Advancing the Modeling of Student Performance through the Inclusion of Physiological Performance Measures

Abstract

Sophisticated virtual environments and computer simulations provide realistic training environments and web-based delivery mechanisms enable students to train virtually anywhere, anytime. Consequently, the ability to automate instructional functions such as assessing and diagnosing student performance, providing instructional feedback, and appropriately advancing students through a given curriculum is vital to the effectiveness of these technologies. While simulations provide a rich environment for training complex tasks, they introduce a complex assessment environment, which creates challenges in the accurate and efficient diagnosis of trainee needs as a single behavior can be interpreted in several ways. Additionally, student state variables such as affect, personality, and motivation contribute to the numerous interpretations of a single student behavior. Therefore, accurate diagnosis of student learning needs is a daunting task; which has resulted in various investigations of simulation-based performance assessment techniques, but no single recommended best practice or guidelines. An adaptive learning research program (Perrin, Dargue, & Banks, 2003; Perrin et al., 2007) has developed a standards-based student modeling capability. This capability is based on root cause analysis of the underlying causes of student learning needs based on evaluation of fundamental knowledge mastery. As this approach is based on industry standards, this student modeling capability can be extended to include additional variables related to student performance such as student affect. In 2001, Sheldon demonstrated the feasibility and effectiveness of utilizing physiological measures to integrate student state variables into a student modeling capability. At the time of this research, physiological measurement devices used sensors that required the user to restrict his movements in order to ensure integrity of the data recorded and to not disturb the wiring that tethered him to a computerized recording device. Physiological measuring technologies have significantly advanced since this time, such that wireless, accurate measurement devices are available, thus allowing for integration with a training environment. The focus of this lecture is on bridging the student state and standards-based student modeling methodologies to provide an improved student modeling capability.