Relating Error Diagnosis and Performance Characteristics for Affect Perception and Empathy in an Educational Software Application

While the benefits of computer-assisted learning are acknowledged within the educational community, numerous new educational software applications are being developed. Educational software can become very effective if it is adaptive and individualised to the student. However, one important aspect of students that has been overlooked so far in such software applications is affect. This paper describes how system observations of students’ behaviour while they interact with an educational application may provide important evidence about students’ emotions while they learn. Observations mainly concern the way students respond to assessment questions in terms of the quality and correctness of their answers. The system’s inferences about students’ emotions are used by the system to adapt interaction to each individual student’s needs taking into account their character and mood.