Visio-motor prediction of collision.

Visually presented moving stimuli were used in a collision prediction experiment aimed at revealing the ability of human subjects to estimate the velocity of motion. Spatial extent, velocity and exposure duration of movement were varied so that the correlation between these characteristics of the stimuli was low (about 0.5). The moving stimulus disappeared at a certain place on the screen. By pressing a button the subject had to estimate the moment the stimulus would reach a target in the visual field. A linear dependence between estimated and actual duration of concealed motion was obtained. It is concluded that spatial extent and duration of motion do not influence the performance of the task. Velocity estimation and individual characteristics of the mechanisms controlling the motor response are supposed to determine the linear dependences obtained for all subjects tested.