| Kristina Höök
SICS, Box 1263, S-164 28 Kista, Sweden
| Nils Dahlbäck
IDA, Linköping University, S-581 83 Linköping, Sweden
(Presented at the workshop "CHI 97 Workshop on Navigation in Electronic Worlds" to be held in Atlanta, March 23-24th, 1997.)
One of the main problems with systems design is dealing with individuals. There is now enough knowledge of individual differences and navigation to say that these differences are important and significant. In particular, we know that users' spatial ability is relevant to their ability to make use of computer systems. For example, Vicente and Williges (1988) found that spatial ability affected users ability to navigate a large file structure. Benyon and Murray (1993) found that a combination of spatial ability and computer experience affected users ability to interact with a database system. More recently we found a similar result concerning navigation in hypermedia (Dahlbäck, et al., 1996, Höök et al. 1996). The best user was almost 20 times faster than the slowest in using the system to solve six problems. The correlation between this performance measure and one aspect of spatial cognition was as high as r = .56 (p < .005).
Another interesting aspect of this study was that the results indicate that navigation in the physical world and in hypermedia system is not the quite the same thing, or rather that different cognitive abilities come into play in the different tasks. We found a strong correlation between one set of spatial measures and the ability to navigate in the hypermedia system, and a correlation with another orthogonal spatial ability and map-reading ability. Since it has been shown previously that the latter is correlated with navigational ability in the physical world, this suggests that we here are dealing with two related but not identical abilities. We do not want to draw any strong conclusions from just one study, but we do find the results interesting and worthy of further investigation.
Given that we a (limited) understanding of the correlation between cognitive ability and ability to navigate in information spaces, how can we design information spaces and interfaces to information space so that low-ability users can find their way? One way is to design navigational aids based on this knowledge. This is one direction on our current work. Another way is to explore a radically different navigational metaphor, based on a personalised and social navigational paradigm. We believe that there are two advantages to this: one is that the communication is done verbally rather than via some graphical interface or abstract language, and the other is that the information becomes personalised to the information seeker. One way of describing the difference between the two approaches is that the first is based on navigational aids such as maps, the second on asking for the way or hiring a guide to find the way to the desired goal.
One such aspect is that information systems do not have to be single, stable spaces; they can be adaptive. For example, there is a growing interest in adaptive hypermedia as a means to overcome the "lost-in-hyperspace" problems (Brusilovsky, 1996). The adaptivity can affect the links in a page through annotating them, hiding them or proposing one or some of them as appropriate nodes to move to next. These adaptations can be based on a user model in the system that keeps track of some of the individual differerences. However, this research has yet to demonstrate which problems can be solved by which solutions. For example, adaptive hypermedia systems that hide parts of the nodes can be used to reduce the information space. However, a low-spatial user may still have problems, perhaps not with the overview of the space but with local decisions such as where to go next. If the adaptive system proposes where to go next it won't necessarily prevent the user from feeling lost.
Adaptive annotations and adaptive interfaces (for example changing a graphical map into a verbal set of instructions) can be of use if carefully designed and matched with individuals differences which have a real impact on the usability of the system. Appropriate metaphors and visualisations of the information space can aid object identification and object clustering.
We are also currently exploring the possible use of work done in different quarters, e.g. film and theatre. There are a nuber of devices used by film directors in giving the viewer a sense of direction and location relationships between different scenes. The use of narrative offers another solution as suggested by Laurel and others (Laurel, 1991). Through a mixture of stereotypical narratives learnt by everybody and ways of moving between scenes and lines of actions, film or theatre can seduce us to think that we know how to orient ourselves and also to look in certain directions for interesting things.
As mentioned above, we are also working on a radical alternative to these approaches basing navigation on a social rather than geographical footing. "Social navigation" could be one way forward to help users who cannot be helped by maps and other such devices. Such users might prefer to converse with interface agents, to discuss their information needs rather than having to find their own way to the information. This notion can be extended to filtering agents (Maes, 1994) which organise and structure information, leaving the user to evaluate the relevance of the contents.
In our work we use both experimental studies of navigation (like the cognitive study mentioned above, but also evaluation of design improvements for hypermedia systems), as well as the design (or re-design) of systems based on the approach outlined above.
Nils Dahlbäck, Ph.D., has previously worked on spatial cognition and individual differences in cognitive strategies within the framework of Johnson-Laird's theory of Mental Models. More recently he has worked on the relation between cognitive abilities and hypermedia navigation (for references, see below).
Dahlbäck, N., Höök, K., and Sjölinder, M. (1996) Spatial Cognition in the Mind and in the World - the case of hypermedia navigation, The Eighteenth Annual Meeting of the Cognitive Science Society, University of California, San Diego, July, 1996.
Höök, K. (1991), An Approach to a Route Guidance Interface, Licentiate Thesis, Dept. of Computer and Systems Sciences, Stockholm University, ISSN 1101-8526.
Höök, K. (1997), Evaluating the Utility and Usability of an Adaptive Hypermedia System, In Proceedings of the Conference on Intelligent User Interfaces (IUI'97), Orlando, Florida.
Höök, K. and Karlgren, J. (1991), Some Principles for Route Descriptions Derived from Human Advisers, Proceedings of the Thirteenth Annual Conference of the Cognitive Science Society, Chicago.
Höök, K., Marie Sjölinder and Dahlbäck, N., (1996) Individual Differences and Navigation in Hypermedia, In Proceedings of the eigth European Conference on Cognitive Ergonomics (ECCE-8), 1996.
The rest of the references can be found here.