Cormie et al 2007 Validation of Power Measurement Techniques in Dynamic

103

The authors are with the Neuromuscular Laboratory, Depart-ment of Health, Leisure and Exercise Science, Appalachian State University, Boone, NC.

Journal of Applied Biomechanics, 2007; 23:103-118. © 2007 Human Kinetics, Inc.

Validation of Power Measurement Techniques in Dynamic Lower Body Resistance Exercises

Prue Cormie, Jeffrey M. McBride, and Grant O. McCaulleyAppalachian State University

The objective of this study was to investigate the validity of power measurement techniques utilizing various kinematic and kinetic devices during the jump squat (JS), squat (S) and power clean (PC). Ten Division I male athletes were assessed for power output across various intensi-ties: 0, 12, 27, 42, 56, 71, and 85% of one rep-etition maximum strength (1RM) in the JS and S and 30, 40, 50, 60, 70, 80, and 90% of 1RM in the PC. During the execution of each lift, six different data collection systems were utilized; (1) one linear position transducer (1-LPT); (2) one linear position transducer with the system mass representing the force (1-LPT+MASS); (3) two linear position transducers (2-LPT); (4) the force plate (FP); (5) one linear position transducer and a force plate (1-LPT+FP); (6) two linear position transducers and a force place (2-LPT+FP). Kinetic and kinematic variables calculated using the six methodologies were compared. Vertical power, force, and velocity differed significantly between 2-LPT+FP and 1-LPT, 1-LPT+MASS, 2-LPT, and FP meth-odologies across various intensities throughout the JS, S, and PC. These differences affected the load–power relationship and resulted in the transfer of the optimal load to a number of different intensities. This examination clearly indicates that data collection and analysis pro-cedures influence the power output calculated as

well as the load–power relationship of dynamic lower body movements.

Key Words: optimal load, force, velocity, jump squat, power clean

There are many investigations on power output during lower body resistance exercises (Baker et al., 2001; Bourque & Sleivert, 2003; Esliger & Sleivert, 2003; Haff et al., 1997; Izquierdo et al., 1999, 2001, 2002; Kawamori et al., 2005; McBride et al., 1999; Siegel et al., 2002; Sleivert et al., 2004; Stone et al., 2003; Winchester et al., 2005). Specifically, these investigations have attempted to determine at what load (typically described as a percentage of 1RM) power output is maximized (Table 1). For example, the load that elicits the greatest power output in the jump squat varies from 0% of 1RM (McBride et al., 1999) to approximately 60% of 1RM (Baker et al., 2001; Esliger & Sleivert, 2003; Sleivert & Taingahue, 2004). Such large disparities in the optimal load reported in the current research have led to ambiguity surrounding the load–power relationship. As a result, it is difficult to draw any conclusions from the literature and form specific training guidelines for the development of muscular power. Although varying subject populations used in different studies may be partially responsible for the observed discrepancies in the literature, the methodology used for determination of power has also been suggested to be a contributing factor (Dugan et al., 2004).

104 Cormie, McBride, and McCaulley

Throughout the literature a number of various apparatus have been utilized for collecting and ana-lyzing power data. One commonly reported method is the calculation of power through displacement data, which, in most cases, is collected by means of a single linear position transducer (LPT) (Alemany et al., 2005; Baker et al., 2001; Baker & Nance, 1999; Bourque and Sleivert, 2003; Cronin & Henderson, 2004, Esliger & Sleivert, 2003, Falvo et al., 2005; Izquierdo et al., 1999, 2001, 2002; Rahmani et al., 2001; Siegel et al., 2002; Thomas et al., 1996; Weiss et al., 2004, 2005). Double differentiation of displacement data permits the determination of acceleration, which is then coupled with the system mass to calculate force output. The force–time curve is then multiplied times the velocity–time curve derived from the displacement data, resulting in a

power–time curve for the movement analyzed.The use of a force plate to calculate power

from vertical ground reaction forces is another common method utilized in power research (Dele-cluse et al., 2005; French et al., 2004; Haff et al., 1997; Iossifidou et al., 2005; McBride et al., 1999, 2002; Sands et al., 2005). This modality relies on the impulse–momentum relationship to determine velocity and allows for power to be calculated through a forward dynamics approach. However, the force plate methodology has typically been utilized to compare performance among various types of body weight jumps or to monitor vertical jump performance following an intervention rather than examining the influence of various loads on power output.

Recent investigations have compared force and

Table 1 Comparison of Literature Reporting the Optimal Load in the JS, S, and PC.

Validation of Power Measurement Techniques 105

power variables calculated using the single LPT and force plate methodologies (Chiu et al., 2004; Cronin & Henderson, 2004; Hori et al., 2005; Rahmani et al., 2001). Hori et al. (2005) cited significant dif-ferences in peak force, velocity, and power during the hang snatch when comparing the use of a single LPT to calculations derived from a force plate. In contrast, some comparisons between displacement-based modalities (i.e., LPT) and force plates have reported no difference in force data during the JS (Chiu et al., 2004) and body weight jumps (Cronin & Henderson, 2004), as well as power data during concentric half squats (Rahmani et al., 2001) when derived by the two methods. However, methodolo-gies that rely solely on either kinematic data (1-LPT) or kinetic data (FP) have yet to be compared to a methodology relying on both kinetic and kinematic variables (i.e., utilizing both an LPT and a force plate; 1-LPT+FP). This technique has been utilized frequently in research examining power across vari-ous intensities in the jump squat (Lyttle et al., 1996; McBride et al., 1999, 2002; Wilson et al., 1993) and bench press throw (Lyttle et al., 1996; Newton et al., 1996). Based on data from this laboratory currently in press, differences exist in power output during a jump squat at both 30 and 90% of 1RM when com-paring calculations from 1-LPT versus 1-LPT+FP methods. These recent findings oppose previous validation studies comparing displacement-based systems to force plate power outputs (Chiu et al., 2004; Cronin & Henderson, 2004; Rahmani et al., 2001). Furthermore, there is a paucity of research examining the ability of a single LPT and other methodologies to determine power output across a loading spectrum in various types of exercises involving both horizontal and vertical motion.

The inability of methodologies utilizing a single LPT to account for horizontal movements in the calculation of power and the influence on the load–power relationship is unknown. A two-dimensional displacement and force system involves the use of two LPTs and a force plate (2-LPT+FP). The two LPTs and the bar form a triangle in which the dimen-sions of the triangle are known, permitting both vertical and horizontal movements to be included in the calculation of vertical velocity (Figure 1). The system is advantageous to the single LPT technique because it tracks displacement in the vertical and the horizontal planes, allowing for a more accurate calculation of power. Although previous research

has not utilized this methodology, the use of two LPTs and a force plate is hypothesized to represent a more accurate measure of displacement, velocity, and power in multidimensional movements.

Although many methods for calculating power during dynamic lower body resistance exercises exist, information pertaining to the most effec-tive method of measuring power output in athletic movements is not available. The standardization of data collection and analysis procedures used to determine muscular power is vital for the replica-tion of experiments and thus necessary to gaining an enhanced understanding of power output. Therefore the purpose of this experiment is to examine the dif-ference between each of the major methodologies currently used while validating these techniques to a new methodology utilizing two LPTs and a force plate.

Methods

Subjects

Ten Division I male athletes participated in this study (age: 20.00 ± 1.49 years; height: 178.03 ± 4.68 cm; weight: 88.75 ± 15.09 kg; body fat: 10.54 ± 4.44%; full S 1RM: 171.00 ± 22.05 kg; PC 1RM: 112.75 ± 13.15 kg; S 1RM–to–body weight ratio: 1.95 ± 0.23). The sample comprised football and track and field athletes who had a minimum of 4 years of resistance training experience and displayed competent technique in the three lifts examined. The participants were notified about the potential risks involved and gave their written informed consent, approved by the Institutional Review Board at Appalachian State University.

Study Design

Participation in this investigation involved a total of four testing sessions. Maximal strength was assessed during the first session through 1RM testing in the squat and the power clean by performing a series of warm-up sets and several maximal lift attempts until a 1RM for each respective lift was obtained. This protocol has been used extensively in our laboratory (Winchester et al., 2005). The 1RM testing was followed by an extensive familiarization session in which subjects became accustomed to performing the lifts explosively throughout the loading spec-trum. After at least 72 hours of recovery, subjects


completed three exercise-specific testing sessions in a randomized order: (a) jump squat session (JS), (b) squat session (S), and (c) power clean session (PC). Each of these testing sessions involved the comple-tion of the respective lift at seven intensities: 0, 12, 27, 42, 56, 71, and 85% S 1RM for the JS and S and 30, 40, 50, 60, 70, 80, and 90% PC 1RM for the PC.

Because body mass must be moved in addition to the external load during the JS and S, the resulting force, velocity, and power were determined by the athlete’s ability to accelerate the total system mass (i.e., external load + body mass). However, the lower leg and feet remain relatively static during the phase of the lift at which peak power typically occurs, spe-

Figure 1 — The procedure used in determining displacement from two LPTs (a—LPT A displacement; b—LPT B displacement; c—distance between the two LPTs; x—horizontal displacement; y—vertical displacement).


cifically just before takeoff. As a result, the mass of the shanks (12% of body mass; Piscopo and Bailey, 1981) were excluded from the system mass used in determining the loading conditions and subsequent power calculations (Dugan et al., 2004). Therefore, loading conditions were determined as the sum of a percentage of 1RM plus the subject’s body mass minus their shank mass (Dugan et al., 2004; McBride et al., 2002). Subjects performed trials in a randomized fashion with 30, 40, 50, 60, 70, 80, and 90% of maximal dynamic strength (MDS = 1RM + [body mass − shank mass]). These loading condi-tions correspond to 0, 12, 27, 42, 56, 71, and 85% of 1RM in the JS and S. In contrast, the PC involves the bar moving independently of the body so the peak force, velocity, and power of the movement is dependent on the subject moving the external load applied to the barbell. Therefore the loading conditions of the PC—30, 40, 50, 60, 70, 80, and 90% of 1RM—differ from those of the JS and S. The order of loading within each testing session was randomized with at least 3 days of recovery required between sessions. Subjects were instructed to move the resistance as explosively as possible in an attempt to achieve maximal power output with each trial (Behm and Sale, 1993). A minimum of two trials at each load were completed, with additional trials performed if incorrect form or submaximal effort was used. Adequate rest was permitted between all trials (3 min).

Data Collection Procedures

All testing (1RM, JS, S, and PC) was performed with the subjects standing on a force plate (AMTI BP6001200; Watertown, MA) with the left and right side of the barbell attached to three LPTs (Celesco PT5A-150; Chatsworth, CA). The data from the right side of the bar was used for analysis. The combined retraction tension of the six LPTs was 49.1 N; this was accounted for in all subsequent calculations. Analog signals from the force place and three LPTs (right side of the bar) were collected for every trial at 1,000 Hz using a BNC-2010 interface box with an analog-to-digital card (National Instru-ments PCI-6014; Austin, TX). LabVIEW (National Instruments, Version 7.1) was used for recording and analyzing the data. Signals from the three LPTs and the force plate as well as data derived using double differentiation underwent rectangular smoothing

with a moving average half-width of 12. From laboratory calibrations, the LPTs and force plate voltage outputs were converted into displacement and vertical ground reaction force, respectively.

Data Analysis Procedures

Vertical force, velocity, and power from every trial was calculated by six methodologies (1-LPT, 1-LPT+MASS, 2-LPT, FP, 1-LPT+FP, 2-LPT+FP; see explanation below). Peak force, velocity, and power were determined as the maximal value achieved during the entirety of the concentric phase of the lift. Average force, velocity, and power were calculated from the beginning of the concentric phase to the time at which peak vertical force, velocity, or power occurred in the JS and S. Similarly, average values for the PC were determined from the beginning of the second pull to the time at which peak force, velocity, or power occurred. The reliability of the six methods was assessed through comparison of peak power between two trials at the optimal load for the JS, S, and PC (0, 56, and 80% 1RM, respec-tively). All methods displayed intraclass correlation coefficients (ICC) above the minimum acceptable criterion of 0.70 (Baumgartner and Chung, 2001) and were significant at an alpha level of 0.05 (1-LPT - JS = 0.87, S = 0.96 and PC = 0.76; 2-LPT - JS = 0.83, S = 0.90 and PC = 0.83; 1-LPT+FP - JS = 0.96, S = 0.95 and PC = 0.99; 2-LPT+FP - JS = 0.95, S = 0.94, PC = 0.98; 1-LPT+MASS - JS = 0.90, S = 0.98, PC = 0.97; FP - JS = 0.94, S = 0.86, PC = 0.88).

Kinematic Methods

1-LPT. Kinetic and kinematic variables were calculated from displacement data through double differentiation processes. Bar displacement was measured using one of the three LPTs mounted directly above the subject on a custom-built power rack. The LPT produced a voltage signal representative of the degree at which the LPT was extended allowing for displacement–time data to be calculated (Baker et al., 2001; Chiu et al., 2004; Cronin et al., 2004; McBride et al., 2002). Instantaneous vertical velocity (v) of the bar throughout the movement was calculated using the displacement (x) and time (t) data at each sample, (v = ∆x/∆t). The body and the bar moved together as a unit in the JS and S; therefore, it was assumed that the bar velocity calculated was equivalent


to that of the entire system (Dugan et al. 2004). Acceleration of the system (a) was calculated using a second-order derivative of the displacement data (a = ∆x/∆t2). Force (F) produced during the lift was determined by adding the acceleration of the system and acceleration resulting from gravity (a

g = 9.81m/s2) and then multiplying the total

acceleration to the mass of the system (SM = external load + body mass), F = SM ⋅ (a + a

g). Following

these calculations, power (P) was determined by multiplying force and velocity at each time point (P = F ⋅ v) (Dugan et al. 2004).

1-LPT+MASS. The procedure utilized to calcu-late power through the 1-LPT+MASS method has been described previously (Jennings, 2005; Coelho et al., 2003a, 2003b). Processes outlined in the 1-LPT method were used to determine the velocity of movements evaluated. The unique aspect of this measurement system is the manner in which force is accounted for. Force was equivalent to the product of the system mass and acceleration due to gravity (F = SM ⋅ a

g), thus making force a constant throughout

the movement measured. Subsequently, power was reported as the product of this force multiplied by the velocity–time curve.

2-LPT. Utilizing two LPTs allows for both verti-cal and horizontal movements to be measured and used to determine the vertical displacement of the bar. The LPTs where mounted above and anterior (LPT A), and above and posterior to the subject (LPT B), forming a triangle when attached to the barbell (Figure 1). Combining known displacements with the displacement measurements from LPT A and B in the equation defined in Figure 1 allows for the calculation of vertical displacement, simultane-ously accounting for any horizontal movement of the bar as well. Identical procedures for calculating of velocity, acceleration, force, and power using the 1-LPT were applied to the displacement data derived from 2-LPT.

Kinetic Methods

FP. The FP methodology has been used extensively throughout the literature to evaluate power produced during dynamic movements, especially the vertical jump (Delecluse et al. 2005, French et al. 2004, Ios-sifidou et al. 2005, McBride et al. 1999, McBride et al. 2002, Sands et al. 2005, Haff et al. 1997). The

FP method can be used to determine power output because the initial vertical velocity of the system was always zero. Vertical ground reaction forces (F) were divided by the mass of the system (SM) at each time point (i) in order to determine acceleration (a), a

(i) = F

(i)/SM. Acceleration due to gravity was

subtracted from the calculated acceleration data to ensure that only the acceleration produced by the subject was used to determine velocity. Instanta-neous vertical velocity of the system’s center of mass was determined by the product of acceleration and time data at each data point: v = ∆a ⋅ ∆t, where ∆a = a

(i−1) − a

(i) and ∆t = t

(i−1) − t

(i). The derived velocity

data was combined with the original force values in order to calculate the power output of each lift.

Kinematic–Kinetic Methods

1-LPT+FP. This methodology involved the com-bination of displacement data and vertical ground reaction forces for calculation of power. Displace-ment data were collected via a single LPT while a force platform measured the vertical force output of the system. The velocity of the system was deter-mined using a first-order derivative of the displace-ment data. Power was calculated from the product of the velocity and force data.

2-LPT+FP. In an attempt to account for horizon-tal movement affecting the vertical displacement measured during dynamic multidimensional move-ments, two LPTs were utilized (Figure 1). Signals from each LPT were combined to determine verti-cal displacement, which was subsequently used in combination with time to calculate velocity (Figure 1). Power was calculated by coupling this velocity with force data collected through the use of a force plate.

Statistical Analyses

Repeated measures ANOVA was used to determine whether significant differences in vertical velocity, force, and power existed between the methodolo-gies. Additional comparisons were made between these techniques under the different loading con-ditions in order to determine their impact on the load–power relationship. Analyses were followed by Bonferroni post hoc comparisons. The criterion level for all analyses was defined by p ≤ 0.05.


Results

The load–power relationship during the JS was simi-lar for 1-LPT, 2-LPT, FP, 1-LPT+FP, and 2-LPT+FP methodologies, with the optimal load occurring at 0% of 1RM and power output decreasing linearly as intensity was increased (Figure 2, Table 2). Owing to the fact that the 1-LPT+MASS consistently under-valued force (Figure 4) and consequentially power output (Figure 3), the nature of the load–power relationship was changed (Figure 2). The optimal load was identified at 42% of 1RM, and no differ-ences in peak power were observed between this load and any other intensity (Table 3). Although the nature of the load–power relationship was not affected by the 1-LPT and 2-LPT methods (Figure 2, Table 2), these techniques significantly elevated force output (Figure 4) and subsequently overvalued power output across various loads (Figure 3).

The load–power relationship of the S varied considerably between methods, including kinetic data (FP, 1-LPT+FP and 2-LPT+FP) and kinematic-only methods (1-LPT, 1-LPT+MASS, and 2-LPT) (Figure 2, Table 3). A variety of optimal loads rang-ing from 0 to 71% of 1RM were identified using the various methodologies (Table 3). Similar to JS results, 1-LPT+MASS consistently undervalued force and power output, whereas 1-LPT and 2-LPT methods overestimated force and power output (Figure 3 and 4).

Similar to the squat, the load–power curve of the PC shifted between methods, including kinetic data and kinematic-only methods (Figure 2, Table 2). The optimal load was reported at 80% of 1RM for all methods, except 2-LPT and 1-LPT in which 30% of 1RM was identified as the optimal load (Table 3). These changes stem from an overvaluation of force output derived by these methods (Figure 4). Utilizing 1-LPT+MASS and FP analysis tech-nique resulted in significantly lower power output throughout the loading spectrum (Figure 3). These changes reflected the undervaluation of force data with 1-LPT+MASS and velocity data derived from the FP (Figure 4 and 5).

Discussion

The current investigation establishes that power values change depending on the data collec-tion and analysis procedures used. Specifically,

methodologies relying solely on kinematic data consistently elevated (1-LPT, 2-LPT) as well as underestimated (1-LPT+MASS) power output across various loads in the JS, S, and PC. Further-more, the FP technique under-represented velocity and power output during the three lifts examined and especially during movements involving the bar traveling independently of the body (PC). The key finding of this investigation was that these misrep-resentations affected the load–power relationship in the JS, S, and PC.

The major concern with the 1-LPT and 2-LPT methodologies is the necessity for extensive data manipulation. These systems utilize a forward dynamics approach to determine force from dis-placement data. Even though these calculations are based on sound mathematical principles, the method-ologies are restricted through the collection of kine-matic data to determine the power output of dynamic movements. Inherent in such manipulations are the amplification of noise and the associated possibility of compromising the integrity of the derived data set (Wood, 1982). As a result, these methodologies reliant on double differentiation procedures involve many different filtering and smoothing operations. Therefore, there is an increased potential for error in the calculated variables, which affects the validity of the derived power values (Dugan et al., 2004). An additional disadvantage associated with methods relying solely on kinematic data is the inability to account for the movement of the body that occurs independently of the barbell, an important factor in weightlifting (Garhammer, 1993). Displacement-based systems track only bar displacement. As a consequence, the force–time curve during the PC is determined independent of the action of the body that is producing the force acting on the bar. Thus, the resulting force and power calculations are rep-resentative of the bar and not the entire system; this contributes to the elevated power values observed in the PC (Hori et al. 2005). These factors may explain the overestimation of power output with 1-LPT and 2-LPT methodologies.

The major concern with the 1-LPT+MASS method revolves around the means by which force is derived. This method may be flawed because force is determined without accounting for acceleration of the system mass (Baker & Nance, 1999; Coelho et al. 2003a, 2003b; Jennings et al., 2005), a process that rejects standard biomechanical procedures

110

Figure 2 — Load–power relationship in the JS (A), S (B), and PC (C) as derived by the six methods examined (significance reported in Table 2).

111

Figure 3 — Relative difference in peak power (A) and average power (B) between 2-LPT+FP and all other methodologies. *Significantly (p ≤ 0.05) different from peak power calculated by 2-LPT+FP technique.

112

Figure 4 — Relative difference in peak force (A) and average force (B) derived from the force plate and all other methodologies. *Significantly (p ≤ 0.05) different from peak force calculated by 2-LPT+FP technique.

113

Figure 5 — Relative difference in peak velocity (A) and average velocity (B) derived from 2-LPT and all other methodologies. *Significantly (p ≤ 0.05) different from peak velocity calculated by 2-LPT+FP technique.

114

Tab

le 2

P

eak

Po

wer

Ou

tpu

t A

cro

ss t

he

Lo

adin

g S

pec

tru

m fo

r A

ll M

eth

od

olo

gie

s E

xam

ined

. *S

ign

ifica

nt

dif

fere

nce

fro

m

po

wer

at

the

op

tim

al lo

ad w

ith

in a

met

ho

d (

bo

ldfa

ce d

eno

tes

op

tim

al lo

ad).

115

Table 3 Comparisons of the Optimal Load Between Data Collection and Analysis Methodologies in the JS, S, and PC. *Significantly (p ≤ 0.05) different from power output at the optimal load calculated by the 2-LPT+FP technique.


(Dugan et al. 2004). Therefore, the possibility of elevated force and power data observed in other displacement-based systems is outweighed to such an extent that the 1-LPT+MASS methodology consistently underestimates power across all loads in each of the lifts examined. These findings are consistent with previous studies utilizing this meth-odology that reported lower power values than those typically observed throughout the literature (Baker et al., 2001, Jennings et al., 2005).

There are two main factors associated with the FP methodology that lead to the miscalculation of power output. First, the FP technique relies on inverse dynamics to calculate velocity from ground reaction forces, a process that requires extensive data manipulation. Similar to kinematic methods reliant on double differentiation, noise amplification and the subsequent risk of erroneous data restricts the force plate’s ability to accurately assess power output (Wood, 1982). An additional disadvantage associated with the FP approach is the inability to account for barbell movement that occurs indepen-dently of the body. As a consequence, velocity is underestimated, which results in the under-repre-sentation of power output (Haff et al., 1997; Hori et al., 2005).

In addition to the misrepresentation of power output, these methodologies have an effect on the load–power relationship in the JS, S, and PC. Whereas the optimal load within a specific move-ment may vary depending on the training status of the athlete (Stone et al., 2003), data collection and analysis techniques clearly contribute to variations within the literature (Hori et al., 2005). Two separate investigations of the JS in elite athletes (professional rugby league players, Baker et al., 2001, and nation-ally competitive sprinters, power, and Olympic lifters, McBride et al., 1999) reported the optimal load to occur at 55–59% of 1RM and 0–30% of 1RM, respectively. The major differences between these studies are the data collection and analysis procedures: Baker et al. (2001) used 1-LPT+MASS, whereas McBride et al. (1999) employed the 1-LPT+FP methodology. Similar relationships exist throughout PC (Haff et al., 1997; Winchester et al., 2005) and bench press literature (Cronin et al., 2001; Newton et al., 1996). Furthermore, an investigation of the S in elite athletes (downhill skiers) utilizing the 1-LPT method revealed the optimal load to occur at an external load of 60 kg, the lightest load exam-

ined (external loads examined were 60, 80, 100, 120, 140, 160, and 180 kg; Rahmani et al., 2001). Peak power at this load (4,719 ± 521 W) was similar to the 1-LPT optimal load of current study (4,215.07 ± 1,227.11 W), which occurred at 0% of 1RM. This comparison suggests that the load–power relation-ship of the skiers could have been considerably different if another methodology was utilized.

The current data indicate that the combination of kinetic and kinematic equipment must be utilized in order to obtain the most valid representation of muscle power generation during dynamic move-ments. In contrast to the original hypothesis, no differences were observed between 1-LPT+FP and 2-LPT+FP methodologies and thus both provide an accurate assessment of mechanical power output in the JS, S and PC. However, it is vital to delineate the most valid method from a theoretical approach. The 1-LPT+FP method does not allow for the accurate measurement of displacement in multi-dimensional movements due to the failure to account for horizon-tal motion. This in turn may contribute to an over-estimation of displacement and thus power output especially in movements that involve a significant horizontal component (i.e. Olympic lifts). Many studies utilizing one dimensional measurement systems avoid this issue by restricting the subject to movements in the vertical plane through the use of a Smith machine or similar device (Bosco et al., 1995; Cronin et al., 2001; Izquierdo et al., 1999, 2002; Mayhew et al., 1992; McBride et al., 1999; Rahmani et al., 2001; Siegel et al., 2002). Even though eliminating measurement errors resulting from horizontal movement, such studies are limited to motion only within the vertical plane and are therefore limited to certain athletic movements. Alternatively, 2-LPT+FP parallels videography and V-scope methods given that it examines displace-ment in both vertical and horizontal planes (Stone et al., 2003; Winchester et al., 2005). Therefore, the 2-LPT+FP method permits accurate measurement of vertical displacement and power in multidimen-sional movements.

In conclusion, the current study indicates that methods reliant solely on kinematic or kinetic data have limitations when used for the determination of power output during the JS, S, or PC. Both 1-LPT and 2-LPT techniques overestimate power owing to the increased force output derived through the double differentiation of bar displacement (Hori et


al., 2005; Wood et al., 1982). The 1-LPT+MASS method rejects standard biomechanical procedures by determining force without consideration of the acceleration produced throughout a movement (Dugan et al., 2004). Furthermore, the accuracy of the FP method is restricted by the inability to accurately monitor movement velocity (Hori et al., 2005). The use of these methodologies has resulted in incomplete reporting of the load–power relation-ship, leading to the different identification of the optimal load at approximately 60% of 1RM in the jump squat (Baker et al., 2001; Esliger & Sleivert, 2003; Sleivert & Taingahue, 2004). The optimal load, as determined by the combination of kine-matic and kinetic data, is 0% of 1RM in the jump squat, 56% of 1RM in the squat, and 80% of 1RM in the power clean for strength and power athletes. It is vital to move toward standardization of data collection and analysis procedures involved with mechanical power output in order to gain a clearer comprehension of the load–power relationship as well as insight into the mechanisms involved with power production and adaptation to training.

ReferencesAlemany, J.A., Pandorf, C.E., Montain, S.J., Castellani, J.W.,

Tuckow, A.P., & Nindl, B.C. (2005). Reliability assess-ment of ballistic jump squats and bench throws. Journal of Strength and Conditioning Research, 19, 33-38.

Baker, D., & Nance, S. (1999). The relation between running speed and measures of strength and power in professional rugby league players. Journal of Strength and Condition-ing Research, 13, 230-235.

Baker, D., Nance, S., & Moore, M. (2001). The load that maxi-mizes the average mechanical power output during jump squats in power trained athletes. Journal of Strength and Conditioning Research, 15, 92-97.

Behm, D.G., & Sale, D.G. (1993). Intended rather than actual movement velocity determines velocity-specific training response. Journal of Applied Physiology, 74, 359-368.

Bosco, C., Belli, A., Astrua, M., Tihanyi, J., Pozzo, R., Kellis, S., et al. (1995). A dynamometer for evaluation of dynamic muscle work. European Journal of Applied Physiology and Occupational Physiology, 70, 379-386.

Bourque, S., & Sleivert, G.G. (2003). Determinants of load at peak power during maximal effort squat jumps in endur-ance and power trained athletes. Fredericton: University of New Brunswick.

Baumgartner, T.A., & Chung, H. (2001). Confidence limits for intraclass reliability coefficients. Measurement in Physical Education and Exercise Science, 15, 92-97.

Chiu, L.Z.F., Schilling, B.K., Fry, A.C., & Weiss, L.W. (2004). Measurement of resistance exercise force expression. Journal of Applied Biomechanics, 20, 204-212.

Coelho, C.W., Hamar, D., & Araujo, C.G.S.D. (2003a). Physi-ological responses using two high-speed resistance training protocols. Journal of Strength and Conditioning Research, 17, 334-337.

Coelho, C.W., Velloso, C.R., Brasil, R.R.D.L.O., Vaisman, M., & Araujo, C.G.S.D. (2003b). Muscle power increases after resistance training in growth-hormone-deficient adults. Medicine and science in sports and exercise, 34, 1577-1581.

Cronin, J.B., McNair, P.J., & Marshal, R.N. (2001). Developing explosive power: A comparison of technique and training. Journal of Science and Medicine in Sport, 4, 59-70.

Cronin, J.B., & Henderson, M.E. (2004). Maximal strength and power assessment in novice weight trainers. Journal of Strength and Conditioning Research, 18, 48-52.

Delecluse, C., Roelants, M., Diels, R., Koninckx, E., & Ver-schueren, S. (2005). Effects of whole body vibration training on muscle strength and sprint performance in sprint-trained athletes. International Journal of Sports Medicine, 26, 662-668.

Dugan, E.L., Doyle, T.L.A., Humphries, B., Hasson, C., & Newton, R.U. (2004). Determining the optimal load for jump squats: a review of methods and calculations. Journal of Strength and Conditioning Research, 18, 668-674.

Esliger, D.W., & Sleivert, G.G. (2003). The neuromechanics of maximal effort squat jumps. Fredericton, AB: University of New Brunswick.

Falvo, M., Moore C., Weiss, L., Schilling, B., Ermert, R.C., Fry, A., et al. (2005). Reliability and precision of force, power, and velocity measures obtained during jump squats. Jour-nal of Strength and Conditioning Research, 19, e14.

French, D.N., Gomez, A.L., Volek, J.S., Rubin, M.R., Ratamess, N.A., Sharman, M.J., et al. (2004). Longitudinal tracking of muscular power changes of NCAA Division I collegiate women gymnasts. Journal of Strength and Conditioning Research, 18, 101-107.

Garhammer, J. (1993). A review of power output studies of Olympic and power lifting: Methodology, performance prediction, and evaluation tests. Journal of Strength and Conditioning Research, 7, 76-89.

Haff, G.G., Stone, M., O’Bryant, H.S., Harman, E., Dinan, C., Johnson, R., et al. (1997). Force-time dependent character-istics of dynamic and isometric muscle actions. Journal of Strength and Conditioning Research, 11, 269-272.

Hori, N., Newton, R.U., Nosaka, K., & McGuigan, M.R. (2005). Comparison of system versus barbell force, velocity and power in the hang snatch. Journal of Strength and Condi-tioning Research, 19, e19.

Iossifidou, A., Baltzopoulos, V., & Giakas, G. (2005). Isokinetic knee extension and vertical jumping: are they related? Journal of Sports Science, 23, 1121-1127.

Izquierdo, M., Ibanez, J., Gorostiaga, E., Garrues, M., Zuniga, A., Anton, A., et al. (1999). Maximal strength and power characteristics in isometric and dynamic actions of the upper and lower extremities in middle-aged and older men. Acta Physiologica Scandinavica, 167, 57-68.

Izquierdo, M., Häkkinen, K., Ibanez, J., Garrues, M., Anton, A., Zuniga, A., et al. (2001). Effects of strength training on muscle power and serum hormones in middle-aged and older men. Journal of Applied Physiology, 90(4), 1497-1507.


Izquierdo, M., Häkkinen, K., Gonzalez-Badillo, J.J., Ibanez, J., & Gorostiaga, E.M. (2002). Effects of long-term training specificity on maximal strength and power of the upper and lower extremities in athletes from different sports. European Journal of Applied Physiology, 87, 264-271.

Jennings, C.L., Viljoen, W., Durandt, J., & Lambert, M.I. (2005). The reliability of the FitroDyne as a measure of muscle power. Journal of Strength and Conditioning Research, 19, 859-863.

Kawamori, N., Crum, A.J. Blumert, P.A., Kulik, J.R., Childers, J.T., Wood, J.A., Stone, M.H., & Haff, G.G. (2005). Influence of different relative intensities on power output during the hang power clean: identification of the optimal load. Journal of Strength and Conditioning Research, 19, 698-708.

Lyttle, A.D., Wilson, G.J., & Ostrowski, K.J. (1996). Enhancing performance: maximal power versus combined weights and plyometrics training. Journal of Strength and Condi-tioning Research, 10, 173-179.

Mayhew, J.L., Ware, J.S., Johns, R.A., & Bembem, M.G. (1992). Changes in absolute upper body power following resistance training in college males. Journal of Applied Sport Science Research, 6, 187-194.

McBride, J.M., Triplett-McBride, T., Davie, A., & Newton, R.U. (1999). A comparison of strength and power characteristics between power lifters, Olympic lifters and sprinters. Jour-nal of Strength and Conditioning Research, 13, 58-66.

McBride, J.M., Triplett-McBride, T., Davie, A., & Newton, R.U. (2002). The effect of heavy- vs. light-load jump squats on the development of strength, power, and speed. Journal of Strength and Conditioning Research, 16, 75-82.

Newton, R.U., Kraemer, W.J., Häkkinen, K., Humphries, B.J., & Murphy, A.J. (1996). Kinematics, kinetics and muscle activation during explosive upper body movements. Jour-nal of Applied Biomechanics, 12, 31-43.

Piscopo, J., & Bailey. J. (1981). Kinesiology: The science of movement. New York: John Wiley and Sons, p. 128.

Rahmani, A., Viale, F., Dalleau, G., & Lacour, J.R. (2001). Force/velocity and power/velocity relationships in squat exercise. European Journal of Applied Physiology, 84, 227-232.

Sands, W.A., Smith, L.S., Kivi, D.M., McNeal, J.R., Dorman, J.C., Stone, M.H., et al. (2005). Anthropometric and physi-cal abilities profiles: US national skeleton team. Sports Biomechanics, 4, 197-214.

Siegel, J.A., Gilders, R.M., Staron, R.S., & Hagerman, F.C. (2002). Human muscle power output during upper- and lower-body exercises. Journal of Strength and Condition-ing Research, 16, 173-178.

Sleivert, G., & Taingahue, M. (2004). The relationship between maximal jump-squat power and sprint acceleration in ath-letes. European Journal of Applied Physiology, 91, 46-52.

Stone, M.H., O’Bryant, H.S., McCoy, L., Coglianese, R., Lehm-kuhl, M., & Schilling, B. (2003). Power and maximum strength relationships during performance of dynamic and static weighted jumps. Journal of Strength and Condition-ing Research, 17, 140-147.

Thomas, M., Fiatarone, M.A., & Fielding, R.A. (1996). Leg power in young women: Relationship to body composition, strength and function. Medicine and science in sports and exercise, 28, 1321-1326.

Webb, A., Smith, L., Weiss, C., Moore, B., Schilling, B., Ermert, R.C., et al. (2005). Reliability and precision of multiple expressions of hang power clean external power. Journal of Strength and Conditioning Research, 19, e36.

Weiss, L.W., Fry, A.C., Magu, B., Moore, C., Chiu, L., Buchanan, K., et al. (2004). Relative external loads elicit-ing maximal concentric force and power during non-coun-termovement squats. Journal of Strength and Conditioning Research, 18, e20.

Weiss, L.W., Moore, C., Schilling, B., Ermert, R.C., Fry, A., Chiu, L., et al. (2005). Reliability and precision of multiple expressions of hang power clean bar velocity. Journal of Strength and Conditioning Research, 19, e36.

Wilson, G. J., Newton, R.U., Murphy, A.J., & Humphries, B.J. (1993). The optimal training load for the development of dynamic athletic performance. Medicine and Science in Sports and Exercise, 25, 1279-1286.

Winchester, J. B., Erickson, T.M., Blaak, J.B., & McBride, J.M. (2005). Changes in bar-path kinematics and kinetics after power-clean training. Journal of Strength and Condition-ing Research, 19, 177-183.

Wood, G.A. (1982). Data smoothing and differentiation pro-cedures in biomechanics. Exercise and Sport Sciences Reviews, 10, 308-362.

Documents

Cormie et al 2007 Validation of Power Measurement Techniques in Dynamic