The effect of practice on predicting elbow flexion movement time was studie
d. Participants (N = 18) performed 400 elbow flexion trials to a target in
the horizontal plane. The trials were distributed equally over four session
s. The goal was to decrease the movement time (MT) for the same degree of a
ccuracy. The electromyographic (EMG) activity of the biceps and triceps bra
chii was monitored with standard Beckman Ag/AgCl surface electrodes. The EM
G measures formed two variable sets within one prediction equation. One var
iable set was composed of the onset of muscle activity relative to the star
t of movement (motor time) and the duration of muscle activity. The other v
ariable set consisted of the mean amplitude value of the entire burst and o
f the first 30 ms (Q(30)) of activity. As the maximal speed of limb movemen
t increased, the duration of`muscle activity (motor time and EMG duration)
decreased, and the magnitude of muscle activity (MAV and Q(30)) increased.
Most of the change in the duration of muscle activity occurred in Session 1
, while the magnitude of muscle activity continued to increase until Sessio
n 3. Multiple participants adjusted the magnitude of muscle activity to inc
rease limb movement speed. As practice continued, alterations in the durati
on of muscle activity became more important, while the magnitude changes we
re less involved. Late in learning, both dimensions of muscle activity were
used to decrease MT. We suggest that the interplay between the magnitude a
nd duration of dimensions of muscle activity may be due to: (a) cognitive f
actors related to the division of attention in a motor skill, (b) an increa
se in the frequency of motor unit firing that affects both dimensions of mu
scle activity or (c) some combination of (a) and (b).