The Reliability of Three Devices Used for Measuring Vertical Jump Height

SAVE OFFLINE

THE RELIABILITY OF THREE DEVICES USED MEASURING VERTICAL JUMP HEIGHT JAMES L. NUZZO, JONATHAN H. ANNING,

AND

FOR

JESSICA M. SCHARFENBERG

Department of Exercise and Rehabilitative Sciences, Slippery Rock University, Slippery Rock, Pennsylvania

ABSTRACT Nuzzo, JL, Anning, JH, and Scharfenberg, JM. The reliability of three devices used for measuring vertical jump height. J Strength Cond Res 25(9): 2580–2590, 2011—The purpose of this investigation was to assess the intrasession and intersession reliability of the Vertec, Just Jump System, and Myotest for measuring countermovement vertical jump (CMJ) height. Forty male and 39 female university students completed 3 maximal-effort CMJs during 2 testing sessions, which were separated by 24–48 hours. The height of the CMJ was measured from all 3 devices simultaneously. Systematic error, relative reliability, absolute reliability, and heteroscedasticity were assessed for each device. Systematic error across the 3 CMJ trials was observed within both sessions for males and females, and this was most frequently observed when the CMJ height was measured by the Vertec. No systematic error was discovered across the 2 testing sessions when the maximum CMJ heights from the 2 sessions were compared. In males, the Myotest demonstrated the best intrasession reliability (intraclass correlation coefficient [ICC] = 0.95; SEM = 1.5 cm; coefficient of variation [CV] = 3.3%) and intersession reliability (ICC = 0.88; SEM = 2.4 cm; CV = 5.3%; limits of agreement = 20.08 6 4.06 cm). Similarly, in females, the Myotest demonstrated the best intrasession reliability (ICC = 0.91; SEM = 1.4 cm; CV = 4.5%) and intersession reliability (ICC = 0.92; SEM = 1.3 cm; CV = 4.1%; limits of agreement = 0.33 6 3.53 cm). Additional analysis revealed that heteroscedasticity was present in the CMJ when measured from all 3 devices, indicating that better jumpers demonstrate greater fluctuations in CMJ scores across testing sessions. To attain reliable CMJ height measurements, practitioners are encouraged to familiarize athletes with the CMJ technique and then allow the

Address correspondence to Jonathan H. Anning, [email protected] 25(9)/2580–2590 Journal of Strength and Conditioning Research Ó 2011 National Strength and Conditioning Association

2580

the

athletes to complete numerous repetitions until performance plateaus, particularly if the Vertec is being used.

KEY WORDS Vertec, Just Jump system, contact mat, Myotest, accelerometer

INTRODUCTION

V

ertical jump height is commonly assessed in the strength and conditioning field. Specifically, sports scientists have used this measurement to calculate lower-body power (8,23) and to discriminate between levels of sport playing ability as indicated by player rankings (22) and comparisons of starters vs. nonstarters (2,6,25). Furthermore, vertical jump height has also been assessed in non-athletic populations and appears to be a valuable measurement in identifying lower-limb functionality in elderly persons (20) and in obese and nonobese children (21). Because of the prevalence of the vertical jump height measurement in the strength and conditioning field, it is important that the methodologies used to quantify this measurement are first found reliable in a laboratory setting. Various methodologies can be employed to determine vertical jump height (13). In a laboratory, vertical jump height can be determined using a force platform or motion analysis system. However, because of the great expenses associated with this type of equipment, and the training needed to operate it, more cost-efficient and simplistic devices have been developed and are now commercially available to strength and conditioning coaches. Examples of these devices include the Vertec (Sports Imports, Hilliard, OH, USA), Just Jump System (Probotics Inc., Huntsville, AL, USA), and Myotest (Myotest Inc., Switzerland). These devices are beneficial to the strength and conditioning coach because they are relatively inexpensive, require minimal training for operation, do not require extensive data analysis, and provide immediate results. The Vertec (12,14,22,31) and Just Jump System (12,14,16,17) devices have been used previously in assessing vertical jump height. The Vertec assesses vertical jump height by measuring the difference between the fully extended standing reach height and the maximal vertical jump-andreach height (22). The Just Jump System determines jump height by measuring flight time. The flight time is determined by microswitches located within the mat, which are sensitive

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research to the liftoff of the feet from the mat and to the landing of the feet back on to the mat (13). The flight time measurement can then be entered into a standardized equation, which estimates the jump height (jump height = [t2 3 g] O 8, where g = acceleration rate due to gravity (9.81 m s22) and t = flight time). More recently, the Myotest accelerometer has been advertised as a device for quantifying kinetic and kinematic variables to monitor athletic performance. The Myotest determines vertical jump height through force measurements, which are the product of the acceleration of the jump and the body mass of the individual. The body mass of the individual must be entered into the device before jumping. From the force data, flight time is calculated by the differences in 2 time points: the end of the concentric phase (i.e., when the pushing force equals the force of the body weight), and the landing (i.e., when the landing force equals the force of the body weight). Based on the flight time, the height of the jump is then calculated by the Myotest. The reliability of measuring the height of the countermovement vertical jump (CMJ) with a device similar to the Vertec (31) and the Just Jump System (16,17) has been previously investigated. From these investigations, it has been concluded that both devices demonstrate adequate intersession reliability (i.e., session-to-session reliability) based on high intraclass correlation coefficients (ICCs) and low coefficient of variations (CVs) (16,17,31). Unfortunately, because of the various CMJ techniques that have been employed in these studies, it is difficult to conclude which of the 2 devices demonstrates the best intersession reliability. It should also be noted that these devices have been measured concurrently within the context of a single investigation (12,14); however, in these investigations, the reliability of the devices was not assessed. Additionally, intrasession reliability (i.e., trial-totrial reliability within in a given testing session) has not been assessed with these 2 devices. In regard to CMJ height measurements from the Myotest, there are currently no published data on either intrasession or intersession reliability. Because these 3 devices have never been tested simultaneously within the context of a single investigation, there is still an unanswered question regarding which device is the most reliable. Answering this question could help strength and conditioning professionals determine which device will be the most sensitive for detecting changes in CMJ height over the course of a training program. Thus, the purpose of this investigation was to assess the intrasession and intersession reliability of the Vertec, Just Jump System, and Myotest for measuring CMJ height.

| www.nsca-jscr.org

by comparing the 3 CMJs performed in each session, and the intersession reliability was assessed by comparing the maximum CMJ values attained in session 1 and session 2. The 3 measurement devices served as the independent variable. The dependent variable was CMJ height, and this was measured from all 3 devices simultaneously. Subjects

Forty male (age: 19.7 6 1.5 years; body height: 180.2 6 6.2 cm; body mass: 79.6 6 8.9 kg) and 39 female (age: 19.5 6 1.3 years; body height: 163.7 6 6.8 cm; body mass: 60.7 6 8.7 kg) university students participated in this investigation. The majority of the participants were exercise science students who had various backgrounds in recreational and competitive sports; however, a physically active background, or experience with vertical jumping, was not required. Informed consent was obtained from all subjects before testing. Subjects also completed a health risk stratification form to ensure safety and to identify underlying physiological conditions that may have impacted the test results. Approval from the Slippery Rock University Institutional Review Board was obtained before the start of the investigation. Informed consent was obtained from all subjects before testing. Procedures

During the first testing session, body height and weight were assessed using a physician’s scale. Subjects then completed a 5-minute warm-up on a cycle ergometer at 0.5 kp. After

METHODS Experimental Approach to the Problem

To determine the reliability of the Vertec, Just Jump System, and Myotest, male and female university students participated in 2 testing sessions, which were separated by 24–48 hours. During each session, subjects completed 3 maximal-effort CMJs. The intrasession reliability of the devices was assessed

Figure 1. Standing reach height for the Vertec.

VOLUME 25 | NUMBER 9 | SEPTEMBER 2011 |

2581

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Vertical Jump Reliability completing the warm-up, reach height for the Vertec was established using the following body position: erect stance, both feet together and flat on the ground, both arms fully extended overhead, and the head and eyes level (12) (Figure 1). Instructions on the CMJ technique were then provided. This technique required subjects to start in an upright position with the feet parallel to each other and hip to shoulder width apart. After hearing an auditory beep from the Myotest device, subjects performed a quick countermovement by flexing the hips and knees. After reaching their preferred depth of descent, subjects explosively extended at the knees and hips, and plantar flexed at the ankles in an effort to attain a maximal jump height. During the concentric and flight phases of the jumps, subjects were required to maintain a level head position (i.e., not looking upward at the Vertec vanes) while reaching upward with both hands simultaneously (12) (Figure 2). After the CMJ technique was explained, subjects completed 2 warm-up CMJs. These 2 jumps served the purpose of a CMJ-specific warm-up involving maximal-effort explosive muscle contractions, and in addition, provided the test administrators with an opportunity to correct any flaws in the CMJ technique. Subjects then completed 3 maximal-effort CMJs with a 1minute rest period between each trial. For all 3 devices to measure CMJ height simultaneously, the Just Jump Mat was placed on the floor below the Vertec vanes, and the Myotest device was attached to an adjustable belt, which was secured

around the subjects’ waists. The spacing between the Vertec vanes was 1.27 cm, and the Myotest device had a sampling frequency of 200 Hz. The same procedures were duplicated during the second testing session. In addition, subjects were not permitted to participate in high intensity physical activity within 48 hours of the first testing session or between the 2 testing sessions. Statistical Analyses

Numerous analysis techniques have been suggested for establishing the reliability of measurements within the realm of sports medicine and sports science (1,10,11,30). Unfortunately, no consensus on this issue has been reached. Thus, the current investigation employed a comprehensive assessment of reliability by using statistics that assess systematic error (bias) (e.g., dependent t-test, 1-way repeated measures analysis of variance [ANOVA]), relative reliability (e.g., ICC), and absolute reliability (e.g., SEM, CV, limits of agreement) (1). In the current investigation, male and female data were analyzed separately using SPSS Version 15.0 (SPSS Inc., Chicago, IL, USA). It should be noted, that for both males and females, the CMJ height scores (including individual trial scores from each session and maximum scores from each session) from all 3 devices, were normally distributed. Normality of the data was accepted if the ratio of the skewness to the standard error of skewness was ,2.0 (29). A criterion alpha level of p # 0.05 was established for the analyses. To examine intrasession reliability (i.e., reliability across the three trials in a session), the means and SDs for trial 1, trial, and trial 3 of the CMJ heights were calculated. Subsequently, a 1-way repeated measures ANOVA with Tukey post hoc test was used to assess systematic error (e.g., learning, fatigue) between trials 1, 2, and 3. Then, the ICC, SEM, and CV were calculated. Based on the specific study design, the ICC was calculated using the following equation (30):

ICC ¼

MSs  MSe ; kðMSt  MSeÞ MSs þ ðk þ 1ÞMSe þ n

where MSs is the subjects mean square, MSe is the error (residual) mean square, k is the number of measurements being compared, MSt is the trials (between items) mean square, and n is the sample size. Subsequently, the SEM was calculated using the following equation (1):

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi SEM ¼ SD 1  ICC; where SD is the average of the SDs from trials 1, 2, and 3. The CV was then calculated using the following equation (10):

CV ¼

SEM ; Mean

Figure 2. Vertical jump technique.

where Mean is the arithmetic average of the 3 trials.

2582

the

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

Figure 3. Bland–Altman plot.

Figure 4. Graphical representation of heteroscedasticity.

To examine intersession reliability, the means and SDs for the maximum CMJ heights from session 1 and from session 2 were calculated. Then, a dependent t-test was used to assess systematic error between the maximum CMJ heights from the 2 sessions. Subsequently, the ICCs, SEMs, and CVs were calculated. It should be noted that the SEM for intersession reliability was calculated with SD representing the average of the SDs from session 1 and session 2. Additionally, the CV for intersession reliability was calculated with the mean representing the arithmetic average of the session 1 and session 2 maximum CMJ heights. The means and SDs of the

| www.nsca-jscr.org

differences between the maximum CMJ heights from session 1 and session 2 were also calculated. Intersession reliability was further assessed using the 95% limits of agreement of the session 1 and session 2 maximal CMJ differences and Bland–Altman plots (1,3). The Bland– Altman plots provide a visual representation of the individual subject means plotted against the individual subject measurement differences. In this study, the mean of the maximum CMJ height measurements from session 1 and session 2 was plotted against the difference between those same 2 measurements (Figure 3). Then, a similar graph was made using the absolute differences between measurements (Figure 4). Subsequently, a Pearson product moment correlation coefficient was used to quantify the level of heteroscedasticity between the CMJ heights and the absolute differences (1,19). Heteroscedasticity in a data set occurs when higher performers demonstrate more variability in scores across testing sessions when compared to lower performers. If the correlation was close to zero, and the differences were normally distributed, the presence of heteroscedasticity was rejected (1,19). Normality of the data was rejected if the ratio of the skewness to the standard error of skewness exceeded 6 2.0 (29). If the presence of heteroscedasticity was accepted, the raw score maximal CMJ heights from session 1 and session 2 were transformed using the natural logarithmic scale (Ln) (1,18,19). The log transformed means of the maximum CMJ heights from both sessions, and the log transformed means and SDs of the differences between the 2 sessions, were then determined. To determine the limits of agreement from the transformed data, antilogs (ex) of both the mean of the differences and the SD of the differences were calculated (1,19). The antilog of the mean of the transformed differences is a ratio of the measurement bias, whereas the antilog of the VOLUME 25 | NUMBER 9 | SEPTEMBER 2011 |

2583

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Vertical Jump Reliability

TABLE 1. Session 1 and session 2 intrasession reliability in females.* Device S1-Vertec S2-Vertec S1-JJ System S2-JJ System S1-Myotest S2-Myotest

Trial 1 (cm)

Trial 2 (cm)

Trial 3 (cm)

ICC

SEM (cm)

CV (%)

30.0 29.3 36.7 36.4 30.0 29.6

29.5 29.2 35.9 36.8 29.6 30.3

31.3 6 6.3† 30.6 6 6.5‡ 36.6 6 6.0† 37.0 6 6.0 30.0 6 4.6 30.3 6 4.5§

0.89 0.87 0.93 0.90 0.91 0.91

2.1 2.3 1.6 1.9 1.4 1.4

6.9 7.6 4.4 5.2 4.6 4.5

6 6.2 6 6.1 6 6.2† 6 6.0 6 4.5 6 4.7

6 6.7 6 6.5 6 6.2 6 6.0 6 4.8 6 4.5*

*ICC = intraclass correlation coefficient; SEM = standard error of measurement; CV = coefficient of variation; S1 = session 1; S2 = session 2; JJ = just jump. †Significantly different (p # 0.05) from trial 2. ‡Significantly different from trials 1 and 2. §Significantly different from trial 1.

TABLE 2. Session 1 and session 2 intrasession reliability in males.* Device S1-Vertec S2-Vertec S1-JJ System S2-JJ System S1-Myotest S2-Myotest

Trial 1 (cm)

Trial 2 (cm)

46.8 46.7 54.8 55.0 44.2 44.4

47.7 6 48.5 6 54.9 6 55.4 6 44.1 6 44.4 6

6 10.8 6 8.9 6 9.1 6 8.3 6 7.5 6 6.4

10.6 9.5‡ 9.2 8.0 7.5 6.4

Trial 3 (cm)

ICC

SEM (cm)

CV (%)

49.2 6 10.4† 48.5 6 9.2‡ 56.1 6 9.1† 56.1 6 9.0 44.8 6 7.4 44.8 6 6.8

0.94 0.94 0.93 0.92 0.95 0.95

2.7 2.2 2.3 2.3 1.7 1.5

5.5 4.6 4.2 4.2 3.9 3.3

*ICC = intraclass correlation coefficient; SEM = standard error of measurement; CV = coefficient of variation; S1 = session 1; S2 = session 2; JJ = Just Jump. †Significantly different (p # 0.05) from trials 1 and 2. ‡Significantly different from trial 1.

TABLE 3. Intersession reliability of the raw data for females and males.*†‡ Device F-Vertec F-JJ System F-Myotest M-Vertec M-JJ System M-Myotest

S1 Max CMJ (cm)

S2 Max CMJ (cm)

ICC

SEM (cm)

CV (%)

31.95 37.80 30.81 50.04 57.10 45.80

31.65 38.25 31.14 49.78 57.25 45.72

0.80 0.92 0.92 0.90 0.84 0.88

2.7 1.7 1.3 3.0 3.6 2.4

8.6 4.4 4.1 5.9 6.3 5.3

6 6.2 6 6.0 6 4.4 6 10.1 6 9.2 6 7.5

6 5.9 6 6.0 6 4.4 6 9.1 6 9.0 6 6.6

S2-S1 Diff (cm) 20.30 0.45 0.33 20.26 0.15 20.08

6 3.91 6 2.42 6 1.80 6 4.22 6 5.14 6 3.49

Limits of agreement

PPM

p Value

20.30 6 7.66 0.45 6 4.74 0.33 6 3.53 20.26 6 8.27 0.15 6 7.13 20.08 6 4.06

0.29 0.19 0.17 20.07 0.26 0.16

0.071 0.258 0.294 0.660 0.108 0.315

*S1 = session 1; S2 = session 2; ICC = intraclass correlation coefficient; SEM = standard error of measurement; CV = coefficient of variation; PPM = Pearson product moment correlation coefficient (correlation between the S2 and S1 absolute differences and the mean of S1 Max CMJ and S2 Max CMJ ); p value = probability value for the PPM; F = female; M = male; JJ = Just Jump. †S1 Max CMJ, S2 Max CMJ, and S2-S1 Diff. are reported as mean 6 SD. Limits of agreement are reported as the S2-S1 Diff. mean 6 (1.96 3 S2-S1 Diff. SD). ‡No significant differences between the maximum CMJ heights from S1 and S2 were detected.

2584

the

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

| www.nsca-jscr.org

TABLE 4. Intersession reliability of the log transformed data for females and males.*† S1 Ln CMJ

S2 Ln CMJ

3.445 3.620 3.418 3.891 4.032 3.811

3.437 3.633 3.429 3.890 4.036 3.812

F-Vertec F-JJ System F-Myotest M-Vertec M-JJ System M-Myotest

S2 2 S1 Diff. 20.008 0.013 0.011 20.001 0.003 0.001

6 0.122 6 0.062 6 0.057 6 0.091 6 0.085 6 0.075

Ratio limits of agreement

PPM

p value

0.99 (3/O1.27) 1.01 (3/O1.13) 1.01 (3/O1.12) 1.00 (3/O1.20) 1.00 (3/O1.18) 1.00 (3/O1.16)

0.16 0.04 0.01 20.42 0.15 20.03

0.339 0.817 0.917 0.006 0.358 0.903

*S1 = session 1; S2 = session 2; PPM = Pearson product moment correlation coefficient (correlation between the S2 and S1 absolute differences and the mean of S1 Ln CMJ and S2 Ln CMJ); F = female; M = male; JJ = just jump. †S1 Ln CMJ and S2 Ln CMJ are reported as the means. S2 2 S1 Diff. are reported as mean 6 SD. Ratio Limis of agreement are reported as the antilog of the S2-S1 Diff. mean 6 the antilog of (1.96 3 S2-S1 Diff. SD).

SD of the transformed differences is a ratio that represents the level of agreement. Because the difference between 2 log measurements is equivalent to the ratio between the same 2 measurements (loge[X1] 2 loge[X2] = log[X1/X2]), the limits of agreement of the transformed data are referred to as ratio limits of agreement (1,19). The ratio limits of agreement should then contain 95% of the observed ratios (19). Finally, comparisons of the measurements recorded from each of the 3 devices were made. The maximum CMJ height measurements from session 2 were compared using a 1-way repeated measures ANOVA with Tukey post hoc test.

RESULTS Reliability

The intrasession reliability results for males can be found in Table 1. The intrasession reliability results for females can be

TABLE 5. Frequency table describing how many subjects achieved maximum CMJ height during specified trials in the 2 testing sessions.*† Session 1

Session 2

Trial Trial Trial Trial Trial Trial 1 2 3 1 2 3 F-Vertec F-JJ System F-Myotest M-Vertec M-JJ System M-Myotest

17 12 21 13 11 16

7 8 9 9 11 10

14 18 8 18 18 14

10 9 11 9 11 13

11 16 17 19 14 11

18 14 11 12 15 16

*F = female; M = male; JJ = Just Jump. †It should be noted that some subjects achieved their maximum CMJ height on multiple trials. The data in this table represent the initial trial in which the maximum CMJ height was achieved.

found in Table 2. The intersession reliability results for males and females, before the logarithmic transformation, can be found in Table 3. The intersession reliability results for males and females, after the logarithmic transformation, can be found in Table 4. Heteroscedasticity

The differences between the maximal CMJ height measurements from session 1 and session 2 were considered normally distributed as the 3 devices had ratios of the skewness to the standard error of the skewness, which ranged from 21.30 to 1.20 in males and 0.29–1.03 in females. However, heteroscedasticity in the CMJ height measurement was ultimately accepted because the correlations between the CMJ heights and the absolute differences was greater than or equal to 0.16 in 5 out of the 6 conditions (female Vertec, male Just Jump System, female Just Jump System, male Myotest, and female Myotest) (Table 3). Log transformations were effective in reducing the level of heteroscedasticity (Table 4). Comparison of Devices

Differences between the maximum CMJ height measurements from the 3 devices were discovered. In males, the maximum CMJ height from the Just Jump System (57.3 6 9.0 cm) was significantly greater than that of both the Vertec (49.8 6 9.1 cm) and the Myotest (45.7 6 6.6 cm). Similarly, in women, the maximum Just Jump System (38.3 6 6.0 cm) CMJ height was significantly greater than both the Vertec (31.7 6 5.9 cm) and the Myotest (31.1 6 4.4 cm). There were no significant differences between the Vertec and Myotest CMJ heights for either sex.

DISCUSSION The purpose of this investigation was to assess the intrassession and intersession reliability of the Vertec, Just Jump System, and Myotest in measuring CMJ height. In both males and females, it was discovered that intrasession and intersession reliability of the CMJ height measurement was best VOLUME 25 | NUMBER 9 | SEPTEMBER 2011 |

2585

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Vertical Jump Reliability when measured by the Myotest. Additionally, heteroscedasticity was discovered with all 3 devices, suggesting that better jumpers demonstrate greater fluctuation in CMJ height measurements when the measurements are taken from 2 testing sessions separated by 24–48 hours. Reports on the intrassession reliability of the CMJ are scarce. In the current investigation, it was discovered that for both males and females, the Myotest demonstrated the best intrasession reliability, followed by the Just Jump System, and then the Vertec. This trend is evident when one considers the SEM and CV from each of these devices. Notably, the CV among females assessed with the Vertec (session 1 = 6.9% and session 2 = 7.6%) is much higher than the Just Jump System (session 1 = 4.4% and session 2 = 5.2%) and the Myotest (session 1 = 4.5% and session 2 = 4.6%) (Table 1). The CV for the female Vertec CMJ height is even noticeably higher when compared to that of males who were measured with the same device (session 1 = 4.6% and session 2 = 5.5%) (Table 2). This can most likely be attributed to systematic error, specifically learning, associated with the use of the Vertec. Systematic error was formally evaluated using the repeated measures ANOVA across the 3 trials, and from this analysis, it was revealed that regardless of sex, trial 3 CMJ heights were significantly higher than trial 1 and trial 2 CMJ heights. Although this systematic error was present in both males and females, the CVs indicate that it affected the consistency of the female measurements more than the male measurements. It is speculated that this finding may be attributed to the previous training experiences of the male and female subjects. Previous training experiences were not quantified in this investigation; however, verbal communications with the subjects before and after testing indicated that the males were more experienced with explosive vertical jumping, and several male subjects had previous experience using the Vertec. On the other hand, the female subjects tended to be less experienced with explosive vertical jumping, and few had previous experience using the Vertec. Table 5, which is a frequency table that indicates how many subjects achieved maximum CMJ height at specified trials, provides further insight into the CMJ height measurements across the 3 trials. Most noticeable from this table are the findings that many subjects do not attain their maximum CMJ height until the third trial. For example, when females were tested with the Vertec in the second session, 18 of 39 participants (46%) attained their maximum CMJ height on the third trial. This same trend can be identified with all 3 measurement devices for both sexes. These findings demonstrate the importance of administering a minimum of 3 trials when assessing the maximum CMJ in recreationally active males and females. As it pertains to intrassession systematic error, one limitation to the current testing protocol should be recognized. Before performing the 3 CMJ trials, subjects performed 2 warm-up CMJs in which the purpose was to provide the subject with a CMJ-specific warm-up involving

2586

the

maximal-effort explosive muscle contractions, and to provide the test administrators with an opportunity to correct any flaws in CMJ technique. However, in an effort to encourage subjects to simply focus on creating maximum lower-body power while maintaining the general CMJ technique requirements, they were not positioned underneath the Vertec device. Thus, subjects did not perform any Vertec-specific warm-up jumps. In retrospect, it may have been more appropriate for subjects to perform Vertec-specific warm-up jumps to provide further familiarization with the testing devices and to decrease the amount of systematic error across trials. Before to this investigation, the intersession reliability of a device similar to a Vertec (31) and the Just Jump System (16,17) had been assessed. Both devices demonstrated high ICCs and low CVs, thus implying an adequate level of intersession reliability (16,17,31). Young et al. (31) reported an ICC of 0.94 and CV of 3.8% when measuring CMJ height from a device similar to the Vertec. Moir et al. (16) reported an ICC of 0.93 and CV of 2.4% when measuring CMJ height with the Just Jump System, whereas Moir et al. (17) reported an ICC of 0.93 and CV of 4.0% in males, and an ICC of 0.90 and CV 6.0% in females. However, several differences in study methodologies may not make the results from the aforementioned studies directly comparable to the current investigation. For example, Young et al. (31) studied 17 male subjects across 4 testing sessions with a CMJ technique in which 1 arm contacted Vertec vanes that were separated by 1 cm. In comparison, the current investigation studied 40 male and 39 female subjects across 2 sessions with a CMJ technique in which both arms reached upward toward Vertec vanes that were separated by 1.27 cm. Furthermore, Moir et al. (16) studied 10 male subjects across 5 testing sessions and used a CMJ technique in which the hands were positioned on the hips, and Moir et al. (17) used a CMJ technique in which the hands were placed around the neck. Thus, because of the different sample sizes studied, and the different CMJ techniques implemented, further discussion and comparison to the above mentioned studies is not appropriate. Instead, a discussion pertaining to the results of the current study and the application of these results to the strength and conditioning field may be more beneficial. Because of the conflicting views and lack of established criteria regarding reliability analysis, interpreting the results from the current investigation is at present difficult. In the past, common criteria have included an ICC . 0.90 and a CV , 10% (1). These criteria are met in many of the conditions in the current investigation; however, some authors have cautioned against these criteria and have suggested interpreting the results within the analytical goal of the study (i.e., determining an acceptable amount of measurement error based on practical use) (1). Although the primary aim of this study was to simply quantify intrasession and intersession reliability to determine which of the 3 devices was most reliable, and not necessarily to determine an acceptable

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research amount of measurement error for each device, an effort will be made to put the study results into a practical context that future researchers can use to address CMJ reliability analyses further. In practice, these measurement devices should be sensitive enough to detect real changes in CMJ height after participation in a training program. If the intersession measurement error of a device is high, then the device may not be sensitive enough to detect a real change (i.e., the real change becomes lost within the error). For example, in a recent study involving elite male volleyball players, a 12month training program was found to elicit a 4-cm improvement in CMJ height (24). This amount of change is similar to that reported in less-trained subjects (2–6 cm) over shorter training periods (4,5,7,9,15,27,28). Thus, it would be the hope of the test administrator that there is ,4-cm error in CMJ height between testing sessions. To assess whether or not this was the case in the current investigation, an example is provided below. For females measured on the Just Jump System, the mean 6 SD for maximal CMJ height was 37.80 6 6.0 cm for session 1 and 38.25 6 6.0 cm for session 2, and no systematic error was detected (p = 0.267) between the 2 sessions. The ICC (0.92), SEM (1.7 cm), and CV (4.4%) were calculated. In the opinion of some researchers, these results alone may be indicative of sufficient intersession reliability. If one considers a female subject whose CMJ height from the Just Jump System was measured to be 30 cm during session 1, it might be assumed that their session 2 CMJ height would be 30 6 1.7 cm (i.e., somewhere between 28.3 and 31.7 cm). Therefore, it would appear that this device is sensitive enough to detect a 4-cm improvement in CMJ height. However, as has been explained elsewhere (1,26), the SEM only accounts for approximately 68% of the observed differences between the 2 sessions. As a result, some authors have suggested multiplying the SEM by 2 to establish a higher level of confidence (95%) (26). In this case, one could then expect the same female subject to have a session 2 CMJ height of 30 6 3.4 cm (i.e., somewhere between 26.6 and 33.4 cm). When compared to the limits of agreement of 0.45 6 4.74 cm (Table 3), 95% of the differences between maximum CMJ height measurements from session 1 and session 2 would lie between 24.29 and 5.19 cm (Figure 3). Thus, one could be 95% confident that the same female subject could score as low as 25.71 cm or as high as 35.19 cm during session 2. Based on these calculations, it appears that when the 95% confidence levels are applied to both the SEM and limits of agreement, the Just Jump System may, in some cases, not be sensitive enough to detect changes of 4 cm. Regardless of the confidence level applied, both the SEM and limits of agreement assume that the differences between session 1 and session 2 CMJ heights are normally distributed and that heteroscedasticity is not present (1). Although these differences did meet the expectations of normality based on the ratio of the skewness divided by the standard error of the skewness (1.02), a slight level of heteroscedasticity was

| www.nsca-jscr.org

detected as the correlation between the mean CMJ heights and the absolute differences was r = 0.19 (Figure 4; Table 3). Thus, the data were transformed using the natural logarithmic scale, and the mean 6 SD of the differences between session 1 and 2 was calculated to be 0.013 6 0.062 (Table 4). The transformation of the data was successful in reducing the level of heteroscedasticity because the correlation between the mean CMJ heights and the absolute differences was reduced to r = 0.04 (Table 4). Then, the ratio limits of agreement were calculated by taking the antilog of the mean of the differences (ex [0.013] = 1.01) and the antilog of the SD of the differences multiplied by 1.96 (ex [1.96 3 0.062] = 1.13) (Table 4). The ratio limits of agreement for females measured on the Just Jump System were 1.01 3/O 1.13 with 95% of the ratios contained within 0.89 and 1.14. Therefore, if a female subject’s CMJ height from the Just Jump System was 30 cm during session 1, her score in session 2 could, in the worst case scenario, be as low as 26.7 cm (30 cm 3 0.89) or as high as 34.2 cm (30 cm 3 1.14). The benefit of applying the ratio limits of agreement is that the limits vary depending on the performance level of each subject, and in a heteroscedastic data set; this seems to be appropriate. Unfortunately, the ratio limits of agreement, just like the SEM multiplied by 2, and the standard limits of agreement, still appear to be too wide for practical use even when the most reliable device in this study, the Myotest, is considered. Some authors have suggested that the 95% confidence limits are too stringent and that they may not be realistic when applied to some variables of athletic performance (10). Based on the results from this study, this viewpoint is supported, particularly when one considers that the practical purpose of the vertical jump test is to monitor the rather small changes (2–6 cm) in CMJ height over the course of a training program (4,5,7,9,15,24,27,28). One potential solution is to use confidence limits that are more practical, and to accomplish this, Hopkins (10) has suggested simply dividing the limits of agreement by 2 (10). If this technique is performed with the above limits of agreement of 24.29 and 5.19 cm, the limits would then be 22.15 and 2.60 cm. With less confidence, we could then expect the same female subject to score as low as 27.85 cm or as high as 32.60 cm during session 2. Whether or not this same procedure would be appropriate to apply to the ratio limits of agreement of a heteroscedastic data set is yet to be determined. Although much debate exists regarding the most appropriate statistics to interpret in reliability analyses, it may be more beneficial to consider ways in which study designs can be changed to improve reliability in general. By establishing well-designed studies, researchers may be able to provide recommendations regarding acceptable amounts of measurement error based on the practical use of the measurement. In retrospect, we are able to recognize 4 study design factors, which could have helped to narrow the 95% confidence limits from the current investigation: further familiarization, additional CMJ trials, a more specific subject VOLUME 25 | NUMBER 9 | SEPTEMBER 2011 |

2587

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Vertical Jump Reliability sample, and a larger sample size. Although no significant differences between the means of the maximum CMJs from session 1 and 2 were detected for either sex, some individual score differences were, for example, as high as 9 cm on the Vertec. It is reasonable to assume that further familiarization with the CMJ technique during session 1, or via a completely separate familiarization session, would have further assisted in obtaining more consistent scores, particularly in the less-trained subjects. The current investigation also employed the minimal number of trials that have been recommended in reliability studies (10). Based on the data presented in Table 5, it is presumed that some subjects may have continued to improve their CMJ height beyond the third trial. The completion of .3 trials may have been more appropriate. In addition, the current investigation assessed recreationally active subjects because of the convenience of recruiting such a sample. Unfortunately, a recreationally active subject sample is fairly vague and may include individuals with extensive experience in vertical jumping (e.g., recreational basketball and volleyball players), while at the same time include individuals with little or no experience with vertical jumping (e.g., recreational joggers and cyclists). A more specific subject sample, such as only recreational basketball players, may have assisted in narrowing the limits and reducing the amount of heteroscedasticity. Finally, the size of the subject samples in the current investigation was at the minimum level recommended for reliability studies (1,10). An increase in the sample size (.40 subjects) would have presumably helped narrow the confidence limits as well. Because of these discussed study design limitations, we are unable to confidently determine if any of the 3 devices are sensitive enough to detect real changes in CMJ height after participating in a training program. It can, however, be concluded that of the 3 devices in this study, the Myotest was the most reliable and the Vertec was the least reliable (both intra and intersession reliability). One of the primary issues with attaining reliable CMJ heights from the Vertec is that it requires much more skill from the participant. The participant must be able to coordinate the arm swing such that the arms are fully extended and in contact with the vanes at the moment that the participant has attained their greatest displacement from the floor. Another area of concern with the Vertec is the sensitivity of its measurements. The CMJ heights from the Vertec can only be measured in 1.27-cm increments because of the spacing between the vanes. Thus, if the tips of the fingers are in the space between 2 vanes, the jump-and-reach height is erroneously reflected as the highest vane contacted rather than the potentially quantifiable area between the 2 vanes. Furthermore, it is important to remember that attaining a jump height from the Vertec involves 2 measurements (the standing reach height and the jump-and-reach height). Both of these measurements could be a potential source of error in the attainment of the final CMJ height.

2588

the

The primary concern with attaining reliable CMJ height measurements from the Just Jump System is the manipulation of flight time because of the landing technique. Technique alterations such as simultaneous exaggerations of hip flexion, knee flexion, and ankle dorsiflexion would have the greatest potential to artificially increase the flight time and the subsequent jump height. The Myotest device does not appear to be effected by the same limitations as the Vertec and Just Jump System, and this is most likely the reason for its superior reliability. However, the Myotest does require that an accurate body mass be entered into the device before jumping to calculate force. Because the force–time curve is used to calculate the flight time, an inaccurate body mass measurement before jumping could affect the reliability of the Myotest across testing sessions. This study appears to be the first to attempt to quantify the level of heteroscedasticity in the CMJ height measurement. Heteroscedasticity in a data set occurs when higher performers demonstrate more variability in scores across testing sessions when compared to lower performers. One previous investigation has discovered that this phenomenon is quite frequent in sports science measurements that are recorded on the ratio scale (e.g., back strength, leg strength, grip strength) (19). In the current investigation, it was discovered that the only time heteroscedasticity was not present was when males were measured with the Vertec. This was evident from the weak, negative correlation (r = 20.07) between the mean CMJ heights and the absolute differences (Table 3). The reasoning behind this specific finding is speculative at this point, but it may be related to the movement coordination associated with the Vertec. Measuring CMJ height from the Vertec requires that the subject forcefully extend at the hips, knees, and ankles to leave the ground, and then while in the air, swing the arms overhead to contact the vanes at the peak of the jump. In the male sample, the higher vertical jumpers are assumed to have had an initially high level of coordination associated with this movement pattern. This coordination may have been because of specific training experiences in the past (e.g., volleyball spiking, basketball rebounding) along with previous experience using the Vertec or other jump-andreach assessments. Thus, the likelihood of a learning effect from session 1 to session 2 in the higher vertical jumpers is presumed to be small for the males. As a result, the higher vertical jumpers demonstrated their typical level of variability across sessions. On the other hand, the lower male vertical jumpers are assumed to have had an initially lower level of coordination between the jump and arm swing, and thus, there was a greater potential for a learning effect from session 1 to session 2. In other words, the normally low variability observed in the scores of lower performers was most likely inflated because of a potential learning effect associated with using the Vertec. In fact, many of the lower jumpers demonstrated a positive change in their CMJ scores during session 2. These differences from learning in the lower vertical jumpers then appear to be about the same as the

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research typical absolute intersession differences that are seen in higher vertical jumpers. Consequently, when males were assessed with the Vertec, heteroscedasticity was not present. However, when considering females, this same tendency most likely did not occur when measured with the Vertec (r = 0.29) because the vast majority, regardless of whether they were a higher jumper or a lower jumper, had limited or no previous experience coordinating such a movement pattern. Although the primary aim of this investigation was to assess the reliability of the vertical jump testing devices, it was also of some value to determine whether or not the devices recorded similar CMJ heights. In both males and females, it was discovered that the CMJ heights recorded from the Just Jump System were significantly higher than those recorded from both the Vertec and the Myotest. Because of the absence of a criterion device for measuring CMJ height (e.g., motion analysis system), it cannot be determined which device provided the most valid measurement. One previous study found the Just Jump System to be a more valid tool for measuring CMJ height than the Vertec (14). However, it is important to understand that the Just Jump System and Vertec are measuring the displacements of different biomechanical constructs, and thus, it is not surprising to see differences in CMJ height between the 2 devices. The Just Jump System estimates CMJ height based on flight time, which is associated with the displacement of the jumper’s center of mass. Conversely, the Vertec measures the difference between the displacements of the standing reach height and the jump-and-reach height. Therefore, it is not surprising to find that the CMJ height from the Just Jump System is most similar to that of a motion analysis system in which a reflector is placed near the body’s center of mass (14). Although both the Just Jump System and the Myotest use flight time to determine the CMJ height, these times are attained differently. It is suspected that these different methodologies are the source of the difference between the Just Jump System and the Myotest CMJ heights. The lack of a significant difference between the Vertec and Myotest appears to be merely coincidental. Further research is needed to validate the Myotest against criterion measures to determine which biomechanical construct of displacement it most closely mimics. In conclusion, the results from this study indicate that in recreationally active males and females, intrasession and intersession reliability of the CMJ height measurement is best when measured by the Myotest. For researchers who seek to continue to investigate the reliability of the CMJ, 4 recommendations are being offered based on this study findings: (a) extensive familiarization with the CMJ technique, particularly when using the Vertec, (b) a minimum of 3 maximal-effort trials or until performance plateaus, (c) specific subject samples, and (d) large sample sizes (.40 subjects). Adherence to these recommendations may help in decreasing the amount of intrasession and intersession systematic error and may also help in decreasing the level of heteroscedasticity

| www.nsca-jscr.org

in the CMJ measurement. In addition, these recommendations may help in establishing an acceptable level of measurement error for devices used to measure CMJ height.

PRACTICAL APPLICATIONS Strength and conditioning coaches can improve the reliability of CMJ height measurements by reducing systematic error. This can be accomplished by thoroughly familiarizing athletes with the CMJ technique and by permitting the athletes to complete a minimum of 3 trials. Thorough familiarization, via a separate familiarization session, is recommended specifically when the CMJ height will be measured from the Vertec. It is also important to recognize the type of athlete and the training status of the athlete being tested when determining how much familiarization is necessary. Athletes who are unfamiliar with explosive CMJs, and who are overall less trained, should undergo the greatest amount of familiarization. Additionally, it is recommended that strength and conditioning coaches permit their athletes an opportunity to perform a minimum of 3 trials or until performance plateaus, and from those values, report the maximal CMJ height. Specifically, in the case of the Vertec, the completion of more than 3 trials is strongly recommended. Differences in the CMJ height values from the various devices were identified in this investigation, thus, strength and conditioning coaches are reminded not to use these devices interchangeably. The device used in pretesting should also be used in any subsequent posttesting. Although the Myotest was found to be the most reliable of the 3 devices, its use cannot be fully recommended until further research assesses the validity of its measurements. When determining whether to use the Vertec or Just Jump System, the strength and conditioning coach should consider (a) the sex of the athlete that they are testing, and (b) the displacement construct that they are most interested in measuring. If the athlete is a male, the coach can use either the Vertec or the Just Jump System because both devices have about the same level of intersession reliability. The coach would then need to determine which construct they would rather measure (i.e., if interested in quantifying the difference between the standing reach height and the jump-and-reach height, use the Vertec; if interested in quantifying the displacement of the center of mass displacement, use the Just Jump System). However, for females, the Just Jump System is noticeably more reliable than the Vertec. Thus, the use of the Just Jump System instead of the Vertec is recommended until the female athlete becomes more familiar with the Vertec. It is important to remember that these recommendations may not be applicable to all athletic populations as the current investigation only studied recreationally active college students.

ACKNOWLEDGMENTS We would like to acknowledge the Teaching, Learning, and Technology Roundtable of Slippery Rock University for the funding that made this research possible. In addition, we would VOLUME 25 | NUMBER 9 | SEPTEMBER 2011 |

2589

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Vertical Jump Reliability like to thank Scott Boyer, Casey Braendle, Nick Marts, and Kelly Stenger for their assistance in the data collection process.

REFERENCES

16. 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 males. J Strength Cond Res 18: 276–280, 2004.

1. Atkinson, G, and Nevill, A. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med 26: 217–238, 1998.

17. Moir, G, Shastri, P, and Connaboy, C. Intersession reliability of vertical jump height in males and females. J Strength Cond Res 22: 1779–1784, 2008.

2. Black, W, and Roundy, E. Comparison of size, strength, speed, and power in NCAA division 1-A football players. J Strength Cond Res 8: 80–85, 1994.

18. Nevill, AM. Why the analsyis of performance variables recorded on a ratio scale will invariably benefit from a log transformation. J Sports Sci 15: 457–458, 1997.

3. Bland, JM, and Altman, DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: 307–310, 1986.

19. Nevill, AM, and Atkinson, G. Assessing agreement between measurements recorded on a ratio scale in sports medicine and sports science. Br J Sports Med 31: 314–318, 1997.

4. Cormie, P, McCaulley, GO, and McBride, JM. Power versus strength-power jump squat training: influence on the load-power relationship. Med Sci Sports Exerc 39: 996–1003, 2007.

20. Pijnappels, M, van de Burg, JCE, Reeves, ND, and van Dieen, JH. Identification of elderly fallers by muscle strength measures. Eur J Appl Physiol 102: 585–592, 2008.

5. Fatouros, IG, Jamurtas, AZ, Leontsini, D, Taxildaris, K, Aggelousis, N, Kostopoulos, N, and Buckenmeyer, P. Evaluation of plyometric exercise training, weight training, and their combination on vertical jumping performance and leg strength. J Strength Cond Res 14: 470–476, 2000.

21. Riddiford-Harland, DL, Steele, JR, and Baur, LA. Upper and lower limb functionality: are these compromised in obese children? Int J Pediatr Obes 1: 42–49, 2006.

6. Fry, AC, and Kraemer, WJ. Physical performance characteristics of American collegiate football players. J Appl Sport Sci Res 5: 126–138, 1991. 7. Gehri, DJ, Ricard, MD, Kleiner, DM, and Kirkendall, DT. A comparison of plyometric training techniques for improving vertical jump ability and energy production. J Strength Cond Res 12: 85–89, 1998. 8. Harman, EA, Rosenstein, MT, Frykman, PN, Rosenstein, RM, and Kraemer, WJ. Estimation of human power output from vertical jump. J Appl Sport Sci Res 5: 116–120, 1991. 9. Holcomb, W, Lander, JE, Rutland, RM, and Wilson, GD. The effectiveness of a modified plyometric program on power and the vertical jump. J Strength Cond Res 10: 89–92, 1996. 10. Hopkins, WG. Measures of reliability in sports medicine and science. Sports Med 30: 1–15, 2000. 11. Hopkins, WG. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 41: 3–12, 2009. 12. Isaacs, LD. Comparison of the vertec and Just Jump Systems for measuring height of vertical jump by young children. Percept Mot Skills 86: 659–663, 1998. 13. Klavora, P. Vertical-jump tests: a critical review. Strength Cond J 22: 70–75, 2000. 14. Leard, JS, Cirillo, MA, Katsnelson, E, Kimiatek, DA, Miller, TW, Trebincevic, K, and Garbalosa, JC. Validity of two alternative systems for measuring vertical jump height. J Strength Cond Res 21: 1296– 1299, 2007. 15. 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.

2590

the

22. Sawyer, DT, Ostarello, JZ, Suess, EA, and Dempsey, M. Relationship between football playing ability and selected performance measures. J Strength Cond Res 16: 611–616, 2002. 23. Sayers, SP, Harackiewicz, DV, Harman, EA, Frykman, PN, and Rosenstein, MT. Cross-validation of three jump power equations. Med Sci Sports Exerc 31: 572–577, 1999. 24. Sheppard, JM, Chapman, DW, Gough, C, McGuigan, MR, and Newton, RU. Twelve-month training-induced changes in elite international volleyball players. J Strength Cond Res 23: 2096–2101, 2009. 25. Stuempfle, KJ, Katch, FI, and Petrie, DV. Body composition relates poorly to performance tests in NCAA division III football players. J Strength Cond Res 17: 238–244, 2003. 26. Thomas, JR, Nelson, JK, and Silverman, SJ. Research Methods in Physical Activity (5th ed.). Champaign, IL: Human Kinetics, 2005. 27. Tricoli, V, Lamas, L, Carnevale, R, and Ugrinowitsch, C. Short-term effects of lower-body functional development: Weightlifting vs. vertical jump training programs. J Strength Cond Res 19: 433–437, 2005. 28. Turner, AM, Owings, M, and Schwane, JA. Improvement in running economy after 6 weeks of plyometric training. J Strength Cond Res 17: 60–67, 2003. 29. Vincent, WJ. Statistics in Kinesiology (3rd ed.). Champaign, IL: Human Kinetics, 2005. 30. Weir, JP. Quantifying test–retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res 19: 231–240, 2005. 31. Young, W, Macdonald, C, Heggen, T, and Fitzpatrick, J. An evaluation of the specificity, validity and reliability of jumping tests. J Sports Med Phys Fitness 37: 240–245, 1997.

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.