Position Paper for CHI 2002 Workshop: Automatic Capture, Representation and Analysis of User Behavior

This is a position paper stating my interest and qualification for attending the CHI 2002 Workshop on Automatic Capture, Representation, and Analysis of User Behavior. The workshop will focus on (1) the development of computerized tools for recording, studying and visualizing new forms and quantities of empirical user data, and (2) how these tools and data will contribute to theories and principles for building easier-to-use and easier-to-learn computer interfaces. Relevant experience Relevant experience includes having empirically validated detailed research hypotheses about visual search of computer menus (Hornof, 1999; Hornof & Kieras, 1997; Hornof & Kieras, 1999) and layouts (Hornof, in preparation; Hornof, in press) by building cognitive models using the EPIC architecture (Kieras & Meyer, 1997). Research hypotheses have included that people search unfamiliar menus looking for a known target with a somewhat random and somewhat systematic search strategy, considering more than one item at a time. Previous theories proposed either a purely systematic or a purely random search, and that people consider one item at a time. Recent modeling confirms these findings for larger visual layout and also demonstrates that, when a useful visual hierarchy is in place, as when items are in groups with useful group headings, people will conduct a highly efficient search first of the headings and then within the group containing the target. I have collected eye movement data to evaluate these theories and am currently in the process of developing automated tools to represent and analyze this eye movement data with respect to the specific theories of the models.