T2 T3

Figure 7.1. Examples of the micro-scheduling of practice for motor-learning studies. (a) Practice of five trials of a single task in a massed and distributed schedule. (b) Practice of three different tasks for eight trials each in schedules with varying levels of contextual variety. T1: Task 1; T2: Task 2; and T3: Task 3.

Figure 7.1. Examples of the micro-scheduling of practice for motor-learning studies. (a) Practice of five trials of a single task in a massed and distributed schedule. (b) Practice of three different tasks for eight trials each in schedules with varying levels of contextual variety. T1: Task 1; T2: Task 2; and T3: Task 3.

dozens of seconds to weeks in duration. A benefit for distributed practice has been shown in a variety of motor tasks from pursuit rotor (Bourne and Archer, 1956) to typing (Baddeley and Longman, 1978). Lee and Genovese (1989), however, found that a distributed practice schedule did not benefit learning of a discrete tapping task compared to a massed schedule. In addition, others have suggested that a distributed practice schedule may not always be optimal and that boundary conditions likely exist related to task type, overall task complexity, and the length of the inter-trial interval (Donovan and Radosevich, 1999).

Distributing practice of several items during a rehabilitation session, however, would be extremely time consuming if the inter-trial interval contained only empty time. An alternative would be to insert other tasks within the vacant "space" so that multiple tasks could be practiced simultaneously. Contextual interference (CI) is "the interference effects in performance and learning that arise from practicing one task in the context of other tasks" (Schmidt and Lee, 1999, p. 412). It is well accepted that a random practice schedule (Fig. 7.1(b)) provides greater CI than a blocked schedule and is generally thought to be better for learning motor skills in healthy adults (Shea and Morgan, 1979; Lee and Magill, 1983; Shea et al., 1990; Sekiya et al., 1994, 1996; Wright and Shea, 2001). It is important to note that a random practice schedule also provides space between repeated trials of the same task. For example, a blocked schedule for the practice of three tasks would provide 0 intervening items between presentations of the same task.

A random schedule, in comparison, would provide a range from 0 to 4 intervening items between presentations of a single item. Therefore, a random practice schedule provides greater spacing, or number of intervening items, than a blocked schedule.

Why might a random practice schedule be better for learning than a blocked practice schedule? One hypothesis is that the processing demands for learning are different for these two practice schedules. In a blocked schedule, the learner knows which item will be presented on the next trial. In a random schedule, the learner does not know which item will be presented next. Therefore, a random schedule is thought to require greater attention and deeper processing (Shea and Zimny, 1983; Lee and Magill, 1985; Lee and Maraj, 1994). However, Lee and Magill (1983) showed that a serial schedule, where the task schedule was predictable but not blocked, was equivalent in effect to that of a random schedule. The predictability of the next task was not the critical factor in this study, but rather the process of determining the movement requirements, and preparing and generating a new solution with each trial. This process required the learner to be more actively engaged in the practice session under random or serial task schedules compared with the blocked task schedule. This analysis suggests that it is the engagement in information processing surrounding task performance and not just the task performance itself that is critical for effective motor learning. In this case, what is learned enables the individual to perform the task with skill effectively with differing speeds and directions and under a variety of environmental contexts.

Recently, though, some authors have questioned whether the CI effect generalizes to all tasks and conditions. It is not clear if random practice is always better for learning more difficult, real-world tasks (Wulf and Shea, 2002; Brady, 2004). Albaret and Thon (1998) found such an effect when studying the learning of a drawing task. Random practice was better for learning the easy version but not the more difficult version of the task. Applying a more difficult practice schedule to an already difficult task may make the processing demands too high for adequate learning (Wulf and Shea, 2002). The skill level of the learner is another variable that may play a role in designing optimal practice schedules for learning. It has been suggested that novices may benefit from a blocked schedule while those who have more experience may benefit from a random schedule (Guadagnoli and Lee, 2004). Some have supported this idea (Shea et al., 1990) while others have not (Ollis et al., 2005).

Only two studies have examined the effect of practice schedules on neural recovery from stroke. Hanlon (1996) investigated the effect of practice schedules on learning a functional reaching task with the hemiparetic UE in individuals post-stroke. Subjects were assigned to either a blocked practice, random practice, or no practice control group. While there was no significant difference between the two practice groups for the number of trials to reach criterion (three consecutive successful trials), the random group had a significantly greater number of successful trials on the 2- and 7-day retention tests compared to the blocked group. Thus, this study provides some evidence that random practice may be more beneficial than blocked practice to improve function of the paretic limb in patients with stroke. However, because the functional task used was a complex five-step sequential task that involved grasping and moving a cup, success of the task was only measured by the cup not being dropped, and no kinematics or force measures were taken, it is hard to know which motor control parameters benefited from the random schedule. Furthermore, the random practice group practiced additional movements with the hemiparetic UE between presentations of the criterion task giving subjects in that group overall more practice in moving the arm, thus confounding a precise interpretation of the results.

In a study by Cauraugh and Kim (2003), subjects post-stroke performed three different movements (wrist/finger extension, elbow extension, and shoulder abduction) with the hemiparetic UE. In the blocked and random practice groups, movements were completed with a combination of active muscle activity and neuromuscular stimulation. A control group performed the same movements without stimulation. After four sessions lasting 90 min each over a

2-week period, subjects in both experimental groups performed better than controls on several behavioral measures but there was no difference in performance between the blocked and random groups. The lack of difference between groups practicing under different schedules is partially explained by the nature of the task, which was more a force production task than a functional task (see Winstein et al., 2004) and did not require the demands of skill acquisition.

Stroke severity and lesion location are important factors to consider when discussing motor learning in this population. Some work has suggested that individuals with more severe motor impairments may not demonstrate the same degree of short-term performance change (Dancause et al., 2002; Cirstea et al., 2003) or learning (Pohl et al., 2001) as individuals with less severe symptoms. This is a relatively open area for additional research to better understand the relationship between stroke severity, concomitant motor control deficits, and motor learning. The effect of lesion location on motor skill learning is not well understood either. Recent work has shown that if the primary locus of the stroke-affected area is the basal ganglia, there is more difficulty using explicit information to benefit implicit motor learning than age-matched controls (Boyd and Winstein, 2004). However, this is only a beginning. More research is needed to determine if differences in motor learning exist for individuals with specific stroke locations and levels of severity. For example, the optimal practice schedule may need to be tailored differently to accommodate individuals with more severe motor impairments compared with those with less severe deficits. More importantly, in order to effectively incorporate motor-learning principles into rehabilitation treatments and to design effective protocols, the interaction of stroke severity, lesion location, and conditions of practice will need a concentrated and focused research program.

Was this article helpful?

0 0

Post a comment