MILITARY MEDICINE, 171, 8:749, 2006
Navy Recruits: Fitness Measuring, Validation, and Norming
Guarantor; Damir Sekulic, PhD
Contributors: Damir Sekulic, PhD; Capt Boris Males, Navy (Ret.); Djurdjica Miletic, PhD
During the standard Army Physical Fitness Test (APFT) used
on Croatian Navy recruits (CNR), we found the standard APFT
battery to be both excessively demanding and time-consuming. The aim of the present study was to validate alternate
APFT tests (A-APFT) that would be appropriate for CNR. In the
sample of 295 CNR (mean age, 23 ± 2.6 years), we applied the
standard APFT (2-minute sit-ups and push-ups, and 2-miienin) and A-APFT (30-second sit-ups and push-ups, 1,500meter run, and 2-minute squats). Using linear correlations and
linear regression equations, we established validity of the AAPFT compared to the APFT. Thirty-second push-ups, but not
30-second sit-ups, can be considered as a valuable alternative
to the same 2-minute test, while the 1500-meter run may be
considered as a valid replacement of the 2-mile run. The modified standards for the A-APFT and some suggestions for the
enhancement of the physical fitness training in CNR are presented.
Introduction
itness in the military is a time-honored and unquestioned
F
axiom. In addition to combat duties and the typical set of
tasks that are routinely performed, the standard of fitness significantly influences both general well-being and health status.''^ The Army Physical Fitness Test (APFT) is a three-event
physical performance test widely used by NATO members to
assess muscular endurance and cardiorespiratory fitness. It is a
simple way to measure a soldier's abiliiy to move his/her body
by using a few major muscle groups and the cardiorespiratory
system. The APFT is generally performed semiarmually and consists of 2 minutes of push-ups, 2 minutes of sit-ups, and a
2-mile run.^ Eaeh part of the test is given a maximum score of
100 points for a potential perfect score of 300. The minimum
passing grade on the APFT is 60 points in eaeh event, for a total
of 180. An extended scale is provided for those who excel in
physical fitness. The Croatian Army (CA), although not a NATO
member, uses NATO standards for APFT. The CA comprises
both full-time professionals and recruits who have enlisted for a
mandatory 8-month period of military service.'' Military training
for recruits in CA is performed in a three-phase program: basicelementary military training (2 months), specialist training (2
months), and flnal training (4 months).^ At the beginning and at
the end of each period, APFT is performed (a total of six times
during the recruit's 8-month military service). However, it became obvious that the current APFT is not adequate for the
Croatian Navy recruits (CNR) mainly because the tests were
both too demanding and time-consuming. Furthermore, a problem arose with the aerobic-capacity measuring procedure (2mile run). Among the CNR, the average maximum time for a
Department of Kinesiology. Faculty of Natural Sciences. Mathematics and Education, University of Split. N. Tesla 12. 21000 Split. Croatia.
This manuscript was received for review in April 2005. The revised manuscript was
accepted for publication in November 2005.
Reprint & Copyright © by Association of Militaiy Surgeons of U.S., 2006.
group of 10 soldiers to complete the 2-mile run exceeded 25
minutes (see Table I), making the test, to say the least, inappropriate. Also, this particular testing procedure is not practical in
some military units (e.g., in the Navy during service on small
boats and for embassy military staif, etc.). APFT-suggested 800yard swimming is inappropriate as a possible substitute because a limited number of swimming pools exist and military
units are typically sited in several locations. The same problems
are encountered with the professional noncombat CA miiitary
personnel. They regularly perform a physical training program
and APFT primarily for the prevention of possible health-related
problems^ but also to facilitate a higher level of working efficiency. However, we are of opinion that the previously explained
problems (tests too demanding and time-consuming) remain.
The aims of the study were: (1) to construct and validate
alternative test procedures in exchange for those specifically
inadequate in CNR and (2) to suggest a scoring standard for the
newly constructed and metrically acceptable test procedures in
CNR.
Construction and validation of the new military tests are not
unusual in military establishments''^ but as far as we know,
there are no published replacement tests in the recent literature
that would solve the previously described problems.
Subjects and Variables
CNR (N = 295, all males, mean age, 23 ± 2.6 years) served as
the sample of subjects. All subjects were in good health and
preceding the experiment gave tJieir informed consent, Testing
was performed after basic 2-month military training, where significant improvement in the physical fitness status of the CNR
was already demonstrated.^
The sample of variables consisted of two sets. The first set
eomprised standard APFT procedures: maximum number of the
push-ups performed in a 2-minute interval (PUa minuieJ- niaximum number of the sit-ups performed in a 2-minute interval
(SUj ^^^J, and a 2-mile run (2M). The second set consisted of
potentid alternative tests: push-ups in a 30-second interval
(PUgosec)- sit-ups in a 30-second inten'al (SUgosed. squats in a
2-minute interval (SQUATg mmutJ- and a 1,500-meter run
(1500m).
Methods
Each examinee was tested on the standard APFT. The test
sequence was as follows: PUa ^^^^,,. SU2 ^^,tes. and 2M. During
the following week over a 2-day period, the subjects performed
the alternative test sequence: PUsosec- SUgosec and 1,500m on
day 1 and the SQUAT^ ,,^^,^ on day 2.
We calculated descriptive statistics for all variables. To establish relationships between standard APFT and the replacement
tests and validity of the replacement tests, we applied (1) linear
749
Military Medicine. Vol. 171, August 2006
Fitness Measuring. Validation, and Norming in Navy Recruits
750
correlation and (2) linear regression analysis.^-^ All coefflcients
were considered significant at p < 0.05. Statsoffs Statistica
(version 6.0. Tulsa. Oklahoma) was used for all statistical procedures and calculations.
Results
As can be seen from the data presented in Table 1, the "average" CNR undeniably fails on standard APFT procedure (sum of
three events is 138). which supports the contention that the
scoring standards currently used are inadequate for CNR. For a
simple comparison. Knapik et al."* recently presented data
where subjects (2.303 men; 20.4 ± 3.3 years: different military
professionals) achieved on average 20.4 repetitions (reps)
(PUj minuies- respectively lower comparing to average result presented herein), 62.5 reps (SUj minuies: respectively higher), and
14.9 minutes (2M: respectively better). Still, low APFT scores
presented in Table I do not surprise us because; (1) we studied
draftees-not professional soldiers, and (2) recruit military service in the Republic of Croatia is compulsory for all healthy
males, thus explaining the poor "fitness-income" on average,
but also points at objectively high fitness standards defined by
APFT.
nificant (r = 0.36). The correlation coefflcient between SQUAT
and 2M is negative (r = -0.32), because of the opposite metrics
of these two variables. The real nature and applicability of discussed correlation coefficients should be explained more precisely when we have first studied the regression results (Figs.
1-4)
In Figure 1. the relationship between the standard and alternate PU tests is presented. Alternate test [PUaoseJ explains a
significant proportion (72%) of the standard APFT test (PUa minutes) variance. The regression equation presented in Figure 1
allows one to calculate the result on the standard APFT test
(PUj minuteJ using the result on the alternate test (PUsoseJ- Therefore, in this moment, we can conclude that the standard APFT
procedure regularly performed for the purpose of establishing
the muscular endurance of the chest, shoulder, and triceps
muscles (PUj minuiJ can be validiy replaced by the suggested
shortened test procedure (PUsoseJ in CNR.
In Figure 2. the relation between SU30sec ^n^l SU2 minutes is
presented. Since only 13% ofthe standard APFT test variance is
explained using the shortened test version (SUgosec). it is evident
00
PU2MIN =-1,776+1,2568 XPU30SEC
R = 0.851
EXPL.VAR = 72%
70
Discussion
In Table 11. linear correlation coefflcients, calculated between
standard and alternative test batteries, are presented. As expected, all coefflcients are statistically significant, given the relatively large number of subjects (N = 295) and therefore the
large number of degrees of freedom [df= N - 1). Meanwhile,
only two coefficients reached an acceptable numerical value,
explaining the common variance of the standard and alternate
PU tests (r = 0.85) and standard and alternative aerobic-endurance tests (r = 0.52). The relationship between standard and
alternative SU variables is numerically negligible, although sig-
;O 0 0
^'i.o
CM
9-
30
i
«
0.
20
L2
0
5
10
TABLEI
SUaniin (reps)
2M (min:sec)
PUao^ (reps)
SUao^e (reps)
SQUATj^^ (reps)
1500 m (min:sec)
15
20
25
30
95K confidence
35
40
|-
45
50
PUSH-UP 30 SEC (repetitions)
DESCRIFTIVE STATISTICS"
PUs^in (reps)
i,^'^:
8
Fig. 1. Linear regression scatterplot-prediction ofthe PUSH-UP^ __^ using the
N
M
SD
MS
295
295
295
295
295
295
295
30.48
37.39
17:39
25.64
18.10
65.07
07:13
11.56
10.73
2:01
7.78
3.43
18.66
00:54
50
51
48
SU2MIN = 17.113 + 1.1200XSU30SEC
R = 0,358
EXPL.VAR = 13%
60
- -0
6
. 8
3 0
CORRELATION MATRIX-PEARSON"S COEFFICIENTS
0
0 ,w
•
0-
- ' •
0
8
6
O
^
•
0 6 o
o ' "'
o
0,-
0
0
0
r
. .0
\ \
• • _ ^ - ^ ^ -
D
J^ 20
8 o o- 0
0
8
o
e
a'
1-h 1=
- ^ o- fi
18
8
" N. Number of subjeets: M. mean test result: MS. mean score
according to the current APFT standards.
TABLED
y
0
o
I o- o-
Q
0
O 0
10
12
0
14
16
\° 0
>
IS
»
1 txB5% conftdenca
22
24
26
|
28
30
SIT-UP 30SEC (repetitions)
Fig. 2. Linear regression scatteiplot-prediction of the SUj ,^m^ using the
''Values a r e p < 0.05.
Military Medicine, Vol. 171. August 2006
Fitness Measuring, Validation, and Norming in Navy Recruits
751
(secor
1700
that the proposed alternative (SUaos^p) cannot be used as a valid
replacement for the standard APFT procedure measuring the
2M1LE =1190,9-2,042 x SQUAT
leoo
muscular endurance of the abdominal and hip-fiexor muscles
R =-0.3172
1500
(SUj ^ ^ in CNR. We are ofthe opinion that the problem is in the
EXPL.VAR = 9%
alternative test duration. Briefly, the short duration (30 sec1400
onds) of the proposed alternative test defines poor metrics (poor
test sensitivity mainly). There is a certain possibility that some
subjects (especially those not familiar with the test) under- or
1200
overestimated their capabilities and did not reach the reliable
result on the SUgosec- which led to low correlation and highestimation error.
1000
The relationship between the standard (2-mile run) and ap900
plied alternative aerobic-endurance tests (1500m) is presented
in Figure 3. The studied replacement test (1500m) explains the
SOO
standard test variance at a significant level (27%) and therefore
700
should be used as a valid replacement for a standard APFT
10
20
30
40
50
60
70
80
90
100
110
120
aerobic-endurance test (2M).
2-MIN-SQUAT (repetitions)
The importance of aerobic endurance in the militaiy is unFig. 4. Linear regression scatterplot-prediction of the 2M using the SQUAT,
questionable."'^ However, as said in the "Introduction," one of
the APFT problems we noticed in the Croatian Navy is the aerobic-endurance measurement In the naval units during service ance using the SQUAT2 minuiM result defines a large prediction
on the small boats. It is not rare for Navy officers and cadets to error and thus should be avoided. Therefore, we cannot estabnot leave vessels for a 6-month period. During this particular lish considerable validity of the SQUAT2 n^u,es test for the purservice, there is no chance to perform the standard APFT pro- pose of the aerobic-endurance measurement. Inoue and Nacedure (2M) or a suggested alternative (1500m). Therefore, we kao'^ found very similar relationships between different squat
tried to establish an additional alternate test procedure tests and aerobic-endurance variables (from 4.7% to 8.1% vari(SQUATj niinuieJ- which can be performed in the limited space, ance explained; significant Pearson's correlation). They conknowing the physiological and biomechanical basis of the aer- eluded that VOjmax can be estimated using the squat tests (1 obic-endurance test procedures.'^'*" Although there are known and 1.5-minute variants). However, we do not support their
alternate aerobic-endurance test procedures (like different opinion. In short, significant correlation is important—but it is
"step-tests" and/or APFT suggested 800-yard swimming), the not the only element which should be observed in the study of
main reason we chose and examined SQUAT; tmmiKs was the test validity, mainly because it is highly dependent on degrees of
possibility of (1) testing in the restricted space (not characteris- freedom (explained previously in "Results"). For example, in our
tic for 800-yard swimming) and (2) testing the group of subjects study, the correlation coefficient between PUgoser and 2M is staat the same time (not characteristic for the step-test proce- tistically significant too (r = 0.30; 9% ofthe variance explained),
dures'^). Consequently, we tried to establish the potential value but there is no doubt that it would be (at least) inappropriate to
ofthe SQUAT2 minutes in the estimation of aerobic endurance. But, suggest the PUgosec for the estimation of the aerobic-endurance
the results presented in Figure 4 do not support the idea of the capacity. As a certain support to our approach, the article reSQUAT2 minutes ^s a possible replacement aerobic-endurance test cently published by Borsboom et al.'^ is interesting. This article
procedure. In short, results on the SQUAT test explain 9% ofthe advances a simple conception of test validity: a test is valid for
criterion (2M). Consequently, the prediction of aerobic endur- measuring an attribute if (1) the attribute exists and (2) variations in the attribute causally produce variation in the measurement outcomes, in short, the authors stated that validation
1700
research must not be directed at the relationship between the
2MILE = 594,29 + 1,0693 x 1500meters
1600
measured attribute and other attributes but at the processes (!)
R = 0,518
that convey the effect of the measured attribute on the test
1500
EXPL.VAR = 27%
scores. In our ease, the "process" is aerobic endurance, and
1400
there is no doubt that the arm, triceps, and chest strength
(PUgospf) do not convey the effect of the measured attribute (aer1300
obic-endurance capability) on the test scores (2M results).
z
1200
CC
LU
1100
1
1000
900
SOO
700
300
350
400
450
500
550
1500 meters (seconds)
Fig. 3. Linear regression scatterplot-prediction of the 2M using the 1.500m.
At the end. one more problem has to be specified. In the
Croatian Navy, there are no defined consequences when APFT is
failed. On the other hand. NATO forces use simple and effective
counseling methods when APFT faiiure is observed.'^ It includes: (1) participating in the company's special fitness program (conducted concurrently with unit physical fitness training and tailored specifically) and (2) periodical APFT reevaluating. Logically, a soldier is not eligible for favorable
actions like awards and promotions until he (she) takes and
passes a record APFT. We must not forget that the responsibil-
Military Medicine. Vol. 171, August 2006
752
Fitness Measuring. Validation, and Norming in Navy Recruits
TABLEm
Conclusions
SCORING STANDARDS FOR THE NEW PHYSICAL FITNESS TESTS
IN CNR
PU30s*c
1,500 m
Age Group (years)
Score
17-21
100
99
98
97
96
95
94
93
92
91
90
60
10
9
8
7
6
5
4
3
2
1
0
15
22-26
59
58
57
60
56
59
55
18
14
17
13
16
12
15
11
14
10
Age Group (years)
17-21
05:00
05:03
05:06
05:09
05:12
05:15
05:18
05:21
05:24
05:27
05:30
22-26
09:30
09:33
09:36
09:39
09:42
09:45
09:48
09:51
09:54
09:57
10:00
09:09
09:12
09:15
09:18
09:21
09:24
09:27
09:30
09:33
09:36
09:39
05:00
05:03
05:06
05:09
ities of possible leaders have to be considered too. Squad leaders
have to ensure medical evaluation for potential factors contributing to physical limitations, structure a fltness plan to meet
soldier's requirements, and. most importantly, stay involved
and interested in the soldier's improvement. Although they did
not directly study improvements in fitness status, but injury
prevention, Knapik et al.'** recently recognized ability grouping
as a possible enhancement in physical fltness training. Of
course, all of these methods have to be recognized (but also
presented) as corrective, not punitive, in nature and to assist
soldiers and the command to ensure the appropriate fitness
level for the purpose of military readiness.
The results presented and discussed here enable us to define
characteristic scores for those tests and procedures we have
found to be valid. For the purpose of defining standards, we
used frequency tables of the results and excluded "outliers."
keeping Uie upper-lower limit of the results by observing the
expected cumulative percent. We considered subjects as outliers when the expected cumulative percent dropped below 1%
and/or exceeded 99%. Of course, only a short version of the
scorecard is presented in Table III, but we would be pleased to
supply more speciOc details to interested parties.
Military Medicine. Vol. 171. August 2006
In conclusion, the main advantages of the usage of the proposed test procedures applied to CNR are as follows: (1) proposed and validated replacement tests [1500m, PUgosed are significantly (up to four times) shorter in duration compared to
standard APFT (2M and PUj minutJ- (2) Since individual soldiers
are not authorized to administer the APFT to themselves, the
shorter duration of the alternative APFT requires fewer evaluators (compared to standard APFT). (3) The regression equation
we established and presented can be easily used in the calculation of the result on the standard APFT using the alternative
APFT tests. (4) Norms and standards described herein can be
easily used in physical fitness evaluation of CNR.
References
1. Jette H. Kimick A. Sidney K: Evaluating Ihe occupational physical fltness of
Canadian forces infantry personnel. Milit Med 1989; 154: 318-22.
2. Knapik JJ, Canham - C hervak M. Hoedebecke E. et ai: The fitness training unit in
U.S. Army basic combat training: physical fltness. training ontcomes. and Injuries. Milit Med 2001: 166: 356-61.
3. Knapik J: The Army Physical Fitness Test (APm: a review of the literature. Milit
Med 1989: 154: 326-9.
4. Katie R, Males B, Ropac D. Padovan M: Effect of programmed kinesiologic treatment on structural transformation of some strength and endnranee manifestations in Croatian army draftees. Coll Antropol 2002: 26: 229-37.
5. Males B. Sekulic D, Katie R: Morphological and mo tor-endurance changes are
highly related in Croatian navy recruits. Milit Med 2003: 169: 65-70.
6. Knapik J J . Canham-Chervak M, Haurel K. Hoedebecke E. Lanrin MJ. Cuthie J:
Discharges during U.S. Army basic training: injury rates and risk factors. Mlllt
Med 2001: 166:641-7.
7. Bishop PA. Fielitz LR. Crowder TA, Anderson CL. Smith JH. Denick KR: Physiological determinants of performance on an indoor military obslacle course test.
Milit Med 1999: 164: 891-6.
8. Robertson EU. Goss FL, Andreacci JL, et al: Validation ofthe children's OMNI RPE
scale for stepping exercise. Med Sei Sporis Exerc 2005: 37: 290-8.
9. Utter AC. Roberison EU, Green JM. Suminski RR, McAnulty SR Nieman DC:
Validation of the adult OMNI scale of perceived exertion for waiklng/running
exercise. Med Sei Sports Exerc 2004: 36: 1776-80.
10. KnapikJJ, BuUockSH, Canada S. etai: Influence ofan injury reduction program
on injury and fitness outcomes among soldiers. Inj Prev 2004: 10: 37-42.
11. Hoffman JR: The relationship between aerobic fitness and recovery from highintensity exercise In infantry soldiers. Milit Med 1997: 162:484-8,
12. Pope RP. Herberi R, Kirwan JD. Graham BJ: Predicting attrition in basic militaiy
training. Milit Med 1999; 164; 710-4.
13. Harman E: The biomechanics of resistance exercise. In: Essentials of Strength
Training and Conditioning, pp 25-57. Edited by Beaehle TR. Earle RW. Champaign, IL. Human Kinetics, 2000.
14. Ward A, Ebbeiing CB. Ahlquist LE: Indirect methods for estimation of aerobic
power. In: Physiological Assessment of Human Fitness, pp 37-56. Edited by
Maud P. Foster C. Champaign. iL, Human Kinetics. 1995.
15. Sekulic D. Viskic-Stalec N. Rausavljevic PJ: Nonlinear relations between selected
anthropological predictors and psycho-physiological exercise responses. Coll
Antropol 2003; 27: 587-99.
16. Inoue Y. fJakao M: Prediction of maximal oxygen uptake by squat test in men and
women. Kobe J Med Sei 1996:42: 119-29.
17. Borsboom D. Mellenbergh GJ. van Heerden J: The concept of validity. Psychol
Rev2004; U l ; 1061-71.
18. Rush RS: Personal iitness improvement. In; NCO Guide, pp 157-174. Mechanicsburg. PA, Stackpole Books, 1999.