Advances in Physical Education
2012. Vol.2, No.3, 88-94
Published Online August 2012 in SciRes (
Copyright © 2012 SciRes.
The Development and Rasch Calibration of a Scale to Measure
Coordinative Motor Skills in Typically
Developing Children
Giulia Bardaglio1*, Michele Settanni2, Danilo Marasso1, Giovanni Musella1,
Silvia Ciairano1,2
1Motor Science Research Center, University School of Motor & Sport Sciences,
University of Torino, Torino, Italy
2Department of Psychology, University of Torino, Torino, Italy
Email: *
Received April 17th, 2012; revised May 15th, 2012; accepted May 30th, 2012
The purposes of this study were to propose and calibrate an instrument based on systematic observation
intended to measure coordinative motor skills in typically developing children. The instrument, called the
Coordinative Motor Skills Scale (CMSS), was administered to 96 third grade children (50% females,
mean age = 8.4, SD = 0.4) from primary schools in northwest Italy. Data were analyzed employing the
base Rasch model for dichotomous items (Rasch, 1960/1980). The Rasch analysis showed that the meas-
ure of coordinative motor skills is unidimensional and that the instrument was correctly targeted to the
level of ability of the participants. Overall, CMSS was demonstrated to be suitable for assessing coordina-
tive motor skills in normally developing children.
Keywords: Childhood; Movement Coordination; Rasch Model
Movement coordination as expressed in coordinative motor
skills can be defined as the ability to perform complex exercises
accurately and quickly, with in constantly changing conditions
(Starosta, 1989). Furthermore, we know that movement coor-
dination and coordinative motor skills are the external manifest-
tation of functions of the central nervous system and that they
should be measured at the proper time during individual devel-
opment (Starosta, 2000). Coordinative abilities contribute to the
resolution of movement tasks in a rational and creative way in
the various fields of sport and daily life (e.g. climbing stairs,
grasping objects, walking, jumping; Weineck, 2009).
Accordingly to Blume (1981), there are seven coordinative
abilities: motor differentiation, motor connection, balance pres-
ervation, spatial orientation, motor rhythmization, speed reac-
tion, and motor transformation. All these coordinative abilities
occur in three structurally different sets: motor learning, motor
control, and motor adaptation. The motor learning ability is con-
nected with all the other coordinative abilities; the motor con-
trol ability is particularly connected with the ability to perform
and regulate standard movements; finally, the motor adaptation
ability is connected with abilities enabling movement adjust-
ments to changing conditions and situations.
The optimal age for learning motor skills, including the co-
ordinative ones, is difficult to define. However, late childhood,
between the ages of 7 and 11 years, seems to be a “sensitive
period” for the sequential development of gross motor skills
(Gallahue, 1982), and particularly of movement coordination
(Hirtz & Starosta, 2002). Late childhood also corresponds to the
third stage of Piaget’s theory on cognitive development (1964):
that of concrete mental operations. At this stage, children can
understand and respect the rules of a team game and cooperate
with peers in order to apply these rules to ensure the efficacy of
the game (Gallahue).
Although the development of motor skills, including coordi-
native ones, is important in typically developing children, motor
skills are often considered only when there are dysfunctions or
inefficient movement behaviours present. Many studies focus
on fundamental movement skills, but in most cases the instru-
ments used are only able to detect deficiencies in the develop-
ment of movement skills (Hartman, Houwen, Scherder, & Viss-
cher, 2010; Wuang, Lin, & Su, 2009). Among others, the Move-
ment Assessment Battery for Children (M-ABC; Henderson &
Sugden, 1992) and the Test of Gross Motor Development, Sec-
ond Edition (TGMD-2; Ulrich, 2000), are two of the most com-
monly used tools for assessing movement skills, especially in
clinical samples of children with developmental coordination
disorders, in order to monitor children’s motor coordination
(Hartman et al., 2010; Hung & Pang, 2010). Slater, Hillier and
Civetta (2010) recommended the use of the M-ABC test and the
TGMD-2 test for assessing developmental coordination disord-
ers and gross motor performance in children. Vuijk, Hartman,
Scherder and Visscher (2010) recently used the M-ABC test for
comparing the motor skills of children with intellectual dis-
abilities (ID) and of normally developing children. This study
found that children with ID had significantly more borderline
and definite motor problems than the normative sample, and
there was an association between the degree of ID and the per-
formance of manual dexterity, ball skills and balance skills.
The results of this study support the notion that levels of
motor and cognitive functioning are related in children with ID.
Finally, Staples and Reid (2010) used the TGMD-2 test for
*Corresponding author.
comparing the fundamental movement skills of children with
autism spectrum disorders, aged 9 to 12 years, with same age
groups of typically developing children.
We think that investigating the development of coordinative
motor skills also deserves further attention in typically devel-
oping children and particularly when they use these skills in
ecological kinds of tasks, such as team games. In fact the find-
ings collected about these children and within these situations
give us further information about which one of their compo-
nents could be more efficient to train in order to obtain both an
optimal development of motor skills and also even for recovery
in temporary conditions of physical impairment.
Furthermore, the validity of the most common instruments
for assessing the motor coordination of children seems still far
from proven. In most cases, the various authors inform us about
convergent and divergent validity but they do not provide plain
statistic information for an accurate interpretation of the find-
ings (Brown & Lalor, 2009; Evaggelinou, Tsigilis, & Papa, 2002;
Van Waelvelde, Peersman, Lenoir, & Engelsman, 2007). Brown
and Lalor individuated one major weakness of the Movement
Assessment Battery for Children, Second Edition (MABC-2) in
its lack of evidence of construct validity and they suggest fur-
ther analysis with the Rasch Measurement Model. With respect
to the Test of Gross Motor Development (TGMD), Ulrich (1985)
presents the findings of a principal components analysis that
showed three components and four items with high loadings on
two different components. Nevertheless, Ulrich referred to only
one component, and this interpretation was confirmed by Zhu
and Cole (1996). However, the study by Evaggelinou et al.
found that seven items loaded on one factor (i.e., children’s
locomotion skills—their ability to move within a space) and
five items overloaded another factor (i.e., object control skills).
Lee, Zhu, and Ulrich (2005) confirmed the unidimensionality
and the psychometric properties of the TGMD-2 test using a
calibration method based on the Many-Facet Rasch Model.
In sum, there are some indications that we still lack reliable
information on the development of coordinative motor skills in
typically developing children, and also that the Rasch calibre-
tion model (Rasch, 1960/1980) is the most appropriate statistic
technique that can validate the instruments for assessing move-
ment development and performance in children (Chien & Bond,
2009; Wuang et al., 2009). The Rasch model is a unidimensional
scaling based on the relation between the subject’s ability and
item difficulty, and it can overcome the limitations of tradi-
tional approaches based on the Classical Test Theory (CTT).
Rasch modeling has been successfully used for estimating abili-
ties, attitudes, and personality traits in psychological and educa-
tional research. The Rasch model uses a model-based approach
that permits the developing of scales with strong measurement
properties, in particular with regards to generalizability, accu-
racy and statistical validity (Embretson & Reise, 2000). One of
the advantages of Rasch modelling is that it allows for inde-
pendently scaling items and persons along a continuous inter-
val-scaled latent trait (Bond & Fox, 2007; Rasch, 1960/1980). As
a consequence, when data show a good fit to the Rasch model
the latent trait will be measured on a true interval scale (Bond
& Fox), hence obtained measures are eligible for analysis using
statistical techniques that assume the data are measured at in-
tervals (e.g., regression or ANOVA). Another big advantage of
the Rasch model over the CTT is that it permits the analysis of
the measurement invariance of the tested instruments in differ-
ent groups by means of Differential Item Functioning (DIF)
statistics. DIF analysis permits the study of the functioning of
each item belonging to the instrument and to recognize possibly
misbehaving items that could led to biases in the measurement
(Myers, Wolfe, Feltz, & Penfield, 2006).
The purpose of this study is to generate and calibrate a scale
to measure coordinative motor skills (Coordinative Motor Skills
Scale, CMSS) in a group of typically developing children using
the Rasch model and to further validate the scale through the
convergent, divergent, and know-difference validity approach.
In particular we concentrated on a specific team game called
Dodgeball because it is a non specialised team game frequently
used when teaching various team games. It shares some basic
characteristics of other team games, such as a high level of coop-
eration between players, the in-motion regulatory condition of
the environment, the intertrial variability between different repe-
titions of the same skill, and the use of objects (Gentile, 2000).
With respect to convergent and divergent validity, we expected
that our instrument would be related to one of the sub-scales of
the TGMD, and the control object (TGMD) can precisely be
considered one component of the more general construct of
coordinative motor skills (Lombardozzi, Musella, Balducci, &
Barigelli, 2001). Yet, we did not expect that the CMSS would
be linked with the other subscales of the TGMD—namely, the
locomotion subscale—because locomotion is a major basic
motor pattern that normally is already consolidated during the
first cycle of primary school (Ulrich, 1985).
Finally, in order to evaluate the construct validity of the new
scale by the know-difference validity approach we expected that
our instrument would be able to discriminate between the levels
of coordinative motor skills in children who are usually involved
in scarce (fixed at less than two hours weekly) or frequent
(fixed at two hours or more weekly) physical activity. The dis-
crimination between scarce or frequent physical activity was
fixed at two hours per week on the basis of the National Italian
Guidelines on Education (2007) and of other previous studies
(e.g., Klepp, Tell Grethe, & Vellar Odd, 1994). Using this cut-off
we assigned children to the “scarce physical activity” group (less
than two hours weekly) or to the “frequent physical activity”
group (two or more hours weekly).
Inventory Development
Systematic observation tries to reduce the excessive subject-
tivity of general observation, making it an excellent qualitative
tool for educational research and the evaluation of behaviour. In
fact the coordinative motor skills that typically developing chil-
dren usually use in team games are difficult or impossible to
measure through the quantitative and validated tests present in
the literature. Besides this, the level of the subjectivity of sys-
tematic observation can be limited through the use of clear and
simple descriptors of the learning motor activity to be observed.
In this way, the observed situation is broken down into criteria
(namely descriptors) that are the components of the complex
situation. These descriptors are then filled into a checklist and
can be evaluated with a dichotomous scale, signalling the pres-
ence or the absence of the descriptor (Aureli, 1997).
Thus, the Coordinative Motor Skills Scale (CMSS) is a sys-
tematic observation tool. The outcome of this observation is a
measure of coordinative motor skills. In particular, we chose
the coordinative motor skills observed in a specific situation—
Copyright © 2012 SciRes. 89
Copyright © 2012 SciRes.
namely, during a simple team game with a ball (Dodgeball).
Even if there are various differences between team games, the
five most common motor skills of these types of games are pass-
ing the ball, shooting the ball, movement with the ball, move-
ment without the ball in attack and movement without the ball
in defence (Lombardozzi et al., 2001). We chose the game of
Dodgeball because it is a popular game in Italy and it is fre-
quently used during normal classes of physical activity in pri-
mary school. The popularity of this game and the moderate
physical effort that it requires allows boys and girls to play it
together without any difference. In addition we have not used a
codified sport because the choice of a codified sport would not
have matched the aims of physical education in Italian primary
school, which are only to increase coordinative motor skills and
are not focused on one specific sport. Specific aspects of sport
are introduced only in the last two years of primary school,
which is not an age group we considered in this study. Finally,
we did not choose a codified sport because we would have fa-
vored children who also practice this sport outside school and
gender differences also could have arisen. In fact, among peo-
ple who practice sports in Italy, 41% of males practice football
compared to 1.4% of females. On the contrary, 3.3% of males
practice volleyball compared to 10% of females (ISTAT, 2005).
The CMSS was previously tested in one pilot research pro-
ject (Marasso, 2009) that belongs to a huge longitudinal project
called “Educatamente Sport”, which is still in progress. In this
pilot study the systematic observation tool consisted of ten items
obtained by combining the five most common motor skills of
team games with motor control ability and motor adaptation
ability (Blume, 1981). The ten items observed are the pass-motor
control ability, pass-motor adaptation ability, shooting-motor
control ability, shooting-motor adaptation ability, movement with
the ball-motor control ability, movement with the ball-adaptation
ability, attack-motor control ability, attack-motor adaptation abil-
ity, defence-motor control ability, and defence-motor adapta-
tion ability. The exploratory factor analysis indicated that the
two items about movement with the ball (motor control ability
and adaptation ability) did not fit with the scale, probably due
to the excessive difficulty in their evaluation (α = .50 if both are
included). For this reason, these two items were deleted to im-
prove the reliability of the scale (α = .72).
The CMSS consists of eight items. The eight items observed
are: pass-motor control ability, pass-motor adaptation ability,
shooting-motor control ability, shooting-motor adaptation abil-
ity, attack-motor control ability, attack-motor adaptation ability,
defence-motor control ability, and defence-motor adaptation ability.
The descriptors of these eight items are illustrated in Table 1.
These items are reported in a checklist and are evaluated using
a dichotomous item, reporting the presence or the absence of
each descriptor. Ultimately, the entire set of eight items may
belong to a higher-order factor structure, evaluated with values
between zero to eight, which expresses a unidimensional con-
cept of coordinative motor skills in team games.
The data used in this study were collected from 96 children
who attended third-grade classes of primary schools in north-
west Italy. The sample was balanced for gender (50% females);
the mean age of the participants was 8.4 years (SD = 0.4) and the
main socio-demographic information of the participants is similar
to that found in the general Italian population (ISTAT, 2010).
Our research was approved by the Ethical Committee of the
University of Torino. We collected all the required informed
consent from the parents of the children (because the children
are minors) according to the ethical principles for research of
the Italian Psychological Association. We also collected active
consent from the children themselves.
The study was conducted in four public primary schools and
from each school we randomly selected 24 children. These chil-
dren brought a questionnaire home to their parents who com-
pleted the questionnaire about their socio-demographic infor-
mation and the children’s number of hours spent practicing sports
or motor extra-curricular activities. The questionnaire was com-
pleted by 58% of the mothers, 14% of the fathers, and 28% by
both parents. This questionnaire took approximately 15 minutes
to complete.
Data about motor skills were collected in a gymnasium. The
administration of both tests—the TGMD test (Ulrich, 1985) and
the CMSS check-list—required approximately four hours for all
the children from one school. The observation protocol included
a camera and the simultaneous presence of three observers, all
doctoral students, who were experts in physical education and
participated in a specific observer training; each observer was
equipped with two checklists (the TGMD and CMSS tests).
In reference to the CMSS test, the presence or absence of the
descriptors must be reported on the checklist during the obser-
vation of a team game (Dodgeball). For coding purposes, the
Table 1.
Descriptors of 8 items of the Coordinative Motor Skills Scale (CMSS).
Items Descriptors
1. Pass-motor control ability He/she throws the ball, regulating the action, with force and trajectory to assume that the ball arrives in the hands
of the teammate.
2. Pass-motor adaptation ability He/she chooses to pass the ball to the teammate closest to the half way line in reference to the field layout of the
3. Shooting-motor control ability He/she throws the ball, regulating the gesture, with force and trajectory to assume the success of the shooting.
4. Shooting-motor adaptation ability He/she chooses to hit with the ball the opponent closest to the half way line in reference to the field layout of the
others opponents.
5. Attack-motor control ability He/she moves and positions to receive the ball and attack, always keeping eye contact with teammate in possession
of the ball.
6. Attack-motor adaptation ability He/she chooses to move, in the position closest to the half way line in reference to the field layout of the opponents,
in order to receive the ball and attack.
7. Defence-motor control ability He/she is positioned to receive the ball from opponents always keeping the front facing the opponent in possession of
the ball.
8. Defence-motor adaptation ability He/she chooses the defensive position farther than the half way line in reference to the opponent in possession of the
ball that is attacking.
playing time was divided into game actions. Each game action
involved the evaluation of all the items. The analysis of the
prevalence of dichotomous values for each descriptor in each
action give the final dichotomous value for each item. The game
situation was observed by all of the researchers at the same time
and each session of play was immediately coded independently
by the three observers. The proportion of agreement between
the scores assigned by the different coders was considered very
high because it was about 95% (D’Odorico, 1990). Furthermore,
the few disagreements were immediately resolved with the ex-
ternal contribution of a fourth researcher not directly involved
in the observation session. In the case of a disagreement, the
coders were invited to justify their choices and, after a short
discussion, in all cases they came to an agreement (D’Odorico).
The game comprised of six children divided into two teams,
three versus three.
The criterion for composing the teams was random. Before
the start of the game, there was a clear and precise explanation
of the rules of Dodgeball to the children. This period was aimed
at giving children the same explanations and rules even if all
the children already knew the game. Finally, the three doctoral
students observed 6 children at a time (3 vs. 3) for approxi-
mately 20 minutes during the game time period. This required
approximately 80 minutes (4 team game periods, each 20 min-
utes long) to observe and register the behaviour of all the 24
children from one school with respect to the CMSS scale.
Data Analysis
Preliminary Analysis
Before proceeding with the further steps of the analysis, to
evaluate the possibility of school biases we used the one-way
analyses of variance (ANOVA) and the Tukey Post-Hoc test to
analyze the differences between different school settings on the
results of the CMSS test. The absence of significant differences
between classes permitted us to exclude the influences of dif-
ferent school settings.
Model-Data Fit
The data were analyzed using Winsteps software (Linacre,
2005) employing the base Rasch model for dichotomous items
(Rasch, 1960/1980). According to the Rasch model, the prob-
ability of an answer when a person faces an item can be de-
scribed as a function of the person’s position on the latent trait
(here, coordinative motor skills in team games). Given the an-
swers of n persons to k items built to measure the same latent
trait, such person and item parameters can be statistically esti-
mated. These estimates represent a person’s ability and the meas-
ures of item difficulty and are expressed in units called logits.
The software we used conventionally defines the mean of the
item measures as zero. With regard to the main fit statistics
reported here, the infit “is an information weighted fit statistic,
which is more sensitive to unexpected behavior affecting re-
sponses to items near the person’s measure level”; the outfit “is
an outlier-sensitive fit statistic, more sensitive to unexpected
behavior by persons on items far from the person’s measure
level”(Linacre). Infit and outfit values substantially less than
1.0 suggest an overfit that may lead to inflated statistics, whereas
values in excess of 1.0 indicate unmodeled noise (underfit). In
accordance with the literature (Wright & Linacre, 1994), accept-
able ranges of infit and outfit values are between 0.5 and 1.5.
Item Location and Person Measures
The logit unit is the natural logarithm of the odds of a person
being able to perform a particular task or an item being suc-
cessfully carried out. Regarding the CMSS items, higher diffi-
culty measures correspond to the higher level of coordinative
ability needed to perform the task. A comparison of the mean
location score obtained for persons (average ability) with that
of the value of the zero set for the items allows for the evalua-
tion of the targeting of the items to the sample (i.e., how well the
CMSS measures the sample under study). A mean person meas-
ure close to zero would represent the ideal targeting of the items
in terms of their difficulty for the sample. A mean person meas-
ure much higher or lower than zero would indicate that the item
set is mistargeted with respect to the sample (Bond & Fox, 2007).
Differential Item Functioning
Gender-based item invariance was studied by considering the
differential item functioning (DIF) of each item. A DIF item is
one for which there is a different likelihood of endorsement for
members of different subgroups. In regard to the CMSS items,
we expected the difficulty of the items to be the same for both
males and females. Ascertaining the absence of items with a
significant DIF allows for affirming that the differences in co-
ordinative abilities in team games found between males and
females are based on real differences and not on items that
perform in a different way across the two subsamples. The DIF
may be verified by statistically testing dif ferences in estimated
item parameters (paired t-tests).
Convergent Validity, Divergent Validity, and
Know-Difference Validity Evidence
In order to further evaluate the validity of the instrument, con-
vergent, divergent, and know-difference validity were also asses-
sed. With regard to convergent and divergent validity, bivariate
correlations were computed between CMSS scores and TGMD
subscales. In order to further evaluate the construct validity, a
one-way ANOVA was conducted to test the presence of differ-
ences in the levels of coordinative motor skills among groups
with a different number of hours spent practicing sports or mo-
tor extra-curricular activities.
Model-Data Fit
On the whole, the CMSS data fitted the Rasch model well.
Infit and outfit statistics (as reported in Table 2) are maintained
between the recommended thresholds (0.5 and 1.5); hence, they
provide evidence supporting the substantial unidimensionality
of the construct that was measured. The absence of both under-
fitting and overfitting items indicates that all the considered tasks
were adequate to measure the construct and there was no need
to remove or modify the items.
Item Location and Person Measures
Figure 1 shows the distribution along the same continuum of
the measures of both participant abilities and item difficulties.
The measure unit is the logit; point zero corresponds to the
mean item difficulty level. The map clearly shows that the dis-
tribution of item difficulties (M = .00, SD = 1.75) is quite well
Copyright © 2012 SciRes. 91
Table 2.
Summary of the CMSS items.
Measure (logit) SE % of correct performanceInfit Outfit
Attack-motor adaptation ability 2.00 0.31 28 0.78 0.45
Defense-motor adaptation ability 1.91 0.3 29 1.34 1.13
Attack-motor control ability 1.47 0.29 34 0.8 0.73
Defense-motor control ability 0.34 0.27 49 0.69 0.58
Shooting-motor control ability –0.08 0.26 55 1.26 1.36
Pass-motor adaptation ability –1.53 0.27 76 1.22 1.29
Pass-motor control ability –1.53 0.27 76 0.98 1.22
Shooting-motor adaptation ability –2.59 0.33 88 0.91 0.62
Figure 1.
Items and subject measures map.
targeted with respect to the distribution of person abilities (M
= .40, SD = 1.93). Item and student measures were widely dis-
tributed along the logit continuum. Table 2 shows the CMSS
item difficulties (expressed in logits) with standard errors, the
percentage of students who correctly performed each task, and
infit and outfit statistics. Results of the Rasch calibration indi-
cated that item difficulties ranged between –2.59 and 2.00 logits,
with higher scores corresponding to more difficult items. The
most difficult task was the attack-motor adaptation ability (cor-
rectly performed by 28% of the students), while the least diffi-
cult task was the shooting-motor adaptation ability (correctly
performed by 88% of the students). The ability measures of the
students ranged between 4.18 and –2.82 logits.
Differential Item Functioning
Gender-based item invariance was studied by considering the
DIF of each item with respect to gender. The DIF contrasts, com-
puted as the differences between difficulty measures separately
obtained for males and females, ranged between .16 and .91 and
were always non-significant. This result indicates that the items
worked in the same way for both genders.
Convergent Validity, Divergent Validity, and
Know-Difference Validity Evidence
The CMSS and the TGMD measures (Ulrich, 1985) used in
this study were compared to establish validity evidence. A posi-
tive relationship (r = 0.39, p < .001) exists between the two
measures. These measures obtained with the new instrument
(i.e., CMSS) correlated in a very different way with the two
TGMD subscales. Indeed, the TGMD object control subscale
showed a stronger correlation (r = .43, p < .001) with the CMSS.
Instead, the TGMD locomotion subscale did not show a sig-
nificant correlation with the CMSS (r = .14, p = .138). These
results support the convergent and divergent validity of the CMSS.
In order to further test the construct validity of the instrument,
the sample was split into two groups according to the number
of hours spent practicing sport or motor extra-curricular activ-
ity (0 - 2 hours and more than 2 hours), and a one-way
ANOVA was conducted to test the differences in coordinative
abilities, i.e. in the CMSS scores, between the two groups (see
Table 3). Children who spent more than two hours practicing
sports or extra-curricular motor activity obtained significantly
higher scores than children who spent less than two hours in the
same kind of activities: F(1, 98) = 3.89, p < .05, η² = .039.
Discussion and Conclusion
The purposes of this study were to generate and calibrate a
unidimensional CMSS for typically developing children using
the Rasch model and to further validate the scale through the
convergent, divergent, and know-difference validity approach.
The substantial validity of the CMSS was confirmed through
the various steps of the research. The first step of this study was
devoted to the generation of the new instrument. In the second
step the new instrument was calibrated using the Rasch model,
which showed a good level of targeting with the level of the
Copyright © 2012 SciRes.
Table 3.
Know-difference validity evidence.
No of hours N M SD ANOVA
0 - 2 37 –.90 1.71
More than 2 63 .70 2.01
F(1.98) = 3.89
p < .05 η² = .04
participants’ ability. Furthermore, the unidimensionality of the
instrument was verified using the Rasch model. Finally, the test
of convergent validity, divergent validity, and know-difference
validity were all satisfactory.
Referring to convergent validity, the results are in accordance
with theoretical expectations. In fact, we found a positive cor-
relation between the CMSS and the TGMD object control
sub-scale. This finding is justified by the fact that the specific
skills for controlling objects are included in the general con-
struct of coordinative motor skills in team games. In particular,
one of the most important peculiarities of game activities—
especially team games—is the use of an object, usually a ball
(Lombardozzi et al., 2001).
Regarding the divergent validity, as expected the TGMD lo-
comotion subscale was not correlated with the CMSS. In fact,
our participants are children from the second cycle of primary
school, while the major basic motor patterns (like locomotion)
are usually already consolidated during the first cycle of pri-
mary school (Ulrich, 1985). Furthermore, this finding suggests
that locomotion skills may be not included in coordinative mo-
tor skills in the team games of typically developing children.
Finally, as expected, the CMSS was able to discriminate be-
tween different quantities of motor practice (i.e., known-difference
validity). Children who spent more than two hours per week on
motor activities (curricular, extra-curricular, physical activity in
leisure time) obtained significantly higher scores than children
who spent less than two hours practicing motor activities. In-
deed, we expected that extensive physical activity would im-
prove the coordinative motor skills in children between 7 and 11
years, because at this age the high plasticity of the cerebral cortex
provides the opportunity to significantly develop coordinative
motor skills (Weineck, 2009).
Although our results are encouraging, this study also has sev-
eral limitations. First, the relatively small sample size and the
fact that all the participants live in one region of Italy makes it
difficult to generalize results to different populations. A larger-
scale investigation is recommended. Second, the limited num-
ber of dimensions considered and the absence of control for
other parameters (i.e., physical, social, psychological) do not
allow for a complete explanation of a phenomenon as complex
as human movement. Third, further studies are necessary to re-
fine the scale and improve all kinds of validities. In addition,
test re-test reliability is recommended.
Despite all its limitations, the CMSS seems to be a valid in-
strument to assess coordinative motor skills in team games of
typically developing children. The results of calibration, conver-
gent, divergent, and know-difference validity encourage contin-
ued research in this direction. Unlike many instruments, which
appear more suitable for detecting deficiencies in movement
skills development (Hartman et al., 2010; Wuang et al., 2009),
our new instrument (CMSS) can be used for investigating co-
ordinative motor skills in the team games of typically develop-
ing children. As has been reported, we think that investigating
the development of coordinative motor skills also deserves fur-
ther attention in typically developing children and particularly
when these skills are used in ecological kinds of tasks, such as
team games.
The authors acknowledge CRC Foundation of Cuneo, CRT
Foundation and ISEF Foundation of Torino, Italy, for contrib.-
uting to this study. REFERENCES
Aureli, T. (1997). L’osservazione del comportamento del bambino (The
observation of the child’s behavior). Bologna, IT: Il Mulino.
Blume, D. D. (1981). Kennzeichnung Koordinativer Fähigkeiten und
Möglichkeiten ihrer Herausbildung im Trainingprozess. In W. Staro-
sta (Ed.), The importance of movement co-ordination, its structure
and the hierarchy of integrant elements in sport and physical educa-
tion (pp. 14-88). Roma: Centro Studi & Ricerche, Federazione Ital-
iana di Atletica Leggera.
Bond, T. G., & Fox, C. M. (2007). Applying the Rasch model: Funda-
mental measurement in the human sciences (2nd ed.). London: Law-
rence Erlbaum.
Brown, T., & Lalor, A. (2009). The movement assessment battery for
children-second edition (MABC-2): A review and critique. Physical
& Occupational T hera py in Pediatrics, 29, 86-103.
Chien, C. W., & Bond, T. G. (2009). Measurement properties of fine
motor scale of Peabody developmental motor scales-second edition:
A Rasch analysis. American Journal of Physical Medicine & Reha-
bilitation, 88, 376-386. doi:10.1097/PHM.0b013e318198a7c9
D’Odorico, L. (1990). L’osservazione del comportamento infantile (The
observation of child beha v io r ). Milano, IT: Cortina.
Embretson, S. E., & Reise, S. P. (2000). Item response theory for psy-
chologists. Mahwah, NJ: Erlbaum.
Evaggelinou, C., Tsigilis, N., & Papa, A. (2002). Construct validity of
the test of gross motor development: A cross-validation approach.
Adapted Physical A ctivity Quarterly, 19, 483-495.
Gallahue, D. L. (1982). Understanding motor development in children.
New York: John Wiley & Sons.
Gentile, A. M. (2000). Skill acquisition: Action, movement, and neu-
romotor processes. In J. H. Carr, & R. D. Shepherd (Eds.), Movement
science: Foundations for physical therapy (2nd ed., pp. 111-187).
Rockville, MD: Aspen.
Hartman, E., Houwen, S., Scherder, E., & Visscher, C. (2010). On the
relationship between motor performance and executive functioning
in children with intellectual disabilities. Journal of Intellectual Dis-
ability Research, 54, 468-477.
Henderson, S. E., & Sugden, D. A. (1992). Movement assesment bat-
tery for children. Sidcup, Kent: Therapy Skill Builders.
Hirtz, P., & Starosta, W. (2002). Sensitive and critical periods of motor
co-ordination development and its relation to motor learning. Jour-
nal of Human Kinetics, 7, 19-28.
Hung, W. W., & Pang, M. Y. (2010). Effects of group-based versus
individual-based exercise training on motor performance in children
with developmental coordination disorder: A randomized controlled
study. Journal of Rehabilitation Medicine, 42, 122-128.
ISTAT (2005). Lo sport che cambia. I comportamenti emergenti e le
nuove tendenze della pratica sportiva in Italia. The sport is changing.
The emerging behavior and new trends of sport in Italy. Roma, IT:
ISTAT (2010). Annuario statistico italiano 2010 (Italian statistical
year book 2010). Roma, IT: ISTAT.
Klepp, K. I., Oygard, L., Tell Grethe, S., & Vellar Odd, D. (1994).
Twelve year follow-up of a school-based health education programme.
European Journal of Public Health, 4, 195-200.
Lee, M., Zhu, W., & Ulrich D. A. (2005). Many-faceted rasch calibration
of TGMD-2. Research Quarterly for Exercise and Sport, 76, 116.
Copyright © 2012 SciRes. 93
Copyright © 2012 SciRes.
Linacre, J. M. (2005). A user’s guide to Winsteps/Ministeps Rasch-
model programs. Chicago, IL: MESA Press.
Lombardozzi, A., Musella, G., Balducci, F., & Barigelli, E. (2001).
Giochi sportivi (Sports games). Padova, IT: Piccin Nuova Libraria
Marasso, D. (2009). La complessità della valutazione oggettiva nel
gesto motorio (The complexity of the objective evaluation of the mo-
tor gesture). Torino, IT: Scuola Universitaria Interfacoltà In Scienze
Myers, N. D., Wolfe, E. W., Feltz, D. L., & Penfield, R. D. (2006).
Identifying Differential Item Functioning of Rating Scale Items with
the Rasch Model: An Introduction and an Application. Measurement
in Physical Education and Exercise Science, 10, 215-240.
National Italian Guidelines on Education (2007). Indicazioni per il
curricolo—Per la scuola dell’infanzia e per il primo ciclo d’istruz-
ione. Indications for th e c u r r i c u l u m —For school children and fo r t h e
first cycle of education. Roma, IT: Ministero della Pubblica Istruzione.
Piaget, J. (1964). Six études de psychologie (Six psychological studies).
Paris: Folio Essays.
Rasch, G. (1960). Probabilistic model for some in telligence and achieve-
ment tests. Copenhagen: Danish Institute for Educational Research.
Rasch, G. (1980). Probabilistic model for some in telligence and achieve-
ment tests. Chicago, IL: University of Chicago Press.
Slater, L. M., Hillier, S. L., & Civetta, L. R. (2010). The clinimetric
properties of performance-based gross motor tests used for children
with developmental coordination disorder: A systematic review. Pe-
diatric Physical Therap y, 22, 170-179.
Staples, K. L., & Reid, G. (2010). Fundamental movement skills and
autism spectrum disorders. Journal of Autism Development Disorder,
40, 209-217. doi:10.1007/s10803-009-0854-9
Starosta, W. (1989). Poziom koordynacji ruchowej u finalistów Ogól-
nopolskiej spartakiady Mlodzieøy. In W. Starosta (Ed.), The impor-
tance of movement co-ordination, its structure and the hierarchy of
integrant elements in sport and physical education (pp. 14-88). Roma:
Centro Studi & Ricerche, Federazione Italiana di Atletica Leggera.
Starosta, W. (2000). Osnowni koordinacionni sposobnosti, struktura i
znaczenie za postizenieto w turnirnite sportni tanci. In W. Starosta
(Ed.), The importance of movement co-ordination, its structure and
the hierarchy of integrant elements in sport and physical education
(pp. 14-88). Roma: Centro Studi & Ricerche, Federazione Italiana di
Atletica Leggera.
Ulrich, D. A. (1985). Test of gross motor development. Austin, TX:
Pro-ED. Inc.
Ulrich, D.A. (2000). Test of gross motor development: Examiner’s
manual (2nd ed). Austin, TX: Pro-ED. Inc.
Van Waelvelde, H., Peersman, W. Lenoir, M., & Engelsman, B. C. (2007).
Convergent validity between two motor tests: Movement-ABC and
PDMS-2. Adapted Physical Activity Quarterly, 24, 59-69.
Vuijk, P. J., Hartman, E., Scherder, E., & Visscher, C. (2010). Motor
performance of children with mild intellectual disability and border-
line intellectual functioning. Journal of Intellectual disability Re-
search, 54, 955-965. doi:10.1111/j.1365-2788.2010.01318.x
Weineck, J. (2009). L’allenamento ottimale (The optimal training).
Perugia, IT: Calzetti e Mariucci.
Wright, B. D., & Linacre, J. M. (1994). Reasonable mean-square fit
values. Rasch Measurement Tr a nsactions, 8, 370-371.
Wuang, Y. P., Lin, Y. H., & Su, C.Y. (2009). Rasch analysis of the
Bruininks-Oseretsky Test of Motor Proficiency—Second Edition in
intellectual disabilities. Research in Developmental Disabilities, 30,
1132-1144. doi:10.1016/j.ridd.2009.03.003
Zhu, W., & Cole, E. L., (1996). Many-faceted Rasch calibration of a
gross motor instrument. Research Quarterly for Exercise and Sport,
67, 24-34.