Research Report
Measuring Physical Fitness in Children
Who Are 5 to 12 Years Old With a Test
Battery That Is Functional and Easy to
Administer
Background. Valid and reliable measures of children’s physical fitness are necessary for investigating the relationship between children’s physical fitness and
children’s health.
Objective. The objective of this study was to estimate the feasibility, internal
consistency, convergent construct validity, and test-retest reliability of a new, functional, and easily administered test battery for measuring children’s physical fitness.
I. Fjørtoft, PhD, Faculty of Arts,
Folk Culture and Teacher Education, Telemark University College,
3670 Notodden, Norway. Address
all correspondence to Dr Fjørtoft
at: ingunn.fjortoft@hit.no.
A.V. Pedersen, PT, MSc, SørTrøndelag University College,
Trondheim, Norway.
fitness tests across age groups 5 to 12 years.
H. Sigmundsson, PhD, Norwegian
University of Science and Technology, Trondheim, Norway.
Methods. Each of the 9 items in the test battery consists of a compound motor
B. Vereijken, PhD, Norwegian University of Science and Technology.
activity that recruits various combinations of endurance, strength (force-generating
capacity), agility, balance, and motor coordination: standing broad jump, jumping a
distance of 7 m on 2 feet, jumping a distance of 7 m on one foot, throwing a tennis
ball with one hand, pushing a medicine ball with 2 hands, climbing wall bars,
performing a 10 ⫻ 5 m shuttle run, running 20 m as fast as possible, and performing
a reduced Cooper test (6 minutes). The test battery was administered to 195 children
(aged 5–12 years) from 4 schools and kindergartens in Norway.
[Fjørtoft I, Pedersen AV, Sigmundsson H, Vereijken B. Measuring physical fitness in children
who are 5 to 12 years old with a
test battery that is functional and
easy to administer. Phys Ther.
2011;91:1087–1095.]
Design. The study was a cross-sectional descriptive survey applying physical
Results. Overall, the children in each age group were able to perform all of the test
© 2011 American Physical Therapy
Association
items, indicating the suitability of the test battery for children as young as 5 years of
age. With increasing age, total scores improved linearly, indicating the adequate
sensitivity of the test battery for the age range examined in this study. Furthermore,
even with the modest sample size used in this study, total scores were normally
distributed, thereby fulfilling the necessary assumptions of most statistical procedures. For investigating the reliability of the test battery, 24 children (mean age⫽8.6
years) in one class were retested 1 week later. Test-retest correlations were high, with
intraclass correlation coefficients for individual test items and total score ranging
from .54 to .92.
Limitations. The survey was limited to samples of 5- to 12-year-old Norwegian
children. Larger samples in each age group are essential for establishing age- and
sex-specific norms.
Conclusions. These promising results warrant further development of the test
battery, including standardization and normalization based on a large, representative
sample.
July 2011
Volume 91
Number 7
Post a Rapid Response to
this article at:
ptjournal.apta.org
Physical Therapy f
1087
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
Ingunn Fjørtoft, Arve Vorland Pedersen, Hermundur Sigmundsson,
Beatrix Vereijken
Physical Fitness Test Battery for Children
O
For investigating such relationships,
reliable tests that can establish children’s physical fitness in large population samples are needed. Although
several tests for determining physical
fitness in adults are available, these
are generally inappropriate for determining physical fitness in children.12
Extant tests of physical fitness typically
focus on isolated physiological com-
Available With
This Article at
ptjournal.apta.org
• Audio Abstracts Podcast
This article was published ahead of
print on May 26, 2011, at
ptjournal.apta.org.
1088
f
Physical Therapy
Volume 91
ponents, such as muscle strength
(force-generating capacity) or aerobic
endurance, that are tested with more
or less advanced technological equipment in controlled laboratory settings13,14 (see also Kemper and van
Mechelen12). These tests are based
mostly on test batteries for adults and
may be ill-suited for testing children
because they place high demands on
endurance and the willingness and
ability of participants to follow strict
instructions. These characteristics
make the most reliable tests of physical fitness particularly unfeasible for
testing young children. In addition,
laboratory tests based on direct measures of physiological variables are
expensive and require highly trained
experimenters; thus, they are not feasible for use with large groups of
participants.7,12,15
A further disadvantage of most existing tests is that they attempt to
divide a complex attribute into constituent components and measure
each of the components separately.16,17 The theoretical problem
with this endeavor is that researchers do not know what the constituent components of a complex skill
are or how they collectively make up
the complex skill. In other words,
investigators know neither the variables nor the function making up
physical fitness.
In this article, we describe a new,
functional test battery that aims to
provide a reliable, objective quantification of children’s physical fitness.
In contrast to many extant tests, it
does not attempt to define and then
measure constituent components.
Rather, it focuses on compound
activities that recruit various combinations of multiple factors, such as
strength, endurance, motor coordination, balance, and agility.18 –20 Furthermore, the test battery focuses on
common activities that are included
in most children’s everyday play
activities. This design reduces the
Number 7
cognitive component of the test
items and more easily incites and sustains children’s motivation to participate and perform as well as possible. Our final consideration was
that to be applicable in larger studies, the test battery should be easy to
administer and not require specialized training of experimenters or
equipment beyond what is normally
available in most gymnasiums. Furthermore, whereas previous tests
were divided into several age bands
with different test items for each age
band, the test battery that we
describe includes the same test items
for all ages (5–12 years). This design
enables the longitudinal monitoring
of children’s physical fitness.
In the present study, we investigated
the applicability of our test battery
for children who were 5 to 12 years
of age. We had 4 specific goals. First,
we examined the feasibility of the
test battery for children as young as 5
years of age and assessed whether
the test battery could distinguish
performance across all age groups.
The focus was on whether the
youngest children would understand
and correctly perform the various
tasks and whether the tasks were difficult enough to distinguish performance in the oldest age groups (11
and 12 years). Second, we estimated
the internal consistency of the individual test items and the relationship
between individual test item scores
and the total test score. By using test
items that encompassed individual
components of physical fitness in different combinations and with some
overlap, we aimed for a compound
measure of the term “physical fitness.”21 Third, we estimated the convergent construct validity of the test
battery by comparing scores of children in one class on the test battery
with evaluations of physical fitness
by their physical education teacher.
Finally, we estimated the test-retest
reliability of the test battery.
July 2011
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
ver the last few decades, children’s physical activity levels
have dramatically changed.1–3
Outdoor physical play is increasingly
being replaced by less physical
indoor activities,4 – 6 children are
increasingly being driven to school
by car or bus instead of cycling or
walking, and participation in organized sports is declining.6 – 8 In
recent years, the possible consequences of these changes for children’s overall development and
health have attracted much attention
from the media, scientific researchers, and policy makers. However,
longitudinal research on the relationships among physical activity, physical ability, and health is scarce, and
cross-sectional studies have not generated consistent results.3,9 Therefore, still in question are how the
frequency, intensity, and duration of
physical activity in children affect
their physical fitness and how
decreasing levels of physical activity
may be related to possible changes
in physical fitness and to ensuing
health problems later in life, such as
obesity, diabetes, osteoporosis, back
pain, cardiovascular disease, and
cancer.10,11
Physical Fitness Test Battery for Children
Table 1.
Characteristics of Participating Children
Body Mass
Index (kg/m2)
Age
Group
(y)
Total
No. of
Children
No. of
Girls/Boys
X
SD
X
SD
X
SD
X
SD
5
21
11/10
5.5
0.29
114.0
5.63
19.9
2.63
15.4
1.60
6
26
13/13
6.5
0.30
120.4
5.07
24.5
4.20
16.8
2.01
7
34
17/17
7.4
0.28
128.4
6.44
27.5
4.12
16.6
1.46
8
29
14/15
8.5
0.26
132.1
5.61
31.2
4.90
17.8
1.84
Age (y)
Height (cm)
Weight (kg)
24
11/13
9.4
0.30
138.1
6.49
33.7
4.52
17.6
1.58
20
13/7
10.4
0.26
140.7
7.10
33.4
6.35
16.7
1.89
11
16
7/9
11.5
0.36
150.8
6.63
42.4
10.60
18.5
3.22
12
25
15/10
12.4
0.37
156.5
5.92
47.0
7.81
19.1
2.38
Method
Participants
Information about the study and
the test items and informed consent forms were distributed in 2
kindergartens (1 in southern Norway and 1 in central Norway) and 3
primary schools (1 in southern Norway and 2 in central Norway). A
total of 195 children (101 girls and
94 boys) with no obvious abnormalities participated in the project;
they were approximately 5 to 12
years of age (X⫽8.3, SD⫽2.21, minimum⫽5.0, maximum⫽13.1). Table
1 shows the distribution of the children across the age groups and the
anthropometric measures height,
weight, and body mass index.
Test Items and Materials
The test battery consisted of 9 test
items that represent typical everyday
activities for children, namely, jumping, throwing, climbing, and running. Most of the test items have
appeared in other tests or test batteries as well, such as the EUROFIT,14
the Allgemeiner Sportmotorischer
Test für Kinder,22 the Folke Bernadotte Hemmet,23 and the Fitnessgram.24 The test item “climbing wall
bars” was designed specifically for
our test battery.
July 2011
The 9 test items were as follows:
1. Standing broad jump. The child
stands with his or her feet parallel
and shoulder width apart behind
a starting line. Upon a signal, the
child swings the arms backward
and forward and jumps with both
feet simultaneously as far forward
as possible. The test item score
(better of 2 attempts) is the distance between the starting line
and the landing position (measured in centimeters).
2. Jumping a distance of 7 m on 2
feet as fast as possible. The test
item score (better of 2 attempts)
is the time needed to cross the
distance (measured in seconds).
3. Jumping a distance of 7 m on one
foot (the child is free to choose
which foot) as fast as possible.
The test item score (better of 2
attempts) is the time needed to
cross the distance (measured in
seconds).
4. Throwing a tennis ball with one
hand (the child chooses which
hand) as far as possible. The child
stands with the contralateral foot
in front of the ipsilateral foot. The
test item score (better of 2
attempts) is the distance thrown
(measured in meters).
5. Pushing a medicine ball (1 kg)
with 2 hands as far as possible.
The starting position is with the
feet parallel to each other and
shoulder width apart, with the
ball held against the chest. The
test item score (better of 2
attempts) is the distance achieved
(measured in meters).
6. Climbing up wall bars, crossing
over 2 columns to the right, and
climbing down the fourth column
as fast as possible. Each column of
the wall bars is 2.55 m high and
0.75 m wide. The test item score
(better of 2 attempts) is the time
to completion (measured in
seconds).
7. Shuttle run. The test item score is
the time required to run 10 ⫻ 5 m
(measured in seconds). If the
child makes a procedural error,
the performance is interrupted
and the test item is repeated.
8. Running 20 m as fast as possible.
The child starts in a standing position. The test item score is the
time required to run the distance
(measured in seconds). If the
child makes a procedural error,
then the performance is interrupted and the test item is
repeated.
Volume 91
Number 7
Physical Therapy f
1089
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
9
10
Physical Fitness Test Battery for Children
9. Reduced Cooper test. The child
runs or walks around a marked
rectangle measuring 9 ⫻ 18 m
(the size of a volleyball field) for 6
minutes. Both running and walking are allowed. The test item
score is the distance covered in 6
minutes (measured in meters).
Procedure
Children were tested individually.
Each test item was explained and
demonstrated before the child
started. Except for the 3 running
tests, each test item was performed
twice, with the better attempt
scored. If a child made a procedural
error, instructions and demonstrations were repeated, and the child
made a new attempt. If a second procedural error occurred or if a child
could not perform the test item, the
test item was scored as missing. In
the present study, 40 children had a
total of 51 missing scores.
In addition, 24 children in one class
at a primary school in central Norway (mean age⫽8.6 years, SD⫽0.3)
were tested a second time with the
same test battery 1 week later to
establish test-retest reliability. The
class was chosen to reflect most
closely the average age in the entire
sample.
Data Reduction and Analysis
To express the child’s overall performance in one score, we calculated a
total test score. To this end, the test
item scores that were measured as
the time needed to accomplish the
test items were first converted to
1/score, such that higher scores
always indicated better performance
than lower scores. After conversion,
the scores on all test items were nor1090
f
Physical Therapy
Volume 91
To estimate the internal consistency
of the test battery items, we calculated the Cronbach alpha value for
the test battery. In addition, we calculated Pearson coefficients of correlation between individual test item
scores and the total test score and
Pearson coefficients of correlation
between scores on individual test
items. When an individual test item
score was correlated with the total
test score, the individual test item
score was excluded from the total
test score to avoid statistical dependence. For example, when the score
on test item 1 was correlated with
the total test score, the latter was
calculated as the average of z scores
for test items 2 through 9.
The construct validity of a test can be
established by comparing it with a
prior test known to be valid, a
so-called gold standard. For physical
fitness in children, no such gold standard is available. Nevertheless, to
obtain an estimate of the suitability
of the test battery, we asked the
physical education teacher of children in one of the classes that we
tested to rank 10 girls and 10 boys in
the class (mean age⫽8.7 years,
SD⫽0.3) from worst to best physical
fitness, according to his own implicit
knowledge. The teacher had been
trained in physical education and
was experienced in grading the
physical performance of his pupils.
He had no knowledge about our test
battery, its individual test items, or
the children’s scores.
Number 7
To estimate the test-retest reliability
of the test battery, we tested 24 children in one class in central Norway
twice, 1 week apart. We then calculated intraclass correlation coefficients (ICC [2,1]) and 95% confidence intervals for test and retest
scores25 to determine relative reliability as well as standard errors of
measurement and 95% confidence
intervals to determine absolute reliability for both individual test items
and the total test score. The standard
error of measurement was calculated
as the square root of the mean
within-subject variance, and 95%
confidence intervals were calculated
as 1.96 times the standard error of
measurement.26
All statistical analyses were performed with SPSS, version 16.0.1,*
and consisted of Pearson correlation
coefficients, Spearman rho (rank)
correlations, Kolmogorov-Smirnov
tests for normality of the data distribution, and linear regression
analyses.
Role of the Funding Source
This work was commissioned and
supported by the Ministry of Social
and Health Affairs, Oslo, Norway.
Results
Total Test Score
The total test score for each child
was calculated as the average of z
scores for the individual test items.
Figure 1 shows a plot of the total test
score against age for girls and boys
separately. The total test score
increased linearly with increasing
age. With respect to our first goal, 2
observations are relevant. First, even
the 5-year-old children were able to
perform the test items, indicating
that the test battery is not too difficult even for the youngest children.
Second, the total test score did not
level off with age, indicating that the
* SPSS Inc, 233 S Wacker Dr, Chicago, IL
60606.
July 2011
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
The following materials were
needed for administering the test
items: masking tape, ruler, stop
watch, tennis ball, medicine ball (1
kg), wall bars at least 4 columns
wide, and gym mats.
mally distributed, as indicated by
one-sample
Kolmogorov-Smirnov
tests. Subsequently, all test item
scores were transformed into z
scores on the basis of the sample
mean and standard deviation so that
each test item would have the same
weight in the total score as other test
items. The total test score for each
child was then calculated as the average of z scores for all test items successfully performed by that child.
Physical Fitness Test Battery for Children
test battery is still challenging even
for the oldest children. A one-sample
Kolmogorov-Smirnov test indicated
that total test scores were normally
distributed: Z(195)⫽.97, P⫽.302.
Changes in total test score (average of z scores for the 9 test items) with age (n⫽195).
Lines represent separate linear regressions for girls (r⫽.85, 95% confidence interval⫽.81–.89) and boys (r⫽.84, 95% confidence interval⫽.79 –.88).
Table 2.
Pearson Correlation Coefficients and 95% Confidence Intervals for Individual Test Item Scores and Total Test Scorea and Pearson
Correlation Coefficients for Individual Test Items
Correlation With:
Correlation
With Total
Score
95%
Confidence
Interval
Standing
Broad
Jump
(m)
Standing broad
jump (m)
.84
0.79–0.88
1.00
Jumping on 2
feet (s)
.67
0.59–0.74
.58
1.00
Jumping on 1
foot (s)
.67
0.59–0.74
.60
.85
1.00
Throwing a tennis
ball (m)
.72
0.65–0.78
.71
.34
.31
1.00
Pushing a medicine
ball (m)
.80
0.74–0.85
.72
.43
.41
.80
1.00
Climbing wall
bars (s)
.80
0.74–0.85
.76
.47
.47
.72
.83
1.00
Shuttle run (s)
.80
0.74–0.85
.71
.54
.55
.67
.72
.69
1.00
Running 20 m (s)
.88
0.84–0.91
.80
.66
.68
.67
.77
.73
.76
1.00
Reduced Cooper
test (m)
.65
0.56–0.72
.53
.50
.49
.56
.54
.49
.57
.63
Test Item
a
Jumping
on 2
Feet (s)
Jumping
on 1
Foot (s)
Throwing
a Tennis
Ball (m)
Pushing a
Medicine
Ball (m)
Climbing
Wall
Bars (s)
Shuttle
Run (s)
Running
20 m (s)
On the basis of the other 8 test items.
July 2011
Volume 91
Number 7
Physical Therapy f
1091
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
Figure 1.
Internal Consistency of the
Test Battery
To estimate the internal consistency
of the test battery, we first calculated
Pearson coefficients of correlation
between scores on individual test
items as well as between individual
test item scores and the total test
score based on the other 8 test items
(Tab. 2). The results indicated that all
individual test item scores correlated
positively with the total test score,
with correlations ranging from .65 to
.88. Correlations between scores on
individual test items were moderate
to high (.31–.85). The Cronbach
alpha value for standardized items
was high (.93). With respect to our
second goal, these results indicated
that the internal consistency of the
test battery is high and that the dif-
Physical Fitness Test Battery for Children
confirmed by high Spearman rho
correlations between the 2 rank
scores, which were .93 for girls and
.90 for boys.
Correspondence between rankings of 10 girls and 10 boys on a scale from worst (rank
of 1) to best (rank of 10) physical fitness by their physical education teacher and
rankings on the basis of their total test score.
ferent test items indeed tap into similar underlying components, without
correlations being so high as to indicate that test items are redundant.
dren in one class on the basis of their
total test scores with the rankings of
the same children on the basis of an
evaluation of their physical fitness by
their physical education teacher. Figure 2 shows that there was a close
association between the rankings on
the basis of the teacher’s evaluation
and the rankings on the basis of the
total test scores. This association was
Construct Validity of the
Test Battery
We estimated the convergent construct validity of the test battery by
comparing the rankings of 20 chil-
Discussion
In this article, we described a new
test battery aimed at quantifying
physical fitness in children who
were 5 to 12 years of age. Our considerations for the construction of
Table 3.
Means and Standard Deviations of Test and Retest Scores and 95% Confidence Intervals for Intraclass Correlation Coefficients
(ICCs) and Standard Errors of Measurement
Test Score
Test Item
Standing broad jump (cm)
Jumping on 2 feet (s)
Jumping on 1 foot (s)
Throwing a tennis ball (m)
Pushing a medicine ball (m)
Retest Score
95%
Confidence
Interval
Standard
Error of
Measurement
95%
Confidence
Interval
X
SD
X
SD
ICC
(2,1)
122.54
19.41
119.79
18.66
.88
0.74–0.95
6.68
5.28–8.65
3.87
0.59
3.84
0.77
.65
0.34–0.83
0.40
0.33–0.55
3.19
0.37
3.11
0.42
.66
0.37–0.84
0.23
0.18–0.30
11.97
3.56
12.04
3.51
.92
0.83–0.97
0.99
0.81–1.33
3.34
0.47
3.44
0.53
.54
0.18–0.77
0.34
0.28–0.45
Climbing wall bars (s)
12.10
2.83
12.04
3.08
.77
0.54–0.89
1.41
1.17–1.91
Shuttle run (s)
25.87
2.03
25.20
2.64
.69
0.41–0.86
1.32
1.02–1.67
4.62
0.40
4.82
0.49
.71
0.32–0.87
0.25
0.17–0.28
984.82
133.47
942.05
108.23
.72
0.41–0.88
66.06
48.19–80.91
.80
0.59–0.91
0.26
0.22–0.36
Running 20 m (s)
Reduced Cooper test (m)
Total test score (z score)
1092
f
Physical Therapy
0.042
Volume 91
0.405
Number 7
⫺0.001
0.728
July 2011
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
Figure 2.
Test-Retest Reliability of the
Test Battery
We estimated the relative test-retest
reliability of the test battery by using
ICC (2,1)25 between test and retest
scores for both total test scores and
individual test item scores. Absolute
reliability was estimated from the
standard error of measurement,
which was calculated as the square
root of the average within-subject
variance for each test item score and
the total test score.26 Table 3 shows
the means and standard deviations of
test and retest scores and the 95%
confidence intervals for the ICCs and
standard errors of measurement. The
results indicated fair to good reliability for individual test item scores and
the total test score, with ICCs
between test and retest scores ranging from .54 to .92.
Physical Fitness Test Battery for Children
Applicability of the Test Battery
Total test scores increased linearly
with increasing age, as one would
expect given that the constituent
components typically improve in
children with increasing age.27
These results indicate that our test
battery can be used across the entire
age span studied here, that is, from 5
to 12 years. Our test battery was not
too difficult for the youngest children and not too easy or boring for
the oldest children. Regression analysis suggested that sex differences
can be revealed by our test battery as
well, although we did not formally
evaluate this concept in the present
study. Furthermore, total test scores
were normally distributed, indicating that the test battery is applicable
at both ends of the scale and can be
used to classify both children with
extraordinarily good fitness and
those with extremely poor fitness. In
contrast, some other tests, such as
the Movement Assessment Battery
for Children,28 are heavily skewed
toward a normal distribution. Such
tests appropriately identify children
performing below normal but are
July 2011
not suited for discriminating
between children performing normally. A further advantage of total
test scores being normally distributed is fulfillment of the necessary
assumptions of most parametric statistical procedures.
Internal Consistency of the Test
Battery Items
One of our considerations in choosing the test items was that they
should consist of compound activities that recruit various combinations of underlying components,
such as strength, endurance, motor
coordination, balance, and agility. If
we were successful in this aim, the
individual test items should correlate
reasonably well with each other,
without correlations being so high as
to indicate that individual test items
were redundant. The results showed
that this aim was met, with correlations between the test item scores
ranging from .31 to .85. Furthermore, all of the individual test item
scores would be expected to contribute to the total test score, without either very high or very low correlations between individual test
item scores and total test scores. This
expectation was confirmed, with
correlations ranging from .65 to .88.
The internal consistency of the
entire test battery was attested to by
the high Cronbach alpha value (ie,
.93). These results support the internal consistency of our test battery
and provide confidence that our
test battery yields a fair measure of
physical fitness, a complex attribute
that otherwise is not so readily
measured.21,29
Construct Validity of the
Test Battery
Construct validity is arguably among
the most important characteristics of
a test. Construct validity represents
the extent to which a test measures
what it is supposed to measure and
how well it measures that property.30 It also is among the most dif-
ficult characteristics to determine,
unless the new test can be compared
with an existing test known to be
valid. For physical fitness in children,
no such test is available. Although
the construct validity of our test battery is therefore difficult to establish,
we can nevertheless discuss convergent construct validity on the basis
of 2 arguments.
First, 7 of the 9 individual test items
are found in existing tests that have
been used for many years and have
been largely standardized and scrutinized for construct validity; these
tests include the BOT,31 the EUROFIT,14 the AST,22 and the FBH.23
However, this fact is insufficient to
conclude that we have measured
physical fitness. One reason why it is
difficult to measure physical fitness
is the wide range of opinions about
what physical fitness really is and
how it should be measured.21,29 Nevertheless, professionals in the field
typically have no problem recognizing higher and lower levels of fitness
and motor ability in children.
Because our aim was to measure children’s physical fitness in a way that
would be understood by various professionals working to increase children’s physical activity and thereby
improve their physical fitness, we
sought help from such a professional. Although several groups of
professionals can be considered
experts on children’s physical fitness, such as sports coaches, various
health care professionals, and physical education teachers, we chose a
physical education teacher for several reasons. The daily experience of
sports coaches might be biased
toward children with higher levels of
fitness and motor skills, and the daily
experience of health care professionals, such as physical therapists,
might be biased toward children
who are less fit and those who have
poor or impaired motor skills. Most
physical education teachers, on the
other hand, work with the entire
Volume 91
Number 7
Physical Therapy f
1093
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
the test battery were that the battery
should be quantitative and that its
test items should consist of compound activities, based on everyday
activities, that recruit several constituent components of physical fitness,
such as strength, endurance, motor
coordination, balance, and agility.18
Furthermore, the test battery should
be easy to administer and should not
require specialized technical equipment or specially trained personnel.
This design would allow the test battery to be used to test large groups of
children, even entire populations,
and reliably monitor children’s physical fitness over time. In this first
round of testing, the test battery was
administered to 195 children from 4
kindergartens and schools, enabling
us to investigate its feasibility, internal consistency, construct validity,
and test-retest reliability.
Physical Fitness Test Battery for Children
range of children’s competence
every day. Therefore, we selected a
physical education teacher to provide an experience-based test of
construct validity.
Test-Retest Reliability of the
Test Battery
A test cannot be valid unless it is
reliable. That is, with repeated
administration of the test to the same
participants, the results should be
highly comparable and should not be
severely influenced by irrelevant or
chance factors, such as the time of
day, motivation, fatigue, or boredom. We administered our test battery twice to the children in one
class, 1 week apart, and established
test-retest reliability. Correlations
between the test and retest scores
for the separate test item scores
were high; ICCs ranged from .54 to
.92. More importantly, the ICC for
the test and retest scores for the total
test score was .80, and the 95% confidence interval ranged from .59 to
.91. With such a relatively small sample of participants, these results can
be considered satisfactory.
Limitations and Future Directions
Although our test battery for physical fitness in children is clearly in an
early stage, the results presented
here are promising and warrant further development of the test battery.
In the present study, each child was
ranked with respect to the entire
sample across all ages and both
sexes. The next step in the further
1094
f
Physical Therapy
Volume 91
A limitation of the present study was
that both a failed attempt and a missing test item were scored as a missing value. However, it might be more
correct to reserve a missing value for
a test item not attempted and to
assign a minimum score to a test
item that a child attempted but was
not able to perform correctly. For
the establishment of an adequate and
fair minimum score for each test
item, a larger data sample is
required. Only when such a sample
is available will it be possible to reliably deduce the minimum score for
each test item, according to sex and
age category, thereby allowing the
assignment of a minimum score to a
“failed performance.” This scoring
method would further increase the
construct validity of the test battery.
Conclusions
To measure children’s physical fitness, we have developed a test battery that is based on children’s everyday activities and scored on an
interval scale. On the basis of this
first round of investigation, we conclude that the test battery is easy to
administer, appropriate for children
who are 5 to 12 years of age, and
discriminates well across the entire
age range. Furthermore, the results
of the initial investigation of the construct validity and test-retest reliability of the test battery are promising
and warrant its further development
and standardization. Because the test
battery is easy to administer and does
not require specialized technical
Number 7
equipment or specially trained personnel, it can be used to measure
physical fitness in large groups of
children. This application would
enable health care authorities to collect reliable longitudinal data on a
population level. Such data, in turn,
may provide valuable information
about changes in the level of physical fitness of children over time. Furthermore, health care personnel,
such as physical therapists, can use
such data for planning prophylactic
interventions on a group level or for
evaluating the effects of such
interventions.
All authors provided concept/idea/research
design, writing, data analysis, project management, and facilities/equipment. Dr
Fjørtoft, Mr Pedersen, and Dr Sigmundsson
provided data collection. Dr Fjørtoft and Dr
Sigmundsson provided participants. Dr
Fjørtoft provided institutional liaisons and
consultation (including review of manuscript
before submission).
The authors thank the children and staff at
Helgen Primary School and Kindergarten,
Sætre Primary School, and Åsvang Primary
School for their participation in this project;
Vigdis Vedul Moen, Arne Martin Hårstad,
and Ann Kristin Forseth for help in data collection; Bjørg Fallang, Jan Morten Loftesnes,
Thomas Moser, and Svein Arne Pettersen for
helpful discussions on test construction; and
Erling J. Solberg, Kyrre Svarva, Tom Ivar L.
Nilsen, and Rolf Moe-Nilssen for advice on
statistical analyses.
This work was commissioned and supported
by the Ministry of Social and Health Affairs,
Oslo, Norway.
This article was submitted October 27, 2009,
and was accepted March 14, 2011.
DOI: 10.2522/ptj.20090350
References
1 Anderssen L, Harro M, Sardinha L, et al.
Physical activity and clustered cardiovascular risk in children: a cross-sectional
study (The European Youth Heart Study).
Lancet. 2006;368:299 –304.
2 Hands B, Larkin D. Physical fitness and
developmental coordination disorder. In:
Cermak SA, Larkin D, eds. Developmental
Co-ordination Disorder. Albany, NY: Delmar Thomson Learning; 2002:172–184.
July 2011
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
When we compared the intuitive
rankings assigned by the physical
education teacher to 20 of his pupils
with the total test scores of the same
children on our test battery, we
found high correlations both for girls
and for boys. Even though this evaluation was not a formal test of construct validity, it provides additional
face validity to our test battery and
supports the further pursuit of its
development.
development of the test battery is to
collect data on a large sample to
make standardization of the test battery possible, along with the establishment of age- and sex-specific
norms. Once age- and sex-specific
norms are available, a child can be
ranked with respect to his or her
own age group and sex; in other
words, a child can be classified with
respect to relative physical fitness for
a child of that particular age and sex.
Physical Fitness Test Battery for Children
July 2011
13 Bös K, Mechling H. International Physical Performance Test Profile for Boys and
Girls From 9 –17 Years (IPPTP 9 –17).
Berlin, Germany: International Council of
Sport Science and Physical Education;
1985.
14 Adam C, Klissouras V, Ravazollo M, et al.
EUROFIT: European Test of Physical Fitness—Handbook. Rome, Italy: Council of
Europe, Committee for the Development
of Sport; 1988.
15 Safrit M. The validity and reliability of fitness tests for children. Pediatr Exerc Sci.
1990;2:9 –28.
16 Fjørtoft I. Motor fitness in pre-primary
school children: the EUROFIT Motor Fitness Test explored in 5- to 7-year-old
children. Pediatr Exerc Sci. 2000;12:
424 – 436.
17 van Praagh E, Franca NM. Measuring maximal short-term power output during
growth. In: van Praagh E, ed. Pediatric
Anaerobic Performance. Champaign, IL:
Human Kinetics; 1998:155–189.
18 Fjørtoft I, Pedersen AV, Sigmundsson H,
Vereijken B. Testing Children’s Physical
Fitness: Developing a New Test for 4 –12
Year Old Children. Oslo, Norway: The
Norwegian Social and Health Ministry;
2003. Report IS-1256.
19 Haga M. The relationship between physical fitness and motor competence in children. Child Care Health Dev. 2008;34:
329 –334.
20 Haga M. Physical fitness in children with
high motor competence is different from
that in children with low motor competence. Phys Ther. 2009;89:1089 –1097.
21 Hopkins WG, Walker NP. The meaning of
“physical fitness.” Prev Med. 1988;17:
764 –773.
22 Bös K, Wohlman R. Allgemeiner Sportsmotorischer Test (AST 6 –11) zur Diagnose
der konditionellen und koordinativen Leistungsfähigkeit. Lehrhilfen für den Sportunterricht. 1987;36:145–160.
23 Bille B, Brieditis K, Ekström B, Esscher E.
FBH Provet: Erfarenheter från Folke
Bernadottehemmet. Örebro, Sweden:
Motorika; 1992.
24 The Prudential Fitnessgram: Test Administration Manual. Dallas, TX: Cooper
Institute for Aerobics Research; 2001.
25 Shrout PE, Fleiss JL. Intraclass correlations:
uses in assessing rater reliability. Psychol
Bull. 1979;86:420 – 428.
26 Bland JM, Altman DG. Measurement error.
BMJ. 1996;313:744.
27 Burton AW, Miller DE. Movement Skill
Assessment. Champaign, IL: Human Kinetics; 1998.
28 Henderson SE, Sugden D. The Movement
Assessment Battery for Children. Kent,
United Kingdom: The Psychological Corporation; 1992.
29 Caspersen CJ, Powell KE, Christenson GM.
Physical activity, exercise, and physical fitness: definitions and distinctions for
health-related research. Public Health
Rep. 1985;100:126 –131.
30 Anastasi A. Psychological Testing. 3rd ed.
New York, NY: Macmillan; 1968.
31 Bruininks RH. Bruininks-Oseretsky Test of
Motor Proficiency Examiner’s Manual.
Circle Pines, MN: American Guidance Service; 1978.
Volume 91
Number 7
Physical Therapy f
1095
Downloaded from https://academic.oup.com/ptj/article/91/7/1087/2735048 by guest on 31 May 2021
3 Stalsberg R, Pedersen AV. Effects of socioeconomic status on physical activity in
adolescents: a systematic review of the evidence. Scand J Med Sci Sports. 2010;20:
368 –383.
4 Ekeland E, Halland B, Refsnes KA, et al. Er
barn og unge mindre fysisk aktive i dag
enn tidligere? Tidsskr Nor Lægeforen.
1999;199:2358 –2362.
5 Grund A, Dilba B, Forberger K, et al. Relationships between physical activity, physical fitness, muscle strength and nutritional state in 5- to 11-year-old children.
Eur J Appl Physiol. 2000;82:425– 438.
6 Hoos MB, Gerver WJM, Kester AD, Westerterp KR. Physical activity levels in children and adolescents. Int J Obes. 2003;27:
605– 609.
7 Rice MH, Howell CC. Measurement of
physical activity, exercise, and physical fitness in children: issues and concerns.
J Pediatr Nurs. 2000;3:148 –156.
8 Søgaard AJ, Bø K, Klungland M, Jacobsen
BK. En oversikt over norske studier: hvor
mye beveger vi oss i fritiden? Tidsskr Nor
Lægeforen. 2000;120:3439 –3446.
9 Marshall JD, Bouffard M. The effects of
quality daily physical education on movement competency in obese versus nonobese children. Adapt Phys Activ Q. 1997;
14:222–237.
10 Andersen RE, Crespo CJ, Bartlett SJ, et al.
Relationship of physical activity and television watching with body weight and
level of fatness among children. JAMA.
1998;279:938 –942.
11 Blair SN, Kohl HW, Gordon NF, Paffenbarger RS Jr. How much physical activity is
good for health? Annu Rev Public Health.
1992;13:99 –126.
12 Kemper HCG, van Mechelen W. Physical
fitness testing of children: a European perspective. Pediatr Exerc Sci. 1996;8:210 –
214.