The effects of typing demand on emotional stress, mouse and keystroke behaviours

Abstract

Past research found that cognitive effort is related to emotion, which negative emotion may influence task performance. To enhance learning experience, it is important to have an effective technique to measure user’s emotional and motivational affects for designing an adaptive e-learning system, rather than using a subjective method that is less reliable and accurate. Keystroke and mouse dynamics analyses shed light on a better automated emotion recognition method as compared to physiological methods, as they are cheaper, noninvasive and can be easily set up. This research shows that unification of mouse and keyboard dynamics analyses could be useful in detecting emotional stress, particularly stress induced by time pressure, text length and language familiarity. The changes of mouse and keystroke behaviours of the students are found cohere with their task performance and stress perception. However anomalies in mouse and keystroke behaviours present when the students are pushed beyond their capabilities.