Keywords: Simulation, Expert Performance, Training, Skill Acquisition, Aviation, Sport, Surgery.

Introduction

Many methods have been used to study experts. Traditionally, researchers have dissected performance into its constituent parts to isolate basic underlying mechanisms. Although this provides experimental control, task simplification and the use of novel and artificial tasks are antithetical to reproducing the “real-world” demands faced by actual domain experts. Changing the nature of the phenomenon under investigation may lead to a reduction, if not eradication, of the expert advantage. Cognitive anthropologists (see Clancey, Chapter 8) and Naturalistic Decision Making researchers (see Ross et al., Chapter 23), on the other hand, have argued that the most useful method of examining expertise is to capture performance as it occurs in the “natural” environment. However, critics have claimed that although this type of approach allows “real-world” performance to be described, only minimal explanation is possible with regard to the underlying cognitive processes (e.g., Yates, 2001). Brehmer and Dörner (1993) concluded that field examination may not permit any definite conclusions to be drawn, whereas laboratory tasks are often too simplistic to reach any conclusions of interest. This leaves us in the invidious position that what is interesting is not explained and what is explained is not interesting. Simulation in its many guises may offer an excellent compromise.

The range and type of possible simulation environments is vast. Some are referred to as Computer-Aided Virtual Environment (CAVE) systems. Others include high fidelity simulations of complex systems (e.g.