Effects of practice on final position reproduction.

Three subjects practiced fast, accurate 36 degrees elbow flexion movements to a 2.5 degrees target for 14 sessions of 100 trials (total, 1400 trials). Subjects then returned for a 15th experimental session in which they were asked to perform 15 movements under identical conditions to the practice condition. They were then tested under three experimental conditions without visual feedback: (1) identical to the practice conditions, (2) with small shifts in starting position (+/- 3 degrees of the practiced starting position), that were insufficient for subjective discrimination and, therefore, subjects were instructed to repeat the practiced movements; and (3) with a large shift in starting position (range, +/- 15 degrees of the practiced starting position), under the instruction to move to the same target. Experimental conditions 2 and 3 demonstrated that shifts in starting position were partially correlated with shifts in final position. These results are interpreted from the point of view of the equilibrium-point hypothesis of motor control.