Modeling, assessing, and supporting key competencies within game environments

Abstract

The first challenge of accomplishing the goals of any successful instructional system depends on accurately identifying characteristics of a particular learner or group of learners - such as the type and level of specific knowledge, skills, and other attributes. The second challenge is then leveraging the information to improve learning. This chapter is intended to extend current thinking about (a) educationally valuable skills and (b) instructional system design by describing an approach for analyzing key competencies and developing valid assessments embedded within an immersive game. Specifically, we will describe theoretically based research relating to stealth assessment, diagnosis, and instructional decisions, operational within an immersive game environment. Stealth assessment and diagnosis occur during the learning (playing) process, and instructional decisions are based on inferences of learners' current and projected competency states. Inferences - both diagnostic and predictive - will be handled by Bayesian networks and used directly in student models to handle uncertainty via probabilistic inference to update and improve belief values on learner competencies. Resulting probabilities inform decision making, as needed in, for instance, the selection of instructional support based on the learner's current state. © 2010 Springer-Verlag US.