Abstract
80 Ss performed a horizontal adduction movement of the arm attempting to approximate a velocity of 6 or 10 ft/sec with or without a 2-lb. weight in the hand, and Ss were randomly divided into 4 equal subgroups based upon the possible velocity-weight combinations. Ss were blindfolded and received only verbal cues as to the direction of error. After 60 trials on the initial task, each group was divided into 4 subgroups, based upon the original velocity-weight combinations, and performed 60 transfer trials. The two groups approximating the faster velocity were significantly more accurate on the initial task while weight had no effect on accuracy. Negative transfer occurred from the faster movement to the slower, more difficult movement.