The Distribution of Practice

Since that time, however, partially because of shifting theoretical perspectives and the emergence of new topics of research interest (and also perhaps because many scholars thought little new was left to discover), few research studies have been directed toward further extending our understanding of practice distribution effects. There is, though, some evidence that this situation may be changing and a new era of research interest beginning (e.g., Dail & Christina, 2004; Utley & Astill, 2008).

In studying the influence of practice distribution within the area of motor skill Opens in new window learning, researchers have typically contrasted the effects of two types of practice schedules. These are distributed practice and massed practice.

Unfortunately, a good deal of confusion has resulted from the fact that researchers have not always used these terms in the same way, leading to problems when comparing the results of various studies.

This way of defining the two terms is subjective and comparative in nature, and is typically used to contrast practice schedules across several practice sessions or within a single session when discrete skills are studied (because discrete skills are often completed quickly, therefore always necessitating relatively longer rest intervals between practice attempts than the actual time in practice).

At other times, researchers also use a more objective method of defining both distribution schedules, particularly when distributive effects on continuous skills are contrasted within a single practices session.

Based upon the long history of research on practice distribution effects, strong consensus has emerged about the relative merits of distributed and massed practice schedules in some areas, whereas in other areas researchers remain divided (Dail & Christian, 2004; Lee & Genovese, 1988). Continuing controversies revolve mostly around the distributive influences vary for within and between practice sessions effects.

Within-Session Distribution Effects

In regard to the practice of motor skills Opens in new window within a single practice session, there is general consensus among researchers (echoed by practitioners) concerning the comparative effects of distributed and massed schedules on practice performance. Two major reviews of the literature support the conclusion that distributed practice promotes better acquisition performance than massed practice (Donovan & Radosevich, 1999; Lee and Genovese, 1988).

This may appear a somewhat obvious conclusion, because distributed practice, being more restful, leads to less fatigue and would therefore be expected to result in better acquisition performance than massed practice. The buildup of both physiological and psychological fatigue resulting from massed practice, imposing greater demands upon learners, can play a powerful role in depressing the ability to respond skillfully during practice.

In their extensive review of the literature, Lee & Genovese concluded that

Distributed practice is beneficial to both the performance and learning of motor skills, although the effect on performance is greater than the effect on learning. (Lee & Genovese, 1988, p. 282)

These reviewers also noted that almost all of the research reported contrasts the effects of distributed and massed practice on continuous skills. By their nature, continuous skills lend themselves more readily to experimental arrangements than do discrete skills because the amount of time devoted to actual work and rest can be more easily and objectively measured. In fact, in Lee & Genovese’s review, they found that only one of 116 studies reported involved a discrete skill.

In the absence of evidence to the contrary, most experts now believe that the massing of practice for discrete skills depresses practice performance, but not learning, when contrasted with distributed practice.

The few studies investigating distribution effects on discrete skills have initiated considerable debate concerning the relative effects of massed versus distributed practice for discrete skills. Although massing practice does depress practice performance, its effect on learning seems far less clear. The available evidence suggests that distributed practice is no more effective than is massed practice for the learning of discrete motor skills.

Some studies have also reported superior benefits for massed practice compared to distributed practice as the time interval between acquisition and retention tests increases (Dail & Christiana, 2004). Perhaps one reason this issue has remained unresolved, as well as under-investigated, is that no good theoretical explanation has been advanced to explain the differential distribution effects relative to continuous and discrete skills.

Whatever the reasons, though, there exists little evidence on which to draw conclusions concerning the relative effects of massed and distributed practice on the learning of discrete motor skills, though what evidence does exist suggests no disadvantage for massed practice and may even point toward benefits in the long-term retention of discrete skills.

Practice Implications for Within-Session Distribution

The implications of research for the practice scheduling of continuous skills seems clear, which is to distribute practice to provide frequent rest periods while seeing that the total time in rest exceeds that in actual practice.

The situation for the scheduling of discrete skills is not as readily clear, however, and requires careful analysis, observation, and a willingness on the part of instructors to make needed adjustments as practice results dictate. In deciding upon the spacing of practice activities, instructors would be well advised to consider the pragmatic constraints of the time available for practice.

Most practice situations, whether in schools, clinics, sports clubs, or on the job, are limited to predetermined amounts of time. A volleyball unit in a physical education class may be limited to twenty 30-minute sessions over a four-week period, and most physical therapy are limited to a set time and number of appointments by insurance providers.

In such cases, the urgency for scheduling sufficient practice experience within a limited timeframe must be weighed against the possible benefits of extending rest intervals within practice. Granting the ambivalent findings relative to massing practice, it is unequivocal that massing practice does allow for a greater amount of time in actual practice, and the amount of practice time is an essential ingredient for learning.

In devising the practice recipe, instructors must weigh the relative benefits of practice time against the needs of learners for additional rest, as well as those factors that help learners gain from more massed practice. Individuals in good physical condition, who are highly motivated and who have acquired a good grasp of the skills being practice, are most likely to benefit from additional massing of practice (Christina & Corcos, 1988).

Instructors might well keep in mind a conclusion drawn from the available research literature. If distributed practice benefits acquisition performance, its benefits for learning—if they exist—appear relatively slight. The imposition of time constraints on practice must be carefully considered, and the amount of actual practice time or number of practice trials deemed necessary to meet target goals balanced against the known beneficial distributive effects of practice.

Between-Session Distribution Effects

Besides making decisions regarding the spacing of practice within practice sessions, decisions concerning the distribution of practice over sequential sessions must also be made.

• Over what length of time should practice be distributed?
• Should practice sessions be scheduled daily, every other day, several days apart, or even further apart?
• Or, perhaps, would more than one daily practice session be most beneficial?
• Should sessions be short or long?
• Is it better to have shorter sessions frequently or longer sessions over a greater length of time?

These questions concern between-session distribution effects.

Fortunately, the results of research concerning the distributive effects of practice are much more definitive regarding between-session effects than is the case for within-session effects. Though relatively few studies have been reported, the available evidence clearly supports the benefits of distributive practice (Schmidt & Lee, 2005).

Measurements of how long it took individual workers to learn the keyboard as well as to achieve a target goal of typing 80 correct keystrokes per minute were recorded and averaged over the four groups. The results are shown in Table X.

As can be seen in Table X, the workers in the most distributed practice group had the best learning outcomes, while learning was the worst for those workers in the most massed practice group.

The number of hours it took workers to reach the target criterion of 80 correct keystrokes per minute was the measure of motor skill learning in this study. It took workers in the most distributed practice group 55 hours on average to reach this target, while after 80 hours of practice, the total hours available for training, the most massed practice group had failed to reach this target, averaging only 70 keystrokes per minute.

Further retention tests were administered at three, six, and nine months after completion of the original training.

Interestingly, when workers were asked to rate the effectiveness of their training, the most massed group rated their practice experiences the best, while the most distributed group rated theirs as the worst, the exact opposite of the actual training outcomes.

Based upon the results of their study, Baddeley and Longman concluded that shorter practice sessions distributed over a longer time were more beneficial for learning than were longer practice sessions massed over a shorter time period. Similar results have been reported in the few other studies to investigate distributive practice effects for between-session scheduling (1985; Bouzid & Crawshaw, 1987; Shea et al., 2000). For between-sessions scheduling, the consensus of the research literature is that distributed practice benefits learning more than does massed practice.

As a general guideline, shorter practice sessions, distributed over longer periods of time, are more beneficial for learning than are schedules where practice is massed in longer sessions and less times. One possible explanation for the beneficial effects of practice distribution observed over practice sessions has to do with memory consolidation Opens in new window.

Memories are consolidated (made permanent) through a series of biochemical processes. Those process take time to be accomplished. When practice is distributed, there is simply more time for consolidation processes to be completed.

There is also an aspect of serial order effects Opens in new window on learning. That is, those practice experiences occurring at the beginning and ending of practice sessions are the most easily consolidated into memory and therefore the best learned. Shorter and more frequent practice sessions also mean that more practice experiences occur in beginning and ending portions of practice than would be true for longer and fewer practice sessions, resulting in more practice experiences being effectively consolidated.

Spacing out learning experiences also weakens the effects of interference on consolidation processes. Interestingly, evidence has begun to emerge in recent years that sleep may be essential for the consolidation of procedural memories Opens in new window, including those for motor skills (Kuriyama, Stickgold, & Walker, 2004).

One benefit of distributed practice may be that spreading practice out over more days results in more of the actual practice experiences being followed by periods of sleep, leading to more effective consolidation of those experiences in memory. Another explanation for the relative benefits of distributed practice is the suggestion that massing practice may diminish the amount of cognitive effort exerted by learners after a certain point in practice is reached.

Practice experiences may become so repetitious during massed practice that they become monotonous and boring to learners, who will then decrease the amount of cognitive effort they exert, which in turn will decrease the learning outcomes of practice.

However, the causes underlying the benefits of distributed practice over massed practice are eventually sorted out, that distributed practice does promote better learning than massed practice is firmly established in the research literature. The question becomes one of how to translate this knowledge into effective practice designs.

Based upon the available research evidence, the implications for scheduling of between-session practice are fairly straightforward. Practice sessions that are shorter and distributed over a longer period (days/weeks, etc.) promote better learning than do practice sessions that are longer and massed over shorter periods of time. All other things being equal, instructors of motor skills are well advised to distribute practice experiences as much as possible.