ABSTRACT
The intent of the symposium is to provide a forum for discussion about what it is like to try to “do” cognitive science in settings involving real products, real customers, and other constraints in which the bottom line is not necessarily to produce a pristine test of a theory but rather, to compete in the marketplace or perhaps save a life. Cognitive science is full of interesting and complex theories. People have surprising difficulty in retrieving relevant information from external sources. Anyone who has looked for something in the index of a book, the Yellow Pages, or in computer databases has experienced this problem. Using cognitive modeling to predict human performance in real-world settings is an enterprise laden with paradox. If cognitive models predict human performance with computer systems that the system designers consider “intuitive,” then modeling is perceived to be of little value.
