Inferring a Learner´s Cognitive, Motivational and Emotional State in a Digital Educational Game


  • Michael Bedek
  • Paul Seitlinger
  • Simone Kopeinik
  • Dietrich Albert


digital educational games, motivation, emotion, problem representation, non-invasive assessment.


Digital educational games (DEGs) possess the potential of providing an appealing and intrinsically motivating learning context. Usually this potential is either taken for granted or examined through questionnaires or interviews in the course of evaluation studies. However, an adaptive game would increase the probability of a DEG being actually motivating and emotionally appealing. In order to adapt the game to the learner´s motivational and emotional state while engaged with a particular game scenario, an ongoing assessment of these states is required. An explicit assessment, e.g. by questionnaires occurring repeatedly in short time intervals on the screen would probably destroy the learner´s flow experience. Thus, it is necessary to apply an approach that assesses the learner´s current states in a non‑invasive way. In the course of this paper we describe such a non‑invasive, implicit assessment procedure which is based on the interpretation of behavioral indicators. A set of behavioral indicators has been elaborated whereby some of them are derived from the theory of information foraging (Pirolli and Card, 1999). Values for each behavioral indicator (e.g. amount, frequency, seconds, etc.) are gathered after equally long lasting time slices. After each time slice, these values serve as weighted predictors to multiple regression equations for the dimensions of a motivation model, an emotion model and a construct called clearness. The motivation model is based on the two dimensions of approach and avoidance motivation. The emotion model encompasses the dimensions valence and activation. Clearness is defined as appropriate problem representation. A comparison of the resulting values on these dimensions between the current and previous time slices covers fluctuations of the learner`s states over time. The assessment of such changes forms the prerequisite for providing in‑game adaptations which aim to enhance the learner`s state, targeting towards a full exploitation of DEGs pedagogical potential.



1 Jul 2012