Mining Bodily Patterns of Affective Experience during Learning

We investigated 28 learners’ postural patterns associated with naturally occurring episodes of boredom, flow/engagement, confusion, frustration, and delight during a tutoring session with AutoTutor, a dialoguebased intelligent tutoring system. Training and validation data were collected in a learning session with AutoTutor, after which the learners’ affective states (i.e., emotions) were rated by the learner, a peer, and two trained judges. An automated body pressure measurement system was used to capture the pressure exerted by the learner on the seat and back of a chair during the tutoring session. We extracted 16 posture-related features that focused on the pressure exerted along with the magnitude and direction of changes in pressure during emotional experiences. Binary logistic regression models yielded medium sized effects in discriminating the affective states from neutral. An analysis of the parameters of the models indicated that the affective states were manifested by three unique postural configurations and a general increase in movement (when compared to neutral).