Domain independent strategies in an affective tutoring system

There have been various attempts to develop an affective tutoring system (ATS) framework that considers and reacts to a student’s emotions while learning. However, there is a gap between current systems and the theory underlying human appraisal models. The current frameworks rely on a single appraisal and reaction phase. In contrast, the human appraisal process (Lazarus, 1991) involves two phases of appraisal and reaction (i.e. primary and secondary appraisal phases). This thesis proposes an affective tutoring (ATS) framework that introduces two phases of appraisal and reaction (i.e. primary and secondary appraisal and reaction phases). This proposed framework has been implemented and evaluated in a system to teach Data Structures. In addition, the system employs both domain-dependent and domain-independent strategies for coping with students’ affective states. This follows the emotion regulation model (Lazarus, 1991) that underpins the ATS framework which argues that individuals use both kinds of strategies in solving daily life problems. In comparison, current affective (ITS) frameworks concentrate on the use of domain-dependent strategies to cope with students’ affective states. The evaluation of the system provides some support for the idea that the ATS framework is useful both in improving students’ affective states (i.e. during and by the end of a learning session) and also their learning performance.