Winstein and colleagues (1999), using a novel extensionflexion
elbow reversal task to investigate skill learning in the
unaffected arm, found no significant differences in
acquisition, offline forgetting, and retention between
healthy subjects and patients with middle cerebral artery
territory strokes. In the affected hand, Raghavan and
colleagues (2006) demonstrated impaired learning despite
preserved execution ability; stroke patients had impaired
anticipatory scaling of grip force and load force
rate to the object weight in a grasping task, although they
could do so after performing the task with their unaffected
hand. Given the differences in tasks, amounts
of practice, effectors, and patient characteristics in the
above-mentioned studies, it is difficult to draw unifying
conclusions from these results. Patients may retain the
ability to learn certain tasks but not others, depending
on the information and processing requirements of the
task, and training with the unaffected side may offer
a way to acquire the information needed to perform
the task with the affected side.