Using Eye Gaze Data and Visual Activities to Infer Human Cognitive Styles

Recent research provides evidence that individual differences in human cognitive styles affect user performance and experience in diverse application domains. However, state-of-the-art elicitation methods of cognitive styles require researchers to apply explicit, in-lab, and time-consuming "paper-and-pencil" techniques, rendering real-time integration of cognitive styles? elicitation impractical in interactive system design. Aiming to elaborate an implicit elicitation method of cognitive styles, this paper reports two feasibility studies based on an eye-tracking multifactorial model. In both studies, participants performed visual activities of varying characteristics, and the eye-tracking analysis revealed quantitative differences on visual behavior among individuals with different cognitive styles. Based on these differences, a series of classification experiments were conducted, and the results revealed that gaze-based implicit elicitation of cognitive styles in real-time is feasible, which could be used by interactive systems to adapt to the users' cognitive needs and preferences, to better assist them, and improve their performance and experience.