A framework for work-based networked learning.

This article describes the development of a task-based framework for providing learning within the workplace using networked multimedia. A philosophy is adopted for the framework that builds on the recognition that tasks must reflect the social process of learning as considered by Vygotsky (1978). This takes into account the role of dialogue in the building of knowledge for an individual and for a group, following the ideas of the conversational framework (Laurillard, 1993), and allowing for future vicarious forms of learning (Mayes & Dineen, 1999). Tasks are supported by links between people as well as activities, material, and the opportunity to find further resources. The framework was provided as a set of guidelines within a manual and also instantiated as a prototype system that represented the course using XML (eXtensible Markup Language). This representation allowed flexible delivery and the system has been used to support undergraduate courses within a university as well as trials within the semicondu ctor industry.

**********

This article has two main objectives. First, to describe a specific framework developed to support the needs of a project implementing work-based learning and second, to refer that framework to a theoretical base drawing from distributed cognition. The context for most of the work is the Advanced Software for Teaching and Evaluation of Processes (ASTEP) project. The aim of the project was to produce a framework to support the creation of multimedia learning environments, operating across the Internet, which facilitate distance learning for process-based high-technology companies. ASTEP was funded for a period of two years (start date: February 1998) under the Educational Multimedia Task Force action initiated by the European Union. The project produced five exemplar courses developed for different sectors of the semiconductor industry, that is, semiconductor fabrication (generic and proprietary), Microsystems manufacturing, high level design and test.

Initial stages of the project involved identifying a suitable approach and model, which would both suit the needs of the industry, and enhance the learning of the individual. To this end, a framework was developed, which supports a task-centred approach. In general terms, the course material is presented to the student as a series of tasks, each of which is supported with relevant resource material. The intention in developing the framework, was that production of such course material should be consistent and repeatable, and that there should be minimal requirements for specialist computing or multimedia skills on the part of the trainers or the users of the framework. At an early stage it was also decided to draw on previous experience and research to identify a theoretical underpinning for the framework.

The Internet is now seen as the natural solution for the production of new educational packages. This was apparent from the start of this project as an Internet based solution was considered appropriate for each of the various scenarios considered for trials whether they involved campus-based students, students from companies near educational institutions, or education within the workplace. The construction of a framework was chosen as a way to define an overall approach and to identify what was needed to lead from known materials to effective learning. The framework incorporated pragmatic reasoning about what could be achieved and how to allow for flexibility in how the learners could operate. The result was an approach that could be identified with theories of where learning happened but also be described in terms of a manual and a structured implementation. The principles that were placed in the framework are:

1. learning aspects should be considered before presentation and hence presentation should not be overvalued;

2. students should always be aware of the learning objective that they are working toward;

3. students should be actively engaged in a task related to the learning objective;

4. dialogue needs to be supported to help build community; and

5. resources should be seen as available from the world as well as within a course.

The basis for these principles is described later and can be seen as relating to the ideas that are supported within distributed cognition. Principle 1 is important as a move from effort on the experience of an individual in isolation to a part of a learning community. Principles 2 and 3 express the overall approach to learning through tasks; this is then connected to the ideas in distributed cognition through considering the knowledge of the overall system to be placed in the immediate community (principle 4) and also the world of resources now available through the World Wide Web (WWW or Web) (principle 5).

Applying these principles leads to a learning model with an emphasis on tasks as the central organisational structure, that is, task-centric. Additional effort is applied when constructing the course on defining these tasks. It was also found that it was possible to look at initial collections of material placed on the Web and quickly identify areas of weakness by considering whether or not the students would have current tasks known to them.

The work of this article is in accordance with ways that online courses are being developed elsewhere but differs in providing a specific framework and by leading with the tasks as the most important element. As with any relatively extreme position better results may be seen in some compromise with presentational and practical issues. In practice, courses produced following the framework have some additional elements of introduction and exposition. This article describes the formation of the framework and the task-centred approach and how it has been implemented when creating task-based courses.

Basis for the Framework

The first stage in designing the framework was to align the activity with a valid model for the process of learning. In particular, care was taken to avoid selecting presentational and design techniques over techniques that would encourage learners to reflect and discuss. The model for the learning process that was adopted emphasises the use of tasks and was drawn from the experience of assisting developments in computer-based learning within the Institute for Computer-Based Learning at Heriot-Watt University. This has included the running of a module on Computers in Teaching and Learning within a Masters in Human Computing Interaction since 1991. That module has always sought to try out innovative ways of supporting the students both to demonstrate ideas to the students and to put into practice methods that should provide the student with good support.

The use of tasks gives a focus for where learning occurs. Alongside this are other aspects of using the computer in learning such as the different modes and variety of presentation that become possible. These are indeed benefits but must be secondary concerns when designing systems for learning. Much packaged material in the form of CD-ROMs can offer attractive presentation but does not give a clear idea of what a learner should do with the material. This can lead to the simple browsing of material in a linear manner or a more or less random approach to clicking through the material (McKendree, Reader, & Hammond, 1995). Adding tasks enables the learner to be guided through to the appropriate sections and encourages them to interact with the material. In Laurillard, Stratfold, Luckin, Plowman, and Taylor (2000), this was described as providing narrative as a form of guided discovery learning and is demonstrated to lead to better learning experiences for school age learners.

The concept of using tasks within the learning process is supported by consideration of the process of learning. Learning new information is relatively easy when we have a structure that allows us to understand information. As a corollary, learning is difficult when we are faced with building a new structure when new information is received. The aim is to support learning for both the case where new structures need to be built and where previous experience provides existing structures. Rumelhart and Norman (1978) considered the three modes of learning as accretion, structuring, and tuning with the commonest case of accretion being seen as when additional information is added to an existing structure. For work-place learning there are a range of experiences and it is appropriate to consider a design that supports each of these aspects.

For the case of learning in a new area the model needs to be such that an effort to understand can be applied. This effort is provided by the tasks under the learning by doing model whereas presenting information, however attractive, does not support the learner through the effort of understanding. While this argument supports the adoption of a task-based approach for material that will encourage learning by doing, it also leads us to recognise that learning can be easy and even pleasurable if the learner already has the initial structure in place for the material. To support this form of learning the authors propose an extension to the model in which founded learning takes place through access to new resources and the building of a society based on dialogue. Dialogue allows sharing of knowledge, reflection on issues, refinement of knowledge, and mobility between the role of learner and teacher. Such a position can be reviewed in terms of the concepts in distributed cognition and their relevance to network ba sed learning. Distributed cognition, as discussed by Karasavvidis (2001) in the introduction to this issue, encouraged the consideration of the system as involving more than the individual. For the trial domain of our framework this has particular resonance in considering people within the work environment. We could identify in our consideration of requirements existing networks of expertise, experience (for example, as processes are rolled out across different manufacturing locations), and formal and informal networks of colleagues. In the formation of the framework these were considered as supporting arguments for the approach. The authors also recognise the opportunity for more formal analysis using methodologies related to distributed cognition; this is discussed further when the experience is reviewed.

In finalising the framework the authors also drew strong reference from several other sources. The framework considers network-based learning systems to support social interaction and allow users to work on tasks that go beyond individual work. The social aspect of learning builds on the work of Vygotsky (1978) that recognised there is a social process that aids learning through discussion and reflection, which can be seen as internalised conversation. Laurillard's conversational framework (1993) also identified the importance of learners progressing through stages that encourage interaction with the teacher and with their own knowledge. Laurillard suggested four desirable stages in the conversational framework: (a) discussion; (b) adaptation; (c) interaction; and, (d) reflection. In particular, a model where guided discovery is used and the teacher and students become collaborators is suggested. Using tasks in the workplace offered a way to develop collaboration between the learners while weakening the imag e of a teacher in the system.

Mayes (1995) considered a three-stage model for carrying learners through the process of learning. In his model there was further emphasis on the value of dialogue as both a stage in the process and as the route by which learners can progress. In conceptualisation learners have to appreciate the background and concepts they have to learn, in construction they fit new knowledge to their internal understanding, and in application they see how actions can be built with their knowledge. Other labelling is possible for this structure but more important is to apply appropriate ways to carry people from one state to the other. Mayes' and Dineen's work then further addressed how this model can encourage learning from observing others as well as acting within the system (1999). This model is closest to that adopted here and identified two important aspects: that application is part of the process; and, dialogue is the most appropriate way to move from one level to another.

The view of learning supported by the adopted model can be summarised (derived from the work of Mayes) as:

* learning is a by-product of understanding;

* understanding occurs through performing tasks and getting feedback; and

* learning is socially driven and supported by dialogue.

Each of these has to be supported within a unified framework that offers:

* task basis for access to material;

* a resource based approach that supports the users to carry out the tasks; and

* support for dialogue with other learners and with tutors and experts.

The further practical aspect is also identified here in the need for a link between the tools that can support the learning activity of all forms: conceptualisation, construction, application, and dialogue. This provides users of the system (teachers and learners) with orientation so they know what they are doing, when they can do it, with what resources, and who else might be involved.

The framework previously identified has a foundation on theories of learning that are often considered in relation to child development and relationships with teachers. For our case we are also wishing to address adult learning within the work place. This means we should consider additional constraints on the time and place where learning is likely to occur and also additional resources in the form of colleagues and experts. These constraints were addressed informally in the construction of the framework.

The work of the project also built on other experience and practical considerations. Previous work by collaborators in the ASTEP project was assessed when preparing the user needs for ASTEP (McAndrew, Desmulliez, Flockhart, Schnurr, Tomes, Ferreira, Santiago, Ra, Strom, Smith, Gracie, Rorvik, Johnsen, & Mayes, 1998). At Buskerud College in Norway, Ra and Str[empty set]m examined a perceived problem in the use of separate lectures and practical sessions when trying to provide industrially relevant training. That work led to revised approaches to combine lecture material with a more practical base. This was studied over two years and showed that a learning by doing approach was seen by those involved as providing a good solution. Another partner in the consortium (Motorola) had carried out a strategic review of their education programme and concluded that a move to a resource-based approach supported by technology was essential to meet their internal targets. Thus within the consortium there was a strong view that the development of a task based approach to online learning would satisfy real needs.

Analysis of Tasks

The framework calls for the use of tasks to build an appropriate system to support learning. The tasks should be capable of analysis against the overall aims and objectives of the course and against the resources provided, with a mapping made from each task back to the learning objectives. The analysis need not be through computer analysis but through specification of validation criteria and an evaluation guide. In practice this is an important step; unless the tasks fit in with the framework they will not necessarily act in the expected way. For example, the profile of tasks is expected to include those based on dialogue and also linked to the processes that are being studied. The emphasis on processes comes from the domain studied and it may be appropriate to consider other forms of tasks in other domains (Table 1).

In the implementation within the project a checklist of desirable characteristics was determined in the specification phase of the project (MacKinnon, McAndrew, & Flockhart, 1998) This checklist was extended to include the use of additional resources as an aim. This was found to be an important benefit of the approach; building the construction process around tasks meant that the resulting course could easily incorporate new resources. Essentially the Web (or company intranet) is treated as an extended and open resource base providing access to a source of distributed knowledge. New resources can be found by the course author; by those supporting the course locally; or, by each individual student. Such resources are now frequently available for many subject areas and this proved to be the case within the semiconductor industry as, for example, chip manufacturers will make available information and simulations of new designs, or manufacturing practices will be published on a company's intranet. In the prototy pe, support for this activity was added through a tracking mechanism for sites visited associated with each task and users also had access to a notation system that could be used for sharing new sites and other information.

The set of characteristics includes information about users as well as about tasks. A profile for each user was created with the intention of operating in two ways. One use of the profile is to assist collaboration between users by giving useful information about location, aspects they are working on, and background. The other use of the profile was to record activity for later replay. In practice this was primarily used as a way to record additional web sites that were visited but consideration was also given to tailoring the presentation of tasks to the user. For the prototype, tailoring was rejected as there was only a limited range of material but by noting the category of user and the type of user for which each task was designed users could make judgements for themselves.

Course Model

The course model in the ASTEP project is task-centric, this means that all the material presented to the user, and all the actions that they take relative to that material, will be associated with learning tasks. The framework, therefore, considers tasks as the controlling mechanism for the user of materials. A model illustrating this task-based framework is shown in Figure 1.

This shows that the knowledge about a course, including the profile of users, is represented through a set of tasks. These tasks draw on a series of resources, such as those highlighted in Figure 1, which is available throughout the course. To complete a task the user uses the resources and works against an activity to meet the aims and objectives of the course and in doing so will adjust the profile of the user. The user profile could also be used to record the results of preassessments to suggest appropriate course material. The technology and task models are additional factors that inform and support the development, construction, and implementation of the tasks. The tasks are the outcome of the use of the framework to develop the learning environment for a particular domain or subset of that domain. As shown in Figure 1, the course knowledge, resources, and technology are guiding factors relative to any domain and hence determine the tasks to be undertaken by the users. To limit the requirement for specia list multimedia skills in the design of the course a simple classification of task models has been developed. Using this classification will help the consistency and repeatability of the design process. Appropriate media can then be added at a later stage.

A task in this framework is a clearly defined activity that helps the learner towards the goal of increasing their understanding. Tasks can range in scale and may contain sub-tasks that could be performed independently. This provides one dimension for the definition of the task. The other is provided by the category of a task this is related to its distributed characteristics as either an active task, reflective task, collaborative task, or dialogue task. These were defined fully in the project and dialogue tasks highlighted as having a special role in the framework as directly supporting the distributed nature of understanding and the belief that understanding can be aided more by dialogue than by standalone resource material.

In treating the two dimensions as independent Figure 2 is simple but powerful. For example, some of the tasks in the ASTEP system were also built using an experimental version of a simulations toolkit--the Multi-Verse toolkit (2001)--that could support sharing between the students as well as their own active work on the task. Even though such tasks may cross between categories, labelling the tasks in this way does encourage the consideration of each category when carrying out the process of instructional design. The description of each task was further developed into a data model. By recording information as metadata about the task it was possible to represent different sorts of tasks to users, and the metadata will also enable future tools to aid task reuse and construction. The use of such representation for learning objects is the subject of current standards work and investigation of ways to store and share information--for example, the IMS project (2000).

Implementation of the Framework

The previous section has outlined an overall approach for the framework used on the ASTEP project. This framework was then examined to be mapped onto a set of principles that it would display, then a set of tools and objects that needed to be supported and finally onto an implementation. The framework was implemented in two ways. First, as a set of principles within a manual and supported by a workshop, second, as a prototype system that provided a structured way to represent a course consistent with the framework.

The ASTEP project produced five modules. In every case the module was constructed according to the framework. Two of the modules using the prototype approach where the material and tasks are described in a structured description based on eXtensible Markup Language (XML) and then presented using either a client-server system or by parsing to active server pages and HTML. The other three modules used commercial environments such as TopClass (WBT Systems, 2001) and WebCT (2001) to host the materials but with task presentation and descriptions added to the base material. The framework does not exclude using such tools for implementation but does require further identification of the tasks and activities for the learner than typically happens. A two-pass process was used with conversion of material followed by refinement. This matched to the source for the materials from a separate project (INSIGHT/INSIGHT II) and also allowed progress in parallel to development of the framework. Applying the framework in this way leads to extra work on the part of the course designers as the material is adjusted to be suitable for active use at a distance.

The prototype implementation for the ASTEP framework needed to consider several areas of technology and make selections of methods for implementing each example. The overall construction of the prototype course was through marking up of the tasks and course material in XML. XML is a method for producing structured documents where elements can be assigned meaningful labels. XML is itself derived from Standardized General Markup Language (SGML), but simplified in syntax and intended to find more widespread use on the Web. To apply XML to the ASTEP course a Document Type Definition (DTD) was developed that mapped the task-based structure. No attempt was made to produce a complete description for all of the course material. Instead the standard XML DTD for HTML 4 was incorporated into the DTD. This enabled the bulk of materials to be produced using existing authoring experience and authoring tools. The additional markup described how the course fitted functionally and in terms of the activities that a learner wou ld carry out, and provided the necessary metadata for the tasks. Once a course was described in a way that conformed to the DTD it could be validated to be consistent with the required structure using software tools and then processed to produce the course as used by the learner. This introduces some flexibility as the processing of the course can be changed without making any changes to the underlying material. In practice this allowed different versions to be produced for the different network configurations identified within the ASTEP trial sites.

Trials of the prototype system were built around two courses in semiconductor principles and design. One course was based on a generic process available for use across the consortium and the other on a specific Motorola product for use in trials internal to the company. These trials have made use of the flexibility of presentation of material represented structurally by running the same material in different forms. These included placing the material into the database of the client-server system, where there was full communication back to the server; into Active Server Pages (ASP) within the company intranet; or, into plain HTML. There is some impact on the facilities offered (tracking of use was not possible in the plain HTML version) but the overall approach worked across each implementation. The way in which various features within the system were implemented could also be varied, for example different forms of asynchronous and synchronous communication were used depending on the local facilities. The fle xibility of presentation allows commitment to be made to the overall approach and structure almost independent of the presentation technology

In the framework, all material is considered as related to tasks. In the prototype system, this is matched by presentation to the student by way of tasks. Before looking through any learning resources, students will be presented with a task to consider that motivates the acquisition of information from those resources. Thus the student is directly aware of the set of tasks that they are being asked to complete in the course. Within the prototype system, students navigate through the course by moving from task to task. This differs substantially from most traditional courses, in which the student moves amongst chapters within a block of reference material.

At any point, the student is identified with the task that they are currently working on, as well as the reference material that they are looking at and using to complete that task. The view that the student has is shown in Figure 3.

Each task can have subtasks and related tasks can be grouped together in folders. Each task consists of an aim, description of the task, and the question itself that the student is being asked to address. The aim of a task is directly related to an identified learning objective. In the case of a subtask, the aim will identify a subset of information that is required to meet the learning objective of its associated task.

Using the Framework for Authoring

The framework itself was not prescriptive on the complete process of constructing a course. As part of the evaluation approach some of those involved in authoring material analysed the task development process. The following stages were identified as shown in Figure 4.

The approach shown in Figure 4 reveals that the structuring of tasks provided not only the emphasis on tasks within the course that was expected from the framework design, but also led to additional analysis by the authors of the aims and objectives of the course section. The use of a structured representation also meant that they felt able to validate the course. The way in which courses could be validated for eventual approval for accreditation in different locations was considered as part of the studies within the ASTEP project. The provision of a strong structure such as that used in the framework helped in that process. However the evaluation identified a need for more structured authoring tools to support the creation of courses and help enforce the model across areas such as assessment criteria.

Reviewing the Framework

The framework described in this article was successful in identifying and agreeing an approach that was then disseminated across a project consortium working to a strict timetable. The companion implementation also showed strengths in incorporating the elements of the framework in a workable and resilient way to produce a range of instantiations in different technologies. These undoubted benefits of the approach may well have occurred with a different theoretical base as the evaluation data supported the task-based structure but was less clear on aspects of collaboration. In reviewing the results from the evaluation it is perhaps at those points where we are following reasoning supported by distributed cognition and seeking collaboration and reference to others that problems were identified in some trials.

Notably in a trial within industry the preidentified model had been of workplace use at each user's own desk and it had been expected that the users would call on available tutors and collaborate wit their peers. In practice there was greater use outside of working hours and in a way that would not disturb colleagues. The onsite trainer discussed this with the users involved in the trial and discovered a view that using the system was not seen as appropriate in working time. She suggested that a minor change in the trial situation to encourage use of machines already associated with the learning resource centre might have corrected this attitude.

In the formation of the framework distributed cognition is used as a supporting argument for the structure adopted but, as stated earlier, no particular analysis was carried out using methodologies or tools from the field of distribute cognition. If a through study of workplace practice had been performed, weaknesses in the use of the framework in this trial may have been identified. Such studies are difficult tasks in themselves and were not seen as appropriate for this project where the overall target was an effective framework for learning rather than a study of particular aspects within it. Further work by Mwanza (2001) within the Open University, looked at related methodologies from Activity Theory and sought to identify a viable way to apply them to design of work-based systems, such as the creation of a knowledge-sharing environment (McAndrew & Taylor, 2000).

CONCLUSIONS

This article has described a framework that uses a sound basis for the approach to learning and recognises that there is a need to consider the social structure and active base for learning. This is not unique and this article extends earlier work in reporting an application designed for work place learning that takes a task-based framework and maps it to a way of producing online material that should have the desirable characteristics associated with the framework. In practice this is achieved through using a markup approach that enforces a structure based on task descriptions. The description of how this is implemented shows that the framework encourages thinking around tasks and dialogue but cannot enforce good process was to direct effort towards considering how learners wold work with resources, tools, and discussion while designing the course. This is consistent with adopting models where these aspects are considered to have priority over presentational issues.

The emphasis on learner tasks as described by this framework also provides a structure that can be reused as further resources are identified or developed elsewhere. Thus the course design can be based on learners' needs but make use of the growing range of resources that is available in all subject areas.

The development of the prototype system has provided some direction for future research to the authors:

* the use of markup for reuse in different formats;

* the automatic construction of portfolios of learner actions;

* the introduction of low effort flexible simulations;

* the development of learner-centred trails through existing materials;

* the development new assessment models and instruments;

* the development of tools and environments to assist with authoring; and

* the use of methodologies from Activity Theory to analyse situations.

Additional use of the system is also happening within Heriot-Watt University where the Learning Technology Service operates an Innovation Programme to assist lecturers in using new approaches to teaching and learning. Within that programme a tutorial on control systems has combined using the ASTEP prototype with the Matlab simulation package (The Math-Works, 2001). Small scale testing showed positive feedback about the task-based structure while highlighting some difficulties in presentation of different options. Trials within the ASTEP project are presented more fully in reports from the project.

The ASTEP prototype system has since been developed at Heriot-Watt University for use in a large scale online learning programme SCHOLAR (2001), providing complete first-year courses in science and engineering. The reason for continuing to use the system can be attributed both to its suitability for reuse and the appropriateness of the underlying models. Reasoning from the framework has also fed into subsequent European Union (EU) funded projects such as Pearl (Knowledge Media Institute, 2001) and GUARDIANS (Fretwell-Downing Group, 2001).

Acknowledgements

The ASTEP project was funded by the European Union (EU). The project consortium consists of the Department of Computing and Electrical Engineering and Institute for Computer Based Learning at Heriot-Watt University (UK), Motorola (France, UK), SensoNor (Norway), the National Microelectronics Institute (UK), Buskerud College (Norway), University of Porto (Portugal), Applied Materials (Germany, UK), and West Lothian College (UK). This paper particularly builds on work on the prototype carried out by Susie Flockhart, David Cole, Stuart Nicol and Michael Holligan from Heriot-Watt University and Thomas Wyatt from West Lothian College.

Figure 4.

The stages in creation of a course to match the framework

1. Structuring the Course

Identify course modules (folders)

Identify Learning Objectives (scope and content of each module)

Identify Tasks

2. Task Creation

Identifying Tasks from Leaning Objectives

Identifying Tasks from Resources Available

Identify Tasks from Tools

Form the Task Skeleton 

Author task with or without consideration of resources

Place task in context specific to task

Create, amend or reuse resources

Review description in folder 

3. Validate the module for sense, completeness and flow

Table 1

Analysis of Tasks Checklist


Tasks are specified and
 visible to users                X
Users are known and visible
 to other users                  X
Communication is supported
 between users matched to tasks  X
Resources support tasks          X
Activities are supported
 through resources and tools     X
As resources (tools and
 materials) become unavailable
 degradation is graceful         X
As resources are found they
 can be incorporated and used    X

References

Fretwell-Downing Group (2001). GUARDIANS: Gateway for user access to remote distributed information and network services, Sheffield, UK: Authors. [Online]. Available: http://www.fdgroup.co.uk/guardians

Heriot-Watt University (2001). SCHOLAR, Edinburgh, UK: Authors. [Online]. Available: http://scholar.hw.ac.uk/

IMS Project (2000). IMS learning resource meta-data best practices and implementation guide. [Online]. Available: http://www.imsproject.org/ metadata/mdbestv1p1.htm113

Karasavvidis, I. (2001). Distributed cognition and educational practice. Journal of Interactive Learning Research, 13 (1/2).

Knowledge Media Institute (2001). PEARL: Practical experimentation by accessible remote learning, Milton Keynes, UK: Authors. [Online]. Available: http://kmi.open.ac.uk/projects/pearl

Laurillard, D. (1993) Rethinking university teaching: A framework for the effective use of educational technology. London: Routledge.

Laurillard, D., Stratfold, M., Luckin, R., Plowman, L., & Taylor, J. (2000). Affordances for learning in a non-linear narrative medium. Journal of Interactive Media in Education, 2000(2). [Online]. Available: http://www-jime.open.ac.uk/00/2

MacKinnon, L.M., McAndrew, P., & Flockhart, S. (1998). Framework user manual: Deliverable D2.4 of project MM1001. [Online]. Available: http://www.cee.hw.ac.uk/~astep/deliverables/d24v12.pdf

MathWorks (2001). MATLAB 6.1 (software). Natick, MA: Authors. [Online]. Available: http://www.mathworks.com/products/matlab/

Mayes, J.T. (1995). Learning technology and Groundhog Day. In W. Strang, V.B. Simpson, & D. Slater (Eds.), Hypermedia at work: Practice and theory in higher education (pp. 21-37). Canterbury, UK: University of Kent Press.

Mayes, T., & Dineen, F. (1999). Developing tertiary courseware through capturing task directed discussions. Proceeding of ED-MEDIA '99 (pp. 1061-1066), Seattle, WA. Charlottesville, VA: Association for the Advancement of Computing in Education.

McAndrew, P., Desmulliez, D., Flockhart, S., Schnurr, C., Tomes, N., Ferreira, J.M., Santiago, M., Ra, O., Strom, T., Smith, S., Gracie, M., Rorvik, R., Johnsen, S., & Mayes, T. (1998). Analysis of user needs report, generic process: Deliverable D2.1 of project MM1001. [Online]. Available: http://www.ceehw.ac.uk/~astep/deliverables/ds2lv11.pdf

McAndrew, P., & Taylor, J. (2000). Organising for a knowledge network. Proceedings of Human Computer Interaction 2000 (pp. 99-100), Volume 2. BCS, Sunderland.

McKendree, J., Reader, W., & Hammond, N. H. (1995). The "homeopathic fallacy" in learning from hypertext. Interactions, 2(3), 74-82.

MultiVerse [Computer software] (2001). Bellshill, Scotland: Authors. [Online]. Available: http:/www.multiverse.co.uk/

Mwanza, D. (2001). Where theory meets practice: A case for an activity theory based methodology to guide computer system design. In M. Hirose (Ed.), Proceedings of INTERACT'2001: Eighth IFIP TC 13 International Conference on Human-Computer Interaction (pp. 342-349). Oxford, UK: IOS Press.

Rumelhart, D.E., & Norman, D.A. (1978). Accretion, tuning and restructuring: Three modes of learning. In J.W. Cotton & R.L. Klatzky (Eds.), Semantic factors in cognition. Mahwah, NJ: Erlbaum.

Vygotsky, L.S. (1978). Mind in society: The development of higher psychological processes (M. Cole, V. John-Steiner, S. Scribner, & E. Souberman, Eds. & Trans.). Cambridge, MA: Harvard University Press.

WBT Systems (2001). TopClass (software). Waltham, MA: Authors. [Online]. Available: http://www.wbtsystems.com/products/overview3.html

Web CT (2001). Most Popular e-Packs (software). Lynnfield, MA: Authors. [Online]. Available: http://www.webct.com/content/viewpage?name=content_popular_epacks.