Information Research, Vol. 2 No. 2, October 1996

Developing educational hypermedia applications: a methodological approach

José Miguel Nunes and Susan P. Fowell
Department of Information Studies
University of Sheffield, Sheffield, UK


Hypertext has played a role in teaching and learning since the mid 1980's when the first versions of HyperCard appeared. This extremely popular environment and authoring tool presented hypermedia as a technology capable of resolving some of the needs of learners and educators using instructional technologies. Drawing on work from computer science, educational studies and psychology, hypertext and hypermedia designers have since then endeavoured, with varying degrees of success, to establish hypermedia as a credible educational technology.

Hypertext systems have been described as providing a number of benefits for the learner, including: self paced, self selective learning; private learning allowing experimentation in a 'safe' environment; accommodation of different ability levels and types of learner; open access to information; reduced teaching costs; provision of reliable and timely help information; and reduced publication costs. However, despite its potential and some successful cases of its implementation (such as those presented by Ambron and Hooper, 1990), hypermedia technology is not being extensively used by schools and universities. The question that immediately arises is: why isn't hypermedia being used more widely, given the vast amount of interest and discussion about its perceived benefits?

In answering this question, researchers and champions of hypermedia may attribute lack of widespread usage to the inertia of educators and educational organisations. Focusing on the case of higher education (HE), hypertext failure could additionally be attributed to the secondary importance attached to teaching in comparison to research activities by the majority of academics (Darby, 1992) or to the Not Invented Here syndrome pointed out by Laurillard et al. (1993). This syndrome reflects the fact that academics are just about prepared to recommend a colleague's textbook, but would not take on video or computer-based teaching material developed elsewhere, as the teaching philosophy would be unlikely to match their own. This is an important point and should not be dismissed as being simply the protectiveness of teachers to their own teaching. In spite of large amounts of funding being directed towards national hypermedia projects, there are only a few examples where the resulting products have been widely adopted. The role of hypermedia seems to be restricted to hypertextbook type applications and to emerging forms of hypermedia now supported by the World-Wide Web. The predominant model for hypermedia applications continues to focus on the production of information resources and much less on the learning activities that the student/learner participates in.

The current situation shows an extraordinary discrepancy between educators' perceptions of the high value and potential of hypermedia as an educational technology and its real use. Such discrepancy cannot be justified exclusively by the factors described above. There are other more fundamental application design issues which need addressing. The design of hypermedia applications should begin with the production of a conceptual model, which represents the various aspects of the subject matter at different levels of abstraction. Additionally, such a design should follow a software development methodology, integrating the perspectives of the main agents involved in the development process: educators and software developers. Such a methodology must establish the educational requirements for the particular subject matter in the preliminary stage, and then, in subsequent stages develop the application in response to the requirements of the conceptual model and the educational specifications.

Failure to establish an appropriate conceptual model and use a suitable software development methodology, results in poor and ineffective applications. McKendree uses an analogy with the camcorder to characterise the current situation:

"It lets amateurs make movies about themselves which they and their immediate family and friends can enjoy. However, it is unlikely that you or I will want to rent it from the video-store and watch it. The professionals are much better able to design and make something, for a wider audience. [...] It is fine if some lecturers want to take time to hack together some on-line material for themselves and their students. They will probably have the pride and commitment to get them to use it. However, the material they produce will possibly not be as flexible or as widely applicable as something crafted professionally." (McKendree, 1994)

The prevailing conceptual model for educational hypertext applications is the hypertextbook model. This approach supports the traditional textbook approach and enhances its traditional use by offering a link structure to provide elaborations of key concepts. However, while providing some useful learning resources this use of hypermedia does not capture the real power of the technology. Therefore, new conceptual models must be developed to take advantage of the hypermedia philosophy and characteristics, as discussed by Nunes and Fowell (1996).

This paper proposes an hypermedia development methodology with the aim of integrating the work of both educators, who will be primarily responsible for the instructional design, with that of software experts, responsible for the software design and development. Hence, it is proposed that the educators and programmers should interact in an integrated and systematic manner following a methodological approach.

Educational hypermedia development methodology

Hypermedia is particularly appropriate for the production of interactive and exploratory educational applications, where large numbers of links and cross-references are provided and the learner can explore her/his own interests according to previous experience, background and perspective. To be effective, hypermedia applications need to be tailored to suit the particular learning tasks planned.

In order to do this, the learning process itself must first be analysed and understood. Learning is a complex process involving a large range of activities, some active, some passive, some creative, some reactive, some directed, some exploratory (Hammond,1992). Furthermore, as proposed by Nunes and Fowell (1996), academic learning should also be seen as the process of construction of knowledge and the development of reflexive awareness, where the individual is an active processor of information. This type of learning occurs through interaction with rich learning environments, and results from engaging in authentic activities, and by social interaction and negotiation.

This complexity of the learning process suggests the need for situated learning, social negotiation and multiple perspectives on the different aspects of the subject matter, the implication being that a number of different learning strategies must be adopted to assist the learner in the construction of knowledge. The adoption of these different strategies creates learning environments that Grabinger and Dunlap (1995) term Rich Environments for Active Learning (or REALs) [7]. REALs promote learning within authentic contexts, and encourage the growth of learner responsibility, initiative, decision-making, intentional learning and ownership over the acquired knowledge. Additionally, REALs should provide an atmosphere which encourages the formation of knowledge building learning communities that assist collaborative social negotiation of meanings and understandings among the members of the community (peers, tutors, and subject matter experts).

In sum, the REAL must essentially support interactions between the tutor, the learner and her/his peers, subject matter specialists and the learning materials. All these interactions may, or may not, be computer mediated. Furthermore, and as defined by Nunes and Fowell (1996), an educational hypermedia application is a software application specifically produced for a particular educational use, built using the hypermedia philosophy. They are developed to resolve a particular educational purpose or learning need, and are thus limited to the solution of the problems arising from that need. This means that although they might be linked with other hypermedia applications, other software applications, databases or even computer mediated communications facilities, they have clearly established boundaries. In this sense, an hypermedia application is one component of a much broader learning universe - the REAL in which they are embedded.

REALs can be seen as instructional systems, in the sense put forward by Nervig (1990): as sets of interacting, interrelated, structured experiences that are designed to achieve specific educational objectives, but organised into a unified dynamic whole. The design of an hypermedia application, as with any other of the other components, should hence result from the design specifications for the overall REAL. In turn, the design of the REAL results from the process of analysing curricular problems. To design and implement the overall REAL, instructional systems design (ISD) should be used. The importance of this overall ISD rests in assuring that the whole REAL is implemented using the same learning theory. In fact, if not carefully planned, the REAL could result in a mix of eventually conflicting techniques from different theoretical perspectives.

Accordingly, Bednar et al. (1992) defend the notion that effective instructional design and development is only possible if it emerges from deliberate application of a particular theory of learning. Furthermore, the developers must have acquired reflexive awareness of the theoretical basis underlying the design. This will ensure that instruction design, hypermedia design and development, and the hypermedia conceptual models selected are compatible and all use the same learning theory philosophy.

This paper addresses the design and development of hypermedia for higher education (HE). Academic learning is here defined as an active process in which meaning is developed on the basis of experience, in accordance with the constructivist theoretical frame (Nunes & Fowell, 1996). So, to develop hypermedia applications in keeping with a constructivist approach, it is important to have an understanding of the kind of specifications that will result from constructivist instructional design.

Constructivist instructional design

Traditionally, ISD is seen as a process approached from a systems strategy, based on the purpose of the system, using a systematic, data-based process for analysing curricular and instructional problems in order to develop tested, feasible solutions (Nervig, 1990). Conversely, constructivist ISD focuses on the learner and on the learning process rather than solely on the subject matter. Since knowledge is constructed, the learning of a concept must be embedded in the use of the concept. In the traditional ISD, the designer analyses the conditions which bear on the instructional system in preparation for the specification of intended learning outcomes (Bednar et al. , 1992). Content, learner and instructional setting are analysed and the instruction is designed using the concepts of learning objectives and specification of goal outcomes. Constructivist ISD requires the separation of method and content, instructional designers develop learning environments rather then packaged instruction (Kember, 1991).

According to Lebow (1993), this constructivist ISD should be carried out while bearing in mind the seven primary constructivist values: collaboration, personal autonomy, generativity, reflectivity, active engagement, personal relevance and pluralism. From these principles Lebow draws a set of general design principles to be used in the ISD process:

make instruction relevant to the learner by providing a context for learning that supports both autonomy and relatedness;

balance the tendency to control the learning situation with the desire to promote personal autonomy;

support self-regulation through the promotion of skills and attitudes that enable the learner to assume increasing responsibility for the developmental restructuring process;

increase emphasis on the affective domain of learning, treating learning and motivation as part of a unified whole process;

strengthen the learner's tendency to engage in intentional learning processes, especially by encouraging the strategic exploration of errors.
( Lebow, 1993)

Although using a traditional ISD model, as shown in Fig. 1, these design principles lead to a development based on the constructivist philosophy. Since knowledge domains are not readily separated in the world, according to this philosophy, information from many sources bears on the analysis of any particular subject matter and it is not possible to isolate units of information. A central core body of information must thus be defined in the analysis phase, but boundaries of what may be relevant should not be imposed. Instead of dividing the subject matter into logical analysis of dependencies, the constructivist approach turns toward a consideration of what users of that knowledge domain do in real life contexts. The ultimate goal of this approach is to move the learner into thinking in the knowledge domain as if they were an expert user of that domain (Bednar et al.,1992 ). Hence, designers should instead identify the variety of experts on the subject matter and the tasks they do. The designer should than define simplified but still authentic tasks to be experienced by the learner. The goal is to portray authentic tasks, not to define the structure of learning to achieve the tasks, since it is the process of constructing a perspective or understanding that is important and no meaningful construction is possible if all relevant information is prespecified (Bednar et al.,1992 ).

Once identified in the analysis phase, tasks must be designed so that they are situated in real world contexts, are authentic, and provide multiple perspectives on the subject matter. Additionally, some degree of coaching or guidance must be provided, by including meaningful examples and the different perspectives of experts and peers. A central strategy for achieving this is to create collaborative learning environments where both face-to-face and computer mediated communication are available. Access to extra information sources must also be provided to allow different learner's needs to be satisfied whenever needed.

It is in the development phase that all the components of the learning environment are implemented according to the specifications coming from the design phase. Since different types of educational technologies may be needed, to implement all the planned tasks, examples and communication channels, different development methodologies may then be applied. If hypermedia applications are needed for a particular instructional task, then a specific hypermedia development methodology must be used, where the specifications are established during the design phase of the ISD. However, no task is isolated, but rather in a REAL it forms part of a larger context. Hence the testing and evaluation of all components of the learning environment must be done in an integrated manner. The hypermedia application must be system tested and field trialled as an embedded component in the overall learning environment.

In summary, in a constructivist approach, the analysis phase of the ISD establishes a core body of information crucial for the subject matter, identifies the type of experts that use it and the tasks they perform. The design phase specifies a comprehensive set of authentic tasks and the coaching and support to be given to the learner. It also must specify the educational tools required and the function of each one of these tools. Therefore, if an hypermedia application is needed, its specifications arise from the design phase. During the development phase, the different educational applications and tools are developed in parallel and then system tested and field trialled together.

Only educators, instructional designers and educational psychologists are involved until the design phase. In the development phase software specialists may be required. Thus, the development of these hypermedia applications requires software engineering methodologies that allow real co-operation between the two types of agents involved: the educators and the software experts. A software engineering discipline is emerging (Nodenot, 1992), called Educational Software Engineering (ESE), which uses methods mainly applied by educators with an educational point of view but which also encompasses principles of good software engineering. This paper proposes such a methodology specifically for hypermedia application development.

Hypermedia development methodology

The production of educational hypermedia involves collaboration between subject matter and education experts involved in the ISD, and hypermedia development experts involved in the application implementation. Hence, the communication between these agents becomes paramount. These groups usually speak different "languages" and do not readily understand the problems of the other (Moonen, 1986). An efficient educational methodology must thus integrate and support the dialogue between these different groups.

The software development methodology that best supports this requirement is the rapid prototyping approach. A rapid prototype is a simplified and untested equivalent of the actual application, performing all the basic functions specified for the final product (Howell, 1992). As shown in Fig 2, by implementing a prototype first, the hypermedia designers are able to put forward a fully functioning application presenting all the basic features of the final product such as user-interface, link structure and coaching facilities. This is not a diagrammatic approximation or representation, which tends to be looked at as an abstract thing, but an actual implementation of the specifications for the application. These prototypes can be realistically tested and assessed and rapidly changed in an iterative manner until consensus is reached. Evaluation and testing of these prototypes must be done by instructional designers and ideally include pilot tests using target learners.

Furthermore, hypermedia applications are inherently different from other software applications. The volume of actual code produced in scripts is relatively low and emphasis is put in user-interface design, link structure design and definition of contents entry as the different multimedia components. These characteristics along with widespread availability of authoring tools, make it possible for rapid development and testing of prototypes.

The Rapid Prototyping Software Development Cycle presented in Fig. 2, reflects the traditional rapid prototyping philosophy. The Specifications are established during the overall ISD process for the REAL where the application should be integrated. Consequently, as discussed above, the System Test must be done along with the testing all other components of the REAL and as part of its ISD cycle.

After the acceptance of the prototype, a thorough Design of the user-interface, link structure and multimedia integration must be undertaken and documented. To assure that the design is done using the constructivist approach, and therefore is in accordance with the overall philosophy chosen for the REAL, a constructivist conceptual model for hypermedia applications must be adopted. Nunes and Fowell (1996) discuss such a model.

The design phase is followed by Implementation using appropriate authoring and multimedia development tools. The implementation must then be subjected to a comprehensive Test. This testing must of two kinds: technical and instructional. The technical test aims at detecting and correcting technical problems, insufficiencies or inadequate use of the different media. It should be performed mainly by technical experts. The instructional-oriented test must establish if the developed application complies both with the educational philosophy adopted and the initial specifications. It should be performed by the subject matter experts and instructional designers.

Finally a release version of the application must be produced and handed over to the ISD for further system testing. This final version should be accompanied by complete documentation both on-line and paper-based, including user manuals if necessary. If any problems are found, new prototypes are built and the process repeated.

Like any other software methodology, the life-cycle of an educational hypermedia application should not stop after the implementation and hand-over. It should be maintained and continually improved. As suggested by Thomas, problems reported should be acted on at the earliest opportunity and feedback should be actively sought from learners and teachers (Thomas, 1994).


Implementing hypermedia educational applications means much more than just designing a few screens and specifying their sequence. Today, such an approach is not sufficient to support effectively support the learning processes envisaged in constructivist, collaborative or experiential learning philosophies. This paper proposes instruction as the act of supporting the construction of knowledge on a particular subject matter, by improving the learner's ability to use the content domain to carry out authentic tasks, and by providing these tasks with the tools needed to develop the skills of constructing an informed response and for evaluating alternative responses. Hypermedia applications are one such tool available to the instructional designer, which can be used to support constructivist, collaborative and experiential learning.

Understanding hypermedia as an educational technology, and its role within educational practice, is the key to the development of successful learning environments. Moreover, the way to prevent a backlash against the use of this educational technology lies in recognising both the technical and pedagogic components of instructional design and integrating them in a methodologically coherent manner. Rapid prototyping is an ideal approach which facilitates the integration of the different agents in educational software development, the subject matter experts, the instructional designers and the software developers.

However, the gap between expert/professional and non-expert/non-professional developers is narrowing, due to the increasingly more comprehensive and easy-to-use authoring facilities of the modern hypermedia authoring tools. Current authoring tools aim to support both professional quality and do-it-yourself endeavours, so that the developer of an hypermedia educational application is now often the educator her/himself. Nevertheless, the need for an adequate conceptual model and a comprehensive design process should always be present.


How to cite this paper:

Nunes, Josť Miguel Baptista & Fowell, Susan P. (1996)  "Developing educational hypermedia applications: a methodological approach" Information Research, 1(1) Available at:

© the authors, 1996.

Check for citations, using Google Scholar


Web Counter
Counting only since 13 December 2002