student learning in higher education. pp. 195-206 in Proceedings of the Apple University Consortium Conference, University of Western Australia, Perth. 3-6 July. (1995).19. Perez, T. A., Lopisteguy, P., Gutierrez, J. and Usandizaga, I. HyperTutor: From hypermedia to intelligent adaptive hypermedia. pp. 529-534 in H. Maurer (Ed.). Educational Multimedia and Hypermedia, 1995. Association for the Advancement of Computing in Education. Proceedings of Ed-Media 95óWorld conference on Educational Multimedia and Hypermedia, Graz, Austria, 17-21 June. (1995).20. Sims, R. Seven levels of interactivity: Implications for the development of multimedia education and training. pp. 589-594, vol. 2, in M. Ryan (Ed.). Proceedings of the Asia Pacific Information Technology in Training and Education (Apitite) Conference, Brisbane, 28 June - 2 July. (1994).21. Hedberg, J., Harper, B. and Brown, C. Investigating Lake Iluka. Wollongong, Australia; Interactive Multimedia Pty. Ltd. (1993).22. Hedberg, J., G. and Harper, B. Exploring interactive multimedia information landscapes. pp. 301-305 in H. Maurer (Ed.). Educational Multimedia and Hypermedia, 1995. Association for the Advancement of Computing in Education. Proceedings of Ed-Media 95óWorld conference on Educational Multimedia and Hypermedia, Graz, Austria, 17-21 June. (1995).23. Brown, J. S., Collins, A. and Duguid, P. Situated cognition and the culture of learning. Educational Researcher, 18(1), 32-41 (1989).24. Mayes, T., Kibby, M. and Anderson, T. Learning about learning from hypertext. In D. H. Jonassen and H. Mandl (Eds.). Designing Hypermedia for Learning. Berlin, Springer-Verlag. (1990).25. Dowling, C. E. and Kaluscha, R. Prerequisite relationships for the adaptive assessment of knowledge. pp. 43-50 in J. Greer (Ed.). Proceedings of the Artificial Intelligence in Education Conference. Washington DC. (1995).26. Petrushin, V., Sinitsa, K. and Zherdienko, V. Probabilistic approach to adaptive studentsí knowledge assessment: Methodology and experiment. pp. 51-58 in J. Greer (Ed.). Proceedings of the Artificial Intelligence in Education Conference. Washington DC. (1995).27. Marchionini, G. Evaluating Hypermedia-based learning. In D. H. Jonassen and H. Mandl (Eds.). Designing Hypermedia for Learning. Berlin, Springer-Verlag. (1990).28. Ho, P. S. and Whale, G. Paperless assessment: Using multimedia and hyperlinks to improve the commentary in assignments. pp. 342-347 in H. Maurer (Ed.). Educational Multimedia and Hypermedia, 1995. Association for the Advancement of Computing in Education. Proceedings of Ed-Media 95óWorld conference on Educational Multimedia and Hypermedia, Graz, Austria, 17-21 June. (1995).29. Freeman, M. The role of computer based leaning, video conferencing and the internet in teaching and learning: A case study in tertiary business education. Paper presented at the Association of Southeast Asian Institutions of Higher Learning: The role of information technology in education development in the AsiañPacific region. Monash University. 11-14 June. (1996).30. Driver, R. and Easley, J. Pupils and paradigms: A review of literature related to concept development in adolescent science students. Studies in Science Education, 5, 61-84 (1978).31. Pines, A. L. and West, L. H. T. Conceptual understanding and science learning: An interpretation of research within a source-of-knowledge framework. Science Education, 70, 583-604 (1986).32. Sunal, D. W. and Sunal, C. S. Young children learn to restructure personal ideas about growth in trees. School Science and Mathematics, 91(7), 314-317 (1991).33. Solomon, J. Social Influences on the construction of pupilsí understanding of science. Studies in Science Education, 14, 63-82 (1987).