Paulina Lindstrom, Agneta Gulz
An interesting class of intelligent tutoring systems are systems that are able to detect and recognize affective responses and states in students, and use such data to adapt its feedback, guidance and instruction. The article presents affective responses and states considered relevant for learning, and methods that have been proposed or used to detect such responses and states. We focus on the affective state of eureka, a relatively short-term state with corresponding difficulties of being detected. There are good reasons to attempt to catch this affective state, since it can mark when understanding is achieved and signal to an intelligent tutoring system that the student is ready to proceed in the learning process. A study is presented where a combination of the behavioural measure of eye tracking and two psycho-physiological measures skin conductance and pupil size are used in an attempt to identify this affective experience. The results showed significant differences on all measures between subjects who understood certain stimuli, and purportedly experienced eureka, and subjects who did not understand the stimuli. The psycho-physiological measure patterns were however not as dramatic as predicted. Finally, we discuss the potential benefits for students using ITS, that can detect eureka experiences.
Subjects: 6. Computer-Human Interaction.
Submitted: Jan 17, 2008