Perceptual and Motor Skills, 2009, 109, 517-530.

© Perceptual and Motor Skills 2009

WHOLE AND PART PRACTICE: A META-ANALYSIS1

FABIO E. FONTANA OLDEMAR MAZZARDO

Health, Physical Education & Leisure Services Health and Physical Activity
University of Northern Iowa University of Pittsburgh

OVANDE FURTADO, JR. JERE D. GALLAGHER

Kinesiology and Sports Studies Department Health and Physical Activity

Eastern Illinois University University of Pittsburgh

Summary.—Motor learning textbooks provide specific recommendations for

the use of whole and part practice during motor skill acquisition. Magill recom-
mended basing decisions on Naylor and Briggs’ hypothesis of task complexity and
organization, while Schmidt and Wrisberg recommended basing decisions on skill
classification. To evaluate the accuracy of these recommendations, a meta-analysis
was conducted. Through a literature search, 44 articles were located, and 20 pro-
vided sufficient information to calculate effect sizes. Effect sizes were calculated
using means for whole and part practice for acquisition and retention. Although
several analyses were compromised by insufficient studies investigating whole and
part practice, mean effect sizes generally supported motor learning textbooks. To
better verify the empirical validity of recommendations for the use of whole and
part practice, more studies are necessary.

Different motor skills are taught in physical education classes with

the expectation that they will be useful in keeping students active for a
lifetime. Thus, it becomes crucial that motor skills are taught effectively.
One decision that physical educators and coaches must make is wheth-
er to teach a motor skill via part or whole practice. Part practice involves
breaking a skill into smaller units, whereas whole practice involves prac-
ticing a motor skill in its entirety. While part practice may simplify skill
rehearsal, it may also change the biomechanics of the task. Consequently,
it becomes important to develop empirical criteria to determine the appro-
priate use of whole and part practice.
Motor learning textbooks (Magill, 2007; Schmidt & Wrisberg, 2008),
which are widely used in teacher-preparation programs, provide specif-
ic recommendations for the use of whole and part practice during mo-
tor skill acquisition. Magill (2007) suggests basing decisions on the use of
whole and part practice on Naylor and Briggs’ hypothesis (1963). Using
their hypothesis, the decision to use whole or part practice during mo-
tor skill acquisition is based on the complexity and organization of tasks.
1
Address correspondence to Dr. Fabio E. Fontana, University of Northern Iowa, Health,
Physical Education & Leisure Services, 203 Wellness/Recreation Center, Cedar Falls, IA
50614-0241 or e-mail ([email protected]).

DOI 10.2466/PMS.109.2.517-530
518 F. E. FONTANA, ET AL.

Complexity refers to the cognitive demands of the task and is usually in-
ferred based on the number of task components. A golf swing and a dance
routine are examples of tasks with low and high complexity, respective-
ly. Although both are difficult to execute, a dance routine has many more
components than a golf swing. Tasks with a large number of components
have high cognitive complexity. Organization refers to how the different
elements of the task are related. Tasks with high organization have parts
that depend on each other. In other words, performance on one part is de-
pendent on how one executes the previous part.
Naylor and Briggs (1963) postulated that the part practice method
should be used for tasks that are high in complexity and low in organiza-
tion. An aerobics class routine has high complexity and low organization.
The large number of steps in aerobics routines makes this skill complex
since it poses high cognitive demands on the students. At the same time,
since execution of one step does not influence the execution of the next
step, an aerobics routine is considered a task with low organization. Oth-
er examples of tasks with high complexity and low organization are team
sports. In soccer, for example, there are several skills including passing,
dribbling, controlling, and kicking that can be learned separately since
learning one skill has little influence on learning any other soccer skill.
Whole practice, on the other hand, is advisable for tasks with low
complexity and high organization. Shooting in basketball is an example.
Shooting in basketball has a relatively low number of components, but the
components are dependent on each other. Practicing only the release of
the ball without any knee flexion completely changes the feel of the skill.
Other examples of tasks with low complexity and high organization in-
clude individual skills used in sports such as throwing in football, kicking
in soccer, and the tennis forehand.
Another popular motor learning textbook (Schmidt & Wrisberg, 2008)
endorses a slightly different approach. They suggest that decisions about
the use of whole and part practice should be based on task classification.
Discrete tasks, i.e., tasks with a well-defined beginning and end such as
the overhand throw, are more efficiently learned if taught as a whole. For
serial tasks, i.e., tasks consisting of a combination of multiple components
such as dance routines, Schmidt and Wrisberg borrowed the organization
classification developed by Naylor and Briggs (1963). Schmidt and Wris-
berg advise the use of part practice for serial tasks with low organization
and whole practice for serial tasks with high organization. Considerations
were not made for continuous tasks such as swimming or running.
Textbooks in motor learning give very clear recommendations about
the appropriateness of whole and part practice for different learning en-
vironments. Physical education teachers make use of these recommenda-
 USE OF WHOLE AND PART PRACTICE 519

tions as a way to assure effective instruction. The purpose of this research

was to verify whether the recommendations provided by these textbooks
can be confirmed scientifically. More specifically, the following research
hypotheses were tested: (1) whole practice is expected to be more efficient
for tasks low in complexity and high in organization, and part practice for
tasks high in complexity and low in organization (Magill, 2007), and (2)
whole practice is expected to be more efficient for discrete or serial tasks
with high organization compared to serial tasks with low organization
(Schmidt & Wrisberg, 2008).
To test these hypotheses, a meta-analysis was conducted. A meta-
analysis is a statistical review of literature that averages the differences be-
tween conditions for the entire literature to provide a measure of reliabil-
ity on differences observed across studies. The results of a meta-analysis
represent a statistical summary of the effectiveness of different conditions.
In addition to investigating the effectiveness of recommendations
present in two popular motor learning textbooks, the present study in-
vestigated whether effects found in studies conducted in externally valid
settings differed from effects found in laboratory studies—since research
tends to be more often conducted in a laboratory setting, but results typi-
cally are generalized to real settings. Finally, this study examined whether
studies with stronger internal validity demonstrated larger effect sizes for
the expected whole and part results. Internal validity was assessed as a
measure of research quality.
Method
Study Selection
Computer searches were conducted in MEDLINE, PsycINFO, and
SportDiscus. The key terms used were whole and part practice, part-task,
whole-task, fractionization, backward practice, forward practice, segmen-
tation, parts, complexity and organization, forward chaining, backward
chaining, whole task training, and chaining strategies. In addition, refer-
ences from key articles were cross-checked. Only published articles were
included in the meta-analysis.
Data Coding
Relevant characteristics were coded from the studies: task classifica-
tion, task organization, task complexity, type of motor program used, in-
ternal validity, and external validity (Table 1). The selection of the coded
characteristics was based on the hypothesis being tested. In rare cases, a
task was classified as undetermined when not enough details were pro-
vided. Each study was coded individually by two researchers. All discrep-
ancies were discussed and resolved prior to entering the data into SPSS.
 TABLE 1 520
Whole and Part Practice Studies
Reference Task Organization Complexity Task Variables Effect Size
Classification Measured
Acquisition Retention
1. Ash & Holding Musical key- Low High Serial Melodic error and tim- ES1 = –.61
(1990) board task ing error accuracy ES2 = –1.72
2. Frederiksen & White Space Fortress High High Continuous Game score ES1 = –1.77
(1989) game ES2 = –1.92
3. Goettl & Shute (1996)
Experiment 1 Slalom flight High Mixed Continuous Speed and accuracy of ES1 = 1.72
task performance ES2 = 0

Experiment 2 Slalom flight High Mixed Continuous Speed and accuracy of ES1 = 0
task performance ES2 = 0
4. Johns (1989) Typewriting Low High Serial Typing speed ES = 0
5. Knapp & Dixon Juggling High Low Continuous Time to learn 100 con- ES1 = .73
(1952) secutive catches
6. Lersten (1968) Movement Low Low Serial Learning score ES = –.03
sequence task
F. E. FONTANA, ET AL.

7. Ma & Trombly Signature task High Low Continuous Movement time ES1 = 1.90
(2001) ES2 = 2.10
8. Mane, et al. (1989) Video game Low High Continuous Game score ES = –1.3
9. McGuigan & Rifle marksman- Low High Serial Total score (accuracy) ES = .34
MacCaslin (1955) ship
10. Murray (1981) Judo High High Serial Performance in judo ES1 = 2.06
contest
11. Nettelbeck & Kirby Assembling a Low High Serial Error ES1 = –1.04 ES1 = .17
(1976) sewing ES2 = –.69 ES2 = –.05
machine
12. Park, et al. (2004) Movement se- Low High Serial Speed of execution ES = 0 ES1 = 0
quence task ES2 = –.159
(continued on next page)
 TABLE 1 (cont’d)
Whole and Part Practice Studies
Reference Task Organization Complexity Task Variables Effect Size
Classification Measured
Acquisition Retention
13. Peck & Detweiler Virtual subma- Low High Serial Overall performance ES1 = –.87 ES1 = – .17
(2000) rine ES2 = –3.17 ES2 = 1.22
14. Rubin-Rabson Piano playing High Low Serial Number of trials to ES1 = –.06 ES1 = –.03
(1940) learn task ES2 = –.02 ES2 = –.07
15. Smith (1999)
Experiment 1 Aerobic dance Low High Serial Error ES1 = 0 ES1 = –1.55
routine ES2 = 0 ES2 = –2.03
Experiment 2 Aerobic dance Low High Serial Error ES1 = 0 ES1 = 1.16
routine ES2 = 0 ES2 = –.30
16. Stammers (1980) Pursuit tracking High Mixed Continuous Average score ES = 1.57
17. Walls, et al. (1984) Assembly task Low High Serial Error ES = –.48
18. Watters (1990)
Experiment 1 Keyboarding Low High Serial Error ES1 = –.97
ES2 = –1.29
Experiment 2 Keyboarding Low High Serial Error ES1 = –1.26
ES2 = –1.64
19. Watters (1992)
USE OF WHOLE AND PART PRACTICE

Experiment 1 Keyboarding Low High Serial Error ES1 = –.87

ES2 = – 1.16
Experiment 2 Keyboarding Low High Serial Error ES1 = –1.38
ES2 = –1.62
20. Wightman & Virtual flight Undetermined Undetermined Continuous Error ES = –1.84
Sistrunk (1987) landing
521
522 F. E. FONTANA, ET AL.

Task classification referred to whether tasks were discrete, continu-

ous, or serial. Discrete referred to tasks with a well-defined beginning and
end (i.e., overhand throw), continuous referred to tasks that are repetitive
in nature (i.e., swimming), and serial referred to a combination of tasks
(i.e., dance routine).
Task organization was defined by the relationship among task parts.
Tasks were classified as having high organization when performance of
one part depended on the performance of the previous part, and low or-
ganization when parts were independent of one another. Task complexity
was defined by the interaction among levels of expertise, number of task
parts, and cognitive ability. If subjects were described as experts, tasks
were considered simple since they had a lot of experience performing the
tasks. For novice subjects without special needs, the task was considered
simple when it had eight or fewer parts. Tasks that had more than eight
parts were considered complex. Any task for individuals with special
needs was considered to be complex.
Coding variables were created for assessing external and internal va-
lidity of the studies in the meta-analysis. External validity was classified
based on the setting and task used in the experiment. The possible clas-
sifications were real-world setting and task, real-world task and artificial
setting, or both setting and task lacked real-world features. Internal valid-
ity was coded using a 5-point scale. If the sum of scores was between 0
and 3.5, the research study was categorized as having low internal valid-
ity, while scores of 4 or higher denoted high internal validity. Scores were
calculated by summing the points of the following questions: were the
participants carefully described in terms of age, experience level, and sex
(.5)? Were subjects randomly assigned to groups (1)? Were the conditions
clearly described (1)? Were all procedures carefully explained, appropri-
ate, and followed (1)? Was the statistical analysis properly described (.5)?
Were the statistics and design appropriately used (1)? Each of these ques-
tions received either 1 or .5 point, depending upon predetermined crite-
ria.
Effect Size Coding
The dependent variable in a meta-analysis is the effect size, which is
an indicator of the magnitude of the treatment effect in standard devia-
tion units. In this study, the effect sizes calculated represent the difference
in treatment effect between whole and part practice groups. Some studies
included more than one type of part practice or more than one dependent
variable. To avoid giving more weight to these studies, no more than two
effect sizes were calculated for each study. The criterion for choosing effect
sizes was based upon the main hypothesis being tested.
 USE OF WHOLE AND PART PRACTICE 523

Data Analysis
Effect sizes were calculated as the difference between the means for
whole minus part practice groups divided by their pooled standard de-
viation. Pooled standard deviation accounts for sample size variation
among research studies. Effect size was calculated by using the formula
ES = (Me – Mc) / pooled SD, where Me is the experimental mean and Mc is
the control mean. When means or standard deviations were not reported,
effect sizes were calculated using the F value with 1 degree of freedom
as in Equation 1 or by the simplified Equation 2 for equal subject groups:

F (Nc + Ne)
ES = [1]
Errordf Nc · Ne

2 F
ES = [2]
Errordf

According to Hedges and Olkin (1985), effect sizes are positively bi-
ased in small samples; however, a virtually unbiased estimate of the effect
size can be obtained by multiplying each effect size by a correction factor
c. We approximately corrected for this using the formula: c = 1 – (3/4m – 9),
where (m = Ne + Nc – 2). Ne represents the number of experimental subjects,
and Nc is the number of control subjects. Furthermore, each individual
corrected effect size was weighted by the reciprocal of its variance, giving
greater weight to the more reliable (lower variance) estimated effect siz-
es and thus yielding a truer estimate of population effect size (Hedges &
Olkin, 1985). The individual variance Equation 3 is:

Ne + Nc [3]
Var(ESi ) = + (ESi )2 2(Ne + Nc )
Ne Nc

After the calculation of corrected and weighted effect size, a test for
a normal distribution was performed. Following the normal distribu-
tion test, effect size was used as an estimator of overall treatment effect
(Hedges, 1981). For each of the corrected and weighted effect sizes, over-
all means and a 95% confidence interval were calculated. The effect sizes
calculated are independent; thus, not all effect sizes can be coded in all cat-
egories. In other words, not all proposed coding characteristics appeared
in all the coded studies. Univariate ANOVAs and t tests were performed
at the p = .05 level of significance whenever sample size was sufficient.
Results
The literature search yielded 44 studies related to the topic of whole
524 F. E. FONTANA, ET AL.

and part practice. Twenty-four studies were excluded from the meta-anal-
ysis for the following reasons: 15 articles did not provide sufficient infor-
mation (means, standard deviations, and t or F values with df ) to calculate
effect sizes; three articles did not have a whole practice condition; two ar-
ticles combined whole and part practice results; three articles dealt with
a cognitive task rather than a motor task. Finally, one study was excluded
because it was an animal study.
From the 44 studies, 20 met the participation criteria to calculate effect
sizes. Effect sizes were calculated based on the difference between whole
and part practice for acquisition and retention. For acquisition, 20 effect
sizes were computed. For retention, 34 effect sizes were computed. Four
studies had two experiments. Eight studies did not measure acquisition,
and two studies did not measure retention.
Results that were not significant or in which the statistics were not re-
ported were given a zero effect size; three acquisition and four retention
effect sizes were zero. All other significant and nonsignificant results were
calculated. Effect sizes that are positive indicate better performance by the
whole practice group, while negative effect sizes indicate better perfor-
mance by the part practice group.
Comparisons based on external and internal validity were made first
to assess whether studies needed to be grouped into categories prior to
further analysis. The data analysis revealed no significant results. The ef-
fect sizes for real-world experiments were not significantly different from
the effect sizes of laboratory studies during acquisition (t19 = .39, p = .70)
and retention (t32 = –1.19, p = .24). The effect sizes of highly controlled stud-
ies were not statistically different from those of less controlled studies
during acquisition (t19 = .97, p = .35) and retention (t32 = –.86, p = .40). These
results suggest that the effects of whole and part practice on motor skill
acquisition are not dependent on the quality of the study or the extent to
which the study reflects real life conditions. The results also suggest that
further comparisons involving the effectiveness of whole and part practice
should progress without splitting studies based on external and internal
validity criteria.
In the following sections, the effectiveness of whole and part practice
methods is tested. First, the overall comparison between part and whole
practice is made. Finally, the scientific veracity of recommendations con-
cerning the use of whole and part practice present in motor learning text-
books is assessed (Magill, 2007; Schmidt & Wrisberg, 2008).
Overall Comparison
The overall comparison resulted in no difference between whole and
part practice for either acquisition (t19 = .391, p = .7) or retention (t32 = –1.196,
p = .24). The overall results for acquisition and retention suggested that
 USE OF WHOLE AND PART PRACTICE 525

without any further consideration for type of skill, whole and part prac-
tice are equally effective. However, more important than the overall ef-
fect was to verify whether the more specific recommendations provided
in motor learning textbooks were confirmed by the results of experiments
in this area.
To test Naylor and Briggs’ hypothesis, two 2 (Complexity) × 2 (Orga-
nization) analyses of variance (ANOVAs) were conducted on the depen-
dent variable effect size, one for acquisition and another for retention. A
significant interaction between complexity and organization was neces-
sary to support the hypothesis. For acquisition, the interaction between
complexity and organization and the main effect for organization were
not calculated due to insufficient studies. The main effect for complex-
ity was not significant (F1,16 = .11, p = .51). Although the interaction was not
calculated, there was a trend for the mean effect sizes of tasks with low
organization and high complexity to support Naylor and Briggs’ hypoth-
esis. The negative mean effect size for tasks with low organization and
high complexity suggests that part practice is the more effective method
of practice during acquisition and retention (Table 2).
For retention, the interaction between complexity and organization
was not calculated due to insufficient studies. The main effects for com-
plexity (acquisition = F1,25 = .13, p = .71) and organization were not signifi-
cant (F1,25 = 1.55, p = .23). Although the interaction was not calculated, the
trend for the mean effect sizes favored Naylor and Briggs’ hypothesis. As
predicted, mean effect sizes indicated that tasks with low organization and
high complexity benefit most from part practice, and tasks with high or-
ganization and low complexity benefit most from whole practice (Table 2).
TABLE 2
Results for Task Complexity and Organization
Phase Low Complexity High Complexity
M SD n M SD n
Acquisition Low organization –.30 1 –.651 .88 18
High organization .00 1 0
Retention Low organization 0 –.501 1.01 22
High organization 1.182 .99 4 –.87 2.61 3
Note.—M represents Mean Effect Size. Main effects and interactions were not significant.
Based on mean effect sizes, 1part practice was more effective for tasks with high complexity
and low organization, and 2whole practice was more effective for tasks with low complexity
and high organization.

Task Classification
For acquisition, of the 20 effect sizes, 18 were serial with low organi-
zation, two were serial with high organization, and none was continuous
or discrete. Due to the lack of studies, discrete tasks were not analyzed. A
526 F. E. FONTANA, ET AL.

t test comparing serial tasks of high and low organization was not signifi-
cant (t18 = –.23, p = .81). However, as predicted, mean effect sizes indicated
that part practice produced better performance for serial tasks with low
organization, and whole practice for serial tasks with high organization
(Table 3).
For retention, of the 33 effect sizes, 22 were for serial tasks with low
organization, two for serial tasks with high organization, nine for continu-
ous tasks, and none for discrete tasks. Due to the lack of studies, discrete
tasks were not analyzed. There were only two effect sizes for serial tasks
with high organization, and a t test comparing serial tasks of high and low
organization was not significant (t22 = –1.81, p = .08). As predicted, mean ef-
fect sizes indicated that serial tasks with low organization were more ef-
fectively practiced using the part method and serial tasks with high orga-
nization using the whole method (Table 3). Based on a positive mean effect
size (M = .19, SD = 1.68; n = 9), continuous tasks apparently benefited most
from whole practice.
TABLE 3
Results for Task Classification
Phase Serial Tasks With
Low Organization High Organization
M SD n M SD n
Acquisition –.621 .86 18 .792 1.10 2
Retention –.501 1.01 22 1.032 1.01 2
Note.—M represents Mean Effect Size. Student t test comparisons did not reach
significance. 1Negative mean effect sizes suggest that part practice was more effec-
tive for serial tasks with low organization, and 2positive mean effect sizes suggest
that whole was more effective for serial tasks with high organization.

Discussion
Deciding what type of practice is most effective for teaching a spe-
cific motor skill is critical for physical education teachers and coaches, yet
the literature on whole and part practice is scarce. Despite the lack of evi-
dence, popular motor learning textbooks (Magill, 2007; Schmidt & Wris-
berg, 2008) provide specific advice for the use of whole and part practice
during motor skill acquisition. Consequently, it becomes imperative to as-
sess the empirical validity of the recommendations.
Magill’s recommendations were based on Naylor and Briggs’ hypoth-
esis (1963). To fully support this hypothesis, the interaction between com-
plexity and organization should be significant. The interaction could not
be tested due to insufficient studies. Although the interaction was not di-
rectly tested, the results generally supported Naylor and Briggs’ hypoth-
esis. Based on mean effect sizes, part practice was more effective for tasks
 USE OF WHOLE AND PART PRACTICE 527

with high complexity and low organization, and whole practice for tasks
with low complexity and high organization.
Schmidt and Wrisberg (2008) advocated that part practice should be
used to teach serial tasks with low organization and whole practice for
discrete or serial tasks with high organization. Although not significant,
mean effect sizes were in line with textbook predictions. Part practice was
more effective for serial tasks with low organization, and whole practice
was more effective for serial tasks with high organization.
Schmidt and Wrisberg (2008) also suggested that whole practice is
more effective for discrete tasks based on a study by Lersten (1968). In this
study, predictions about discrete tasks were not tested due to the lack of
studies. Disagreement was probably due to the fact that the task used in
the Lersten study was classified as discrete by Schmidt and Wrisberg and
as serial in the present study. The task consisted of two components: cir-
cular and linear. For the circular component, the subject rotated an arm
crank for a complete clockwise revolution. For the linear part, the subject
moved his arm in a straight path for 12 in. before knocking down a tar-
get. None of the other experiments present in the literature seems to have
measured discrete tasks. Consequently, the effectiveness of part practice
for discrete tasks cannot be assessed at the moment. Discrete tasks such as
throwing, catching, and kicking are common in physical education class-
es. It is critical that studies using discrete tasks be developed.
Whole and part practice is a critical issue for physical education
teachers designing effective motor skill acquisition sessions. However, the
analyses of the recommendations contained in motor learning textbooks
were compromised by the scarcity of studies on whole and part practice.
Although several analyses were compromised, mean effect sizes were in
agreement with information contained in motor learning textbooks. To
better verify the empirical validity of recommendations for the use of
whole and part practice, more studies need to be conducted.
REFERENCES* 2

Ash, D. W., & Holding, D. H. (1990) Backward versus forward chaining in the acqui-
sition of a keyboard skill. Human Factors, 32, 139-146.
Barry, G., & Valerie, S. (1996) Analysis of part-task training using the backward-
transfer technique. Journal of Experimental Psychology: Applied, 2, 227-249.*
Blake, K. A., & Williams, C. L. (1969) Retarded, normal, and superior subjects’ learn-
ing of paired associates by whole and parts methods. Psychological Reports, 25,
319-324.*
Briggs, G. E., & Naylor, J. C. (1962) The relative efficiency of several training methods
as a function of transfer task complexity. Journal of Experimental Psychology, 64,
505-512.*

*Articles were retrieved but not used in the meta-analysis.

528 F. E. FONTANA, ET AL.

Fabiani, M., Buckley, J., Gratton, G., Coles, M., & Donchin, E. (1989) The training of
complex task performance. Acta Psychologica, 71, 259-299.*
Fezzani, K., Thon, B., & Lagarde, J. (2000) Can different levels of S-R practice influ-
ence sequence learning? An investigation into the context of a perceptual-motor
sequence learning task. Perceptual and Motor Skills, 91, 463-475.*
Frederiksen, J. R., & White, B. Y. (1989) An approach to training based upon prin-
cipled task decomposition. Acta Psychologica, 71, 89-146.
Goettl, B. P., & Shute, V. J. (1996) Analysis of part-task training using the backward
transfer technique. Journal of Experimental Psychology, 2, 227-249.
Hedges, L. V. (1981) Distribution theory for Glass’s estimator of effect size and related
estimators. Journal of Educational Statistics, 6, 107-128.
Hedges, L. V., & Olkin, I. (1985) Statistical methods for meta-analysis. New York: Aca-
demic Press.
Johns, L. (1989) Hemispheric preference and progressive-part or whole practice in
beginning typewriting. Perceptual and Motor Skills, 68, 359-364.
Keele, S. W., Jennings, P., Jones, S., Caulton, D., & Cohen, A. (1995) On the modular-
ity of sequence representation. Journal of Motor Behavior, 27, 17-30.*
Knapp, C. G., & Dixon, W. R. (1952) Learning to juggle: II. A study of whole and part
methods. The Research Quarterly, 23, 398-401.
Lersten, K. C. (1968) Transfer of movement components in a motor learning task. The
Research Quarterly, 39, 575-581.
Ma, H., & Trombly, C. A. (2001) The comparison of motor performance between part
and whole tasks in elderly persons. The American Journal of Occupational Therapy,
55, 62-67.
Magill, R. (2007) Motor learning: concepts and applications. (8th ed.) Boston, MA: Mc-
Graw-Hill.
Mane, A. M., Adams, J. A., & Donchin, E. (1989) Adaptive and part-whole training in
the acquisition of a complex perceptual-motor skill. Acta Psychologica, 71, 179-196.
McGuigan, F. J. (1960) Variation of whole-part methods of learning. Journal of Educa-
tional Psychology, 51, 213-216.*
McGuigan, F. J., & MacCaslin, E. F. (1955) Whole and part methods in learning a
perceptual motor skill. The American Journal of Psychology, 68, 658-661.
Murray, J. F. (1981) Effects of whole vs part method of training on transfer of learning.
Perceptual and Motor Skills, 53, 883-889.
Naylor, G., & Briggs, J. (1963) Effects of task complexity and task organization on
the relative efficiency of part and whole training methods. Journal of Experimental
Psychology, 65, 217-224.*
Nettelbeck, T., & Kirby, N. H. (1976) A comparison of part and whole training meth-
ods with mildly mentally retarded workers. Journal of Occupational Psychology, 49,
115-120.
Newell, K. M., Carlton, M. J., Fisher, A. T., & Rutter, B. G. (1989) Whole-part train-
ing strategies for learning the response dynamics of microprocessor driven simu-
lators. Acta Psychologica, 71, 197-216.*
Park, J. H., Wilde, H., & Shea, C. H. (2004) Part-whole practice of movement sequenc-
es. Journal of Motor Behavior, 36, 51-61.
 USE OF WHOLE AND PART PRACTICE 529

Peck, A., & Detweiler, M. C. (2000) Training concurrent multistep procedural tasks.

Human Factors, 42, 379-389.
Pisacreta, R. (1982) A comparison of forward and backward procedures for the ac-
quisition of response chains in pigeons. Bulletin of the Psychonomic Society, 20, 233-
236.*
Povel, D. J., & Collard, R. (1982) Structural factors in patterned finger tapping. Acta
Psychologica, 52, 107-123.*
Rubin-Rabson, G. (1940) Studies in the psychology of memorizing piano music: III. A
comparison of the whole and the part approach. The Journal of Educational Psychol-
ogy, 31, 460-476.
Schmidt, R. A., & Wrisberg, C. A. (2008) Motor learning and performance: a problem-
based learning approach. (4th ed.) Champaign, IL: Human Kinetics.
Smith, G. J. (1999) Teaching a long sequence of behavior using whole task training,
forward chaining, and backward chaining. Perceptual and Motor Skills, 89, 951-965.
Spooner, F. (1984) Comparisons of backward chaining and total task presentation in
training severely handicapped persons. Education and Training of the Mentally Re-
tarded, 19, 15-22.*
Spooner, F., Weber, L. H., & Spooner, D. (1983) The effects of backward chaining and
total task presentation on the acquisition of complex tasks by severely retarded
adolescents and adults. Education and Treatment of Children, 6, 401-420.*
Stammers, R. B. (1980) Part and whole practice for a tracking task: effects of task vari-
ables and amount of practice. Perceptual and Motor Skills, 50, 203-210.
Stammers, R. B. (1982) Part and whole practice in training for procedural tasks. Hu-
man Learning, 1, 185-207.*
Summer, J., & Kennedy, T. (1992) Strategies in the production of a 5:3 polyrhythm.
Human Movement Science, 11, 101-112.*
Verwey, W. B. (2001) Concatenating familiar movement sequences: the versatile cog-
nitive processor. Acta Psychologica, 106, 69-95.*
Verwey, W. B., & Dronkert, Y. (1996) Practicing a structured continuous key-pressing
task: motor chunking or rhythm consolidation. Journal of Motor Behavior, 28, 71-
79.*
Walls, R. T., Dowler, D. L., Haught, P. A., & Zawlocki, R. (1984) Progressive de-
lay and unlimited delay of prompts in forward chaining and whole task training
strategies. Education and Training of the Mentally Retarded, 19, 276-284.
Walls, R. T., Zane, T., & Ellis, W. D. (1981) Forward and backward chaining, and
whole task methods: training assembly tasks in vocational rehabilitation. Behavior
Modification, 5, 61-74.*
Walls, R. T., Zane, T., & John, T. E. (1980) Trainers’ personal methods compared to
two structured training strategies. American Journal of Mental Deficiency, 84, 495-
507.*
Watters, R. G. (1989) Human sequenced behavior: a comparison of forward chaining,
backward chaining, and whole task training procedures. Journal of Human Move-
ment Studies, 16, 183-190.*
Watters, R. G. (1990) Human sequenced behavior: a comparison of forward chaining,
backward chaining, and whole task training at different sequence lengths. Journal
of Human Movement Studies, 18, 25-35.
530 F. E. FONTANA, ET AL.

Watters, R. G. (1992) Retention of human sequenced behavior following forward

chaining, backward chaining, and whole task training procedures. Journal of Hu-
man Movement Studies, 22, 117-129.*
Watters, R. G., & Scott, J. M. (1992) Teaching human sequenced behavior with and
without prompts: a comparison of forward chaining, backward chaining, and
whole task procedures. Journal of Human Movement Studies, 22, 239-249.*
Weber, N. J. (1978) Chaining strategies for teaching sequenced motor tasks to men-
tally retarded adults. The American Journal of Occupational Therapy, 32, 385-389.*
Weis, K. M. (1978) A comparison of forward and backward procedures for the acqui-
sition of response chains in humans. Journal of the Experimental Analysis of Behavior,
29, 255-259.*
Wightman, D. C., & Sistrunk, F. (1987) Part-task training strategies in simulated car-
rier landing final-approach training. Human Factors, 29, 245-254.
Wilcox, B. (1974) The teaching of serial tasks using chaining strategies. British Journal
of Psychology, 44, 175-183.*
Zane, T., Walls, R. T., & Thvedt, J. E. (1981) Prompting and fading guidance proce-
dures: their effect on chaining and whole task teaching strategies. Education and
Training of the Mentally Retarded, 16, 125-135.*

Accepted September 23, 2009.