Affective Didactic Models in Higher Education

Introduction

In the past, many didactic models have been developed. Such a model should describe the teaching learning process and the conditions for optimal performance. A didactic model should provide insight in the phenomena of delaying students and dropouts. A diagnostic instrument is needed to detect delaying students as soon as possible and to provide assistance by student-counselling. In many models characteristics of students and Universities have been developed to predict the course of the interaction process of teaching learning of individual students in the future.

Such a diagnostic instrument is needed for the wellbeing of students. In the Netherlands, the financial support of the Ministry of Education is limited to 5 years and the length of a course is also 5 years. This implies that students who need a restart or have study delay lack financial support at the end. Dutch Universities have no entrance exam. The first propaedeutic year should be used for (self-)selection of students. At the end of the first-year students get an advice of the Faculty to stop or continue the study. It proves that about 20% of the students start a technical study without the required mathematical or technical abilities. They have to realise as soon as possible that the selected technical study is no option for them and leave this study as soon as possible or start the (orientation on) the new study as soon as possible. But it also proved that some less gifted students are extreme hard workers and are able to complete the study successfully using a strong motivation and strict study discipline. Another 20% stops the study because the interaction student-educational environment fails. Some gifted students have no experience of hard working in math courses at secondary school but at the University you have to make many study hours to be successful. The hope is that via additional counselling these students are able to survive the dip and early diagnosis is important.

Affective phenomena as emotions, moods and affect have a great impact on the learning process of students. Positive core affect as pleasure, relaxation and energy stimulate students to continue the study process. Negative core affect as displeasure, tension, and tiredness have the opposite effect. Characteristic properties of affect and emotions is that they direct the attention towards the eliciting stimulus, have cognitive appraisal and result in overt behaviour. Affective computing is focussed on the development of systems to assess affect and emotion. We developed systems to extract emotions from non-verbal behaviour as facial expressions, gestures, posture but also from the speech signal. These systems have been used to analyse recorded study behaviour via webcams and microphones.

In this paper, we present two didactic models one underlying regular courses and one for e-learning courses. Based on the first model a questionnaire has been developed which provides a basis for a diagnostic instrument. The items in the questionnaire are questions about the interaction process student with the teaching-learning environment. Usually characteristics of students and the teaching learning environment are used to predict study success or study failure. We assume that the study process is a dynamic process and characteristics may change over time.

The second model for e-learning is focussed on the interaction student-learning material and interaction via social media. Because there is no direct support from the teacher and peers the dropout rate is usually rather high. The support is supposed to come from the learning material. In this model, we focus on the affective components. It proves that it is important that the students enjoy learning and the interaction with the teaching-learning material and environment. They have to feel a respectful member of the digital learning community.

In case of regular courses students are requested to fill in questionnaires during one of the lectures. It proves that during e-learning courses and especially for MOOCs the response to questionnaires included in digital learning material is low. In the e-learning didactic model we studied the interaction process teaching learning on a micro-level. Students have to login, select learning material, make assignments, all these steps can be logged during e-learning and appropriate feedback can be provided. Many learning analytic tools have been developed to analyse the learning behaviour of students during e-learning course. For the whole cohort of students, data analytics tools can be used to assess the progress of students during the course.