ABSTRACT
Fidelity is an important measure that can be used to evaluate the quality of human-in-the-loop simulation. Although the validity of the simulation is not necessarily related to its fidelity, appropriate levels of fidelity will facilitate transfer of training. Simulation fidelity and level of detail in a simulation model are major issues in the development of simulation models. Most applications of fidelity use a rather vague classification, such as high, medium and low. It can be argued that precise measurement of fidelity is required to gain more insight on simulation transfer of training, and fidelity can be quantitatively measured. Currently there is no standard quantification measurement of fidelity. This paper proposes a conceptual framework of fidelity assessment by integrating the current research results and using multi-layer methods (object, attributes and behavior). Mathematical model and assessment procedures are illustrated. Also the relationship between fidelity and human task performance were investigated. This framework shall serve as a springboard for future research on this subject.
