This study was an extension of my work on adaptive interfaces. In this experiment, I examined the psychological effects of software adaptation that was based not on the user's cognitive performance, but on the user's emotional state. Specifically, the study looked at the effects of emotion-based adaptation on user performance and evaluation of the software. In addition, the study addressed the question whether or not the software should tell the user what emotion it had registered.

Subjects were told that a game interface would dynamically adapt based on their level of physiological arousal as determined by their Galvanic skin response, which would be measured throughout the expriment. Results showed that both emotion-based adaptation, and telling the user what his/her emotion is, have significantly affect users' performance and evaluation of the software. However, results were not sufficiently uni-directional to draw clear conclusions, and more research is needed to better understand the psychological effects of affective interfaces.