Discussion

The non-significant results to the Chi-squared tests show that this aspect of thinking style has no effect on the choice of navigational aid in this case. This means that either the hypothesis is wrong, and consequently the theory, or the experiment is flawed.

There are flaws in this experiment. The sample is not representative of the general population, being an undergraduate population, so there may be aspects of their thinking style which are different to that of others. For instance, there are a great many versatile thinkers, which may be an advantage when applying to university. The fact that we had to change the scoring system reflects this and may make our experiment less reliable.

There were some participants for whom English was not a first language, who found it difficult to understand the instructions, which may make their data invalid. There were also many environmental variables, such as ambient light, which were not controlled, but we are interested in a real-world learning environment; a website may be viewed in a variety of different environments, and within this department, the Maclab is the most likely one to be used.

It is possible that the learning material and its structure demand the use of one particular thinking style over another. Wright and Lickorish (1990) found that different navigational aids were preferred when hypertext was structured differently. Also, the participants were required to complete multiple-choice questionnaires, which favours local thinkers (Sternberg, 1997), and may have caused those with a global preference to use the material in a more local fashion, because it is better to suit the style of learning to the style of assessment (Kelly, 1995, Heppel, 1994). From the distribution of routes chosen, however, it looks like the route was chosen at random. There is a slight preference for the linear format, but this may be because it looked less demanding. It may also be the case that orientation towards the specific task, as discussed in the introduction, may have a stronger impact on the choice of navigational aid than dominant thinking style.

One definite problem was that the sample was unfamiliar with the hypertext layout. Lumb, working with third years, did not have this problem, as working with the WWW is a necessary skill on this course. We overlooked the fact that some of the participants might never have seen a hypertext system before. Previously, learning research has always used media and presentation methods that people have used since children and throughout their lives. “A fair evaluation of learning from hypertext can only come from hypertext-literate learners who have developed a useful set of strategies for navigating and integrating information from hypertext.” (Jonassen, 1993), a view supported by Kommers (1990, in Whalley, 1993). Some of the people in our sample may never have touched a computer until they came to university, and would not yet understand their own preferences. In further studies in this department, it would be advisable to use participants with more experience of hypertext systems or to allow some practice time. Maybe amount of experience would be a useful variable to examine in future, “’novices’ and ‘experts’... will have different requirements and expectations and so the hypermedia system must be designed accordingly.” (Barker, 1993) The kinds of presentation the participants were given a choice between may have been familiar from paper tasks or books in the past, but their use in this context would not be.

All this leaves the idea that the hypothesis is wrong, and therefore the theory. The fact that this particular hypothesis is wrong does not necessarily mean that the theory is wrong, as it may have been misinterpreted by us. The aspect of mental self-government which we looked at may not affect the use of navigational aids at all. Since the nature of the task was to read all the information available, all types of thinker may have done this, and local thinkers pulled out isolated facts and global thinkers pulled out patterns, although they showed the same use of navigational aids.

Our experiment also departs from the theory by looking at only one aspect of thinking style. Sternberg insists that everyone has a profile of styles which interact to produce their behaviour. He also uses three separate tests to build up a more complete picture of a person's profile of styles, but this would be impractical in an undergraduate project. Any of the other aspects of thinking styles in the theory might influence an individual's use of a hypertext system; a person with a legislative function might prefer to browse, and an executive one might prefer to follow someone else's guide; a person with a monarchic form might not get lost, because they would always be pursuing the same aim, where an oligarchic or anarchic person might benefit from a device which reminded them occasionally what they were supposed to be doing; people with external scope might not like working with hypertext at all, and might prefer to use the Internet as a discussion forum. There could be any number of effects from all the possible combinations of learning style, and the fact that the one we chose to look at does not exist does not mean that others will not. Perhaps the strength of hypertext is that there is room to explain concepts in a way that suits everyone.

A learner will also bring more than their learning style to a real-life learning situation; Cunningham et al. (1993) find that a learner’s interaction with their system, Intermedia, is “a complex interaction of the user’s motives and goals, the user’s prior knowledge, the user’s understanding of the task, the affordances and constraints of the Intermedia system, the available documents, prior work with Intermedia, self-knowledge and a number of other factors,” (p.38). It would be easier to look at the effect of learning styles if these numerous other factors could be controlled or measured. Another flaw is the system which we used. A hypertext system is only as good as its designers, and we are, at best, novices. The text was written by people who only have experience of writing linear text, which can lose its coherence when “chunked” to make hypertext (Whalley, 1993). It was also not possible to show the whole of the concept map on-screen at once, or to mark nodes which the user had visited, which would have been helpful characterisitics.

Many of the participants found the task dull, as they had recently had an exam on the topic area (some on the same day!) of which we were not aware. It has also been thought that the task was rather long, causing fatigue. Neither was it a particularly interactive task, the kind in which hypertext can be so strong. It merely required participants to learn the material as best they could, which employs no information search strategies at all, and none of the integration that Constructivism suggests is so important. Perhaps a further development would be to set an information search task on the World Wide Web, and to observe how students go about finding the information. It is possible that the information capture technique used for this experiment could be adapted to allow this.

Hypertext has a great deal of potential as a learning medium. As well as a tool amongst others for the student to find the most up-to date information on the Web, it also has the potential for use in distance and student-centred learning. Many authors suggest that it will never be appropriate as a core course resource (Whalley, 1993, McKnight, 1993) and Whalley argues that navigational aids “spoon-feed” the structure which Constructivists believe one should form oneself, encouraging surface learning. On the other hand, Jonassen (1990) believes that a system constructed by an expert can help by providing the semantic framework as an aid to integrating the knowledge. With the advance of technology, hypertext is rapidly becoming a fact of life, and is predicted to be ubiquitous within very few years by the likes of Bill Gates of Microsoft (“the world’s most powerful and treacherous software company” Reid, 1997, p.xvii), who has made it a personal aim to put a personal computer in every American home, and is powerful enough to make it happen.

Hypertext will become increasingly important as education moves towards student-centred learning. Children are growing up now for whom hypermedia is familiar, and the manufacturers of hypermedia learning systems hope that this is going to become more and more common. One thing is certain, the WWW is not going to go away, it continues to expand by the day, and the Internet on which it depends was designed to withstand direct nuclear strike. It will soon be no easier to go through our educational system without understanding how to use a system like this than it is now without being able to read. However, it will never be the only resource available, and it will always be true that “different situations call for different solutions...adequate presentation rests more on good user-centred design than on slavish adherence to one medium,” (p.4, McKnight, 1993).

More and more information is becoming available to the learner, with multiple satellite and digital television services, and Internet connection soon to be available through digital television sets, it will become imperative that learners are able to filter out irrelevant information. “Information management may even be identified as a skill of growing importance in our society,” (Cunningham, et al, 1993). It is certainly a skill that ought to be taught even in low technology learning situations. I am naturally an information browser, and easily get “lost” in the information available in a library, and since “learning can hardly be effective if learners merely ramble through the knowledge base in an unmotivated and haphazard fashion” (Hammond, 1993, p52), I would have benefited from gaining earlier the kind of self-knowledge and self-discipline which I have gained in the past year.

The popular press occasionally complain that modern children have no attention span, and no interest in issues which they (the authors) consider important. “Might we be seeing not a decline in their capability, but a change in the skill set that represents that capability?” (Heppel, 1994). Perhaps children are already ahead of us, and have become skilled information grazers through necessity, and the true Information Revolution lies not in the technology, but in the children who have grown up with it.