EXPLORATION AND LEARNING IN INTERACTIVE SYSTEMS

Peter Polson
Department of Psychology and
Institute of Cognitive Science
and
Clayton Lewis
Department of Computer Science and
Institute of Cognitive Science

University of Colorado

CONTACT INFORMATION

Peter Polson, Campus Box 344, University of Colorado, Boulder CO 80309.

ppolson@psych.colorado.edu, (303) 492 5622, fax (303) 492 7177

Clayton Lewis, Campus Box 430, University of Colorado, Boulder CO 80309

clayton@cs.colorado.edu, (303) 492 6657, fax (303) 492 2844

WWW PAGE

http://www.cs.colorado.edu/~clayton/nsfpage.html

PROGRAM AREA

Usability and User-Centered Design.

KEYWORDS

Cognitive modeling, cognitive architecture, learning, exploration

PROJECT SUMMARY

Our long-term research strategy is to develop design methods for user interfaces and cognitive theory together. Cognitive theory can advance the state of the art of interface design by clarifying the characteristics of interfaces that lead to fast learning and accurate performance. On the other hand, modeling key mental processes in computer use is a significant challenge for current cognitive theory. Building such models helps cognitive theory to grow in useful directions.

Our work on design has led to the development of the Cognitive Walkthrough method and its extensions. The aim of the walkthrough is to deliver some of the benefits of a theoretical analysis of an interface to designers who are not cognitive theorists.

On the theory side we have developed models within Kintsch's construction integration architecture, in collaboration with Muneo Kitajima of Industrial Products Research Institute, that provide an account of error processes, including errors made by highly skilled users. Working with both Anderson's ACT-R architecture and SOAR (in collaboration with Richard Young) we have modelled the process by which users can apply knowledge of familiar user interface conventions to an unfamiliar application. This process is fundamental to the success of modern interface families but is as yet poorly understood.

We are currently working with Kitajima on an extension of our construction-integration model to account for the same behaviors. The approach being pursued draws on work by Kintsch and Greeno that treats some problem-solving as a form of comprehension.

PROJECT REFERENCES

Kitajima, M., and Polson, P. G. (1995) A Comprehension-Based Model of Correct Performance and Errors in Skilled, Display-Based Human-Computer Interaction . International Journal of Human-Computer Studies, 43, 65-99.

Kitajima, M., and Polson, P. G. (1995). Mechanisms of Slips in Display-Based Human-Computer Interaction: A Model-Based Analysis. In Y. Anzai, K. Ogawa. and H. Mori (Eds.) Advances in Human Factors/Ergonomics, Symbiosis of Human and Artifact: Human and Social Aspects of Human-Computer Interaction . Amsterdam: Elsevier, pp. 515-520

Lee, A.Y., Foltz, P.W., and Polson, P.G. (1994) Memory for Task-Action Mappings: Mnemonics, regularity, and consistency. International Journal of Human-Computer studies,, 40, pp 771-730.

Kitajima, M. and Polson, P. (1992) A computational model of skilled use of a graphical user interface. In Proc. CHI'92, New York: ACM, 241-249.

Rehder, B., Lewis, C., Terwilliger, B., Polson, P., and Rieman, J. (1995) A model of optimal exploration and decision making in novel interfaces. In CHI'95 Conference Companion, New York: ACM, 230-231.

Rieman, J., Lewis, C., Young, R. and Polson, P. (1994) Why is a raven like a writing desk? Lessons in interface consistency and analogical reasoning from two cognitive architectures. In Proc. CHI'94, New York: ACM, 438-444.

Wharton, C., Rieman, J., Lewis, C. and Polson, P. (1994) The Cognitive Walkthrough method: A proactitioner's guide. In J. Nielsen and R. Mack (Eds.) Usabiltity Inspection Methods. New York: Wiley, 1994, 105-140.

AREA BACKGROUND

Cognitive architecture is a term of analogy to computer architecture: it refers to a characterization of the underlying facilities of the brain from which particular mental processes and skills are composed. Many proposed architectures have been framed, emphasizing different aspects of cognitive functioning and different conceptions of how such functions are realized. A theme of our work is relating these architectures specifically to mental processes in computer use as a means of identifying sources of difficulty and ease in interactions, and as a means of identifying aspects of cognition that are relevant to HCI but not well covered by current theory.

Considerable progress has been made along these in characterizing skilled performance, including transfer of skill among tasks. Much remains to be done in understanding errors, which are surprisingly frequence even among experienced, knowledgeable users, processes of learning and exploration, and goal formation in non-routinized tasks.

AREA REFERENCES

Baecker, R.M., Grudin, J., Buxton, W.A.S. and Greenberg, S. (1995) Readings in Human-Computer Interaction: Toward the Year 2000, San Francisco: Morgan Kaufmann, collects useful material in Chapter 9.

RELATED PROGRAM AREAS

Our work is relevant to all of the ISP areas, but particularly to Area 4, Adaptive Human Interfaces.

POTENTIAL RELATED PROJECTS

We're interested in comparison activities that would apply different cognitive architectures to a common set of HCI problems. We're also interested in designers' perspectives on the possible role of cognitive models in design, including interface phenomena in need of explication.