Eye Movements Disclose Decisions in Set
D.J. Broekens, Walter A. Kosters, T. De Vries · Research Repository (Delft University of Technology) · 2009
During the past several decades, much research has been done in the field of visual attention and cognitive processing, i.e., how does the human brain process visual stimuli. We show that the card game SET is a good candidate for investigating the relation between top-down versus bottom-up visual information processing. Further, we show that Machine Learning techniques can be effectively used to study this relation. We use these techniques to predict if (a) the player thinks he/she found a set (a particular threecard pattern), and (b) he/she correctly identifies a set, both based on eye movements of the player. Our results indicate the following. First, pop-out plays a role in SET-playing performance. Second, the more we move towards predicting correctness (the player found a correct set), the less important eye movements are and the more important player experience is. Third, the more we move towards predicting the claiming of a set (the behavior), the more important eye movements are and the less important player experience is. This indicates that eye movements disclose whether a player thinks he/she found a set or not, but not whether that player correctly identified a set.