Predictive Validity of Flight Simulators as a Function of Simulator Motion

Abstract

Ninety multi-engine instrument rated pilots were assigned to no motion, sustained linear scaled down analog motion, and washout motion in a GAT II simulator for determining the effects of degree of motion upon the predictive validity of flight simulators. Five instrument and five contact maneuvers were flown in the simulator followed by flight in a Piper Aztec aircraft. Performances were recorded by two observers and the interobserver reliability coefficients were 0.962 and 0.919 for instrument maneuvers and 0.879 and 0.613 for contact maneuvers in the simulator and aircraft, respectively. The condition of no motion resulted in greater error than the other two groups in the simulator, but there were no significant differences in the aircraft. Correlations of aircraft performance from the simulator maneuvers were 0.763 (no motion), 0.911 (sustained motion), and 0.651 (washout motion). Simulator motion did not result in better aircraft performance, and higher predictive validity was found with very basic sustained motion.