An Introduction to Game-Based Competence Assessment Based on Cognitive Diagnostic Models

In Serious Games, in particular in learning and training games, the assessment of competencies and skills is crucial for monitoring learning progress, tailoring learning experiences, and providing individual formative feedback. A sound psychometric diagnostic of competencies is not trivial, however. Conventional scoring techniques have severe shortcomings in terms of accuracy and the degree to which actionable information can be drawn from them. In this paper we introduce Cognitive Diagnostic Models and in particular Competence-based Knowledge Space Theory as theoretical underpinnings of in-game competence assessments. We exemplify the approach by a gamified mathematics learning scenario named Mathiade and illustrate the steps of developing, implementing, and evaluating competence models.