1

Perceptual and Motor Skills, 2005, 101, 131-148

A NEW MEASURE OF PERCEIVED MOTOR COMPETENCE FOR CHILDREN

AGED 4 TO 6 YEARS1

LUIS MIGUEL RUIZ PÉREZ

Castilla La Mancha University. Toledo

JOSÉ LUIS GRAUPERA SANZ Alcalá de Henares University. Madrid

1 Address correspondence to Prof. Dr. Luis M. Ruiz, Facultad de Ciencias del Deporte, Universidad de Castilla La Mancha, Campus Universitario “Antigua Fábrica de Armas”, Avda. Carlos III s/n. 45071 Toledo, España, mail (luismiguel.ruiz@uclm.es)

2

A NEW MEASURE OF PERCEIVED MOTOR COMPETENCE FOR CHILDREN

AGED 4 TO 6 YEARS

3

Summary: The purpose of this study was to develop and validate an instrument for the assessment of perceived motor competence in young children aged 4 to 6 years old. The structure of the Children’s Perception of Motor Competence Scale (CMPI) was analyzed through exploratory and confirmatory factor analysis with 495 children between 4 and 6 years of age. A second order model was selected and consisted of one scale of Perceived General Motor Competence and two subscales: Perceived Gross Motor Competence and Perceived Fine Motor Competence. The number of items was 22. This instrument seems to possess acceptable internal reliability: global scale (α = .81, Perceived Gross Motor Competence (α = .80) and Perceived Fine Motor Competence (α = .65). Children manifested accuracy in the assessment of the competence. Perceived motor competence was related to actual motor competence as measured by the Movement ABC Test and by an observational scale used by Physical Education teachers. There were no gender differences. The Children’s Perception of Motor Competence Scale could be considered an interesting assessment instrument for physical and movement educators for identifying current self-perceptions of motor competence in young children.

4

Competence is defined as a general capability of an individual to interact effectively

with his environment (White, 1959). Competence is a perceived mastery of skills in different

domains: motor, cognitive and social. Play, curiosity and exploratory behaviors of children are

based on the need to interact effectively with the environment. It is a functional notion which

refers to the effective participation of a subject in his or her environment. The literature about

the concept of competence makes a broad distinction between a cognitive approach,

emphasizing the development and mastery of skills in general, and motor skills in particular,

and a motivational approach in which competence is related to behavioral and personality

development (Connolly & Bruner, 1974; Van der Steen & Vermeer, 1987).

According to Harter’s competence theory (Harter, 1978), there are several

achievement domains: Cognitive competence, which concerns school and intellectual

results, being successful in class, being bright, appreciating good school results; social

competence, which describes the relationship of the child with his or her peers, having a lot

of friends, being pleasant, being someone in the class; and physical competence, which

brings together the results in games and sporting activities which need great motor

commitment, being good at sport, improving easily in physical education, preferring to

practice rather than be a spectator in physical and sporting tasks. Children do not perceive

their competence as being identical in all domains. The feeling of competence is the

complex result of these three components.

With regard to the term motor competence, different authors have used it to express

what common sense and daily observation of physical education sessions make it possible

to witness, i.e. the behavior of schoolchildren when they carry out a complex motor task

(Connolly, 1980; Keogh & Sugden, 1985). Physical education sessions involve a constant

interaction between the context, the task and the pupil, and children must move effectively

and efficiently to achieve and solve constantly changing objectives and problems. Children

are motivated to have an effect on their environment and to engage in attempts at mastery.

When attempts result in competent or successful motor performance, children experience

intrinsic pleasure, thereby maintaining or enhancing their competence (Shapiro, Yun &

Ulrich, 2002).

5

Being competent in physical education implies a dynamic and complex process

characterized by a progression of change in the control of one's actions with others or with

objects in the environment, a process which means acquiring the capability to obtain the

objectives one sets oneself that are often impossible to measure with classical procedures of

skills evaluation. This concept has a global aspect comprising knowledge, procedures,

attitudes and feelings which intervene in the many motor behaviors which the subject

carries out in his or her environment and with others and, on the other hand, has a concrete

aspect which is translated into specific competencies which the schoolchild should be able

to demonstrate at each stage of the curriculum, and which permit evaluating and

determining his or her motor status in physical education (Ruiz, 1995).

As Delignières (1992) indicates, competence in physical education means the

possession of a repertoire of pertinent responses for situations which are to a large extent new;

to possess, in fact, a body of knowledge, procedures and attitudes which permit autonomous

practice. Schoolchildren learn to be competent because they learn to better interpret situations

which demand performance and because they develop the necessary resources to respond

adequately to the demands of the situation, which means developing a feeling of ability to act,

of confidence in being able to cope with the problem situations which arise and of joy at being

able to transform one's environment (Harter, 1981).

The construct of perceived competence or ability has been highlighted in both

psychological and sport psychological research as one that is intimately related to

motivational indicators such as choosing to participate, sustaining effort, and continuing

interest in tasks or activities (Harter, 1982; Ulrich, 1987; Hagger, Asford, & Stambulova,

1998). Harter and Pike (1984) found that children aged 4 to 7 years make reliable

judgments about their competence across the different domains (cognitive, physical, social

and behavioral) if they are pictorially depicted as concrete observable behaviors (Harter,

1990), but they can have problems in differentiating competences across these four

domains. If we apply this model to the learning of motor skills and physical education, it

may be supposed that enhanced motor skills will result in feelings of motor competence

that could promote participation in physical education classes.

6

Perceptions of motor competence by children

An analysis of the different measures of perceptions of competence, self-concept,

self-regard, self-esteem, etc., show different perspectives and approaches. Villa &

Auzmendi (1992, pp. 39-54) reviewed these measures and found that the majority were

made for older children. If we consider our main interest, there are not many measures of

self-perceptions of motor competence in children younger than 6 years old, though a few of

them have some items or subscales designed to measure this dimension.

The Martinek-Zaichkowsky Self-concept Scale for Children (Martinek &

Zaichkowsky, 1977) and The Pictorial Scale of Perceived Competence and Social Acceptance

for Young Children (Harter & Pike, 1984) are two of the scales that present physical and

motor competence items together with pictures about other domains, but these instruments

were not developed for physical education settings. Ulrich & Collier’s (1990) Modified

Pictorial Scale of Perceived Physical Competence for Children was developed specifically to

measure perceived gross motor competence in children between 7 and 12 years of age in the

physical education setting. This measure consists of different items representing skills like

kicking a ball, rolling a ball, throwing a ball, etc. These scales contain pictures that depict a

competent child (boy or girl) on one side and another not very skilled child on the other side.

For each skill, children are asked to select the child in the picture most like him or her

(Shapiro et al, 2002).

The notion of motor competence involves a motivational and affective dimension

which should not be overlooked and which mediates children’s performance and influences

the choice of tasks, persistence in practice and performance in movement contexts (Griffith &

Keogh, 1982). Therefore, one could ask: what are the perceptions of their physical or motor

competence in young children?, what role do these perceptions play in school physical

education? and can we get information about these self-perceptions in young children?.

The purpose of this investigation was to provide evidence for validity and reliability of

a new scale, the Children’s Perception of Motor Competence Scale (CMPI) for measuring

young children’s perceptions of gross and fine motor competence (4 to 6 years old) in the

physical education settings.

7

Method

Participants

A total of 495 children, ranging from 4 to 6 years (247 boys and 248 girls) provided

data for this study: 142 children were four years old (78 boys and 64 girls), 147 children

were five years old (74 boys and 73 girls) and 206 children were six years old (95 boys and

111 girls). All of them attended state and private schools in Madrid, Spain. Principals and

parents were informed about this study and they accepted to participate in it.

Children’s Perception of Motor Competence Scale (CMPI)

The first task in the development of this scale was to analyze physical education

classes of young children (4 to 6 years old), to classify the kind of tasks and skills that

children usually practice in these contexts. The first selection was based on a report that

twenty school physical education teachers made about the most common tasks at this

physical education level. The analysis of this report shown that 33 skills were the most

common. As a second step, these skills were considered and validated by three physical

education experts who accepted that 23 of the skills and behaviors were typical and

representative of physical education and movement sessions in this age range.

These skills were prepared as a booklet with pictures, where boys and girls were

shown performing, or not performing, these diverse basic perceptual-motor tasks: running,

jumping, balancing, swimming, throwing, catching, posting coins, threading, etc., in a

bipolar fashion as previously explained. At the same time, scenes related to the subject and

the teacher were added These items were administered to a group of children of this age

range to assess the understanding the drawings and expressions. After analysis, item

number 4 (kicking task) was eliminated because the drawings were confusing for these

children. The final version had 22 items; the first item served as an example to explain how

to respond. To respond to the scale, children pointed out with his or her finger or a pencil,

which boy or girl in each scene is most like what the child felt or would do themselves (Fig.

1).

INSERT FIG.1 HERE

8

Movement ABC Test and Children’s Motor Competence Observation Scale

In order to examine the relationship of perceived and actual motor competence, after

completing the scale, subject’s actual motor competence was measured using the Movement

ABC test (Henderson & Sudgen, 1992). This test was designed for use with children aged 4

to12 years and older. A total of 32 items are divided into four sets of eight tasks, each set

intended for use with children of specific ages. The first set of items, labeled Age Band 1,

is designed for use with 4 to 6 year-old children. The second set, Age Band 2, for 7 and 8

years old children, the third set for 9 to 10 year-olds and the fourth set for children 11 year

old and older. The test is identical in its structure in each age band. Three items involve the

use of the hands, two items require the children to catch or throw a bean bag or small ball

and three items assess static and dynamic balance. In this investigation, items of Age Band

1 were used for 4 to 6 year-olds (Table 1).

INSERT TABLE 1 HERE

Other strategy was to ask physical education teachers to observe children’s motor

competence in classes, by means of the Children's Motor Competence Observation Scale

(ECOMI) designed as an observational tool (Ruiz, Graupera, & Gutiérrez, 2001). This scale

depicts different items representing situations of differing complexities by grouping in two

factors, General Motor Competence and Motor Control, which frequently occur in physical

education sessions and indicate children’s motor competence. The presentation of each of the

items used a Likert scale of 4 points, in which 1=the absence of the behavior and 4=presence

(Table 2).

INSERT TABLE 2 HERE

Procedures

The Children’s Perception of Motor Competence Scale (CMPI) was administered

individually in a separate part of the school (office or empty classroom) and lasted a

variable length of time. Each boy or girl was shown each of the 22 scenes and asked which

boy or girl of the scene he or she identified most with. After answering this scale, subjects

performed individually the Movement ABC test (MABC) and during the next two weeks

their teachers observed them in the physical education classes and completed the Children's

Motor Competence Observation Scale (ECOMI).

9

Results

Exploratory Factor Analysis

Exploratory factor analysis was conducted on the scale to examine the internal

structure of the instrument and its principal dimensions. The maximum likelihood method was

used, making all calculations with the PRELIS 2.51 program (Sörbom & Jöreskog, 1999). One

main component was extracted from this analysis (Eigenvalue = 8.58; % Variance = 40.84)

and a second component was extracted with lower factor loadings (Eigenvalue=1.92; %

Variance=9.15). All factor loadings in the first factor were higher than .40. These data suggest

that the items of the CMPI scale measure a single construct than we could call: Perception of

Motor Competence (Table 3).

INSERT TABLE 3 HERE

Factor loadings of the majority of items in the second factor was lower than .30, and

when these loadings were higher, still were far below loadings on the first factor. These

higher coefficients in the second factor were of opposite sign, item 13 (rope jumping), item

10 (jumping over a rope) and item19 (Climbing a gym ladder) showed positive loadings,

but item 16 or item 18 (threading beads and posting coins) had negative correlations. It is

possible that there is a secondary factor that comprising the fine-motor items, in contrast

with the gross motor competence items. In conclusion, Factor 1 accounted for 40% of the

variance, and factor loadings on all items were higher than for the second factor (except

item 13). This suggested that the Children’s Perception of Motor Competence Scale is one-

dimensional and measures a general first-order factor (Carmines & Zeller, 1979). As an

alternative model, it is possible to accept that this general factor corresponds to a second-

order dimension with the inclusion of other first-order factors, differentiated by perceptions

of competence in gross and fine motor tasks.

Confirmatory factor analysis

The adequacy of the factor structure of the Children’s Perception of Motor

Competence Scale was examined using confirmatory factor analysis. Two alternative models

were examined to assess their effectiveness in representing the perception of motor

competence responses in young children. These analyses were made with the LISREL 8.51

program (Jöreskog & Sörbom, 1996). The single-factor model specifies all the CMPI items

10

loading on one first-order factor (Table 3). Such a model tests a one-dimensional

conceptualization of the Perceived Motor Competence construct; that is, it tests the

assumption that all items measured a single construct, rejecting the theory that perceived

motor competence in young children consists of distinct gross and fine motor elements.

Examination of goodness-of-fit measures showed that the model had a satisfactory fit to data,

as the non-normed fit index, comparative fit index and the root mean squared error of

approximation reach the desirable cut-off values (Chi square not significant: p=.33, GFI>.90,

RMSEA<.05; AGFI near .90; NNFI=1, CFI=1 and PGFI=.75) (Table 4). This one-

dimensional goodness-of-fit model was analyzed with a sample of 200 children, as different

authors recommend for estimating these kind of models (Hair et al., 1998) and that is proposed

as being the critical sample size (Hoelter, 1983).

Higher order factor analysis model was considered as an alternative, with a second-

order factor that assessed Perceived Motor Competence and two first-order factors that

differentiated the perceived motor competence into fine or manipulative tasks and gross

motor tasks. This model is represented in Fig. 2 with correlation coefficients between the

factors and variables.

INSERT FIGURE 2 HERE

The second order model had all of the main fit requisites, but showed better indexes

than the one-dimensional model. Goodness-of-fit model was analyzed in this case with the

same 200 children. Then, absolute fit measures, Chi-square value was lower (p=.87), GFI

higher (.93), standardized RMR and RMSEA better, and the ECVI was a little lower,

something that is particularly interesting when we talk about alternative models. Indexes of the

incremental fit measures are equivalents into the two models but the AGFI of the higher-order

model is better, and surpass the recommended value of .90 (AGFI=.91), this measure in the

one-dimensional model was lower (AGFI=.89). The indexes of parsimonious fit measures

were very similar into the two models (Table 4). As a consequence of this comparison, we

selected the second-order factor model as the theoretical structure of The Children’s

Perception of Motor Competence Scale. The measure comprises one scale of Perceived Motor

Competence (PMC) and two subscales, Perceived Gross Motor Competence (PGMC) and

Perceived Fine Motor Competence (PFMC). The score would be calculated by summing items

responses for the subscales.

11

INSERT TABLE 4 HERE

The final estimation process was a cross-validation procedure of dividing the sample

into two parts with the objective of validating the results of the selected second-order factor

analysis. Goodness-of-fit measure of the second sample were very similar to the first sample.

The absolute fit measures showed a Chi-square value highest but very satisfactory (p=.76), the

GFI, standardized RMR and RMSEA were very similar and the ECVI was identical. Indexes

of the incremental fit measures were equivalents into the two models, AGFI y CFI were

identical with the two samples, indexes of parsimonious fit measures were very similar into

the two samples, though PNFI was slightly lower in the second sample (Table 4). In

conclusion, internal validity of the second-order factor model was satisfactory and indicative

of its power.

External Validity

One of the main aspects in the development of a test of perceived competence is the

level of accuracy in representing the actual competence of the subjects. In this case, it was

necessary to consider the relationships among the results of Children’s Perception of Motor

Competence Scale (CMPI) and the Movement ABC Test (Henderson & Sudgen, 1992) and

teachers’ perceptions of children’s actual competence as measured by the Motor

Competence Observation Scale (ECOMI; Ruiz et al., 2001). These results showed that

scores on the scales of the CMPI, Perceived Gross Motor Competence, Perceived Fine

Motor Competence and the total score had significant relationships with some of the actual

motor competence variables. These relationships were low and moderate (ranging from .10

to .31) (Table 5).

INSERT TABLE 5 HERE

The Perceived Gross Motor Competence Scale had significant relationships with the

two ECOMI subscales, General Motor Competence and Motor Control and with the global

result of the scale. There were significant relationships with two components of the

Movement ABC test, Gross-motor Skills and Ballistic Skills, and with the total score, too.

These coefficients ranged between .31 and .12. The Perceived Fine Motor Competence

Scale had significant relationships with the two components of the Movement ABC test

that employed hands in their performance, Ballistic Skills and Fine-motor skills, and with

12

the total score of the ECOMI scale. These coefficients were lower and ranged from .14 to

.10. The global index of the Perceived Motor Competence had significant relationships

with all items of the General Motor Competence dimension of the ECOMI scale, and with

Balance skills and Ballistic skills measures of the Movement ABC test. These coefficients

were between .20 y .10. It is interesting to emphasize, specifically, correlations between the

general and gross-motor competence, and Balance skills of the two motor instruments, the

Movement ABC and the ECOMI scale, with the Perceived Gross Motor Competence

subscale of the CMPI (ranging from .26 to .31). Finally, multiple correlations were

calculated between the Children’s Perception of Motor Competence Scale (CMPI) with the

five partial measures of the two motor tests. This result was R= .35 (p< .001), 12%

common variance.

Reliability

Evidence of reliability was determined using internal consistency reliability and

complementarily, was calculated as the composite reliability of the two subscales (Jöreskog

& Sörbom, 1996).The estimate of the internal consistency of the global Children’s

Perception of Motor Competence Scale (CMPI) indicated that this instrument was reliable

(α = .81) and showed that the responses of these young children were consistent. Reliability

estimates of the two subscales were lower: Perceived Gross Motor Competence .78 and

Perceived Gross Motor Competence Scale .65. The reliability estimates of the composite

reliability coefficients were very similar to alpha coefficients (Table 6).

INSERT TABLE 6 HERE

As a complement of these analyses, the reliability of the global scale was calculated

(alpha coefficients) for the different age and gender samples of this study. There were no

differences in the consistency of the responses across age groups, and there were no gender

differences (Table 7). All of these coefficients were near .81 or between .79 and .82. This is

particularly important because these results suggest that children as young as 4 years old can

respond to this kind of scale with so reliability as older children.

INSERT TABLE 7 HERE

13

Item analysis

Item analysis demonstrated a good homogeneity for the total Perceived Motor

Competence scale, all items having correlations item-total higher than .30. Only item 6 had a

lower coefficient (Table 8). It is important to remember that all items had loadings higher than

a .40 on the first factor of the exploratory analysis (Table 3), and all items had item-subscale

correlations higher than .60 on the Perceived Gross Motor Competence subscale, .71 on the

Perceived Fine Motor Competence and .43 on the Perceived Motor Competence (except item

6 that was .39) in the confirmatory analysis (Fig. 2).

INSERT TABLE 8 HERE

Differential Item Functioning (DIF) was calculated by two methods, the Mantel-

Haenszel method (Holland & Thayer, 1988) and logistic regression method (Swaminathan &

Rogers, 1990). The first method analyzed the possibility of a differential functioning by

gender in each item. Only two items (9.5%) showed this effect, one in favor of girls (item 2)

and the other in favor of boys (item 14). These items form part of the general scale but not of

the two subscales. In conclusion, gender’s Differential Item Functioning effect in the general

score was not significant (Table 8). When we analyzed the gender effect with the logistic

regression method, we observed that there was no coincidence with the results of the other

method because this test is stronger when the differential item functioning is not uniform

(Rogers & Swaminathan, 1993). In this case there were no significant differential effects in the

two items identified by the other test (item 2 and 14), nevertheless, this procedure found three

items with a non uniform Differential Item Functioning. The effect must be considered small

because it was not detected with the other test. Item 5 is slightly favorable to girls and the

other two items (numbers 16 and 22) to boys. In our opinion there is no global gender effect.

In order to analyze if item differences by gender had the same direction than the total

score, it was estimated a non uniform Differential Item Functioning (interaction gender by

scale), as a result of this analysis the same three items showed an opposite direction to the total

scale, thus this effect were balanced and an small proportion of items (14%), had a satisfactory

homogeneity and good factorial loads, and it was not necessary to eliminate them.

14

Differences by age group and gender in CMPI scales.

As a part of this analysis the effects of gender and age group were assessed. Three

unvaried ANOVA (SPSS 11.01) was applied with the two independent factors: gender and

age, and Perception of Gross Motor Competence, Perception of Fine Motor Competence and

the global factor, Perception of Motor Competence of the CMPI Scale as dependent variables,

since CMPI subscales were related among them. Table 9 gives means and standard deviations

contrasted by gender and age for the three dependent variables.

INSERT TABLE 9 AND TABLE 10 HERE

There were no significant effects by gender in the every three subscales. However, if

we consider age, there were significant effects for each subscale, and a significant interaction

gender by age effect in relation to the Perception of Motor Competence scale (Table 10).

Perception of motor competence increase with age, and particularly between four and five

years old, differences were significant on the three measures (Table 11). This effect was larger

in girls between 4 and 5 years old, specifically on the general Perception of Motor

Competence, where the interaction effect was significant.

INSERT TABLE 11 HERE

Discussion

The focus of this study was on a first construct validation of the measurement of 4 to 6

year-olds children perceived motor competence based on the responses from the Children’s

Perception of Motor Competence Scale (CMPI). It is interesting that young children can

respond consistently if they are confronted with scenes of their life like a gymnasium,

swimming pool or physical activity settings. Another interesting result was that there were no

differential functioning by sex. It is a question of debate and occasionally causes criticisms

when measures do not have male and female versions. In this case, it is possible to say that

children observed and decided about these scenes in terms of the skill levels and not in terms

of the sex of the figures.

In relation to gender differences, the results of this study were different with other

studies made in the U.S.A. (Harter, 1985) or in the Netherlands (Van Rossum et al., 1999). In

these studies perceived motor competence was higher in boys than in girls, but these

investigations were made with children older than children of this study.

15

This study also assessed the relationships between perceived and actual motor

competence in this age range. Some studies have showed that there are positive and significant

correlations between actual and perceived motor competence (see McKiddie & Maynard,

1997), by contrast others had not found this relationships (Goodway & Rudisill, 1997; Yun &

Ulrich, 1997; Raudsepp & Liblik, 2002). All of these studies were made with children of ages

over 7 years. Dutch adaptations of Harter’s measures obtained similar findings with positive

correlations, .36 for younger children, between perceived motor competence and the

performance on a motor skills test. It is interesting to point out that the results of multiple

correlations in this study with young children was .35, therefore we can say that the

covariation between perceived and actual motor competence in younger children was about

12% (Van Rossum et al., 1999). In our study, children 4 to 6 years of age showed accuracy in

evaluating their own motor competence, Children’s Perception of Motor Competence Scale

(CMPI) results were related with their actual competence measured by a motor test and a

motor competence observational scale.

The practical significance of this study is related to the interest in self perceptions

among motor behavior scientists, psychologists and physical education teachers. The popular

notion of perceived motor competence as tied to self-esteem and healthy psychosocial

function, has empirical support. It is possible to think that low perceived motor competence in

these ages could be the origin of future motivational problems, emotional difficulties in

physical activity contexts and sports participation as in the case of Developmental

Coordination Disorders (Causgrove & Watkinson, 2002). Physical education teachers could

use the scale to assess self-perceptions of motor competence in their pupils and design

developmental tasks and activities to improve skills and self-perceptions.

Given that motor self-perceptions are an important component of self-esteem, more

research is needed in the near future, as to the potential of the Children’s Perception of Motor

Competence Scale (CMPI) as a measure instrument in younger children. As with any measure,

establishing validity is a never-ending process, and hence further research is warranted to

replicate the current results. At this time, however, these data indicate that it has promise.

16

References

Carmines, E. G., & Zeller, R. A. (1979) Reliability and Validity Assessment. Newbury Park,

CA: Sage.

Causgrove, J., & Watkinson, E. (2002) Considering motivation theory in the study of

developmental coordination disorders. In S. Cermark & D. Larkin (Eds.)

Developmental coordination disorders. Albany, NY: Delmar, Thompson Learning. Pp.

185-199.

Connolly, K. (1980) The development of motor competence. In C. Nadeau, W. Haliwell,

K.M. Newell & G. Roberts. (Eds.), Psychology of motor behavior and sport-1979.

Champaign: Human Kinetics. Pp. 229-252 .

Connolly, K. & Bruner, J. (1973) Competence: its nature and nurture. In K. Connolly & J.

Bruner (Eds.), The growth of Competence. London: Academic Press. Pp. 3-7.

Delignières, D. (1992) Risque préférentiel, risque perçu et prise de risque. In J. Famose (Dir.),

Cognition et performance. Paris: INSEP. Pp. 79-103.

Goodway, J., & Rudisill, M. (1997) Perceived physical competence and actual motor skill

competence of African American preschool children. Adapted Physical Activity, 14,

314-326.

Griffith, N., & Keogh, J. (1982) A model of movement confidence. In J. Scott Kelso, & J.

Clark (Eds.), The development of movement control and co-ordination. New York:

Wiley. Pp. 213-236.

Hagger, M., Ashford, B., & Stambulova, N. (1998) Russian and British children’s physical

self-perceptions and physical activity participation. Pediatric Exercise Science, 10,

137-152.

Hair, J., Anderson, R., Tatham, R., & Black, W. (1998) Multivariate data analysis (Fith

edition). Upper Saddle River, NJ: Prentice-Hall, Inc.

Harter, S. (1978) Effectance motivation reconsidered: toward a developmental model. Human

Development, 21,34-64.

Harter, S. (1981) The development of competence motivation in the mastery of cognitive and

physical skills: is there still a place for joy?. In G. Roberts, & D. Landers (Eds.),

Psychology of motor behavior and sport. Greenwich, CT: JAI Press. Pp. 3-29.

17

Harter, S. (1982) The Perceived Competence Scale for Children. Child Development, 53, 87-

97.

Harter, S. (1985) Manual for the self-perception profile for children. Denver: University of

Denver.

Harter, S. (1990) Causes, correlates and the functional role of global self worth: A lifespan

perspective. In R. Stemberg, & J. Kolligian, Jr. (Eds.) Competence considered. New

Haven, CT: Yale Unive. Press. Pp. 67-97.

Harter, S., & Pike, R. (1984) The Pictorial Scale of Perceived Competence and Social

Acceptance for Young Children. Child development, 55, 1969-1982.

Henderson, S., & Sudgen, D. (1992) Movement Assessment Battery for Children. London:

Psychological Corp.

Hoelter, J. (1983) The analysis of covariance structures: Goodness-of-fit indices. Sociological

Methods and Research, 11, 325-344.

Holland, P., & Thayer, D. (1988) Differential item performance and the Mantel-Haenszel

procedure. In H. Wainer, & H. I. Braun (Eds.) Test validity Hillsdale, NJ: Erlbaum.

Pp.129-145.

Jöreskog, K., & Sörbom, D. (1996) LISREL 8: user’s reference guide. Chicago, IL:

Scientific Software International.

Keogh, J., & Sudgen, D. (1985) Movement skill development. New York: Macmillan.

Martinek, T., & Zaichkowsky, L. (1977) Manual for the Martinek-Zaichkowsky Self-concept

Sacle for Children. Jacksonville, Il.: Psychologists & Educators, Inc.

McKiddie, B., & Maynard, I. (1997) Perceived competence of schoolchildren in Physical

Education. Journal of Teaching in Physical Education, 16, 324-339.

Raudsepp, L., & Liblik, R. (2002) Relationship of perceived and actual motor competence in

children. Perceptual and Motor Skills, 94, 1059-1070.

Rogers, H. & Swaminathan, H. (1993) A comparison of logistic regression and Mantel-

Haenszel procedures for detecting differential item functioning. Applied Psychological

Measurement, 17, 105-116.

Ruiz, L. (1995) Competencia Motriz. Madrid: Gymnos.

18

Ruiz, L., Graupera, J., & Gutiérrez, M. (2001) Observing and detecting pupils with low motor

competence in physical education: ECOMI scale in the gymnasium. International

Journal of Physical Education, XXXVIII, (2), 73-77.

Shapiro, D., Yun, J., & Ulrich, D.A. (2002) Measuring perceived gross motor skill

competence in children. International Journal of Sport Psychology, 33, 391-409.

Swaminathan, H., & Rogers, H. (1990) Detecting differential item functioning using logistic

regression procedures. Journal of Educational Measurement, 27, 361-370.

Sórbom, D., & Jöreskog, K. (1999) PRELIS 2: user’s reference guide. Chicago, IL: Scientific

Software International.

Ulrich, B. (1987) Perceptions of physical competence, motor competence, and participation

in organized sport: their interrelationships among young children. Research

Quarterly for Exercise and Sport, 58, (1), 57-67.

Ulrich, B., & Collier, D. (1990) Perceived physical competence in children with mental

retardation: modification of a pictorial scale. Adapted Physical Activity Quarterly, 1,

338-354.

Van der Steen, M., & Vermeer, A. (1987) Competence and movement training. Bewegen &

Hulpverlening, 4, 92-111.

Van Rossum, J., Musch, E. & Vermeer, A (1999) Perceived motor competence: Self-

referent thinking in physical education. In Y. Vanden, F. Bakker, S. Biddle, M.

Durand, & R. Seiler (Eds.) Psychology for physical educators. Champaign, IL.:

Human Kinetics. Pp. 187-210.

Villa, A., & Auzmendi, E.(1992) Medición del autoconcepto en la edad infantil (5-6 años).

Bilbao: Mensajero.

White, R. (1959) (1959) Motivation reconsidered: the concept of competence. Psychological

Review, 66, (5), 297-323.

Yun, J., & Ulrich, D. (1997) Perceived and actual physical competence in children with

mild mental retardation. Adapted Physical Activity Quarterly,14, 285-297

19

Fig. 1. Selected items of the Children’s Perception of Motor Competence Scale: left, item 7; right, item 22.

20

Table 1: Test Items of Age Band of the MABC Test.

DOMAIN ITEM DESCRIPTION MEASURE Manual- dexterity skills Posting coins Place coins through the slot in a bank

box as quickly as possible (preferred and non- preferred hand)

Time (s)

Threading beads Thread beads on a string as quickly as possible.

Time (s)

Bicycle trail Draw with a pen a single continuous line, following the trail without crossing its boundaries

Number of errors

Ball skills Catching a bean bag Catching a bean bag with two hands Number of catches

Rolling a ball into goal Roll a ball along the floor between two jumping stands.

Number of goals

Static/dynamic balance skills

One-leg balance Stand on one leg with the free leg bent at the knee (preferred and non preferred leg)

Time (s)

Jumping over the cord Jump over a cord with feet together Successful jump

Walking heels raised (tip-toe) Walk along a line, with heels raised, without stepping off the line.

Number of correct consecutive steps (up to 15)

21

Table 2. Items of the Children's Motor Competence Observation Scale (ECOMI).

GENERAL MOTOR COMPETENCE ITEMS 1. S/he runs about the gymnasium without bumping into classmates or objects. 2. S/he catches a ball with both hands in a controlled manner (thrown from a distance of 2 - 3 meters) 3. S/he can keep time while performing (clapping to music, playing a tambourine to music, walking in

time to music) 4. S/he maneuvers with agility round the obstacle courses. 5. S/he moves like children of her/his own age. 6. S/he is able to bounce a ball with one hand continuously while standing still. 7. S/he is able to bounce a ball continuously with one hand while moving around. 8. S/he is able to run and stop to avoid bumping into a classmate or object. 9. S/he has an overarm ball throwing pattern and performs it skillfully.

10. S/he learns the skills in the physical education program well. 11. S/he takes part in sports and ball games with her/his classmates in a competent manner. 12. S/he performs automatically the same type of movements as those performed by her/his classmates

MOTOR CONTROL ITEMS 1. S/he hops forward on the left foot in a controlled manner at least 10 times without stopping. 2. The same on the right foot at least 10 times without stopping 3. S/he hits a ball back with a racket or bat in a controlled manner. 4. S/he can balance on one foot for more than 30”. 5. S/he is able to jump continuously on the spot on one foot in a space of 50 cm x 50 cm a maximum of

50 times. 6. S/he can catch a tennis ball with two hands in a controlled manner. 7. S/he can catch a tennis ball with one hand in a controlled manner.

22

Table 3. Loadings for Unrotated Orthogonal Factors for the Children’s Perception of Motor Competence Scale.

Item descriptions Factor 1 Factor 2 Item 2. One-leg balance .58 .29 Item 3. Children running .61 .00 Item 5. Walking on boxes .74 -.30 Item 6. Throwing a ball .43 .16 Item 7. Catching a ball .72 -.11 Item 8. Playing into the water .53 .18 Item 9. Bouncing a ball .68 .21 Item 10. Jumping over a rope .57 .37 Item 11. Playing with others a game .52 -.02 Item 12. Collaborating with others in a game .63 -.33 Item 13. Rope jumping .43 .53 Item 14. Dressing the sport suit .61 .13 Item 15. Walking along a line .79 .02 Item 16. Threading beads .63 -.49 Item 17. Playing with a hockey stick .65 .28 Item 18. Posting coins .72 -.33 Item 19. Climbing a gymnasium ladder .68 .37 Item 20. Going to physical education class .70 -.40 Item 21. Playing with his/her teacher .78 -.50 Item 22. Rolling a ball into goal .59 .29 Item 23. Rolling the body (somersault) on the mat .67 .19 Eigenvalue 8.58 1.92 % Variance 40.84 9.15 % Cumulative Variance 40.84 49.99

23

Fig. 2: Confirmatory Factor Analysis: Second-order Factor Model.

24

Table 4: Confirmatory Factor Analysis: Goodness-of-fit Measures of the Alternative Models.

Factor model One-dimensional Second-order

Goodness-of-fit measure First sample

(N=200) First sample

(N=200) Second sample

(N=200) Absolute fit measures

Degrees of Freedom 189 187 187 Likelihood-ratio Chi-Square 197.09 (p=.33) 165.44 (p=.87) 173.09 (p=.76) Goodness of Fit Index (GFI) .91 .93 .92 Standardized RMR .053 .050 .051 Root Mean Square Error of Approximation (RMSEA) .015 >.001 >.001 Expected Cross-Validation Index (ECVI) 1.41 1.38 1.38

Incremental fit measures

Adjusted Goodness of Fit Index (AGFI) .89 .91 .91 Non-Normed Fit Index (NNFI) 1.00 1.00 1.02 Normed Fit Index (NFI) .74 .77 .74 Comparative Fit Index (CFI) 1.00 1.00 1.00 Incremental Fit Index (IFI) 1.00 1.00 1.02 Relative Fit Index (RFI) .71 .74 .71

Parsimonious fit measures

Parsimony Normed Fit Index (PNFI) .67 .68 .66 Parsimony Goodness of Fit Index (PGFI) .75 .75 .75

25

Table 5: Correlation Coefficients Among the CMPI Scales and the Main Scores of the Test MABC and ECOMI, and Multiple Correlation among the Children’s Perception of Motor Competence Scale and the Five Subscales of Motor Tests.

CMPI Scale Perceived Motor

Competence Perceived Gross

Motor Competence

Perceived Fine Motor

Competence

ECOMI Scale

General motor competence .20 † .31 † Motor control .10 * .17 † Total .16 † .25 † .13 †

MABC Test

Balance skills .12 † .26 † Ballistic skills .10 * .21 † .14 † Fine-motor skills .10 * Total .03 .12 * .07

Multiple correlation .35 † * p<.05; † p<.01

26

Table 6. Reliability (Alpha Coefficients) of the total CMPI Scale (PMC) and of the Two Subscales and Composite Reliability.

CMPI scale Alpha Composite reliability

Subscales Perceived Gross Motor Competence .78 .80 Perceived Fine Motor Competence .65 .63

Total scale

Perceived Motor Competence .81

27

Table 7: Perceived Motor Competence Reliability in the Total Sample and by Gender and age.

Perceived Motor Competence Group sample Alpha N Gender Girls .80 247 Boys .82 248

Age (years) 4 .81 142 5 .79 147 6 .80 206

Total sample .81 495

28

Table 8: Item Analysis of the Children’s Perception of Motor Competence Scale: Homogeneity (corrected item-scale correlation) and Differential Item Functioning (Gender Effect and Interaction of Perceived Motor Competence by Gender).

Differential Item Functioning Analysis Method Logistic Regression Mantel-Haenszel Effects Effect Gender Interaction

PMC by Gender Gender

Item

Corrected Item-scale Correlation Wald p Wald p Chi-square p

2 .40 2.80 .10 1.48 .22 7.04 * .01 3 .42 1.10 .30 .71 .40 1.03 .31 5 .51 4.70 * .03 4.27 * .04 .43 .51 6 .28 .42 .52 .18 .67 2.31 .13 7 .50 .62 .43 .93 .33 .22 .64 8 .36 .93 .33 .94 .33 .19 .66 9 .47 .00 .97 .09 .76 .57 .45

10 .42 .29 .59 .62 .43 .26 .61 11 .36 .35 .55 .00 .99 1.59 .21 12 .42 .41 .52 .28 .60 .01 .95 13 .30 .47 .49 1.22 .27 3.55 .06 14 .42 .00 .97 .28 .60 4.60 * .03 15 .56 2.91 .09 3.20 .07 .01 .93 16 .50 7.77 * .01 7.17 * .01 .78 .38 17 .48 1.90 .17 2.30 .13 .30 .59 18 .56 1.04 .31 .51 .47 3.50 .06 19 .40 .04 .83 .30 .59 1.92 .17 20 .55 .50 .48 .60 .44 .13 .71 21 .66 3.01 .08 3.28 .07 1.23 .27 22 .44 8.13 * .01 6.13 * .01 1.93 .17 23 .46 .52 .47 .38 .54 .43 .51

* Differential Item Functioning with p < .05

29

Table 9: Descriptive Statistics of the Children’s Perception of Motor Competence subscales by gender and age.

Perceived Gross Motor Competence

Perceived Fine Motor Competence

Perceived Motor Competence

Gender Age (years)

Mean Standard Deviation

Mean Standard Deviation

Mean Standard Deviation

N

Boy 4 3.46 .91 3.72 .72 17.82 3.33 78 5 3.72 .54 3.81 .66 19.20 2.61 74 6 3.55 .73 3.80 .65 18.39 3.29 95 Total 3.57 .74 3.78 .67 18.45 3.15 247

Girl 4 3.25 1.15 3.53 .87 16.38 3.86 64 5 3.59 .68 3.95 .23 18.81 2.60 73 6 3.53 .77 3.85 .49 18.79 2.55 111 Total 3.48 .87 3.79 .58 18.17 3.13 248

Total 4 3.37 1.03 3.63 .79 17.17 3.64 142 5 3.65 .62 3.88 .50 19.01 2.60 147

6 3.54 .75 3.83 .57 18.61 2.91 206 Total 3.52 .81 3.79 .63 18.31 3.14 495

30

Table 10: Unvaried ANOVAs of Children’s Perception of Motor Competence subscales by Gender and Age, and Interaction Effects.

Dependent variable Effect Sum of squares

df Squared mean

F p Eta squared

Power

Gender 1.67 1 1.67 2.58 .11 .01 .36 Age 6.46 2 3.23 4.97 .01 * .02 .81 Gender*Age .82 2 .41 .64 .53 .00 .16 Error 317.27 489 .65

Perceived Gross Motor Competence

Total 6470.00 495

Gender .00 1 .00 .00 .97 .00 .05 Age 5.19 2 2.59 6.73 .00 * .03 .92 Gender*Age 2.00 2 1.00 2.59 .08 .01 .52 Error 188.46 489 .39

Perceived Fine Motor Competence

Total 7290.00 495

Gender 27.45 1 27.45 2.98 .08 .01 .41 Age 293.69 2 146.85 15.95 .00 * .06 1.00 Gender*Age 71.29 2 35.65 3.87 .02 * .02 .70 Error 4502.59 489 9.21

Perceived Motor Competence

Total 170873.00 495 * p<.05

31

Table 11: Post-hoc Multiple Comparisons by Age.

95% CI CMPI Scale Age differences

Mean differences

Standard Error p

Lower Upper 5-4 yrs. old .29 † .09 .008 -.51 -.06 Perceived Gross Motor

Competence 6-5 yrs. old -.11 * .09 .569 -.09 .32

5-4 yrs. old .24 * .07 .003 -.42 -.07 Perceived Fine Motor Competence 6-5 yrs. old -.05 .07 1.000 -.11 .21

5-4 yrs old 1.84 † .36 .000 -2.70 -.98 Perceived Motor Competence 6-5 yrs old -.40 .33 .668 -.39 1.19 Bonferroni criterion. * p<.05; † p<.01