Introduction
Physical performance assessment is crucial in both the sporting and rehabilitation fields. Rather than the performance itself, assessment usually involves the use of a variety of functional tests. The underlying logic is that these measurements provide a quantification of muscular capability relevant in the diagnosis of muscle function and/or athletic performance. Of particular interest is power, because muscular power output is the main determinant of success in many activities and sports. For example, powerful muscular actions are necessary in throwing, jumping, and striking activity. In addition, sudden bursts of power are needed when rapidly changing direction or accelerating during various sports or athletic events (football codes, basketball, gymnastics). In terms of measuring functional power of the lower limb, assessing the leg extensor musculature via some sort of jump is often the method of choice. The jumps used for this type of assessment include single hop for distance (2,8,30 2,8 2,8,18 2,8,18 18 9 17 39 18 8 37 33
Given the wide variety of jumps available, the question arises as to which jump(s) may provide the best diagnostic information to strength and conditioning coaches, clinicians, and researchers. Bilateral vertical jumps are used widely to test athletes and seem to be valid to test activities or sports where 2-legged vertical jumps are important (e.g., rugby lineout, volleyball block). However, many jumps and most propulsive forces (e.g., running and agility) are generated in a unilateral fashion. Therefore, using unilateral assessment would seem to have the advantage of reproducing specific movement patterns and, consequently, may offer better rehabilitation and training information. For example, limb symmetry can be identified, and, for rehabilitation purposes, measurement of the noninjured limb can serve as the biological baseline to which the injured limb should return (18 3,23,34,36 17,23,29,39
Typically, assessment is used to define an athlete's profile, identify talent, monitor performance, identify predictors of performance, determine asymmetries, and/or indicate readiness to return from injury. With this in mind, we must choose tests that assess independent qualities. Testing batteries need to provide the best prognostic/diagnostic value in the shortest amount of time; therefore, having many jumps that have a great deal of shared variance is counterproductive to the aforementioned aims of testing. Whether different jump tests assess independent qualities or are strongly interrelated is unknown and still needs to be determined. It may be that various unilateral jumps better predict performance in different sport-specific populations. For example, lateral jump capability may better predict performance in sports that have a lateral change-of-direction (COD) component such as racquet sports or ice hockey players. However, whether this is actually the case remains unexplored.
As mentioned previously, it is common practice to use vertical or horizontal bilateral jump displacement as indirect measures of leg extensor power and, subsequently, as predictors of speed and COD performance (17,31,32 23 r = −0.73 to −0.86) than the vertical jump tests (r = −0.52 to −0.73). Three studies investigated the relationship between COD performance and unilateral jump in a vertical direction, with inconsistent results (r = −0.23 to −0.71) (17,29,39 r = −0.65) between these 2 variables (29
Methods
Experimental Approach to the Problem
A sample of convenience was used to determine the reliability of various unilateral jump assessments, their interrelationship, and their usefulness as predictors of acceleration (10 m) and COD performance. To achieve this, the unilateral vertical, horizontal, and lateral jump abilities as well as 10-m sprint times and COD performance of 80 men and women were assessed using contact mat and timing light technology. Reliability (% differences, coefficient of variation [CV], and intraclass correlation [ICC]) and correlation analyses were performed on the sample as a whole and on the sample separated by gender to answer the questions of interest.
Subjects
A total of 80 participants (44 men; 36 women) participated in this study. Participants recruited for this study were physical education university students who were involved in recreational physical activities. Their age, body mass, and height were 20.9 ± 4.5 years, 78.1 ± 10.5 kg, and 180.1 ± 7.0 cm for the men and 19.7 ± 2.0 years, 62.1 ± 7.8 kg, and 166.9 ± 6.0 cm for the women (mean ± SD ). The human research ethics committee of Edith Cowan University approved all procedures before commencing the study. All participants were required to sign an informed written consent form before participating in the study.
Testing Procedures
Testing was performed within 1 session. Before testing, each subject's age, height, and mass were recorded. Participants completed a standardized warm-up that consisted of 5 minutes of jogging at a self-selected pace followed by a series of dynamic warm-up drills (walking knees to chest, lunge walks, lateral lunge, high knee, butt kicks, and high skips) and 3 × 10-m sprints at approximately 75, 85, and 100% of each individual's maximal sprint speed. After the warm-up, participants performed 3 practice trials for each of the jump assessments. After the practice trials, 3 test trials were performed in a randomized order for each test, with the 2 best trials recorded, averaged, and used for analysis. The jump assessments consisted of a single-leg vertical countermovement jump (VCJ), single-leg horizontal countermovement jump (HCJ), and single-leg lateral countermovement jump (LCJ) performed with the dominant leg. During all jump assessments, participants were instructed to keep their hands placed on their hips. After the jump assessments, the sprint assessments were performed. These consisted of a 10-m sprint, a dominant-leg COD sprint, and an alternate-leg COD sprint. Participants were allowed 2 practice trials followed by 2 test trials for the sprint and COD assessments. Leg dominance was defined as the leg the participants preferred to use to kick a ball. At least 60 seconds of rest was taken between jumps, sprints, and agility assessments, with the longest tests performed last to minimize the effects of fatigue.
Assessments
Vertical Countermovement Jump
Each subject's VCJ was calculated from flight time (4
Horizontal Countermovement Jump
For the HCJ, each participant began by standing on the designated testing leg with the toe at the starting line and hands on the hips. Each participant was instructed to sink to a self-selected depth as quickly as possible and then jump as far forward as possible and land on 2 feet. The distance jumped was measured to the nearest 0.01 m with a tape measure
Lateral Countermovement Jump
For the LCJ, each participant began by standing on the designated testing leg with the foot at the starting line and hands on the hips. Each participant was instructed to sink to a self-selected depth as quickly as possible and then jump laterally to the inside as far as possible and land on 2 feet. The distance jumped was measured to the nearest 0.01 m with a tape measure
Ten-Meter Sprint
Sprint time was measured with timing light system consisting of a dual-beam modulated visible red-light system with polarizing filters (Swift Performance) placed at the start and end of the 10 m. The starting position was standardized for all participants. Each participant started in a split stance position with the preferred foot 30 cm from the starting line and the other foot in line with the heel of the preferred foot.
Change of Direction-Dominant Leg
The timing lights system was placed at the start and at 10 m to measure COD-dominant leg (CODD) time. The starting position was standardized as for the 10-m sprint. Each participant was instructed to run to a marker situated 7.5 m from the starting line, perform a 180° using the dominant leg to push off, return to a marker 5 m away, perform another 180° turn using the dominant leg to push off, and continue running past the final timing light situated 7.5 m away. A total distance of 20 m was covered (Figure 1 ).
Figure 1: Change-of-direction test. D = change of direction with the dominant leg; A = change of direction with the alternate leg; dotted line = the point at which change of direction occurs.
Change of Direction-Alternate Leg
The COD-alternate leg (CODA) time was similar to the CODD, but 1 COD was performed with the dominant leg and the other one was performed with the alternate leg (Figure 1 ).
Statistical Analyses
After data collection, mean and SD were calculated for all results. The 2 best trials for all assessments were averaged for an individual subject mean, and subject means for each assessment were averaged to provide a group mean for men and women. The reliability of the jump assessment procedures was calculated using 3 different statistical methods. Percent differences and a CV (SD /mean × 100) between the 2 best trials were calculated for all test variables to determine the stability of measurement among trials. The ICCs were also calculated to determine the reliability of measures between trials using a single measure, 2-way mixed effect model ICC with absolute agreement. Confidence intervals (90%) of the difference between trials were calculated for all 3 reliability statistics by the statistical packages. Pearson product-moment correlations and multiple regressions were used to determine relationships between jump performance and sprint/agility times. The forward stepwise regression began with no variables in the equation and, thereafter, entered the most "significant' predictor at the first step and continued to add or delete variables until none “significantly” improved the fit. Minimum tolerance for entry into the model and alpha-to-enter/remove were set at 0.01 and 0.15, respectively. From this analysis, the best single- and double-predictor statistical models of sprint and agility performance were derived for men and women. Regression diagnostics were used to examine normality, variance, colinearity, outlier effects, leverage, and influence. The subjects were pooled for analysis and then separated by gender as scatter plots revealed correlations based on bipolar data, which gave rise to artificially high correlations. Coefficients of determination (R 2 ) were used to represent the goodness of fit of the predictor models and also to represent the shared variance between the variables of interest.
Results
The mean, SD , and within-trial reliability statistics for all the outcome measures of interest are presented in Table 1 . The results of this analysis have been presented as men, women, and pooled (men + women) data. In terms of the men-women comparisons, for the anthropometric measurements there were differences of 21 and 5.6% between genders in mass and height, respectively. The difference between genders for the COD and sprint tests ranged from 11.4 to 12.4%, and the difference between jumps ranged from 17.6 to 27.2%. The largest between-gender differences was noted in VCJ height (27.2%).
Table 1: Reliability measures and 90% confidence limits for men (n = 44) and women (n = 36).
With regard to the reliability statistics, percent change between trials ranged in absolute terms from 0.1 to 3.8%, with the largest variability found in the women's VCJ assessment (3.8%). The largest CV was observed for the VCJ of both genders (CV = 6.7-7.2%), whereas the sprint and COD assessments had the smallest variability (CV = 0.8-2.8%). All ICCs were more than 0.85 apart from the men's CODA test. All pooled ICCs were higher than the gender-separated ICCs.
No anthropometric measurements were significantly related to any of the performance variables. The intercorrelations between variables for both men and women can be observed in Table 2 . The jumps were, for the most part, highly correlated to each other for both genders (r = 0.46-0.66; p < 0.01); however, the shared variance (R 2 ) between jumps was less than 44%.
Table 2: Relationship between measures and significance for men (n = 44) and women (n = 36).
In terms of the relationship between the jumps and sprint performance, all jumps except for the women's LCJ were significantly related to the 10-m sprint time. Correlations were moderate to high in magnitude, and the best predictor of 10-m sprint time for men was the HCJ (r = −0.65) and, for women, the VCJ (r = −0.44). For COD performance, the LCJ was not significantly related to either of the COD tests. For the most part, the HCJ assessment was the best predictor (r = −0.46 to −0.59) of COD performance, with a shared variance of less than 35%.
The relationship between COD times of the dominant and alternate legs for both men and women was high; however, the shared variance (R 2 ) between all these tests ranged from (30 to 46%). The 10-m sprint times were highly correlated (r = 0.57-0.64; p < 0.01) to the CODA test but not significantly related to the CODD test for men and women.
The best single predictor of COD and sprint performance for men was the HCJ, accounting for 21-42% of the shared variance between the measures. The introduction of VCJ performance to the 10-m statistical model improved the common variance for that variable by only 6.6-48.6%. For women, the best single predictors of sprint and COD performance were the VCJ and HCJ, but the shared variance between variables was less than 34%. The predictive model is shown in Table 3 .
Table 3: Best predictor models for acceleration and change-of-direction performance.
Discussion
The difference between genders for the COD and sprint tests ranged from 11.4 to 12.4% and from 17.6 to 27.2% for the jump assessments. There are a number of physical and physiological distinctions between men and women that could possibly explain these results, including differences in upper- and lower-body anthropometry (38,40 28 6 19 21 r = 0.03-0.19). Differences in jumps, sprints, and COD performance between genders may be better explained by other factors discussed previously, further discussion of which is outside the scope of this article.
Many studies reporting reliability of various measures pool (men + women) their data for the statistical analyses. As can be observed from Table 1 , this practice artificially inflates the magnitude of the statistics. That is, because a correlational analysis such as the ICC is sensitive to the heterogeneity (spread) of the sample, the difference in jump performance between men and women increases the heterogeneity and, subsequently, inflates the ICC results. To the authors' knowledge, the current study was the first to investigate the variability of acyclic jump assessment between genders. It can be observed from Table 1 that, for this particular sample, the men subjects exhibited less movement variability across all jumps compared with the women.
A primary purpose of this study was to determine the reliability of 3 unilateral jumps and the reliability of 10-m sprint and 2 types of COD test. All jump measurements were found to be reliable, with VCJ results being the less consistent in both genders (see Table 1 ), but still highly acceptable. These findings are similar to those of previous studies that have investigated the reliability of unilateral VCJ (18,23 2,16,23,30
With respect to the reliability of the 10-sprint test (−0.1 to 0.6% mean change; CV = 1.7-1.9; ICC = 0.87-0.91), the stability between trials can be considered highly acceptable and in accordance with reliability data previously reported for 10-m sprints (ICC = 0.86-0.93; CV = 1.2-2%) (10,12,27 Table 1 ). This may be expected and, possibly, explained by the greater use and subsequent strength and coordination of the preferred or dominant leg. Reliability measures of numerous COD tests have been reported in the literature, and regardless of the duration of the test, the number of CODs, or the direction in which most of the forces were applied, all COD tests showed similar reliability (ICC = 0.8-0.96; CV = 1-5%) (1,7,9,13-15,22,25
Testing batteries need to provide the best prognostic/diagnostic value in the shortest amount of time; therefore, having many jumps that have a great deal of shared variance is counterproductive to implementing an efficient test battery of practical and/or clinical utility. Of interest in this study was whether the different jump tests assessed independent qualities or were strongly interrelated. For the most part, the 3 different jumps were moderately correlated to each other for both genders (r = 0.46-0.66); however, the shared variance (the correlation value squared = R 2 ) between jumps was less than 44%, indicating that the vertical, horizontal, and lateral tests are, for the most part, measuring different leg power qualities and should not be used interchangeably. The strength and conditioning coach therefore needs to select tests that have the greatest specificity to the activity of interest or run a full battery of jump tests.
Of interest was whether the jump assessments could predict functional performance. With regard to sprint performance, it was thought that unilateral horizontal and vertical jump performance should best predict sprint performance, given that sprinting in a straight line is the resultant of vertical and horizontal forces. The best single predictors of 10-m sprint performance were the HCJ for men (R 2 = 42%) and the VCJ for women (R 2 = 18%). The introduction of the VCJ performance to the 10-m statistical model for men only accounted for a 6.6% improvement in predictive ability. For men, it seems that the HCJ is a better predictor of acceleration performance. These results are supported by the findings of Maulder and Cronin (23 R 2 = 0.55). However, it should be noted that there is very little difference between the magnitude of these correlations (HCJ and VCJ) and sprint performance for both genders. Perhaps of more interest is that there is a great deal of unexplained variance between the vertical and horizontal jump measures and acceleration performance for men (54%) and women (82%). This may be explained by the types of jumps used (countermovement or stretch-shorten cycle [SSC]), whereas the contribution of the SSC to the beginning of a sprint is considered minimal (26
In terms of the predictors of COD performance, it was thought that the LCJ may better predict this quality, given the requirements of this test. However, the HCJ was found to be the best single predictor of both COD tests in men (R 2 = 0.21-0.24) and the best single predictor of CODA in women (R 2 = 0.34). The VCJ was the best single predictor for the CODD in women (R 2 = 0.24), but HCJ had almost the same predictive ability (R 2 = 0.22). The poor predictive ability of the LCJ may be attributed to the test design: there were only 2 CODs involved in the test, whereas most other movements in the test were in a straight line, and therefore the HCJ was, once more, the better predictor of this type of movement. Negrete and Brophy (29 32 r = −0.61 to −0.79) compared with the relationship of the T-test and double-leg vertical jump (r = −0.27 to −0.71) in both genders. Given the results of the present and previous studies, it may be tentatively claimed that jumps that involved the combination of both horizontal and vertical ground-reaction forces may better predict COD ability.
Interestingly, the shared variance between the dominant and alternate legs in the COD tests was less than 50% for the men (45%) and women (29%). It would seem that the ability to change direction with 1 leg is relatively independent of a COD with the other leg, especially in the women of this study. This has interesting diagnostic and training implications for the strength and conditioning coach.
The final comparison of interest was between the 10-m sprint and COD performance. Once more, a test battery needs to incorporate assessments that give information on different aspects of performance. If a COD and sprint test share a lot of common variance, there seems little need for the 2 tests. As it can be observed from the results of this study, the straight-line 10-m sprint and COD test seem to be assessing relatively independent qualities, because the shared variance between CODD and 10-m sprint performance for men and women was approximately 4%. Interestingly, the common variance between the CODA test and 10-m sprint was much higher (32-41%). Results of previous studies have demonstrated the same inconsistency, showing low to high correlations between straight sprint (10 m to 40 yd) and COD with various numbers of CODs (1-9; r = 0.05-0.69) (5,11,20,24,31,32 35
Practical Applications
As can be observed from the tables, pooling gender data during any type of correlational analysis (Pearson and ICC) results in an inflated correlation because of the heterogeneity or bipolar scatter of the sample. This pooling of data should be avoided, and, if the sample size allows, the correlations should be represented separately, as in this study. It seems that single-leg lateral jump assessment was just as reliable as the other 2 measures of jump performance as represented by percent change, CV, and ICC. The practitioner can use the average of 2 trials for this type of jump with reasonable confidence about the stability of the measurement. Interestingly, the shared variance between jumps for men and women was less than 50%, indicating that the jumps are relatively independent of one another and represent different leg strength/power qualities. This needs to be taken into account when selecting tests for diagnostic or prognostic purposes. The ability of the jumps to predict sprint and COD performance was limited. However, if 1 jump assessment were selected to predict sprint and COD performance in a test battery, the HCJ would seem the logical choice, given the results of this study. It would seem that the ability to change direction with 1 leg is relatively independent of a COD with the other leg, especially in the women of this study. Many of the findings in this study have interesting diagnostic and training implications for the strength and conditioning coach. It should be noted that correlations can only give insights into associations and not into causes and effect; therefore, the practical applications described previously need to be interpreted with this in mind.
References
1. Alricsson, M, Harms-Ringdahl, K, and Werner, S. Reliability of sports related functional tests with emphasis on speed and agility in young athletes.
Scand J Med Sci Sports 11: 229-232, 2001.
2. Bolga, L and Kestula, D. Reliability of lower extremity functional performance tests.
J Orthop Sports Phys Ther 26: 138-142, 1997.
3. Booher, L, Hench, K, Worrell, TW, and Stikeleather, J. Reliability of three single-leg hop tests.
J Sport Rehabil 2: 165-170, 1993.
4. Bosco, C, Luhtanen, P, and Komi, PV. A simple method for measurement of mechanical power in jumping.
Eur J Appl Physiol Occup Physiol 50: 273-282, 1983.
5. Buttifant, D, Graham, K, and Cross, K. Agility and speed in soccer players are two different performance parameters. In:
Science and Football IV . W. Spinks, T. Reilly, and A.J. Murphy, eds. London: Taylor & Francis, 2001. pp. 329-332.
6. Chow, RS, Medri, MK, Martin, DC, Leekam, RN, Agur, AM, and McKee, NH. Sonegraphic studies of human soleus and gastrocnemius muscle achitecture: gender variability.
Eur J Appl Physiol 82: 236-244, 2000.
7. Christou, M, Smilios, I, Sotiropoulos, K, Volaklis, K, Pilianidis, T, and Tokmakidis, SP. Effects of resistance training on the physical capacities of adolescent soccer players.
J Strength Cond Res 20: 783-791, 2006.
8. Clark, NC, Gumbrell, CJ, Rana, S, Traole, CM, and Morrissey, MC. Intratester reliability and measurement error of the adapted crossover hop for distance.
Phys Ther Sport 3: 143-151, 2002.
9. Cronin, J, McNair, PJ, and Marshall, RN. The effects of bungy weight training on muscle function and functional performance.
J Sports Sci 21: 59-71, 2003.
10. Cronin, JB, Green, JP, Levin, GT, Brughelli, ME, and Frost, DM. Effect of starting stance on initial sprint performance.
J Strength Cond Res 21: 990-992, 2007.
11. Darper, J and Lancaster, M. The 505 test: a test for agility in the horizontal plane.
Aust J Sci Med Sport 17: 15-18, 1985.
12. Duthie, GM, Pyne, DB, Marsh, DJ, and Hooper, SL. Sprint patterns in rugby union players during competition.
J Strength Cond Res 20: 208-214, 2006.
13. Gabbett, T, Georgieff, B, Anderson, S, Cotton, B, Savovic, D, and Nicholson, L. Changes in skill and physical fitness following training in talent-identified volleyball players.
J Strength Cond Res 20: 29-35, 2006.
14. Gabbett, TJ. A comparison of physiological and anthropometric characteristics among playing positions in sub-elite rugby league players.
J Sports Sci 24: 1273-1280, 2006.
15. Gabbett, TJ. Performance changes following a field conditioning program in junior and senior rugby league players.
J Strength Cond Res 20: 215-221, 2006.
16. Greenberger, H. The test-retest reliability of a one-legged hop for distance in healthy young adults.
J Orthop Sports Phys Ther 19: 62, 1994.
17. Hoffman, JR, Ratamess, NA, Klatt, M, Faigenbaum, AD, and Kang, J. Do bilateral power deficits influence direction-specific movement patterns?
Res Sports Med 15: 125-132, 2007.
18. Hopper, D, Goh, S, Wentworth, L, Chan, D, Chau, J, and Wooton, G. Test-retest reliability of knee rating scales and functional hop tests one year following anterior cruciate ligament reconstruction.
Phys Ther Sport 3: 10-18, 2002.
19. Komi, PV and Bosco, C. Utilization of stored elastic energy in leg extensor muscles by men and women.
Med Sci Sports 10: 261-265, 1978.
20. Little, T, and Williams, AG. Specificity of acceleration, maximum speed, and agility in professional soccer players.
J Strength Cond Res 19: 76-78, 2005.
21. Markovic, G and Jaric, S. Movement performance and body size: the relationship for different groups of tests.
Eur J Appl Physiol 92: 139-149, 2004.
22. Markovic, G, Jukic, I, Milanovic, D, and Metikos, D. Effects of sprint and plyometric training on muscle function and athletic performance.
J Strength Cond Res 21: 543-549, 2007.
23. Maulder, P and Cronin, J. Horizontal and vertical jump assessment: reliability, symmetry, discriminative and predictive ability.
Phys Ther Sport 6: 74-82, 2005.
24. Mayhew, JL, Piper, F, Schegler, TM, and Ball, TE. Contributions of speed, agility and body composition to anaerobic power measurement in college football players.
J Appl Sport Sci Res 3: 101-106, 1989.
25. McBride, JM, Triplett-McBride, T, Davie, A, and Newton, RU. The effect of heavy- vs. light-load jump squats on the development of strength, power, and speed.
J Strength Cond Res 16: 75-82, 2002.
26. Mero, A. Force-time characteristics and running velocity of male sprinters during the acceleration phase of sprinting.
Res Q Exerc Sport 59: 94-98, 1988.
27. Moir, G, Button, C, Glaister, M, and Stone, MH. Influence of familiarization on the reliability of vertical jump and acceleration sprinting performance in physically active men.
J Strength Cond Res 18: 276-280, 2004.
28. NASA.
Anthropometric Source Book Volume I: Anthropometry for Designers . NASA Reference Publication 1024. Washington, DC: NASA, 1978.
29. Negrete, R and Brophy, J. The relationship between isokinetic open and closed chain lower extremity strength and functional performance.
J Sport Rehabil 9: 46-61, 2000.
30. Paterno, MV and Greenberger, HB. The test-retest reliability of a one legged hop for distance in young adults with and without ACL reconstruction.
Isokinet Exerc Sci 6: 1-6, 1996.
31. Pauole, K, Madole, K, Garhammer, J, Lacouse, M, and Rozenek, R. Reliability and validity of the T-test as a measure of agility, leg power, and leg speed in college-aged men and women.
J Strength Cond Res 14: 443-450, 2000.
32. Peterson, MD, Alvar, BA, and Rhea, MR. The contribution of maximal force production to explosive movement among young collegiate athletes.
J Strength Cond Res 20: 867-873, 2006.
33. Ross, M. Test-retest reliability of the lateral step-up test in young adult healthy subjects.
J Orthop Sports Phys Ther 25: 128-132, 1997.
34. Ross, M, Langford, B and Whelan, P. Test-retest reliability of 4 single-leg horizontal hop tests.
J Strength Cond Res 16: 617-622, 2002.
35. Sheppard, JM, and Young, WB. Agility literature review: classifications, training and testing.
J Sports Sci 24: 919-932, 2006.
36. Simpson, R, and Cronin, J. Reliability of a unilateral horizontal leg power test to assess stretch load tolerance.
Meas Phys Educ Exerc Sci 3: 169-178, 2006.
37. Tkac, M, Hamar, D, Komadel, L, and Kuthanova, O. Measurement of anaerobic power of the lower limbs by a method of repeated vertical jumps.
Sports Med Train Rehabil 1: 317-325, 1990.
38. Walsh, MS, Waters, J and Kersting, UG. Gender bias on the effects of instruction on kinematic and kinetic jump parameters of high-level athletes.
Res Sports Med 15: 283-295, 2007.
39. Young, WB, James, R, and Montgomery, I. Is muscle power related to running speed with changes of direction?
J Sports Med Phys Fitness 42: 282-288, 2002.
40. Zatsiorsky, VM and Selujanov, VN. The mass and inertia characteristics of the main segments of the human body. In:
Biomechanics VIII . Matsui, BH and Kabayashi, K. eds. Champaign: Human Kinetics, 1983. pp. 1152-1159.
