Simulator fidelity is commonly considered to be the degree to which the equipment provides an accurate reproduction of the real world (Hays, 2006). Typical fidelity discussions about displays concern things like pixel density, contrast ratios, latency, and field of regard; if they are about motion, the topics of latency, frequency, and degrees of freedom often come up; and for audio, it's common to discuss frequency response and dynamic range. These are good and important subjects; but, there are other important aspects of fidelity.

These other aspects are the motivation behind this paper. They take the form of layers and they can be dichotomized into a framework that looks at the training environment and the human using the training environment. The framework thus includes layers for engineering considerations for designing, building, and using the training environment, as well as, psychological considerations for understanding the effects of fidelity manipulations on the human's experience and on the environment's training effectiveness.

The Training Environment Fidelity layers meet the Human Fidelity layers at the point where displays, motion systems, and physical controls meet the human visual, auditory, and muscular systems. The training environment fidelity layers also include a description of the stimulus- and response-producing systems, the mathematical and computational models that drive them, behavioral models of entities that will be encountered in the environment, and the underlying scenarios. The human layers also include a higher-level perceptual and motor layer that describes object and spatial fidelity, an intuitive layer that describes the degree to which the simulator is suitable for developing highly practiced automatic skills, a cognitive layer that describes the plausibility of more complex situations, a social layer that describes the participant's interactions with other real and synthetic participants, and a pedagogical layer that describes features likely to aid training such as simplification and highlighting for novices.

In this paper we describe the framework and show its successful application to a simulator intended to develop advanced flight skills.