What do Scanpaths Tell Us About Cognitive Processes? An Investigation in a Problem Solving Domain

Scanpaths are the specific sequence of fixations elicited by someone when viewing a scene or object. Not only do they illustrate which areas of the scene that were fixated on, they also capture the viewer's change in attention overtime. Dynamics of visual attention contain different movement patterns that are not otherwise measured with basic methods of eye data analysis. In this thesis, we employ MultiMatch, a scanpath analysis tool that provides quantitative measure of similarity between two scanpaths, to examine visual patterns for two different data sets. These data sets came from studies that presented students with math problems and varied instructional material to manipulate student's solving strategies. We apply an analysis method corresponding to grouping scanpaths across and within-conditions to determine whether the MultiMatch tool can distinguish between the presentation formats. We further our analysis by providing initial interpretation guidelines through a brief scanpath simulation model. Results are discussed in context of the hypothesized cognitive processes that the conditions were designed to elicit.