Advances in Physical Education
2013. Vol.3, No.4, 190-196
Published Online November 2013 in SciRes (http://www.scirp.org/journal/ape) http://dx.doi.org/10.4236/ape.2013.34031
Open Access 190
Reliability, Validity and Minimal Detectable Change of a New
Multi-Change of Directionagility Test for Soccer Players
Mehdi Ben Brahim1,2, Rym Bougatfa3, Mohamed Amri1
1Laboratory of Physiology, Faculty of Sciences of Tunis, University of Tunis, Tunis, Tunisia
2Faculty of Sciences, Bizerte, Tunisia
3Laboratory Adaptations Cardio-Circulatory, Respiratory, Metabolic and Hormonal to Muscular Exercise,
Faculty of Medicine Ibn El Jazzar, University of Sousse, Sousse, Tunisia
Received June 18th, 2013; revised July 18th, 2013; accepted July 25th, 2013
Copyright © 2013 Mehdi Ben Brahim et al. This is an open access article distributed under the Creative Com-
mons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, pro-
vided the original work is properly cited.
This study examined the test-retest reliability, validity and external responsiveness of a new multi-change
of direction agility test (NMAT) designed for soccer players. Forty-four Tunisian soccer players were re-
cruited and were divided into two groups according to their playing levels (International, n = 21 and Na-
tional, n = 23). Following familiarization, athletes performed squat jump (SJ), countermovement jump
(CMJ), running speed test (5 m and 20 m), 15-m agility run (Agility-15 m), 15-m ball dribbling (Ball-15
m), and NMAT and ball dribbling NMAT (Ball-NMAT) in 2 sessions, 48 h apart. The intraclass correla-
tion coefficient and SEM values were .96 (CI 95%: .94 - .98) and .05 seconds for NMAT and .97 (CI
95%: .94 - .98) and .09 seconds for Ball-NMAT, respectively. The smallest worthwhile changes were
greater than their SEM for both NMAT and Ball-NMAT. The MDC95 values were .15 seconds and .25
seconds for NMAT and Ball-NMAT, respectively. Both NMAT and Ball-NMAT were respectively cor-
related with Agility-15 m (r = .78; p < .001) and Ball-15 m (r = .81; p < .001). Similarly, significant cor-
relations were observed between both NMAT and Ball-NMAT and leg power and straight sprint (.01 < p
< .001). International-level soccer players were better than national-level in all tests including NMAT and
Ball-NMAT (.01 < p < .001). The areas under their receiver operator characteristics curve were > .7 (.85;
CI 95%: .71 - .94 and .91; CI 95%: .78 - .97 for NMAT and Ball-NMAT, respectively). These results in-
dicated that NMAT provides excellent absolute and relative reliabilities. The NMAT can distinguish soc-
cer athletes of different competitive levels. Thus, the NMAT may be suitable for field assessment of spe-
cific agility of soccer players.
Keywords: Agility; Field-Testing; Change of Direction; Soccer; Relative Reliability; Absolute Reliability;
Responsiveness
Introduction
Soccer is the most popular sport in the world, especially
among children. It consists of intermittent high-intensity exer-
cises that involve various types of runs with rapid changes of
directions, starts, stops, jumps and kicks (Reilly, Bangsbo, &
Franks, 2000; Stølen, Chamari, Castagna, & Wisløff, 2005).
Accordingly, players are predisposed to possess well developed
aerobic fitness and anaerobic power, coupled with good agility
(Sheppard & Young, 2006) to be capable of maintaining high
power during fast movements over the entire match (Mohr,
Krustrup, & Bangsbo, 2005). In this type of sport, players are
required to accelerate, decelerate, and change direction
throughhout the game in response to a stimulus, such as an
opposing player’s movements or the movement of the ball
(Sheppard & Young, 2006). Team game players need thus to be
exceptional movers in forward, lateral, back, and multidirec-
tional movements in a considerable reduced area (Bloomfield,
Polman, & O’Donoghue, 2007). Agility, commonly defined as
an individual’s ability to change direction while at speed, has
been deemed an identifiable athletic quality in the development
of individual and/or team success in field and court sports
(Young, James, & Montgomery, 2002; Sheppard & Young,
2006). Field testing is a key component to measure player per-
formance in all sports, which provides coaches and condition-
ing staff information to evaluate player performance and meas-
ure desired training effects. Movement skills are often sport-
and sometimes position-specific (Hasegawa, Dziados, Newton,
Fry, Kraemer, & Hakkinen, 2002), suggesting that test selection
should be related to sport-specific characteristics or posi-
tion-specific movement patterns. A number of tests have been
purported to assess agility including the 505 agility test (Ellis,
Gastin, Lawrence, Savage, Buckeridge, Stapff et al., 2000),
T-test (Haj-Sassi, Dardouri, Haj Yahmed, Gmada, Mahfoudhi,
& Gharbi, 2009), and the Edgren side step test (Harman, Gar-
hammer, & Pandorf, 2000). Other tests commonly are used,
such as Illinois agility test (Hachana, Chaabène, Nabli, Attia,
Moualhi, Farhat et al., 2013). Nevertheless, it is generally ac-
cepted that many current tests used to measure agility perform-
ance within field based team-sports are not matched with
known game-day movement characteristics (Sheppard &

M. B. BRAHIM ET AL.
Young, 2006). In this context, Mujika et al. (2009) proposed a
specific agility test for soccer players which take into account
the specific movement patterns. While the value of these tests is
acknowledged, there are some limitations. In fact, these tests do
not take into account the sideways and backward running and
changes of direction met in soccer game. These appropriate
soccer movement characteristics should be understood when
selecting drills or exercises in athlete’s training process. Central
to the importance of features’ knowledge of the various actions
met during soccer games, we develop a new multi-change of
direction agility test (NMAT). It is generally admitted that any
test should respond to several standard criteria, including reli-
ability, validity and external responsiveness (Hopkins, 2000;
Beekhuizen, Davis, Kolber, & Cheng, 2009) for its scientific
acceptance. Therefore, the aim of this study was to evaluate the
reliability, validity and external responsiveness of the NMAT
designed for soccer players. In addition, we examined the rela-
tionship between the performance of this test and those of the
15-m agility run and vertical jump and speed tests. We hy-
pothesized that the NMAT would provide stable test-retest
scores, and low minimal detectable change. It would have a
stronger relationship with the speed and 15-m agility run per-
formances and all jump ability tests.
Methods
Participants
A total of forty-four male soccer players (21 Internationals
and 23 Nationals) of two different performance levels took part
in this study (mean ± [SD] age: 17.4 ± .6 year; Height: 181.1 ±
4.7 cm; Body mass: 76.5 ± 7.8 kg; Body fat: 13.5% ± 2.1%).
They practise soccer 11 months a year, for at least 7 years (7.7
± 1.4) at a rate of 5 sessions with one competitive game per
week, in addition to their school physical education. In general,
soccer training sessions lasted ~1 h 30 min, including about 15
- 20 min of warming up, low-intensity games and stretching
exercises, 15 - 25 min of technical soccer exercises (kicking
actions, dribbling, jumping, and running with fast accelerations
and decelerations), 20 - 30 min of match practice, and 10 min
of active recovery. None of the participants reported any cur-
rent or on-going neuromuscular diseases or musculoskeletal
injuries specific to the ankle, knee, or hip joints, and none of
them were taking any dietary or performance supplements that
might be expected to affect performance during the study.
Written informed consent was received from all subjects after
verbal and written explanation of the experimental design and
potential risks of the study. The study was conducted according
to the Declaration of Helsinki and the protocol was fully ap-
proved by the local Ethic Committee of the University before
the commencement of the assessments. All the participants
were fully accustomed with the procedures used in this research
and were informed they could withdraw from the study at any
time without penalty.
Procedure
The first phase of this study aimed to establish the relative
and absolute reliability of the NMAT Agility test in a group of
44 soccer players. Each subject completed the NMAT twice
separated by at least 48 hours. All subjects were familiarized
with the NMAT protocol before data collection.
To avoid the effect of diurnal variations, the NMAT was
completed at the same time of the day after a 15-minute warm-
up including jogging, sprinting, lateral displacements, dynamic
stretching, and jumping. The second phase of our investigation
aimed to examine the relationship between the NMAT and
other physical fitness components that were anaerobic perfor-
mance and explosiveness. All subjects performed the speed,
agility, jump tests. For the jump tests, subjects were allowed to
perform 3 trials. All subjects performed each test with at least 3
minutes of rest between all trials and 5 minutes between tests to
ensure adequate recovery. Vertical jump performances (peak
height) were measured by using the Opto-jump system (Micro-
gate SARL, Italy).
Anthropometry
Each subject was weighed and its stature determined. To es-
timate the adiposity, skinfold thickness were measured at four
sites on the left-side of the body (triceps, biceps, subscapular
and suprailiac) using a Harpenden skinfold calliper (British
Indicators Ltd., Luton). All measurements were taken by the
same investigator.
New Multi-Change of Direction Agility Test (NMAT)
In this test, players’ velocity in a 25 m agility run was meas-
ured using the same photocell gates system. The player begins
to make a lateral displacement of 2.5 m linear and then turns
back in traversing the same distance (2.5 m) and arriving at the
starting point where he conducts a running back 2.5 m followed
by a 3-m race before. At this stage, the athlete will begin a race
with change of direction of 1 m chained by a linear stroke of
1.35 m and then crosses a barrier (fence) with a height of .5
meters. The test will end a run of 5 m linear (Figure 1).
Squat Jump
The subject started from a semi-squat position with the hands
Figure 1.
Schematic representation of the NMAT. 1) 2.5 m; 2) 2.5 m; 3) 2.5 m; 4) 3 m; 5) 2 m; 6) 4 m; 7) 2 m; 8) 1.5 m; 9)
5 m.
Open Access 191

M. B. BRAHIM ET AL.
held at the hips to avoid upper limb body contribution and
jumped upward as high as possible. This test was used to esti-
mate muscle power under concentric condition. A successful
trial was one where there was no sinking or countermovement
before the execution of the jump. The intra-class correlation
coefficient (ICC) of the squat jump in our study was .96 (95%
confidence interval [CI]: .91 - .98) with no significant differ-
ences between the 2 trial scores (p = .62, effect size [ES] = .05
[trivial]).
Countermovement J ump
The subject began from an upright standing position, per-
formed a very fast preliminary downward eccentric action fol-
lowed immediately by a jump for maximal height. Hands re-
mained at the hips for the entire movement to eliminate any
influence of arm swing. The ICC of the countermovement jump
(CMJ) in our study was .97 (95% CI: .92 - .99) with no signifi-
cant differences between the 2 trial scores (p = .91, ES = .01
[trivial]).
Running Speed Test
The time needed to cover 5 m (5 mSS) and 20 m speed (20
mSS) was measured with an infrared photoelectronic cell (Cell
Kit Speed Brower, USA). The participants were motivated to
run as fast as they could, and the best (fastest) 5 m and 2 m
sprint time were selected for analysis.
15-m Agility Run (Agility-15 m)
This test was performed according to the protocol previously
described by Mujika et al. (2009). In this test, players started
running 3 m behind the initial set of gates. After 3 m of line
running, players entered a 3-m slalom section marked by three
sticks 1.6 m of height and placed 1.5 m apart, and then cleared
a .5-m-height hurdle placed 2 m beyond the third stick. Players
finally ran 7 m to break the second set of photocell gates, which
stopped the timer. Each player performed two maximal Agil-
ity-15 m interspersed with 3 min of passive recovery, and the
fastest time achieved was recorded.
15-m Ball Dribbling (Ball-15 m)
This test was performed according to the protocol previously
described by Mujika et al. (2009). During this test, players were
required to dribble a ball while performing the test. After the
slalom section of the test, the ball was kicked under the hurdle
while the player cleared it. The player then freely kicked the
ball towards either of two small goals placed diagonally 7 m on
the left and the right sides of the hurdle, and sprinted to the
finish line. Each player performed two maximal Ball-15 m
interspersed with 3 min of passive recovery, and the fastest
time achieved was kept for analysis.
Statistical Analyses
Data are shown as mean ± SD. Normality was analysed using
the Shapiro-Wilktest. All variables presented a normal distribu-
tion. To investigate systematic bias, a paired Student’s t-test
was conducted to test hypothesis of no difference between the
sample mean score for the test versus the sample mean score
for the retest. Estimates of effect size were calculated to assess
meaningfulness of differences. Effect sizes of 1.2, between 1.2
and .6, between .6 and .2, and .2 have been considered as large,
moderate, small, and trivial, respectively (Hopkins, 2005). The
ICC was used to examine the relative reliability of both NMA-
Tand Ball-NMAT. The SEM and 95% limit of agreement (LOA)
method (Bland & Altman, 1995) were calculated as an indica-
tion of the absolute reliability of both NMAT and Ball-NMAT.
Toestablish the usefulness of the NMAT and Ball-NMAT, the
SWC was determined (Atkinson & Nevill, 1998). The sensitiv-
ity of the test was assessed by comparing the SWC and SEM,
using the thresholds (Liow & Hopkins, 2003). If the SEM is
smaller than the SWC, the ability of the test to detect a change
is “good”; if the SEM equals SWC, then the test is “satisfac-
tory”, but if the SEM is greater than the SWC, then the test is
rated as “marginal”. Knowledge of the SEM allows the calcula-
tion of the MDC95. The MDC reflects the 95% CI of the differ-
ence in scorebetween paired observations, calculated as MDC95
= SEM ×2 × 1.96 (Beckerman, Roebroeck, Ankhorst, Be-
cher, Bezemer, & Verbek, 2001).
Heteroscedasticity was assessed using a zero-order correla-
tion coefficient between the means of the subject’s test and
retest scores and the absolute differences between the subject’s
test and retest scores. Comparison of anthropometric variables
and test performances between international and national play-
ers was assessed using in dependent t-test. The receiver opera-
tor characteristics (ROC) curve with area under the curve (AUC)
was used to evaluate external responsiveness of both NMAT
and Ball-NMAT according to performance levels. The area
under the ROC curve was interpreted as the probability of cor-
rectly discriminating between soccer athletes with a good and a
poor outcome. The area of .5 is interpreted as no discriminatory
accuracy and 1. as complete accuracy (De Vet, Bouter, Beze-
mer, & Beurskens, 2001).
Results
Both NMAT and ball-NMAT were observed to have accept-
able relative and absolute reliability. Residual data for both
NMAT and Ball-NMAT test and retest trials comparison were
normally distributed (p = .36 and p = .42, respectively). Mean
scores (SD) of both NMAT and Ball-NMAT, ICC, SEM and
MDC95 values between test and retest are given in Tables 1 and
2. Both NMAT and Ball-NMAT showed a high degree of rela-
tive reliability between the test-retest sessions. ICC values
were .96 (95% CI, .94 - .98) for NMAT and .97 (95% CI, .94
- .98) for Ball-NMAT (Table 1).
The heteroscedasticity coefficients for both NMAT and
Ball-NMAT were not significant (r = −.05 [95% CI, −.34 to .25;
p = .74] and r = .15 [95% CI, −.15 to .43; p = .33], respectively).
The mean difference (bias) ± the 95% LOAs were .02 ± .16
seconds and −.01 ± .26 seconds for NMAT and Ball-NMAT
respectively. Heteroscedasticity diminished for both NMAT
and Ball-NMAT when test and retest data were Log trans-
formed (r = −.02 [95% CI, −.31 to .28; p = .91] and r = .06
[95% CI, −.24 to .35; p = .69], respectively). The mean differ-
ence (bias) ± the 95% LOAs of the Log transformed data were
of .0007 ± .0076 seconds for NMAT and of −.0003 ± .0091
seconds for Ball-NMAT (Table 2). Taking antilog of these
LOAs gave a mean bias of 1.001 with an agreement component
of ×/÷ 1.008 for NMAT and a mean bias of 1.000 with an
agreement component of ×/÷ 1.009 for Ball-NMAT. Thus, the
95% of the ratios for the log transformed test score divided by
Open Access 192

M. B. BRAHIM ET AL.
log transformed retest score should be contained between .993
(1.001 ÷ 1.008) and 1.009 (1.001 × 1.0086) for NMAT and
between .991 (1.000 ÷ 1.0091) and 1.009 (1.000 × 1.0091) for
Ball-NMAT.
Descriptive data from testing performance is listed in Table
3, including NMAT, Ball-NMAT, 20 m straight sprint, JS and
CMJ performed separately by international and national level
groups. Independent sample t-test revealed that international
level soccer players had significantly better performance in
whole tests, especially NMAT and Ball-NMAT (.01 < p < .001;
Table 3). A receiver operator characteristic (ROC) curve analy-
sis was calculated between international and national level soc-
cer players. The NMAT was considered having very good dis-
criminant ability. The areas under the ROC curve were .85
(95% CI, .71 to .94; p < .001) for NMAT and .92 (95% CI, .79
to .98; p < .001) for Ball-NMAT (Figure 2).
Table 4 showed that both NMAT and Ball-NMAT were re-
spectively correlated with Agility-15 m and Ball-15 m (r = .78
and r = .81; all p < .001, respectively). In addition, both tests
were significantly correlated with maximal speed and lower
limb power (r = .51 to r = .6, .01 < p < .001).
Discussion
The purpose of this study was to assess the reliability, valid-
ity and sensitivity of the NMAT with and without ball as well
as to examine the relationship between this test and Agility-15
m and both sprint and vertical jump. The main design idea of
the NMAT was to consider the forward, lateral, back, and mul-
tidirectional movement patterns encountered during soccer
games. In the present study, we found a high reliability of the
NMAT with and without ball was high across the two meas-
urement trials. Several recent studies usually investigated the
reliability of field tests by using 2 commons indices including
the ICC values and 95% LOA method (Haj-Sassi, Dardouri,
Haj Yahmed, Gmada, Mahfoudhi, & Gharbi, 2009; Haj Sassi,
Dardouri, Gharbi, Chaouachi, Mansour, Rabhi et al., 2011;
Hachana, Chaabène, Nabli, Attia, Moualhi, Farhat et al., 2013).
These studies considered the two methods as the most appro-
priate and objective for reliability assessment. ICCs across the
two trials in our study were .96 and .97 for NMAT and Ball-
NMAT, respectively. These values were in the same range of
relative reliability value indices reported in other agility tests
(Haj-Sassi, Dardouri, Haj Yahmed, Gmada, Mahfoudhi, &
Gharbi, 2009; Haj Sassi, Dardouri, Gharbi, Chaouachi, Man-
sour, Rabhi et al., 2011; Hachana, Chaabène, Nabli, Attia,
Moualhi, Farhat et al., 2013). Haj-Sassi et al. (2009) reported
an ICC of .92 to .96 across two modified agility T-test trials in
men and women physical education students. In the same con-
text, Hachana et al. (2013) found an ICC of .96 across two Illi-
nois agility run trials in 89team sports players. It is commonly
accepted that an ICC over .9 is considered high for physiologi-
cal field tests (Vincent, 1995); so, our results demonstrated a
high reliability of both NAMT and Ball-NMAT.
Table 1.
Performance characteristics and results of relative reliability of both NMAT and Ball-NMAT.
Test (s) Retest (s) Mean difference ICC (95% CI) dz
NMAT (s) 9.56 ± .32 9.54 ± .29 .01 ± .08 .96 (.94 - .98) .06
Ball-NMAT (s) 11.92 ± .5 11.93 ± .51 .01 ± .13 .97 (.94 - .98) .02
Note: ICC = intraclass correlation coefficient; dz = Cohen’s d for the paired sample t-test.
Table 2.
Performance characteristics, minimal detectable change and results of absolute reliability of both NMAT and Ball-NMAT.
Bias 95% LOA Ratio LOA SEM seconds SWC seconds MDC seconds
NMAT .02 .16 .0007 ± .0076 .05 .06 .15
Ball-NMAT −.01 .25 −.0003 ± .0091 .09 .1 .25
Note: LOA: limits of agreement; SEM: standard error of measurement; SWC: smallest worth-while change; MDC: minimal detectable change.
Table 3.
Results of physical testing.
International (n = 21) National (n = 23) Combined (n = 44)
5 mSS (second) 1.05 ± .1 1.09 ± .12** 1.07 ± .09
20 mSS (second) 2.85 ± .09 3.1 ± .17*** 2.98 ± 0.9
Agility-15 m (second) 3.69 ± .16 3.8 ± .31*** 3.75 ± .15
Ball-15 m (second) 4.64 ± .3 4.78 ± .36*** 4.74 ± .28
NMAT (second) 9.33 ± .27 9.7 ± .22*** 9.52 ± .31
Ball-NMAT (second) 11.48 ± .42 12.24 ± .42*** 11.88 ± .5
SJ (cm) 39.83 ± 1.75 30.07 ± 4.78*** 34.84 ± 2.22
CMJ (cm) 47.06 ± 2.12 34.46 ± 3.61** 40.29 ± 2.71
Open Access 193

M. B. BRAHIM ET AL.
Figure 2.
Receiver operating characteristics curve for both NMAT and ball-NMAT. NMAT:
Sensitivity 91%, specificity 67% and criterion was >9.36 s; Ball-NMAT: Sensitiv-
ity 96%, specificity 76% and criterion was >11.91 s.
Table 4.
Relationship between both NMAT and Ball-NMAT with other tests’
performance.
NMAT Ball-NMAT
r p r p
V20 .51 <.01 .54 <.01
Agility-15 m .78 <.001
Ball-15 m .81 <.001
SJ .6 <.01 .55 <.01
CMJ .6 <.01 .58 <.01
The relative reliability of both NMAT and Ball-NMAT has
been also confirmed in our study by the 95% LOAs. In our
study, the bias ± the 95% LOAs of the NMAT and Ball-NMAT
were .02 ± .16 seconds and −.01 ± .26 seconds, respectively.
The antilog of these LOAs could be expressed as the mean bias
of 1.001 ×/÷ 1.0081 for NMAT and of 1.000 ×/÷ 1.009 for
Ball-NMAT. Thus, the 95% of the ratios for the log trans-
formed test score divided by log transformed retest score
should be contained between .993 (1.001 ÷ 1.008) and 1.009
(1.001 × 1.0086) for NMAT and between .991 (1.000 ÷ 1.0091)
and 1.009 (1.000 × 1.0091) for Ball-NMAT. As practical con-
siderations, when an athlete from the experimental group per-
formed respectively 9.5 seconds and 11.9 seconds on the
NMAT and the Ball-NMAT, on the retest he could perform a
score as high as 9.5 × 1.009 = 9.59 seconds, or as low as 9.5
× .993 = 9.43 seconds for NMAT and a score as high as 11.9 ×
1.009 = 12.01 seconds, or as low as 11.9 × .991 = 11.79 sec-
onds (Cooper, Baker, Tong, Roberts, & Hanford, 2005). Ac-
cording to Atkinson and Nevill (1998), it is important to use the
MDC95 as a criterion todetermine whether a real change has
occurred between testand retest. In this study, the MDC95 for
the NMAT and the Ball-NMAT have been .15 and .25 seconds,
respectively. When changes in both NMAT and Ball-NMAT
test-retest score are ±.15 and ±.25 seconds, true changes can be
associated. Thus, in agreement with recent studies of Haj-Sassi
et al. (2009) and Hachana et al. (2013), we could consider these
LOAs acceptable.
One of the most aims of the NMAT is to select athletes. To
that end, it should be able to discriminate athletes of different
competitive and fitness level. This kind of validity is commonly
established by testing differences between groups of subjects of
different competitive level (Impellizzeri & Marcora, 2009). In
our study, a significant difference has been found between both
NMAT and Ball-NMAT performances of international and
national groups. In clinimetrics, alternative methods such as the
receiver operator characteristics (ROC) curve are gaining
popularity and can be used to validate the discriminant ability
and the responsiveness of physiological and performance tests
(Mannion, Elfering, Staerkle, Junge, Grob, Semmer et al.,
2005). Thereby, a performance test was considered responsive
if its area under the ROC curve was ≥.7 (Mannion, Elfering,
Staerkle, Junge, Grob, Semmer et al., 2005). Indeed, the area
under the ROC curve represents the probability of correctly
discriminating international from national soccer players using
both NMAT and Ball-NMAT. Accordingly, in the present
study we found that the areas under the ROC curve were >.7
(.85 [95% CI, .71 to .94] and .92 [95% CI, .79 to .98] for
NMAT and Ball-NMAT, respectively). The test scores able to
differentiate between international and national soccer players
are >9.36 seconds for NMAT and >11.91 seconds for Ball-
NMAT. This cut-off value gives true positive rates (sensitivity)
of 91 and 96% and false positive rates (1-specificity) of 67 and
76% for NMAT and Ball-NMAT, respectively. Therefore, this
kind of statistical tool suggest that both NMAT and Ball-
NMAT have excellent discriminant ability if its purpose is to
differentiate between international and national soccer players.
Pearson product-moment correlations have been also calcu-
lated between the NMAT and Agility-15 m, vertical jump and
straight sprint tests (Table 4). The coefficient of determinant
(R2) highlighted that NMAT and Agility-15 m as well as
Ball-NMAT and Ball-15 m share 61% and 66% common vari-
ance, respectively. These results indicate that the NMAT could
be used to evaluate change of direction speed, and thus soc-
cer-specific agility. Accordingly, compared with other testing
methods, the NMAT could be considered as the most appropri-
ate test for assessing specific-agility in soccer. Since, this test
Open Access 194

M. B. BRAHIM ET AL.
takes into consideration the soccer-specific characteristics and
movement patterns. In addition, both NMAT and Ball-NMAT
were correlated to SJ, CMJ and 20 mSS. Our results were
comparable with other studies that examined the relationship
between agility tests (change of direction speed tests) and ver-
tical jump and straight sprint tests (Haj-Sassi, Dardouri, Haj
Yahmed, Gmada, Mahfoudhi, & Gharbi, 2009; Hachana,
Chaabène, Nabli, Attia, Moualhi, Farhat et al., 2013). Haj Sassi
et al. (2009) have shown that a modified agility T-test was only
correlated to free CMJ and 10-m straight sprint in female but
not in male athletes. On the other hand, Young et al. (1996)
have revealed a low and non-significant correlation between the
CMJ test and 20-m change-of-direction test. Similar results
have been reported by Webb and Lander (1983). They have
reported a low and non-significant correlation between the “L”
change-of-direction run test and vertical jump. Recently,
Hachana et al. (2013) have shown significant correlation be-
tween Illinois change-of-direction agility run and agility T-test
and between Illinois change-of-direction agility run and CMJ
and acceleration and speed. In this context, Thomas and Nelson
(2001) stated, “When common variance between the two vari-
ables is less than 50%, it indicates that they are specific or
somewhat independent in nature”. Based on these results, it
seems that change of direction speed and straight sprint were
two specific determinant qualities on performance. Given the
complexity of agility, it appears that the complex control motor
and coordination of several muscle groups could contribute
considerably to the change of direction speed performance
(Young, Hawken, & McDonald, 1996).
Conclusion
In conclusion, the NMAT can be considered as a sport-spe-
cific field test designated to evaluate change of direction and
agility performances in soccer athletes. This test provided good
absolute and relative reliability and successfully discriminated
karate athletes by competitive level. Its performance is signifi-
cantly related to speed and lower limbs power. Considering that
reliability and discriminant ability of a test are two important
aspects, the NMAT can be used to monitor soccer training pro-
grams, especially those directed to improve change of direction
and agility of soccer athletes.
REFERENCES
Atkinson, G., & Nevill, A. (1998). Statistical methods for assessing
measurement error (reliability) in variables relevant to sports medi-
cine. Journal of Sports Medicine, 26, 217-238.
http://dx.doi.org/10.2165/00007256-199826040-00002
Beckerman, H., Roebroeck, M. E., Ankhorst, G. J., Becher, J. G, Beze-
mer, P. D., & Verbek, A. L. (2001). Smallest real difference, a link
between reproducibility and responsiveness. Journal of Quality of
Life Research, 1, 571-578.
http://dx.doi.org/10.1023/A:1013138911638
Beekhuizen, K. S., Davis, M. D., Kolber, M. J., & Cheng, M. S. (2009).
Test-retest reliability and minimal detectable change of the hexagon
agility test. Journal Strength Conditioning Research, 23, 2167-2171.
http://dx.doi.org/10.1519/JSC.0b013e3181b439f0
Bland, J. M., & Altman, D. G. (1995). Comparing two methods of
clinical measurements: A personal history. International Journal of
Epidemiology, 24, 7-17.
http://dx.doi.org/10.1093/ije/24.Supplement_1.S7
Brughelli, M., Cronin, J., Levin, G., & Chaouachi, A. (2008). Under-
standing change of directionability in sport: A review of resistance
training studies. Sports Medicine, 38, 1045-1063.
http://dx.doi.org/10.2165/00007256-200838120-00007
Cooper, S. M., Baker, J. S., Tong, R. J., Roberts, E., & Hanford, M.
(2005). The repeatability and criterion related validity of the 20 m
multistage fitness test as a predictor of maximal oxygen uptake in ac-
tive young men. British Journal of Sports Medicin e, 39, 19-26.
http://dx.doi.org/10.1136/bjsm.2004.013078
Dawson, B., Hopkinson, R., Appleby, B., Stewart, G., & Roberts, C.
(2004). Player movement patterns and game activities in the Austra-
lian football league. Journal of Science and Medicine in Sport, 7,
278-291. http://dx.doi.org/10.1016/S1440-2440(04)80023-9
De Vet, H. C., Bouter, L. M., Bezemer, P. D., & Beurskens, A. J.
(2001). Reproducibility and responsiveness of evaluative outcome
measures. Theoretical considerations illustrated by an empirical ex-
ample. International Journal of Technology Assessment in Health
Care, 17, 479-487.
Ellis, L., Gastin, P., Lawrence, S., Savage, B., Buckeridge, A., Stapff,
A., Tumilty, D., Quinn, A., Woolford, S., & Young, W. (2000). Pro-
tocols for the physiological assessment of team sports players. In C. J.
Gore (Ed.), Physiological tests for elite athletes (pp. 128-144).
Champaign: Human Kinetics.
Gabbett, T., & Benton, D. (2009). Reactive agility of rugby league
players. Journal of Scie n ce an d Medicine inSport, 12, 212-214.
Gabbett, T. J. (2009). Physiological and anthropometric correlates of
tackling ability in rugbyleague players. Journal of Strength and Con-
ditioning Research, 2 3 , 540-548.
http://dx.doi.org/10.1519/JSC.0b013e31818efe8b
Hachana Y., Chaabène, H., Nabli, M. A., Attia, A., Moualhi, J., Farhat,
N., & Elloumi, M. (2013). Test-retest reliability, criterion related va-
lidity and minimal detectable change of the Illinois agility test in
male team sport athletes. Journal of Strength and Conditioning Re-
search, in press. http://dx.doi.org/10.1519/JSC.0b013e3182890ac3
Haj Sassi, R., Dardouri, W., Gharbi, Z., Chaouachi, A., Mansour, H.,
Rabhi, A., & Haj Yahmed, M. (2011). Reliability and validity of a
newrepeated agility test as a measure of anaerobic and explosive
power. Journal of Strength and C o nditioning Research, 2 5, 472-480.
http://dx.doi.org/10.1519/JSC.0b013e3182018186
Haj-Sassi, R., Dardouri, W., Haj Yahmed, M., Gmada, N., Mahfoudhi,
M. E., & Gharbi, Z. (2009). Relative and absolute reliability of a
modified agility T-test and its relationship withvertical jump and
straight sprint. Journal of Strength and Conditioning Research, 23,
1644-1651. http://dx.doi.org/10.1519/JSC.0b013e3181b425d2
Harman, E., Garhammer, J., & Pandorf, C. (2000). Administration,
scoring, and interpretation of selected tests. In T. R. Baechle, & R. W.
Earle (Eds.), Essentials of strength training and conditioning (pp.
287-317). Champaign: Human Kinetics.
Hasegawa, H., Dziados, J., Newton, R. U., Fry, A. C., Kraemer, W. J.,
& Hakkinen, K. (2002) Periodized training programs for athletes. In
W. J. H. Kraemer (Ed.), Strength training for sport (pp. 69-134).
Ames: Blackwell Science.
Hopkins, W. G. (2000). Measures of reliability in sports medicine and
science. Sports Medicine, 3, 1-15.
http://dx.doi.org/10.2165/00007256-200030010-00001
Hopkins, W. G. (2005) A new view on statistics: Log transformation.
http://sportsci.org/resource/stats/index.html/Log
Impellizzeri, F. M., & Marcora, S. M. (2009). Test validation in sport
physiology: Lessons learned from clinimetrics. International Journal
of Sports Physiology and Performance, 4, 269-277.
Jones, P., Bampouras, T. M., & Marrin, K. (2009). An investigation
into the physical determinants of change of direction speed. Journal
of Sports Medicine and Physical Fitness, 49, 97-104.
Liow, D. K., & Hopkins, W. G. (2003). Velocity specificity of weight
training for kayak sprint performance. Medicine and Sciences in
Sports Exercise, 35, 1232-1237.
http://dx.doi.org/10.1249/01.MSS.0000074450.97188.CF
Mannion, A. F., Elfering, A., Staerkle, R., Junge, A., Grob, D., Semmer,
N. K., Jacobshagen, N., Dvorak, J., & Boos, N. (2005). Outcome as-
sessment in low back pain: how low can you go? European Spine
Journal, 14, 1014-126. http://dx.doi.org/10.1007/s00586-005-0911-9
Marcovic, G. (2007). Poor relationship between strength and power
qualities and agility performance. Journal of Sports Medicine and
Open Access 195

M. B. BRAHIM ET AL.
Open Access 196
Physical Fitness, 47, 276-283.
Mohr, M., Krustrup, P., & Bangsbo, J. (2005). Fatigue in soccer: A
brief review. Journa l o f S p o r t s S c i e nces, 23, 593-599.
http://dx.doi.org/10.1080/02640410400021286
Reilly, T., Williams, A. M., Nevill, A., & Franks, A. (2000). A multid-
isciplinary approach to talent identification in soccer. Journal of
Sports Sciences, 18, 695-702.
http://dx.doi.org/10.1080/02640410050120078
Reilly, T., Bangsbo, J., & Franks, A. (2000). Anthropometric and
physiological predispositions for elite soccer. Journal of Sports Sci-
ences, 18, 669-683. http://dx.doi.org/10.1080/02640410050120050
Sheppard, J. M., Young, W. B., Doyle, T. L. A., Sheppard, T. A., &
Newton, R. U. (2006). An evaluation of a new test of reactive agility
and its relationship to sprint speed and change of direction speed.
Journal of Science and M edicine in Sport, 9, 342-349.
http://dx.doi.org/10.1016/j.jsams.2006.05.019
Sheppard, J. M., & Young, W. B. (2006). Agility literature review:
Classifications, training and testing. Journal of Sports Sciences, 24,
919-932. http://dx.doi.org/10.1080/02640410500457109
Stølen, T., Chamari, K., Castagna, C., & Wisløff, U. (2005). Physiol-
ogy of soccer: An update. Journal of Sports Medicine, 35, 501-536.
http://dx.doi.org/10.2165/00007256-200535060-00004
Thomas, J. R., & Nelson, J. K. (2001). Research methods in physical
activity (4th ed.). Champaign, IL: Human Kinetics.
Vaeyens, R., Lenoir, M., & Williams, A. M. (2007). Mechanisms un-
derpinning successful decision making in skilled youth soccer play-
ers: An analysis of visual search behaviors. Journal of Motor Behav-
ior, 39, 395-408. http://dx.doi.org/10.3200/JMBR.39.5.395-408
Vincent, W. J. (1995). Statistics in Kinesiology. Champaign, IL: Hu-
man Kinetics.
Webb, P., & Lander, J. (1983). An economical fitness testing battery
for high school and college rugby teams. Sports Coach, 7, 44-46.
Williams, A. M., Davids, K., Burwitz, L., & Williams, J. G. (1994).
Visual search strategies in experienced and inexperienced soccer
players. Research Quarterly for Exercise and Sport, 65, 127-135.
http://dx.doi.org/10.1080/02701367.1994.10607607
Williams, A. M. (2000). Perceptual skill in soccer: Implications for
talent identification and development. Journal of Sports Sciences, 18,
737-705. http://dx.doi.org/10.1080/02640410050120113
Young, W. B., James, B., & Montgomery, I. (2002). Is muscle power
related to running speed with changes of direction. Journal of Sports
Medicine and Physical Fitness, 42, 282-288.
Young, W. B., Hawken, M., & McDonald, L. (1996). Relationship
between speed, agility, and strength qualities in Australian Rules
football. Streng th Conditi o ning C oach, 4, 3-6.