Training Effectiveness of Whole Body Flight Simulator Motion: A Comprehensive Meta-Analysis

We present a meta-analysis of 24 effect sizes from transfer of training experiments using whole body simulator motion as an independent variable. Three moderator variables were investigated: experiment design, task type, and subjects' experience. Due to the large heterogeneity of the included experiments, we used a random-effects model. Correction for measurement error was applied. The results revealed an overall transfer effect in favor of motion (d = 0.51). Effects were weaker in true transfer (d = 0.10) than in quasi-transfer with different (d = 0.73) and identical (d = 1.19) motion, stronger for helicopter (d = 0.86) and disturbance tasks (d = 0.84) than for maneuvering fixed-wing aircraft (d = 0.07), and stronger for subjects without flight experience (d = 1.57) than for pilots with intermediate (d = 0.53) and expert (d = −0.01) experience. In conclusion, motion seems important for flight-naive individuals learning tasks with external disturbances or control of vehicles with low dynamic stability, but not for experts learning fixed-wing aircraft maneuvering tasks. Effects are attenuated in true transfer as compared to quasi-transfer.