Designing Hypertext Multimedia Educational Software

Don Lehman
Department of Medical Technology
University of Delaware
Newark, DE 19716

I. INTRODUCTION

Before undertaking the creation of hyperlinked educational multimedia, the author needs to have a design plan. Creating without a design plan is like building a house without blueprints. Besides considering the content, the author must make decisions on the user interface, the type of media used (e.g., text, pictures, audio, video), and when and where to apply hyperlinks. The cognitive processes and experiences of the user should also be considered. The author has control over the content and the links, but the user's discretion determines the sequence. This article discusses issues of designing to facilitate the user's access to the content in educational software.

II. MULTIMEDIA DESIGN

A. Historical Context
In advanced courses, students are expected to master complex concepts and transfer acquired knowledge to new situations. Understanding must go beyond the presented information. In order  to comprehend text or a concept, students must construct the meaning in terms they personally understand. The presented information must be combined with other information (most prominently the prior knowledge of the learner) to form an adequate representation of the meaning. Extra cognitive processing to integrate new information with prior knowledge is sometimes called elaborative processing. Different instructional media can encourage elaborative processing, but because hyperlinked multimedia incorporates a variety of media, it can facilitate elaborative processing [1].

What is hypertext?  Theodore Nelson, the man who coined the term hypertext, was quoted as saying, "By hypertext, I mean nonsequential writing - text that branches and allows choices to the reader, best read at an interactive screen" [2]. Landow adds that hypertext is a medium linking verbal and nonverbal information. Hypertext is more reader-driven than printed text: readers are able to access information that is relevant to themselves.

In traditional textbooks, learners are expected to read through the information in a manner designated by the author. With hypertext, learners have more control on how the information is presented. Learners can follow links and navigation buttons created by the author. It is up to the author to create these links in a logical organized fashion to prevent the user from getting distracted and confused.

It has been assumed that learner control is an important aspect of effective learning, and that is the purported benefit of hypertext. Studies have found that students with larger quantities of learner control rate the instruction more favorably [3], [4]. However, the more favored instruction does not necessarily translate into more efficient learning. In addition, empirical findings yield mixed results with respect to the learning benefits of learner control compared to programmed-control instruction [5], [6], [7]. The majority of these studies were done several years ago; in the meantime, computer technology has become more powerful and made it easier to create hypertext applications. New learning theories have been presented to guide authors in creating hypertext applications, and thus new studies should be conducted to evaluate the contributions of these theories.

Do multimedia and computer assisted instruction (CAI) improve learning?  When compared to traditional instruction, CAI improved student scores and attitudes toward learning, and decreased learning time [8], [9]. In addition, computer technology may improve academic achievement, motivation, and time on task [10]. If a program is well designed, students will want to spend time learning, and their scores should improve.

However, not everyone feels CAI is an effective teaching medium. Clark [11], [12] claims that media do not influence learning. He feels any perceived advantage of CAI can be explained by other hypotheses: 1) content differences between instructional methods with different media were not controlled and, therefore, the medium was not necessarily the cause of any significant effect; 2) because the learning method was new, it will automatically be more exciting and the novelty effect will only temporarily improve learning, and; 3) it is the method of instruction that fosters learning, not the medium.

B. Learning Theories
Spiro [13] uses the word "ill-structuredness" to describe conceptual complexity and case-to-case irregularity in knowledge domains. In order for students to understand a difficult case, they must appreciate the complex interaction among several concepts. Spiro argues that all domains involving the application of knowledge to unconstrained naturally occurring situations, or cases, are substantially ill structured.

Many learning theories stress the importance of retrieving organized packets of knowledge, or schemas, from memory to organize presented information. Spiro argues that conceptualization of ill-structured domains renders the use of prepackaged schemas inadequate. Knowledge will have to be used in too many different ways for them all to be anticipated in advance. Therefore, emphasis must be shifted from the retrieval of intact knowledge structures to the construction of new understandings, to the situation-specific assembly of prior knowledge drawn from diverse pre-existing mental representations. In other words, instead of retrieving a previously packaged solution from memory, one must bring together from various knowledge sources an ensemble of information needed to understand or solve the problem at hand.

Cognitive Flexibility Theory (CFT) was developed to teach ill-structured domains. It refers to a particular constructivist theory that integrates learning theory, mental representation, and instruction. The basis of CFT is allowing students to revisit the same material at different times, in rearranged contents, for different purposes, and from different conceptual perspectives. For full understanding, content must be covered more than once. A single explanation of a complex concept would miss salient knowledge facts that could be important in a different context. Thus, simply repeating a process is not sufficient; that would be oversimplification. However, re-examining a case in the context of comparison to another case can lead to new insights.

CFT recommends the use of "landscape criss-crossing" for instruction [14], [15]. The content is re-edited to produce a particular kind of criss-crossing of the conceptual landscape.  The conceptual landscape needs to present a large set of case examples of a particular conceptual structure being taught. In this way, the learner can see a range of conceptual applications of the case. Hypertext and multimedia can support the use of the CFT.

III. INSTRUCTIONAL DESIGN

A. User Interface
The user interface controls how the user interacts with the software. Ebersole [16] defines the interface as where two worlds meet--in this case humans and computers. The Windows and Macintosh operating systems have a graphical users interface (GUI). Multimedia with hyperlinks can produce an interactive interface, requiring the user to be more active than traditional media such as books and television.

Cognitive overhead for users requires keeping track of hyperlinks, where they are in the software, and how to get back to a place they have been. Too much cognitive overhead can have a negative effect on learning. Thus, for example, navigation should appear effortless to the user. Hyperlinks can make educational software appear fragmented, adding to user confusion. To increase coherence, some authors recommend placing links at the end of blocks of texts or in side bars [16]. Another way to minimize the appearance of fragmentation is to provide the user hints or clues as to where hyperlinks will lead. This will give users more information before they decide to follow a link, so they can decide if following the link is worth the journey.

It is important to provide users with a variety of easy-to-use and understandable navigational buttons to prevent them from getting lost and to minimize cognitive overhead. Consistency can help accomplish this. Authors should place similar buttons or links in the same location on different screens. In addition, the same actions should result in the same effect. Other techniques to minimize cognitive overhead include the use of back buttons, maps, and bookmarks.

To increase the user's understanding of the interface and navigational tools, an appropriate metaphor for navigation should be incorporated. Commonly used metaphors are books, a desktop, travel, and stack of cards [16]. These metaphors bring familiar real-world concepts to the complexity of hyperlinks.

B. Dual Coding
Multimedia has the advantage of presenting material in different ways, and some media communicate specific information better than others. For example, text appears better than sound for communicating verbal information [1]. Pictures generally help people learn more effectively than text, except when items are conceptually similar or if items are presented too fast for learners to create verbal labels. However, pictures may be limited when communicating abstract ideas.

Multimedia combinations may help users learn by processing information through more than one channel. This is termed dual-coding. For example, encouraging learners to use both verbal and pictorial channels appears to be an effective instructional design. Evidence exists that verbal and picture information should be presented together [1]. Students seem to perform better when textual annotations were combined with drawings.

Some multimedia authors believe that pictures improve learner interest and therefore learning; however, authors must avoid the indiscriminate use of images. It appears that adding unrelated pictures does not improve learning, and may in fact decrease learning [1]. Unrelated pictures may be a distraction from the content and intended curriculum.

C. Hypertext Structure
As mentioned above, the overall structure of hypertext can affect learner outcomes. In fact, disorientation may be the major limiting factor of hypertext [17], [18]. The problem seems to result in a measurable decline in performance. Disoriented users may encounter problems deciding where they want to go and how to get there.

McDonald and Stevenson [19] examined the effects of different hypertext topologies and prior knowledge on navigation performance and user disorientation. They examined hierarchical, nonlinear, and mixed topologies. In hierarchical text, the information nodes are linked in a hierarchical fashion in which a node at one level can access only the nodes directly above and below it. In nonlinear (network) text, the nodes form a complex network of connections based on a large number of referential links. Mixed text is basically a hierarchical structure with a number of links allowing users to jump to other branches of the hierarchy.

In both browsing and navigation, mixed hypertext produced the best results, followed by hierarchical, with nonlinear producing the poorest. Overall, knowledgeable participants performed better than nonknowledgeable participants; however, there was no difference between knowledgeable and nonknowledgeable participants with mixed hypertext. Navigation results suggested that mixed hypertext provided the best mixture of freedom and constraint.

The advantages of nonlinear structure are claimed to be 1) information is more readily available to the reader, and 2) network structure allows nonlinear access to information [19]. Readers, therefore, have increased control over the sequence of information. However, this increased control may have negative consequences if users are unable to navigate around unfamiliar and complex information without experiencing disorientation.

Disorientation with hypertext structure may be decreased if a user has prior knowledge of the subject matter. Conversely, disorientation may be heightened in novices [20]. Compared to novices, more knowledgeable users may experience fewer navigational problems in hypertext environments because they have a greater understanding of the conceptual structure of the subject matter, allowing them to impose structure on the hypertext [19].

D. Random Access Instruction
The instructional theory derived from the CFT is termed Random Access Instruction. Hypertext computer applications are well suited for the criss-crossed instruction of the CFT. However, implementing CFT is not just using the computer to connect everything with everything else. The learner could become lost in a confusing labyrinth of incidental connections.

It is important that only those cases and parts of cases pertinent to the focal conceptual structure be presented. Students need to see these examples in a close time frame. Examining cases separated by long time periods is not an efficient way to learn complex concepts. The computer with hypertext is well suited for this task. Numerous examples can be programmed for students' immediate access.

Providing background information on the contexts being explored is another important aspect of CFT. This information needs to be functional and context sensitive. Just providing dictionary definitions, which are subtypes of a word's meaning, is not adequate. Particularized definitions providing supplemental guidance about the way meaning is used in a particular situation are needed. Particularized definitions are functional and context (case) sensitive representations of a concept. Abstract definitions fail to cover conceptual meanings used in ill-defined domains; therefore, supplemental guidance about the way meaning is used in a particular situation is required.

E. Supported Text
It has been estimated that between 10 and 12 percent of students in English-speaking countries cannot read sufficiently to successfully acquire content-area information [21]. Content literacy is the use of reading and writing to acquire new knowledge in a given domain. Reading environments are needed that support content literacy. Electronic text designed to promote improved comprehension (usually by inserting a variety of text enhancements or electronic resources) is called supported text.

There are three major components to supportive text: presentation, keys, and resources. Presentation is the content and the system that presents it. The system consists of the physical artifacts and the interface that operates them. With electronic documents, the system is the computer monitor, mouse, and keyboard. Computer-based text presentations reproduce and transform or expand the function of printed text. The computer version has the advantage of giving students more control over their learning. They can proceed in a nonlinear fashion in a manner that best suits their learning style.

Keys are problematic words, phrases, sentences, or paragraphs that students encounter as they read the text. Keys may be embedded in other keys. Students with different prior experiences and reading ability will find different parts of the text problematic. Authors must try to anticipate keys and provide students with resources in comprehending their meaning. The keys need to be identified in some manner so students know when resources are available. In computer software, keys can easily be underlined, or displayed in a different color.

The resources may be translational, where the text is rewritten into simpler language. They may be content vocabulary, where crucial technical terms and phrases are defined. They may also be references to other text, figures and materials. Resources should be readily accessible so that they do not disrupt the learning process. The use of hypertext and mouse-overs can meet the requirements of resources in supported text.

F. Using Learning Objectives with Hypertext and Multimedia
Effective use of hypertext systems requires a sense of purpose--that is a goal or focus--while reading the hypertext. Posing a question or problem to be solved by learners can help focus learning in environments with large amounts of learner-control. Alternatively, the addition of questions at the end of content nodes can get students to think about what they have just seen.

Research suggests that getting users engaged in problem solving tasks make hypertext systems more efficient [22]. In the absence of learning objectives that encourage readers to organize the text, readers using unstructured hypertext may not put forth the effort necessary to create a complete representation of the content. 

IV. CONCLUSION

The constructive processing of knowledge for transfer must be taken beyond the retrieval of knowledge structures from memory to include flexible situation-specific assembly of the background knowledge by the learners themselves. In addition, flexible learning environments are required for learners to develop problem-solving skills. Hypertext computer systems are ideal for creating these types of environments.

Authors must be careful when designing hypertext applications. Learner control is a double-edged sword. As the complexity of a program increases (such as more navigational links), so does the potential for confusion or disorientation of the learner. This can leave the learner lost in a maze of information. Authors must be consistent, provide clearly defined navigational tools, and focus the learner, for example with a problem to be solved.

REFERENCES

1. Najjar, L.J., Principles of Educational Multimedia User Interface Design, Human Factors, 40(2), 311-23, 1998.
2. Landow, G.P., Hypertext: The Convergence of Contemporary Critical Theory and Technology, Baltimore, The Johns Hopkins University Press, 1992.
3. Hannifan, R.D. and Sullivan, H.J., Preferences and Learner Control Over Amount of Instruction, Journal of Educational Psychology, 88, 162-173, 1996.
4. Morrison, G.R., Ross, S.M. and Baldwin, W., Learner Control of Context and Instructional Support in Learning Elementary School Mathematics, Educational Technology Research and Development, 40(1), 5-13, 1992.
5. Goforth, D., Learner Control = Decision Making + Information: A Model and a Meta-Analysis, Journal of Educational Computing Research, 11, 1-26, 1994.
6. Kinzie, M.B., Sullivan, H.J., and Berdel, R.L., Learner Control and Achievement in Science Computer-Assisted Instruction, Journal of Educational Psychology, 80, 299-303, 1988.
7. Niemiec, R.P., Silorski, C., and Waber, H.J., Learner-Control Effects: A Review of Reviews and a Meta-Analysis, Journal of Educational Computing Research, 15, 157-164, 1996.
8. Baker, W., Hale, T., and Gifford, B., From Theory to Implementation: The Mediated Learning Approach to Computer-Mediated Instruction, Learning and Assessment, Educom Review, 32(5), 1997. Available: http://www.educause.edu/pub/er/review/reviewArticles/32542.html Accessed Dec. 21, 2000.
9. Fletcher-Flinn, C.M, and Gravatt, B., The Efficacy of Computer Assisted Instruction (CAI): A Meta-Analysis, Journal of Educational Computing and Research, 12, 219-242, 1995.
10. Okolo, C., Barth, C., and Rieth, H.A., Retrospective View of Computer-Based Instruction, Journal of Special Education Technology. 12(1):1-27, 1993.
11. Clark, R., Reconsidering Research On Learning From Media, Review of Educational Research, 53(4):445-459, 1983.
12. Clark, R., Media Will Never Influence Learning, Educational Technology, Research and Development, 42(1):21-32, 1994.
13. Spiro, R.J., Feltovich, P.J., Jacobson, M.J., and Coulson, R.L., Cognitive Flexibility, Constructivism, and Hypertext: Random Access Instruction For Advanced Knowledge Acquisition In Ill-Structured Domains. In T.M. Duffy and D.H. Jonassen (Eds.), Constructivism and the Technology of Instruction: A Conversation, pp. 57-75, New Jersey, LEA Publishers, 1992.
14. Spiro, R.J., Vispoel, W., Schmitz, J., Samarapungavan, A., and Boerger, A., Knowledge Acquisition for Application: Cognitive Flexibility and Transfer in Complex Content Domains. In B.C. Britton (Ed.), Executive Control Processes, pp. 177-200, Hillsdale, New Jersey, Lawrence Erlbaum Associates, 1987.
15. Wittgenstein, L., Philosophical Investigations, New York, Macmillan, 1953.
16. Ebersol, S., Cognitive Issues in the Design and Deployment of Interactive Hypermedia: Implications for Authoring WWW Sites, Interpersonal Computing and Technology, 5(1-2), 19-36, 1997. Available: http://jan.ucc.nau.edu/~ipct-j/1997/n2/ebersole.html  Accessed December 21, 2000.
17. Kim, H., and Hirtle, S.C., Spatial Metaphors and Disorientation In Hypertext Browsing, Behaviour and Information Technology, 14, 239-250, 1995.
18. McDonald, S. and Stevenson, R.J., Disorientation In Hypertext: The Effects of Three Text Structures on Navigation Performance, Applied Ergonomics, 27, 61-68, 1996.
19. McDonald, S. and Stevenson, R.J., Effects of Text Structure and Prior Knowledge of the Learner on Navigation in Hypertext, Human Factors, 40(1), 18-27, 1998.
20. Hammond, N.V., Hypermedia and Learning: Who Guides Whom? In H. Maurer (Ed.), Computer Assisted Learning, pp. 167-180, Berlin, Springer-Verlag, 1989.
21. Miller, W.H., What Are Learning Disabilities? The Complete Reading Disabilities, The Emerging Worldwide Electronic University Handbook, pp. 1-31, West Nyack, NY, The Center for Applied Research in Education, 1993.
22. Barab, S.A., Young, M.F., and Wang, J., The Effects of Navigational and Generative Activities in Hypertext Learning on Problem Solving and Comprehension, International Journal of Instructional Media, 26(3), 283-299, 1999.

ABOUT THE AUTHOR

Don Lehman, MSc, MT(ASCP), SM(AAM) is Assistant Professor, Dept. of Medical Technology, University of Delaware, Newark, Delaware. His interests include medical microbiology, parasitology