Control Type Identification in Student-Teacher Interaction

The goal of this research is identifying the type of control for student-teacher interaction in the classroom, using a preliminary investigation based on a role-play. This research is used as a start phase in the building of a mathematical model for the student module in a future computer-assisted virtual Affective Tutoring System (ATS). An ATS is a software environment that can understand student emotions, behaviour, skills and needs, and adapt its teaching strategy for an optimal guidance of training. The control of the interaction is considered at any time on the side which is launching the questions. This controlling side is either the teacher or the student, whoever asks the other side questions about the lesson. Our research has a start point in the “Control-Value Theory” of psycho-pedagogy and aims to build statistical models of the control based on statistical regression – for guidance of the student based on pre-lesson and post-lesson tests and for guidance of the student based on teacher actions (using techniques of conversation control). For all these kinds of control, this paper aims to propose a mathematical model based on hidden Markov models (HMM) having, as a statistical base, volunteer role-play scenarios of classroom activities.