It is generally assumed that skills trained and assessed in a simulator will transfer to the line. However, there is a class of maneuvers that demand an immediate response to an unexpected event (e.g., rejected takeoffs) for which such transfer can be questioned and for which there is little or no empirical data to support a transfer assumption. Thus, we have completed a series of studies aimed at investigating the effects of expectancy on performance for unanticipated events in a laboratory situation with undergraduate college students and experienced pilots. Our participants were trained on both normal and rejected takeoffs and the expectancy for a rejected takeoff was manipulated in each study. There were two primary measures of performance on rejected takeoff trials: the amount of time it took the participant to close down the throttle after engine failure and the maximum deviation from center line achieved while bringing the aircraft to a stop. T-tests indicated that there was a significant degradation in throttle performance for both studies (all ps<.05) and in maximum deviation from center line performance for one of the studies (p<.001). Thus, it is questionable whether the assumption that performance on events that occur in high expectancy conditions will transfer to low expectancy conditions is valid.
Stevens, S. M.,
& Goldsmith, T. E.
(2007). Skill Decay on Takeoffs as a Result of Varying Degrees of Expectancy. 2007 International Symposium on Aviation Psychology, 687-690.