Faculty of Sport, University of Ljubljana, ISSN 1318-2269  5Kinesiologia Slovenica, 12, 1, 5–13 (2006)
Gustav Bala* 
Dejan Madić
INFLUENCE OF THE SELECTION 
OF MOTOR TEST SAMPLES ON THE 
PARSIMONY OF MOTOR SPACE
VPLIV IZBIRE VZORCEV MOTORIČNIH 
TESTOV NA PARSIMONIČNOST 
MOTORIČNEGA PROSTORA
Abstract
The present study assesses characteristics of the selection 
of motor test samples from the viewpoint of parsimony 
of motor space. The obtained results point to the need 
to respect elementary principles when defining the 
structures in an analysed sample of variables. Namely, 
a) every hypothetical factor should be evaluated by at 
least three motor tests with good metric characteristics 
and b) every hypothetical research model should be 
balanced, meaning that every factor should be evaluated 
by the same number of motor variables. Considering 
these principles, the relevant characteristics of the 
primary factor, as well as the highest order of a certain 
hierarchical factor model, can be obtained.
Key words: motor test, hypothetical factor model, 
parsimony
Faculty of Physical Education, University of Novi Sad, 
Serbia
*Corresponding author:
Faculty of Physical Education
University of Novi Sad
Lovćenska 16, 21000 Novi Sad, Serbia
Tel.: +38121 450 188
Fax: + 38121 450 199
E-mail: gustavbala@yahoo.com
Izvleček
Pričujoča študija ugotavlja značilnosti izbiranja 
vzorcev motoričnih testov z vidika parsimoničnosti 
motoričnega prostora. Rezultati raziskave so pokazali, 
da je potrebno pri definiranju struktur analiziranih 
vzorcev spremenljivk upoštevati osnovna načela. 
Velja izpostaviti dve načeli: a) vsak hipotetični 
dejavnik mora biti preverjen z vsaj tremi motoričnimi 
testi z dobrimi merskimi značilnostmi in b) vsak 
hipotetični raziskovalni model mora biti uravnotežen, 
kar pomeni, da se vsak dejavnik preverja z enakim 
številom motoričnih spremenljivk. Z upoštevanjem 
teh načel lahko ugotovimo relevantne značilnosti 
primarnih faktorjev in faktorjev višjega reda izbranega 
hierarhičnega faktorskega modela. 
Ključne besede: motorični test, hipotetični faktorski 
model, parsimoničnost 
6 Selection of motor test samples Kinesiologia Slovenica, 12, 1, 5–13 (2006) 
INTRODUCTION
Amongst true kinesiologically-/methodologically-oriented researchers it is well-known and 
common that kinesiological research requires at least 3-5 variables for determining the factor 
structure of anthropological characteristics and abilities in order to gain a parsimonial solution 
of isolated latent variables (factors, dimensions) (e.g., Kurelić, Momirović, Stojanović, Šturm, 
Radojević, & Viskić-Štalec, 1975). Meanwhile, in practice, especially at the level of master’s 
theses and doctoral dissertations, as well some research works produced by insufficiently 
experienced students, researchers or mentors, one finds cases where this methodological 
demand is neglected.
The selection of a sample of motor tests necessary demands an explicitly defined hypothetical 
model of motor abilities that is to be researched or confirmed. When defining a hypothetical 
model created on the basis of results of previous research works or based on practical experi-
ence, one needs to make it operative and express it at the manifest level, i.e. define the proper 
sub-sample of motor tests for each latent motor dimension (factor) and eventually form a 
sub-sample of all motor tests to be applied in the relevant kinesiological research (Gredelj, 
Metikoš, Hošek, & Momirović, 1975). With such a qualitative selection of motor tests it is not 
irrelevant how this problem will be resolved quantitatively, i.e. how many motor tests each 
sub-sample should comprise, that is how many measurement instruments should be used for 
the total sample.
The aim of this paper is to analyse the structures of motor factors, i.e. the parsimony of 
analysed motor space, based on a sample of motor tests selected in such a way that one, two 
or three relevant motor variables are used to define a single hypothetical motor factor. The 
motor factors will not be interpreted in detail because this paper is essentially methodological 
and has the purpose to illustrate theoretical demands concerning the number of motor tests 
and to determine the factor structure of motor space.
METHOD
Participants
The sample of subjects consisted of 260 first- and second-year male students of the Faculty of 
Physical Education in Novi Sad (Serbia) who were 18-22 years of age and selected according 
to their biological, health, motor and psychological development.
Instruments
The sample of motor variables was obtained on the basis of the motor model according to Kurelić 
et al. (1975) and Gredelj et al., (1975):
1) Mechanism for movement structuring 
 I Functional co-ordination of primary motor abilities (C)
  a) The test ‘Agility on the floor’ (CAGFLOOR)
  b) The test ‘Dragging and jumping over’ (CDRAJUMP)
  c) The test ‘Co-ordination with stick’ (CCOSTICK)
2) Mechanism for tonus and synergetic regulation 
 II Balance (B)
  d) The test ‘One foot cross balance – eyes open’ (B1CROPEN)
Selection of motor test samples 7Kinesiologia Slovenica, 12, 1, 5–13 (2006)
  e) The test ‘One foot length-wise balance – eyes open’ (B1LEOPEN)
  f) The test ‘Flamingo’ (BFLAMING)
 III Frequency of simple movements (FQ)
  g) The test ‘Two foot tapping’ (FQ2FOOTT)
  h) The test ‘Plate-tapping’ (FQTAPPIN)
  i) The test ‘One foot-tapping’ (FQ1FOOTT)
 IV Flexibility (FL)
  j) The test ‘Toe touching – sitting straddle’ (FLTOESIT)
  k)  The test ‘Toe touching – standing’ (FLTOESTA)
  l) The test ‘Push off one leg – lying on the side’ (FLPUSH1L)
3) Mechanism for excitation intensity regulation 
 V Explosive strength (E)
  m) The test ‘Standing broad jump’ (ESTANJUM)
  n) The test ‘20m dash’ (E20MDASH)
  o) The test ‘Spring forward from front support on the floor’ (EREFLOOR)
4) Mechanism for excitation duration regulation
 VI General strength (S)
  p) The test ‘Bent arm hang’ (SARMHANG)
  r) The test ‘Horizontal hold lying on the back’ (SHORHOLD)
  s) The test ‘Sit-ups’ (SSIT-UPS).
The coding of the motor tests was as follows: the first letter in the code name was according 
to a hypothetical motor factor while the others were according to the names of the tests. The 
reason for this was to make the interpretation of the analysed factors easier. This means that 
the first letter in the code name C was for co-ordination, B for balance, FQ for the frequency 
of simple movements, FL for flexibility, E for explosive strength and S for general strength.
Table 1: Reliability levels of the motor tests
test α test α
a) CAGFLOOR .92 j) FLTOESIT .93
b) CDRAJUMP .93 k) FLTOESTA .94
c) CCOSTICK .88 l) FLPUSH1L .99
d) B1CROPEN .76 m) ESTANJUM .94
e) B1LEOPEN .90 n) E20MDASH .91
f) BFLAMING .92 o) EREFLOOR .95
g) FQ2FOOTT .92 p) SARMHANG .62
h) FQTAPPIN .85 r) SHORHOLD .97
i) FQ1FOOTT .87 s) SSIT-UPS .92
Legend: α – Cronbach’s reliability coefficient 
The conditions and techniques of measuring used in most of the tests were according to 
Metikoš, Prot, Hofman, Pintar and Oreb (1989); except the tests flamingo (Moravec, 1996), 
push off one leg – lying on the side and spring forward from front support on the floor (Madić, 
8 Selection of motor test samples Kinesiologia Slovenica, 12, 1, 5–13 (2006) 
1995, 2000). Every test was performed three times so that each of them was a composite test 
of three successive items.
The reliability of these tests was computed as Cronbach’s α-coefficient (α). The reliability values 
are shown in Table 1.
It is obvious that all the motor tests have quite good reliabilities, except the tests bent-arm 
hang (SARMHANG ) and one foot cross balance – eyes open (B1CROPEN).
The motor variables were assumed as the first principal component of the correlation matrix 
of every repetition result (three items) in the same composite test.
Procedure
The effect of applying different numbers of motor tests according to hypothetical factors can be 
seen in the final analysis of an explored motor space in a sample of subjects, in both quantitative 
and qualitative ways. This effect will be illustrated in the following cases:
1) the analysis of motor factor structures on a sample of motor tests which were selected so 
that one variable defines one hypothetical motor factor;
2) the analysis of motor factor structures on a sample of motor tests which were selected so that 
two variables define one hypothetical motor factor; and
3) the analysis of motor factor structures on a sample of motor tests which were selected so that 
three variables define one hypothetical motor factor.
In all three analyses, a promax transformation of the principal components of the corresponding 
correlation matrix of all variables was applied. The number of statistically significant principal 
components was obtained on the basis of the Guttman-Kaiser criterion. 
RESULTS
a) One variable defines one hypothetical motor factor
The analysis of motor factor structures on the sample of these motor tests which were selected so 
that one variable defines one hypothetical motor factor gave the following results:
Functional co-ordination of primary motor abilities (C)
a) the test co-ordination with stick (CCOSTICK) 
Balance (B)
b) The test one foot length-wise balance – eyes open (B1LEOPEN)
Frequency of simple movements (FQ) 
c) The test plate – tapping (FQTAPPIN) 
Flexibility (FL)
d) The test toe touching – standing (FLTOESTA) 
Explosive strength (E) 
e) The test standing broad jump (ESTANJUM) 
General strength (S) 
f) The test sit-ups (SSIT-UPS).
Selection of motor test samples 9Kinesiologia Slovenica, 12, 1, 5–13 (2006)
By applying the Kaiser-Guttman’s criteria two factors were obtained which explained 46.21% 
of the common variance of the entire motor space, which consists of applied variables. After a 
promax rotation of the principal components and on the basis of the pattern and structure of 
two isolated factors, the following conclusions can be derived about the nature of the factors 
(the significance saturations are typed in bold) (see Table 2):
a) functional co-ordination of primary motor abilities which assumed the speedy use of trunk 
and leg muscles, and 
b) factors which are saturated with variables for estimating balance, frequency of simple 
arm movements and flexibility, which pointed to the mechanism for tonus and synergetic 
regulation. Between these two motor factors there is a statistically significant correlation. 
Table 2: Pattern and structure matrices of rotated factors and correlation coefficient of factors 
(one variable per factor)
Variable
Pattern Structure
A1 A2 F1 F2
CCOSTICK -0.574 -0.279 -0.645 -0.425
B1LEOPEN -0.280 0.828 -0.070 0.757
FQTAPPIN 0.317 0.484 0.440 0.564
FLTOESTA 0.114 0.560 0.256 0.589
ESTANJUM 0.569 -0.059 0.554 0.085
SSIT-UPS 0.765 -0.109 0.737 0.086
r = 0.254
b) Two variables define one hypothetical motor factor
The analysis of motor factor structures on the sample of these motor tests which were selected so 
that two variables define one hypothetical motor factor produced the following results:
Functional co-ordination of primary motor abilities (C)
a) The test co-ordination with stick (CCOSTICK)
b) The test dragging and jumping over (CDRAJUMP)
Balance (B)
c) The test one foot length-wise balance – eyes open (B1LEOPEN)
d) The test flamingo (BFLAMING)
Frequency of simple movements (FQ)
e) The test two foot tapping (FQ2FOOTT)
f) The test one foot – tapping (FQ1FOOTT)
Flexibility (FL) 
g) The test toe touching – standing (FLTOESTA)
h) The test push off one leg – lying on the side (FLPUSH1L)
Explosive strength (E)
10 Selection of motor test samples Kinesiologia Slovenica, 12, 1, 5–13 (2006) 
i) The test standing broad jump (ESTANJUM)
j) The test spring forward from front support on the floor (EREFLOOR)
General strength (S)
k) The test bent arm hang (SARMHANG)
l) The test sit-ups (SSIT-UPS).
When the structure analysis of motor space was made according to the hypothetical model 
which was estimated with two motor variables per one hypothetical factor, four factors were 
isolated which explained 54.84% of the common variance of all 12 motor variables. With the 
same procedures as in the previous analysis, the interpretation of the structure for these motor 
factors was carried out (see Table 3): 
Table 3: Pattern and structure matrices of rotated factors (two variables per factor)
Variable
pattern structure
A1 A2 A3 A4 F1 F2 F3 F4
CCOSTICK -0.257 -0.606 0.067 0.084 -0.449 -0.671 -0.206 0.107
CDRAJUMP -0.325 -0.257 -0.203 0.078 -0.482 -0.434 -0.382 0.088
B1LEOPEN -0.101 0.019 0.852 -0.094 0.177 0.265 0.820 -0.026
BFLAMING -0.092 -0.001 -0.723 -0.227 -0.298 -0.272 -0.768 -0.271
FQ2FOOTT 0.379 0.352 0.054 -0.361 0.544 0.498 0.262 -0.386
FQ1FOOTT 0.623 -0.190 0.310 0.033 0.651 0.125 0.442 0.004
FLTOESTA 0.767 -0.032 -0.147 0.247 0.691 0.183 0.098 0.175
FLPUSH1L 0.786 -0.068 -0.087 0.010 0.735 0.171 0.135 -0.059
ESTANJUM 0.124 0.422 -0.146 0.460 0.186 0.418 0.064 0.442
EREFLOOR 0.102 0.029 0.116 0.800 0.084 0.106 0.213 0.800
SARMHANG -0.192 0.652 0.260 0.141 0.100 0.673 0.426 0.178
SSIT-UPS -0.194 0.820 -0.070 0.109 0.056 0.732 0.148 0.124
a) frequency of simple movements and flexibility,
b) general strength and functional co-ordination of primary motor abilities,
c) balance, and 
d) explosive strength. The correlation matrix between these four factors shows there is a statisti-
cally significant correlation among the first, second and third factors, as well between the 
second and third factors (see Table 4). 
Table 4: Correlation coefficient matrix of factors
Factor 1 2 3 4
1 1.000 0.341 0.310 -0.080
2 0.341 1.000 0.330 0.005
3 0.310 0.330 1.000 0.070
4 -0.080 0.005 0.070 1.000
Selection of motor test samples 11Kinesiologia Slovenica, 12, 1, 5–13 (2006)
c) Three variables define one hypothetical motor factor
The analysis of motor factor structures on the sample of motor tests which were selected so 
that three variables define one hypothetical motor factor showed the following results:
Functional co-ordination of primary motor abilities (C) 
a) The test agility on the floor (CAGFLOOR)
b) The test co-ordination with stick (CCOSTICK)
c) The test dragging and jumping over (CDRAJUMP)
Balance (B)
d) The test one foot cross balance – eyes open (B1CROPEN)
e) The test one foot length-wise balance – eyes open (B1LEOPEN)
f) The test flamingo (BFLAMING)
Frequency of simple movements (FQ)
g) The test two foot tapping (FQ2FOOTT)
h) The test plate – tapping (FQTAPPIN)
i) The test one foot – tapping (FQ1FOOTT)
Flexibility (FL) 
j) The test toe touching – sitting straddle (FLTOESIT)
k) The test toe touching – standing (FLTOESTA)
l) The test push off one leg – lying on the side (FLPUSH1L)
Explosive strength (E)
m) The test standing broad jump (ESTANJUM)
n) The test 20m dash (E20MDASH)
o) The test spring forward from front support on the floor (EREFLOOR)
General strength (S) 
p) The test bent arm hang (SARMHANG)
q) The test horizontal hold laying on the back (SHORHOLD)
r) The test sit-ups (SSIT-UPS).
This case showed the most acceptable solution for the size and balance of a variable sample 
according to the hypothetical model of motor abilities. Common variance in the motor 
space of 18 motor variables accounted for 57.87%. With the same transformation of principal 
components, i.e. in the promax position, the isolated factors can be defined in the following 
way (see Table 5):
a) frequency of simple movements,
b) flexibility, 
c) general strength, 
d) balance, 
e) functional co-ordination of primary motor abilities (dual factor), and 
f) explosive strength (dual factor). 
12 Selection of motor test samples Kinesiologia Slovenica, 12, 1, 5–13 (2006) 
Table 5: Pattern and structure matrices of rotated factors (three variables per factor)
Variable
pattern structure
A1 A2 A3 A4 A5 A6 F1 F2 F3 F4 F5 F6
CAGFLOOR -0.01 -0.07 0.09 -0.05 0.78 -0.01 -0.20 -0.14 -0.17 -0.16 0.77 -0.10
CCOSTICK -0.47 -0.05 -0.09 -0.03 0.16 0.00 -0.55 -0.22 -0.26 -0.23 0.29 -0.11
CDRAJUMP -0.14 -0.07 0.04 -0.01 0.78 -0.01 -0.32 -0.18 -0.23 -0.17 0.80 -0.13
B1CROPEN 0.42 -0.06 0.16 0.40 0.21 -0.04 0.50 0.12 0.29 0.52 0.03 0.04
B1LEOPEN -0.04 -0.01 -0.16 0.87 -0.04 0.00 0.18 0.10 0.09 0.82 -0.12 0.04
RFLAMING 0.01 -0.04 -0.05 -0.74 0.06 -0.01 -0.25 -0.16 -0.28 -0.77 0.19 -0.09
FQ2FOOTT 0.74 -0.13 -0.06 -0.11 -0.18 0.07 0.70 0.09 0.13 0.10 -0.30 0.17
FQTAPPIN 0.74 -0.03 0.02 -0.03 -0.04 0.06 0.74 0.20 0.21 0.20 -0.21 0.17
FQ1FOOTT 0.55 0.15 -0.04 0.10 0.10 0.06 0.61 0.33 0.12 0.27 -0.04 0.14
FLTOESIT -0.09 0.88 0.06 -0.03 -0.04 0.14 0.21 0.88 0.19 0.11 -0.13 0.25
FLTOESTA -0.05 0.86 0.00 0.07 -0.11 -0.01 0.25 0.86 0.15 0.20 -0.19 0.11
FLPUSH1L 0.41 0.55 -0.02 -0.07 0.05 -0.23 0.51 0.63 0.08 0.10 -0.05 -0.12
ESTANJUM -0.10 0.11 0.10 0.03 -0.06 0.81 0.10 0.21 0.25 0.11 -0.18 0.84
E20MDASH -0.27 0.09 0.07 0.03 -0.05 -0.77 -0.32 -0.07 -0.08 -0.07 0.07 -0.78
EREFLOOR -0.03 0.11 0.48 0.01 0.22 0.11 0.09 0.16 0.44 0.13 0.06 0.17
SARMHANG -0.05 -0.09 0.59 0.26 -0.20 -0.04 0.18 0.02 0.69 0.42 -0.38 0.08
SHORHOLD 0.03 0.08 0.79 -0.11 0.19 0.00 0.17 0.16 0.73 0.10 -0.04 0.11
SSIT-UPS 0.04 -0.08 0.67 -0.15 -0.27 -0.05 0.18 0.01 0.70 0.07 -0.44 0.08
It may be concluded that with the change to the motor test sample so that three motor variables 
define every hypothetical factor almost the entire hypothetical motor model was confirmed. 
Certain discrepancies were probably the result of the specificity of the subject sample, but 
also of some motor tests with poor metrical characteristics (validity and reliability). That was 
certainly the case with variable co-ordination with stick (CCOSTICK), which was a better 
estimator for frequency of simple movements, and with spring forward from front support on 
the floor (EREFLOOR) which had better validity for general strength. The correlation matrix 
showed it contains a large number of statistically significant elements (see Table 6)
Table 6: Correlation coefficient matrix of factors
Factor 1 2 3 4 5 6
1 1.000 0.302 0.240 0.291 -0.219 0.144
2 0.302 1.000 0.134 0.154 -0.090 0.124
3 0.240 0.134 1.000 0.279 -0.295 0.162
4 0.291 0.154 0.279 1.000 -0.153 0.075
5 -0.219 -0.090 -0.295 -0.153 1.000 -0.125
6 0.144 0.124 0.162 0.075 -0.125 1.000
Selection of motor test samples 13Kinesiologia Slovenica, 12, 1, 5–13 (2006)
DISCUSSION
The illustrated cases with different numbers of motor variables for defining the hypothetical 
models of motor abilities showed the need to respect elementary principles when defining the 
structures in an analysed sample of variables. To achieve the parsimony of motor space, i.e. 
the good and valid pattern and structure required for a meaningful logic interpretation, every 
hypothetical motor factor should be evaluated with at least three motor variables. All adequate 
motor tests, in fact the motor measurements, must have good metric characteristics. The third 
principal is that every hypothetical research model should be balanced, meaning that every 
factor should be evaluated by the same number of motor variables (an uneven number of 
variables could produce a biased situation). Assuming all principals are taken into considera-
tion, it is possible to obtain the relevant characteristics of latent factors (motor dimensions, 
abilities) of the primary (first-order) factors, but also at the extraction of higher-order ones in 
some hierarchical model of motor abilities.
This paper illustrates that it is impossible to get latent dimensions (factors) with a meaningful 
logical interpretation by grouping variables on the basis of just one or two of them.
REFERENCES
Gredelj, M., Metikoš, D., Hošek, A., & Momirović, K. (1975). Model hijerarhijske strukture motorickih 
sposobnosti: Rezultati dobijeni primjenom jednog neoklasičnog postupka za procjenu latentnih di-
menzija [Model of a hierarchic structure of motor abilities: The results obtained using a neo-classical 
method for estimating latent dimensions]. Kineziologija, 5(1-2), 7-81.
Kurelić, N., Momirović, K., Stojanović, M., Šturm, J., Radojević, Ð., & Viskić-Štalec, N. (1975). Struktura 
i razvoj morfoloških i motoričkih dimenzija omladine [Structure and development of morphologic and 
motor dimensions of youth]. Belgrade: Institute for Research of the Faculty of Physical Education.
Madić, D. (1995). Konstrukcija i metrijske karakteristike motoričkih testova specifične gipkosti 
gimnastičarki [Construction and metric characteristics of motor tests of specific flexibility of female 
gymnasts]. Unpublished master’s thesis, Novi Sad: Faculty of Physical Education.
Madić, D. (2000). Povezanost antropoloških dimenzija studenata fizičke kulture sa njihovom uspešnošću 
vežbanja na spravama [Relationship between anthropological dimensions of physical education students 
and successful exercising on gymnastic apparatuses]. Unpublished doctoral thesis, Novi Sad: Faculty 
of Physical Education.
Metikoš, D., Prot, F., Hofman, E., Pintar, Ž., & Oreb, G. (1989). Mjerenje bazičnih motoričkih dimenzija 
sportaša [Measurement of basic motor abilities of athletes]. Zagreb: Faculty of physical education.
Moravec, R. (Ed.). (1996). Eurofit – Physique and motor fitness of the Slovak school youth. Bratislava: 
Slovak Scientific Society for Physical Education and Sports.