Adaptive e-learning : motivating learners whilst adapting feedback to cultural background

The personalization of feedback by an Intelligent Tutoring System (ITS) has the potential to greatly improve learner motivation. This PhD investigates how an ITS can adapt to the cultural background of learners when giving feedback. Much published work in ITS is based in Western culture, and resulting systems also correspondingly exhibit this bias. Understanding and adapting to learner culture is important, especially in e-Learning, as evidenced in the awarding gap present for learners of minority cultural backgrounds in many teaching institutions worldwide. In this thesis, we investigate how feedback options are used for learners across cultural backgrounds. The aim is to create a culturally sensitive algorithm for use in an ITS. In this thesis, we use Hofstede's cultural dimensions to model and portray the cultural background of learners. We create and validate 2 cultural stories for each cultural dimension (Individualism, Power Distance, Masculinity, Uncertainty Avoidance, and Long-Term Orientation) to express each dimension at polarized levels (high and low). We use these cultural stories to investigate how people believe the cultural background of a learner should affect the kind of feedback the learner receives. We develop several options for an algorithm that encapsulates these adaptations. We investigate how appropriate and motivating the resulting emotional support options are in a study with participants from five different countries that vary in Power Distance, Individualism and Uncertainty Avoidance. Results provideevidence that the perceived quality of emotional support options depends on culture. Based on the data analysis, a final algorithm is created. This thesis on the one hand shows the importance of considering a learner's cultural background as it doesimpact how appropriate and motivating different feedback options are. On the other hand, it shows that with gooduser-studies, it is possible to construct options that will work for learners from different cultural backgrounds.