Abstract
On a two-choice tracking test of normal subjects, numerous false moves were induced when the directional relation between movement of the hand and the response marker was changed. The mean error-correction times were not consistently less than simple reaction times. When the center of the display was screened to eliminate visual feedback, numerous moves were reversed before the marker was visible. The probability of reversing an error was much greater than that of reversing a correct move. The findings are best explained in terms of a central mechanism that uses efference copy to amend errors of movement.