The great majority of information made available on the Web is not specifically prepared for learning purposes. Rather, it is more often developed with a view to interest and inform a general and varied audience. It is in this respect akin to public communication products and less aligned with instructional products, even though it often finds itself at the interface of both realms. From this ambiguity arises the design problem for web-based learning materials. This paper highlights some of the principal design issues involved and proposes the outline of a design methodology that can help developers in their work of preparing web-based learning materials. It is in this respect but an initial effort, but may well encourage further thinking in this important area of Web development.
Two perspectives in evolution
The Web can be seen from two perspectives: as a vehicle for the distribution of resources, and as a medium of expression-representation with its own specificity. The distribution perspective emphasizes the Web as a network infrastructure that has the great advantage of flexibility and low cost. A simple example is the case of university course materials (syllabus, readings, lecture notes, etc.) being placed on the Web and made available to students. The fact that Web information is readily updatable and that users always access the latest version provides tremendous flexibility by eliminating the outdatedness of printed materials; and the electronic nature of the medium provides cost savings compared to traditional paper-based alternatives.
What are being distributed at the moment on the Web are multimedia documents. However, the advent of Java and other similar languages enable true interactive instructional transactions on the Web and will likely totally reshape the field of instructional software. What this means for the design of such software and how it will affect the larger picture of web-based learning is of tremendous importance to web usage for learning. .
Instructional software, principally through its corrective feedback functionality, guides the learner's gradual understanding or skill development. While knowledge development and production has up till now been the main area of learning on the web, interactive software can well shift that to a wider range of outcomes, including more intensive and prolonged forms of learning like those just mentioned. Instead of snippets of information, we can now consider full modules of skill development as one of the main components of web content. And the fact that interactive modules are accessible online may well reconfigure the production-distribution of learning software, in the same way that information production-distribution has been reshaped by the web as a social vehicle.
The expression-representation perspective emphasizes the Web as a novel, distinct form of publication of informative materials that does more than simply replicate other forms. It is not so much the actual Web documents (Web pages) that distinguish this perspective, but rather the context of usage (rapid and rich access, hypermedia linking, etc.). A crucial question for the field then is how we capitalize on this rich environment by designing optimal Web documents for learning purposes?
It is the dynamic nature of the web that is of importance here. Web documents are generally not static over time: they have a tendency to evolve, both in their content and in their linkages with surrounding documents. There is thus an 'event' dimension to the web that makes it a moving target as a resource for scholarship and possibly for any kind of formal learning (in an institutional, assessable sense). Its flexibility adds to its richness, but may well lead to redefining formal educational activities. Once again, we see the web as a particularly useful vehicle for informal learning rather than for more structured educational efforts, despite very true possibilities there as well.
An important issue that underlies design considerations is that of the directiveness of the learning materials. Under another guise, this issue of learner control has been with the field of computer-based instruction ever since its inception (Duchastel, 1994). On the one hand, some forms of instruction are very didactic and strongly guide the learner to the achievement of very specific learning outcomes. Formal instructional systems design (ISD) is the appropriate methodology for this approach (Gagné et al., 1988). Along the way are forms of instruction that create a context for guided exploration, often supported by more advanced intelligent CAI systems; here, the design methodology of choice may well be one derived from an artificial intelligence software design context (Duchastel, 1991). At the far end of the spectrum, we find fully autonomous learners guided more by their interests and personal learning goals (the situation of informal learning on the Web). It is in this realm that we see a strong need for a design methodology that is adapted to the particular style of interaction that the Web generally entails (Duchastel, 1990; Nielsen, 1995).
Learning context is central to adapting a design process to the web environment. In particular, the issue of level of learning is crucial. In traditional learning contexts, the level of difficulty is carefully controlled through adaptive means so that learning processes are optimized. Simplification and structural support (sometimes called intellectual scaffolding) are instructional means used to achieve this optimum level of learning. These, however, are generally not present in a typical web context, with the consequence that greater cognitive demands are put on the user-learner to make the best use possible of the information resources available.
Intellectual scaffolding is the core of information design for instructional purposes, reducing the complexity of a domain to fit the level of the student and providing support for the novice to navigate appropriately within the potentially overwhelming nature of all the new information. The support provided must address 3 requirements for learning to be effective in the long run. First, it must captivate the learner, keeping attention focused on the learning and not subject to waning or distraction (the latter being a particularly troublesome possibility in the web context). Second, it must address understanding through gradual shaping of the mental models that a learner constructs to structure and thus handle the information complexity one is presented with. And third, practice must be asked for to solidify the information and skills one is shaping and thus ensure long-term retention. These can be considered the imperatives that can guide design for learning on the web.
Personal context: Effort and interest (the effin factor)
The notion of an information-rich learning context greatly affects traditional forms of learning and instruction. An information-rich context means not only one in which a rich store of information exists, but more importantly, one in which there is easy access to that information. The reason the printed resource is so appealing today as a learning resource is quite simply that it is very readily available.
It is useful to ponder the notion of an ever-increasing information-rich learning milieu, for there is something at once fascinating and suspicious about the possibility that merely increasing information accessibility will lead to more and better learning.
Let's start by considering the current situation. Until recently (and even today), the typical student may not have had instant access to all desired information, yet there was a tremendous wealth of information available both in a home's encyclopedia and in the local public library. However, despite this vast wealth of information in the student's educational context, there was relatively little use made of it (certainly much less use than one might initially expect). Why is this ? The answer lies in the relationship that exists, at a psychological level, between personal effort and interest, what can be called the effin factor (an effort-to-interest trade-off).
The effin factor concerns the appeal or lack of appeal of some activity or resource. It involves the effort a learner is willing to invest in pursuing an activity or in accessing a resource, in relation to the learner's current intrinsic interest in the topic. The effin factor deals with the motivation of the learner, but with intrinsic motivation only (Berline, 1960), one related largely to curiosity (because of its relation to knowledge, this is known as epistemic curiosity).
The effin factor is at the center of learning technology. It builds on our natural curiosity about the world around us, and on the potential of technology to make information easy to obtain. We are rather flighty in our curiosity and our attention keeps shifting as we encounter new and interesting things in our dealings with the world. It is important therefore that the satiation of our appetite for information occur at the time of need, otherwise there is a good chance that interest will be lost. Hence the importance of immediate availability of the needed information. Any delay increases the mental effort in keeping the interest alive.
The effin factor embodies a straightforward relationship: the more the effort needed to obtain the needed information, the more that topic must be of interest to the learner. If the required effort is great, only topics of very high interest will be pursued. On the other hand, if the required effort is minimal, then many topics become of interest. As web technology reduces the effort to access information, the potential for interest in many diverse topics grows for individuals. Natural curiosity has at last a chance to blossom.
The central idea developed here is that information is becoming readily accessible in a variety of compelling forms, and that this information-rich context can dramatically affect how students and people in general go about learning. From that notion follow implications for how we design instruction, that is for how we develop learning environments.
An initial design model
The design model presented here is framed by the need to interest the user on a continual basis. This can be thought of in terms of provoking learning, that is enabling it to spontaneously happen in the context of the user's interaction with web resources as this interaction takes place in a somewhat un-planned and perhaps even haphazard way.
The design model is couched within a standard software development life-cycle process based on a general systems approach. Thus, various elements of requirements specification conducted up-front and prototyping, evaluation, construction and support at the back-end help ensure a high level of quality of the products designed following the model.The model's core specificity, however, revolves around 2 central design processes dealing respectively with information specification / representation and problem-solution tasking. These are considered below.
Information specification and representation are two crucial design facets in determining the content to place in front of the learner, that is what to provoke the learner with! Specification involves the what of content determination, while representation involves the how of it. While still being very much an art (captured in talented authorship), these information decisions can be guided at a general level by theoretical perspectives such as information mapping (Horn, 1989) or minimalist design (Carroll, 1990). Structural approaches are also very pertinent in this arena (e.g. Jonassen et al., 1993), as are all the precepts one may adopt from the wide world of multi-media design.
Problem-solution tasking, likewise, remains a careful art in determining the difficulty level of the challenges proposed (or viewed from the opposing angle, in terms of capacity of accepting the challenges). This is the element that activates learning and that keeps in on track, both conceptually and, to a large extent, emotionally. That this element can be truly capitalized on for web learning is becoming more recognized, we believe (see for instance the model proposed in Duchastel (1997).
No design model is universal and it is important to situate the present model in terms of appropriateness. Specifically, how appropriate the model may be for different types of learning needs to be examined specifically for procedural, declarative, and cultural learning. These are somewhat arbitrary ways of carving up the learning pie, so to speak, but nevertheless useful. Procedural and declarative learning relate respectively to skill development and to knowledge assimilation. Both would seem to be forms of learning that could be facilitated by materials developed through a model such as the one presented above.
Cultural learning, on the other hand, is a class of learning that joins informality with broad, wide-ranging exploration and which may well therefore be more in tune with a less structured design approach altogether (more in line with what we witness on the web at the current time, before the interactive element comes strongly into play.
It remains difficult to determine applicability at this early stage of web development, however. As learning materials will be developed for the web, model applicability issues will surface with added importance.
The web context provides a novel setting for learning that is different in a fundamental way from its predecessor non-information-rich setting. It is important to keep in mind that much learning-related decision-making that goes into the design of more traditional learning products involves considerations of efficiencies, such as keeping things reasonable in terms of costs and practicality. As the web removes some of these considerations (even though not entirely) and as it opens new playing fields with frontiers such as virtual reality and at-a-distance interactions, new thinking is called for in making full use of the electronic environment for learning purposes.
This new setting invites reflection and experimentation with new design processes that will facilitate learning from resources to be found in that chaotic and expanding environment. Design may well, we believe, be a blossoming domain of examination and creativity in the next few years.
Berline, D. (1960). Conflict, Arousal, and Curiosity. New York: McGraw Hill.Carroll, J. (1990). The Nurnberg Funnel. Cambridge, MA: The MIT Press.
Duchastel, P. (1990) Examining Cognitive Processing in HyperMedia Usage. Hypermedia, 2, 221-233.
Duchastel, P. (1991) Towards Methodologies for Building Knowledge-Based Instructional Systems. Instructional Science, 20, 349-358.
Duchastel, P. (1994) Learning Environment Design. Journal of Educational Technology Systems, 22 (3), 225-233.
Duchastel, P. (1997) A web-based model for university instruction. Journal of Educational Technology Systems (in press). Http://www.nova.edu/~duchaste/unimodel.html
Gagné R., Briggs L. and Wager, W. (1988). Principles of Instructional Design. (3rd ed.) New York: Holt, Rinehart and Winston.
Horn, R. (1989) Mapping Hypertext. Lexington, MA: Lexington Institute.
Jonassen, D., Beissner, K. & Yacci, M. (1993) Structural Knowledge: Techniques for Assessing, Conveying, and Acquiring Structural Knowledge. Hillsdale, NJ: Erlbaum.
McAdams, M. (1995) Information Design and the New Media. Interactions (October 1995), 39-46.
Nielsen, J. (1995) Hypertext and Hypermedia:
The Internet and Beyond. Boston: Academic Press Professional.