Abstract
Perceived distance tends to be less than physical distance. The error increases with distance and is greater when distance cues are limited and in virtual environments. How then do people perform visually directed actions, such as throwing and blind walking, accurately? This study examined the effect of error feedback on performance in these tasks in virtual environments. There were two cue conditions: lighted room and dark room, and 7 target distances. The experiment had four stages designed to analyze the effects of error feedback training: 1) Measure throwing and blind walking to perceived positions of visual targets without feedback. 2) Provide visual error feedback to train accurate performance in one of the tasks. 3) Measure transfer of learning to the other task. 4) Measure throwing and walking to the perceived positions of the targets. The 4 stages required 16 half hour sessions. Results: All eight subjects initially threw and walked systematically short of the targets, shorter in the dark condition. All subjects improved substantially with feedback training and their average responses became close to accurate. Throw training produced substantial transfer to walking. Walk training produced less transfer to throwing. There were large stage 4 differences. Two throw trained subjects showed no change from stage 1 to stage 4. Two walk trained subjects showed no change in throw responses. In the other cases response distance increased from stage 1 to stage 4 in all four conditions. An analysis based on a simple model leads to the following conclusions. Both types of training change the transform between perceived distance and cognitive distance. Walk training has the additional effect of changing the motor transform between the walking response and distance. Perceptual changes cannot be ruled out for the subjects who showed stage 4 changes in all four conditions.