Both Continuous and Discrete Skills Are Learned Better With Distributed Schedules
+ +
For most instruction and rehabilitation situations, the primary practice distribution concern is how to schedule within and between practice sessions for a specified amount of practice time. As we described earlier in this chapter, many instruction and rehabilitation situations have specified limits for the amount of available practice time. For example, in most clinical applications, a patient may receive treatment for only a limited number of sessions because of health care management restrictions. Also, in teaching and coaching situations, there is often little flexibility in the number of days available for classes or practice sessions. For example, if a teacher has only ten days for a unit of instruction, then the practice schedule must fit that limit. Similarly, if a dancer must perform in a concert that is one month away, then the rehearsal schedule must adjust accordingly. Thus, outside limitations may determine how many days a person should devote to practice. However, the instructor, coach, or therapist still decides the number of practice sessions and the length of each one.
+ +
massed practice a practice schedule in which the amount of rest between practice sessions or trials is very short.
distributed practice a practice schedule in which the amount of rest between practice sessions or trials is relatively long.
+ + +
The Benefit of More and Shorter Sessions
+ +
Research on the question of the length and distribution of practice sessions shows support for the benefit of distributed practice. This means that when experiments have compared a few long practice sessions with more frequent and shorter sessions, the results show that practicing skills during shorter sessions leads to better learning.
+ +
A classic example of research supporting this general conclusion is a study published many years ago by Baddely and Longman (1978). They were attempting to determine the best way to schedule training sessions for postal workers on a mail-sorting machine, which required operating a typewriter-like keyboard. The postal service had allotted a total of 60 hours and 5 days each week for training the workers. Although this available training time could be distributed in a variety of ways in terms of number of training sessions, the researchers distributed this amount of time in four different ways, which are described in table 17.1 in the Practice Schedule column. Two groups trained for one hour in each session. One of these two groups practiced for only one session each day, which resulted in a total training time of twelve weeks, whereas the second group had two sessions each day, thereby reducing the number of weeks in training to 6. Two other groups practiced for two hours in each session. One of these groups had only one session each day, whereas the other had two sessions per day. These latter two groups therefore had six weeks and three weeks of training, respectively. These four schedules illustrate the variety of ways the researchers distributed 60 hours of practice. The most distributed schedule required workers to train for twelve weeks, whereas the most massed schedule allowed them to complete their training in only three weeks. The primary difference between these schedules was in how long each session was and how many sessions occurred each day.
+ +
Table Graphic Jump Location
TABLE 17.1 Results of the Baddeley and Longman Experiment with Practice Distribution Schedules for Training Postal Workers
View Table| |Download (.pdf)
TABLE 17.1 Results of the Baddeley and Longman Experiment with Practice Distribution Schedules for Training Postal Workers
| Practice Schedule | Number of Hours to Type 80 Key-strokes/Minute |
|---|---|
| 1 hr/session–1 session/day (12 weeks training) | 55 |
| 1 hr/session–2 sessions/day (6 weeks training) | 75 |
| 2 hrs/session–1 session/day (6 weeks training) | 67 |
| 2 hrs/session–2 session/day (3 weeks training) | 80+ |
+ +
The results of this study are shown in table 17.1, which describes the number of hours the trainees required to achieve a typing speed of 80 keystrokes per minute, which was the motor performance goal for their training. Notice that only one of the four schedules (the most distributed schedule) resulted in the workers achieving this goal in the allotted training time of 60 hours (they did it in 55 hours). The other three groups required additional practice time. It is interesting that those in the most massed schedule group, which practiced two 2-hour sessions each day, never achieved this goal. After 80 hours of practice they were still doing only a little better than 70 keystrokes per minute.
+ +
Retention tests were given 1, 3, and 9 months after the workers had finished training. After 9 months, the most massed group performed worse on the typing speed test than the other groups, which performed about equally. Finally, the researchers obtained a very revealing result from the trainees' own ratings of the training schedules. Although most workers preferred their own schedule, those in the most massed group preferred theirs the most, whereas members of the most distributed liked theirs the least. Interestingly, these preferences were exactly opposite to the performance test results.
+ +
The results of this experiment indicate that fitting 60 hours of training into 3-week, where there had to be two 2-hour practice sessions each day, was a poor practice schedule. Although those in the most distributed schedule generally attained performance goals in the shortest time, they did not perform any better than two of the other groups on the retention tests. Given all the results, the authors concluded that the one-hour training sessions were more desirable than the two-hour sessions, and that one session per day was only slightly more effective than two sessions per day. However, having two 2-hour sessions each day was not a good training schedule.
+ +
More recent studies have shown similar learning benefits for distributed practice for a variety of motor skills, as the following examples demonstrate. Annett and Piech (1985) found that two 5 trial training sessions separated by one day led to better learning of a computer target-shooting game than one 10 trial session. One trial involved shooting at ten singly presented moving targets. On a retention test given one day after the end of the training session, the distributed group not only had more "hits" but also had less error in the shooting attempts.
+ +
Bouzid and Crawshaw (1987) reported similar results for the learning of word processing skills. Typists who practiced twelve skills during two sessions of 35 and 25 min each, separated by a 10 min break, required less time to learn the skills and had fewer errors on a test than typists who practiced the skills during one 60 min session.
+ +
Shea et al. (2000) showed that distributing practice sessions across days resulted in better learning than massing all the sessions within one day for a continuous dynamic balance task and a discrete key-press timing task. The results for the continuous balance task are shown in figure 17.2. Note that for the first session of trials (each trial involved 90 sec of continuous balancing), both the one-day practice (massed) and the two-day practice (distributed) groups performed similarly. However, during the second practice session the groups began to perform differently. By the end of this session, the distributed group, for whom this session was the next day, had significantly less balancing error. Importantly, this difference continued during the retention test, which each group performed one day after the end of the practice sessions.
+ +
FIGURE 17.2
Results of the experiment by Shea et al. in which one group participated in two practice sessions on one day (circles) and another participated in one session on each of two days (squares). The graph shows the amount of balancing error (RMSE, which was calculated as the amount of deviation, in degrees, from horizontal) for each 90 sec trial on a dynamic balance task. [Figure 3, p. 745 in Shea, C. H., Lai, Q., Black, C., & Park, J. C. (2000). Spacing practice sessions across days benefit the learning of motor skills. Human Movement Science, 19, 737–760.]
+ +
Finally, in a study involving learning to putt in golf, Dail and Christina (2004) had novice golfers practice putting a 3.7 m distance for 240 trials. One group followed a massed practice schedule in which they performed all the trials on one day, with short breaks between blocks of 10 trials. In contrast, another group practiced according to a distributed schedule of 60 trials per day for four consecutive days. The results, which you can see in figure 17.3, showed that at the end of 240 practice trials, the distributed practice group performed at a higher level than the massed schedule group. More importantly, this difference continued one and seven days later on retention tests of 60 trials. It is also interesting to note that at the end of each block of 10 trials during the practice sessions the experimenters asked the participants to predict their performance on the retention test. For this assessment of their own competence (i.e., metacognition), the participants who experienced the distributed schedule more accurately predicted their retention test performance.
+ +
FIGURE 17.3
Results of the experiment by Dail and Christina in which two groups practiced putting a golf ball either in 240 trials in one day (massed practice) or in four days of 60 trials each (distributed practice). The graph shows the results (a lower score is better) at the end of the practice trials and during retention tests 1 and seven days later. [Data from Dail, T. K., & Christina, R. W. (2004). Distribution of practice and metacognition in learning and long-term retention. Research Quarterly for Exercise and Sport, 75, 148–155, figure 1.]
+ +
Taken together, the results of these experiments support the learning advantage of distributed over massed practice schedules when the number and length of practice sessions is the concern. And when considered in terms of the types of motor skills involved in the experiments, the benefit of distributed practice extends to a variety of types of skills, which include discrete and continuous skills as well as open and closed skills. Finally, although the research does not give us a specific number and length of practice sessions that would be optimal for the learning of all motor skills, the conclusion that shorter and more practice sessions lead to better learning than longer and fewer sessions provides an excellent general principle on which to base specific decisions when planning practice, training, or rehabilitation sessions.
+ + +
Explanations for the Distributed Practice Benefit
+ +
There are at least three possible reasons why the distribution of practice sessions across more days leads to better learning than massing the sessions within fewer days. One is that fatigue negatively influences learning for massed practice schedules. Although none of the experiments discussed in this section assessed participants' levels of fatigue, it is possible to suspect that fatigue influenced learning because of the task performance requirements. For example, in the Shea et al. (2000) experiment, participants performed a continuous dynamic balance task for 90 sec on each trial. The massed practice schedule required them to perform 14 trials on the same day with only a 20 min break between trials 7 and 8. On the other hand, participants in the distributed practice schedule performed the second set of seven trials on the following day. Similarly in the Dail and Christina (2004) experiment, participants who experienced the massed practice schedule performed 240 putts in one session, with short rest breaks only after each set of 10 trials. In contrast, those who practiced according to the distributed schedule performed only 60 trials in each session.
+ +
A CLOSER LOOK Relating Practice Distribution and Contextual Interference to Skill Learning Contexts
The concept of contextual interference can be incorporated into practice distribution by organizing practice sessions to include principles related to both. The following are some examples that relate to three different skill learning contexts.
-
Physical education class. If several drills or other kinds of activity are planned for the day's lesson, use a station-organization approach by assigning each skill or activity to a location in the gym or on the field so that there are several stations. Divide the class into groups and assign each to a station. Let the groups stay in their stations for about 12–15 minutes and then rotate to the next station. Continue this rotation approach for the entire period. If the class period is sufficiently long, allow for two or more rotations.
-
Sports-related practice. Practices for team and individual sports typically include several activities. Rather than spend an extended amount of time on any one activity, divide in half the amount of time planned for each activity, and do each activity as two sets during practice. The two sets can be randomly or serially scheduled during the practice session.
-
Physical rehabilitation session. Like sports-related practice sessions, rehab sessions typically involve several activities. If the planned activities allow, apply the approach described for sports-related practice by dividing in half the total amount of time planned for each activity, and do each activity as two randomly or serially scheduled sets during the session.
+ +
Second, the massing of practice within a day or a few days may reduce the amount of cognitive effort used on each trial as practice continues beyond a certain critical amount. We considered this explanation earlier in this chapter when we discussed reasons why more practice beyond a certain amount could lead to diminished learning. The massing of practice trials may institute a practice condition in which performance of the skill on each trial becomes so repetitious that it becomes monotonous or boring. As a result, the learner begins to decrease the amount of cognitive effort involved in each trial, which in turn diminishes the level of learning.
+ +
The possibility that either or both fatigue and reduced cognitive effort accounted for the poorer learning that resulted from the massed compared to the distributed practice sessions can be seen in the results of the Shea et al. (2000) and Dail and Christina (2004) experiments. As you can see in figure 17.2, lower practice performance in both studies did not begin until the last several trials for the massed practice condition. This suggests that as participants continued to practice, the effects of fatigue and/or reduced cognitive effort eventually began to influence their performance in a negative way. And this influence affected not only their practice performance but also their retention test performance, indicating an influence on their learning the skills.
+ +
The third explanation relates to memory consolidation, which is a long-term memory storage process. The memory consolidation hypothesis proposes that to store in memory the relevant information we need to learn a skill, certain neurobiochemical processes must occur. These processes, which transform a relatively unstable memory representation into a relatively permanent one, require a certain amount of time without additional practice of the same skill. The distribution of practice across several days provides a better opportunity for the memory consolidation process to take place than does the massing of practice within a day or a few days (see Brashers-Krug, Shadmehr, & Bizzi, 1996; Shadmehr & Brashers-Krug, 1997; Simmons, 2011).
+ +
LAB LINKS
Lab 17 in the Online Learning Center Lab Manual provides an opportunity for you to experience a comparison of the effects of massed and distributed practice on the learning of a discrete motor skill.
Source: https://accessphysiotherapy.mhmedical.com/content.aspx?bookid=2311§ionid=179410838
0 Response to "Both Continuous and Discrete Skills Are Learned Better With Distributed Schedules"
Post a Comment