Share this article on:

Performance Evaluation of a High-Speed Inertial Exercise Trainer

Caruso, John F1; Hari, Parameswar2; Coday, Michael A1; Leeper, Adam2; Ramey, Elizabeth1; Monda, Julie K3; Hastings, Lori P4; Davison, Steve5

Journal of Strength and Conditioning Research: November 2008 - Volume 22 - Issue 6 - p 1760-1768
doi: 10.1519/JSC.0b013e318187684d
Original Research

Caruso, JF, Hari, P, Coday, MA, Leeper, A, Ramey, E, Monda, JK, Hastings, LP, and Davison, S. Performance evaluation of a high-speed inertial exercise trainer. J Strength Cond Res 22(6): 1760-1768, 2008-A high-speed, low-resistance inertial exercise trainer (IET, Impulse Training Systems, Newnan, Ga) is increasingly employed in rehabilitative and athletic performance settings. Repetitions on an IET are done through a large range of motion because multijoint movements occur over more than one plane of motion, with no limitation on velocities or accelerations attained. The current study purpose is to assess data reproducibility from an instrumented IET through multiple test-retest measures. Data collection methods required the IET left and right halves to be fitted with a TLL-2K force transducer (Transducer Techniques, Temecula, Calif) on one of its pulleys, and an infrared position sensor (Model CX3-AP-1A, automationdirect.com) located midway on the underside of each track. Signals passed through DI-158U signal conditioners (DATAQ Instruments, Akron, Ohio) and were measured with a four-channel analog data acquisition card at 4000 Hz. To assess data reproducibility, college-age subjects (n = 45) performed four IET workouts that were spaced 1 week apart. Workouts entailed two 60-second sets of repetitive knee- and hip-extensor muscle actions as subjects were instructed to exert maximal voluntary effort. Results from multiple test-retest measures show that the IET elicited reproducible intra- and interworkout data despite the unique challenge of multiplanar and multijoint exercise done over a large range of motion. We conclude that future studies in which IET performance measurement is required may choose to instrument the device with current methodology. Current practical applications include making IET data easier to comprehend for the coaches, athletes, and health care providers who use the device.

1Exercise & Sports Science Program, 2Department of Physics & Engineering Physics, 3Department of Chemistry & Biochemistry, and 4Department of Biology, The University of Tulsa, Tulsa, Oklahoma; and 5Impulse Training Systems, Newnan, Georgia

Address correspondence to John F. Caruso, john-caruso@utulsa.edu.

Back to Top | Article Outline

Introduction

Human movement requires the central nervous system to emit a series of neural impulses so that forces that result from muscle fiber recruitment may overcome inertial or gravitational resistance. Sometimes, forces are applied against resistance provided by exercise devices, such as those used in physical conditioning and rehabilitation programs. A recent trend in conditioning programs is to train athletes so that they may learn to exert higher rates of force and power development to meet the physical demands of modern-day competition (27,35). Quantification of exercise performed on such devices is made easier to interpret through proper instrumentation. The reproducibility of exercise data speaks to the merits of the instrumentation process. Because of the unique design and operation of some devices, instrumentation that leads to a high degree of data reproducibility is a challenge to engineers, exercise scientists, and health care professionals. One such device is an inertial exercise trainer (IET, Impulse Technologies, Newnan, Ga).

Whereas standard exercise equipment (barbells, dumbbells, etc.) uses resistance provided by gravity and additional mass, IET users impart momentum to the system and then counter the momentum to complete repetitions. Unlike many exercise devices, the novel operation of the IET enables high-speed, low-resistance repetitions to be performed at multiple joints as movement occurs over more than one plane of motion simultaneously. In addition, the IET imposes no restrictions on exercise velocity and with phasic repetitions, whereby muscles alternatively contract and relax at high speeds (10), permitting increased acceleration. The design of the Iet allows a low-weight sled, mounted on four wheels, to traverse a 1.9-m track with minimal frictional resistance (10). A nylon cord connects an exercise attachment handle to the sled. A series of pulleys above and below the sled permit multiple exercises to be performed at a variety of angles. The pulleys transfer the user's force so that sled resistance is always pulled toward the center of the track. As the sled passes above a pair of centered pulleys, cord movement is instantly reversed and a pulling force is imparted to users commensurate to the initial force exerted. Overhead, side, and front-view IET illustrations appear in Figures 1-3, respectively.

Figure 1

Figure 1

Figure 2

Figure 2

Figure 3

Figure 3

As compared with traditional exercise devices, the IET permits performance of high-speed, low-resistance repetitions. Although the device shown in Figures 1-3 was originally designed for use in physical therapy and rehabilitative settings, the IET does not entail movement of a known mass against the pull of gravity. Thus, the design of the IET may be useful in novel environments such as space flight to counteract the neuromuscular impairments experienced by astronauts. As a result, physiological responses to exercise done on the IET may also be unlike those associated with standard weight training equipment. Yet, before IET exercise performance can be adequately assessed, the device must be instrumented in a manner such that the data generated are reproducible. The purpose of the current study is to examine the reproducibility of exercise data obtained on an instrumented IET through multiple test-retest measures. Because of the design and operation of the IET, we hypothesize that its exercise reproducibility values may be less than for other physical conditioning and rehabilitative devices (1,17,18,21,25,26,29,34,36,37).

Back to Top | Article Outline

Methods

Experimental Approach to the Problem

The greatest challenge to instrument the IET was to capture sled position as it moves at high velocities and acceleration rates. To address this concern, the right and left halves of the IET were equipped with a TLL-2K load cell (Transducer Techniques, Temecula, Calif) attached to one of the pulleys (Figure 4) and an infrared position sensor (Model CX3-AP-1A, automationdirect.com) located midway on the underside of each track (Figure 5). As a sled moved along its track, the load cell and position sensor recorded force output and displacement, respectively. Through integration of the sensor's time response, we estimated average acceleration and force output. Load cell and position sensor data were sent to DI-158U signal conditioners (DATAQ Instruments, Akron, Ohio) and measured by a four-channel analog data acquisition card at 4000 Hz. Instrumentation also included a power source (Model Dual 0-30VDC/3A and 5VDC/3A; Jones & Assoc., Lake Park, Fla). Force and work output were calculated and analyzed with Microsoft Excel. A macro was written to perform the numeric integration of force data. A bilateral IET instrumentation diagram appears in Figure 6.

Figure 4

Figure 4

Figure 5

Figure 5

Figure 6

Figure 6

Back to Top | Article Outline

Subjects

To assess test-retest data reproducibility from the instrumented IET, college-age subjects (32 women, 13 men) made six visits to our laboratory, spaced 1 week apart. Women subject characteristics (average ± SEM) were as follows: height, 1.69 ± 0.01 m; weight, 69.0 ± 1.9 kg. The ranges of height (1.57-1.86 m) and weight (48.1-102.3 kg) data from women subjects are also provided. Men subject characteristics were as follows: height, 1.85 ± 0.02 m; weight, 91.0 ± 5.2 kg. The ranges of heights (1.75-2.00 m) and weights (58.2-120.0 kg) from men subjects are also included. A university-based IRB approved all data collection procedures and the use of human subjects. In the current study, four workouts were preceded by two familiarization sessions to accustom subjects to the IET and reduce their injury risk. Because the IET design and operation enable greater performance variability through the higher velocities and accelerations attained (17,18), it was deemed prudent that subjects perform two familiarization sessions before workouts. For each subject, care was taken to ensure that workouts occurred at similar times of day to limit test-retest variability (9).

Workouts and familiarization sessions were preceded by a 5-minute warm-up on a stationary bicycle against 9.8 N of resistance at a self-selected velocity. Subjects performed an open-chained seated hip and knee extension exercise exclusively with their left leg. A chair that faced the U-shaped IET (Figure 1) maintained body position, with its location held constant for each subject. Workouts entailed two 60-second sets separated by a 90-second rest period and consisted exclusively of either tonic muscle actions, which may be described as synergistic coactivation of agonist and antagonist muscles, or phasic bouts, which involved muscle contraction and relaxation at high acceleration rates (10). For tonic workouts, the nylon cord, which connects an exercise attachment handle to the weight sled, was kept continually taut throughout repetitions. In contrast, phasic actions entailed a cyclic pattern in which the cord both slackened and tightened with each repetition. Subjects performed two tonic and phasic workouts each, with their sequence randomized to prevent an order effect. With an 8.1-kg mass added to the sled and an attachment handle that terminated as a velcro strap around the left foot, subjects performed as many repetitions as possible per set. Subjects were instructed to exert maximal voluntary effort and not to pace themselves, and they received vocal support. Per-set peak force (PF), average force (AF), and total work (TW) variables were collected.

Back to Top | Article Outline

Data Reproducibility

The subjects' PF, AF, and TW values were analyzed for intra- and interworkout reproducibility with several test-retest measures. Intraclass correlation coefficients (ICCs) assessed the degree of agreement between measures (4,12). The coefficient of variation (CV), a measure of variability and reliability (14), determined the degree of data distribution and dispersion among paired data. The CV also indicates the variance about the average when SDs are significantly less than positive mean values (32). The CV was calculated as an SD/mean ratio and multiplied by 100 to derive percentage values. The degree of systematic change appears in Bland-Altman analyses, which plots delta (difference) measurements among paired values as a function of their mean (12,19). Bland-Altman analyses assessed data for heteroscedasticity and outliers (12). Current Bland-Altman plots include dashed lines that represent ± 1.96 SD, which show the limits of agreement and signify test-retest reliability for 95% of a population (3,5,23). The SEM, also known as within-subject variance, noted the variability in a subject's performance (4). The smallest real difference (SRD) revealed the least amount of change needed to indicate real variability in a subject's performance. The SRD was calculated as SEM × 2.77 × 1.414. To express values as percentages, SEM % and SRD % were calculated as SEM/mean and SRD/mean ratios and multiplied by 100. Because SEM, SRD, SEM %, and SRD % are functions of within-subject variance (4), only intraworkout values were provided for these variables.

Back to Top | Article Outline

Results

No subjects were injured from their exercise participation. Current study data were normally distributed, and with 1 week between workouts, there seemed to be no carryover effects from prior workouts (3,8). Earlier work had indicated that the current sample was large enough to detect effect sizes of 5-15% (37). Research has shown that such samples are appropriate for high-speed exercise, which typically incur more data variability (18,37). Performance values appear in Table 1. Development of new motor patterns and exercise-induced fatigue, which comprise the systematic bias seen with measurement error (3), likely added to the current study variability. Yet, despite the higher speeds and greater range of motion, current test-retest values show a high degree of data agreement. Table 2 shows more agreement among intraworkout, as compared with interworkout, values. Figures 7-12 show the Bland-Altman plots and limits of agreement for each performance variable. Like Table 2 data, the Bland-Altman analyses show greater variability among interworkout performance values.

Table 1

Table 1

Table 2

Table 2

Figure 7

Figure 7

Figure 8

Figure 8

Figure 9

Figure 9

Figure 10

Figure 10

Figure 11

Figure 11

Figure 12

Figure 12

Back to Top | Article Outline

Discussion

Previous studies have examined the merits of IET workouts (2,6,22,27,33). A 5-week elbow flexor IET workout protocol evoked significant isokinetic strength gains (2). Significant time (post > pre) gains occurred for concentric and eccentric peak torque measured at 1.05 rad·s−1, as well as eccentric peak torque at 2.09 rad·s−1 (2). In a comparison of concurrent kinematic and electromyographic (EMG) data collected from phasic and tonic IET elbow flexor actions, 12 subjects performed workouts against loads that ranged from 0 to 9.1 kg (33). As repetitions were performed, EMG data were collected from agonist and antagonist muscle groups (33). Higher peak angular velocities, peak platform accelerations, and average and peak antagonistic EMG activity, as well as less range of motion, resulted from phasic actions (33). In an investigation that examined the same muscle group as the current study, applied loads and EMG activity were compared in eight seated quadriceps exercises done against four different types of resistance (22). Vastus medialis/lateralis EMG ratios from eccentric IET actions evoked higher muscle activity than the other quadriceps exercises examined (22). Thus, for rehabilitative purposes directed toward greater quadriceps activity, the IET may be preferable to other training modalities. Recently, performance measures derived from exercise done on two gravity-independent equipment modalities were compared for their ability to predict the variance in blood lactate levels (6). The results show that the IET performance measures were far superior in the prediction of blood lactate variance than another inertial resistance device (6). The merits of IET workouts were also directed toward improvement of high-speed, sports-specific tasks such as a baseball pitch, whereby a small mass is moved as rapidly as possible (27). Thus, the merits of IET have been examined in both rehabilitative and athletic performance settings; the reliability of performance measures derived from this device is of utmost importance to persons in the aforementioned areas.

The IET permits multiplanar exercise at high velocities and accelerations over a large range of motion. Thus, the potential exists for large amounts of performance variability. To limit current study variability, a second familiarization session was employed, as was a greater sample size compared with similar investigations (1,17,18,21,25,26,29,34,36,37). Prior study test-retest reproducibility values were likely enhanced through single-joint and/or plane exercise (1,17,18,21,25,26,29,34,36,37). Other methodologies that likely limited variability included the use of closed-chain devices (17), exercise done over a smaller range of motion (7,34), or exercise done across fewer repetitions and/or fixed velocities (1,18,21,26,29,34,36,37). Given the current study instrumentation and IET operation, the results have yielded new information on the reproducibility of data obtained from a high-speed, low-resistance device that permits multijoint/planar movement without velocity restrictions. Most of the results shown in Table 2 exceeded our hypothesized outcome and yielded reproducibility values that met or surpassed those from other devices (1,17,18,21,25,26,29,34,36,37).

Current results yielded stark performance contrasts (Table 1) from the tonic and phasic workouts. The differences were likely attributable to the manner in which repetitions were performed. Unlike tonic sets, phasic actions caused the nylon cord to visibly slacken with each repetition. As the cord slackened, the involved muscle groups accelerated as they continued to move until the weight sled resistance was incurred. At that time, motor unit recruitment increased to overcome the resistance, and this produced an impulse of muscle activity as well as the higher PF values from phasic workouts (10). In contrast, both tonic and phasic TW values observed large SEM and SRD values (Table 2). This likely occurred because some error indices typically increase as absolute values become larger, whereas data reproducibility normally improves when within-subject variability is reduced (14,23). In the current study, several test-retest measures were examined so that greater knowledge of our sample's within-subject variability could be ascertained.

Our heterogeneous sample included large spreads in individual performance. Although within-subject variability impacts data reproducibility, it also is important to note that test-retest values are sensitive to between-subject variability or the spread of scores among the sample examined (14). A large performance spread may have contributed to current ICC results (12,14). The ICC is defined as a ratio of the covariance between paired values and total variance (11), or a ratio of the true variance of interest to error variance collected with the measurement (24). Some may regard intersubject ICC values as the variance of greatest relevance, yet some place more importance on intrasubject variability (4). Unlike Pearson correlations, ICCs do not 1) inflate the strength of agreement among paired values or 2) show bias toward small samples (15). Prior exercise studies (17,18) done over a single plane for a limited number of repetitions have stated that ICC values of 0.75-0.80 indicate excellent reproducibility, and these studies have referenced earlier research (30) to derive such values. Current ICC values show that all but one performance variable exceeded the 0.75-0.80 threshold. Other strength training investigations yielded comparable (13) or smaller (31) ICC values. Given the manner in which the IET operates, and that most current ICC values exceeded the 0.75-0.80 threshold, intra- and interworkout performance data may be considered to demonstrate excellent reproducibility.

The CV is a measure of test-retest reproducibility that assesses data distribution and dispersion. An early paper notes CV values of 10-15% as typical for biological systems (32). Current CV results fell within that range or were less. Recent studies (16,34,36) have evoked comparable values. Isokinetic knee extensions yielded interworkout CV values of 3.9-5.9, as higher speeds led to greater variability (36). Isokinetic ankle extension tests also yielded similar intrasubject CV values (34). Finally, a recent study has examined data reproducibility from exercises done on an inertia-based device (16). Subjects (n = 16) performed concentric actions against resistance from four submaximal loads (16). Intertrial and session CV values ranged from 2.6 to 16.3 (16). Like ICC, CV is also impacted by sample heterogeneity (14,15). Yet, current CV values were not only comparable with those from prior studies (16,34,36), but, because of the broad range of scores, variability was less than for more homogenous samples (15). Because of their similarity to prior values (16,34,36), the current CV results are thought to demonstrate high test-retest reproducibility.

Bland-Altman plots enable visual inspection for outliers, systematic data changes, and heteroscedasticity (3,5,12,37). Some suggest that it is the most accurate estimator of measurement error, even after examination with multiple test-retest measures (20,24). Bland-Altman analysis plots allow for examination of heteroscedasticity, which occurs when differences among paired data increase as mean values become higher, and this suggests that variability is influenced by the magnitude of the measured value (4,5,12). For each performance variable, Figures 7-12 do not demonstrate heteroscedasticity. Data plotted beyond ± 1.96 SD or outside the dashed lines are considered outliers, as they lie beyond the limits of agreement, or where 95% of a population is expected to reside (3,5,23). In the absence of heteroscedasticity, SEM and the limits of agreement are considered the most popular measures of absolute reliability (3), and, as such, they are included as a test-retest measure in the current study. Figures 7-12 show that most data fell within ± 1.96 SD, and outliers were primarily interworkout measurements.

Prior studies have employed Bland-Altman plots for outlier detection (12,19). Blood pressure value reproducibility was measured with three different devices (19). Absolute intradevice measurement differences at 5, 10, and 15 mm Hg have shown that roughly 8% of the test-retest measurement differences exceeded 15 mm Hg (19). The percentage of blood pressure outliers exceeded those of the current study. However, criteria provided by The British Hypertensive Society deem reproducibility from the three devices as acceptable (19). A study of 50 stroke patients assessed the degree of agreement between test-retest gait performance measures (12). Bland-Altman analysis identified, in each of the six gait measures assessed, roughly 6% of the data as outliers. It was concluded that gait tests showed a high degree of measurement agreement (12). Because Figures 7-12 generally show a smaller percentage of outliers as prior studies (12,19), there seems to be an acceptable degree of test-retest data reproducibility as determined through the Bland-Altman plots.

The SEM is the square root of within-subject variance, or the dispersion of scores when subjects are tested repeatedly (4,28). The SEM values have the added benefit of being expressed in the same units of measurement as the variables under examination (11). Although current SEM % values compare favorably with prior results (4,12), SEM values are a reflection of the absolute magnitude of each performance variable. Current SEM % values allow for variability comparison among the performance variable without the influence of their absolute magnitude. The reproducibility of gait performance tests administered to subjects with hemiparesis after stroke indicated that SEM and SEM % values ranged from 0.07 to 18.6 and 4.8 to 8.2, respectively (12). Sickness impact profiles collected from stroke victims indicated that SEM % ranged from 3.34 to 14.53 (4). Current SEM % results are similar to those from prior studies (4,12). In contrast, the absolute magnitude of the current performance variables likely led to higher SEM values than those observed previously (4,12).

Associated with SEM, SRD is the minimal change required to note a true difference in subjects' performance (12) and is thought to link subjects' responsiveness to data reproducibility (4). Current SRD and SRD % values, which address the degree of systematic error attributable to learning effects (14), compare favorably with similar values obtained from prior studies (4,12). A study that examined isometric strength performance reproducibility in healthy women (n = 10) performed 1 week apart deemed SEM a practical estimate of intrasubject reliability (28). Prior reproducibility research has observed SRD % values of 9.3-40.3 (4) and 13-23 (12), values comparable with those from the current study. In conclusion, it was initially thought that the instrumentation of the IET would pose a considerable challenge. Because of its operation, IET reproducibility values were hypothesized to be less than those for conventional devices. However, because of the current study's instrumentation methodology, whereby data collection occurred at a high sampling frequency, the reproducibility of IET performance variables was deemed acceptable. Future research investigations in which it is essential to quantify performance from an IET intervention may choose to instrument the device as described in the current study.

Back to Top | Article Outline

Practical Applications

Unlike many exercise devices, the novel operation of the IET enables high-speed, low-resistance repetitions to be performed over multiple joints and planes of motion. Because of the low amount of resistance imposed by the IET, its use is popular in rehabilitation programs to aid recovery from injury or to accommodate for pain and limited mobility. In addition, demands for increased rates of force and power development required in sports competition, in which athletes often must move a modest amount of mass as rapidly as possible (27,35), make conditioning programs centered around devices such as the IET an ideal training modality.

Back to Top | Article Outline

Acknowledgments

We wish to thank our subjects for their participation. J.F. Caruso, P. Hari, M.A. Coday, A. Leeper, and J.K. Monda are TURC (Tulsa Undergraduate Research Challenge) program participants at The University of Tulsa. Financial support of this project was provided by a university-based faculty research grant.

Back to Top | Article Outline

References

1. Aagaard, P, Simonsen, EB, Trolle, M, Bangsbo, J, and Klausen, K. Moment and power generation during maximal knee extensions performed at low and high speeds. Eur J Appl Physiol 69: 376-381, 1994.
2. Albert, MS, Hillegass, E, and Spiegel, P. Muscle torque changes caused by inertial exercise training. J Orthop Sports Phys Ther 20: 254-261, 1994.
3. Atkinson, G and Nevill, AM. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med 26: 217-238, 1998.
4. Beckerman, H, Roebroeck, ME, Lankhorst, GJ, Becher, JG, Bezemer, PD, and Verbeek, ALM. Smallest real difference, a link between reproducibility and responsiveness. Qual Life Res 10: 571-578, 2001.
5. Bland, JM and Altman, DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: 307-310, 1986.
6. Caruso, JF, Coday, MA, Monda, JK, Roberts, KP, and Potter, WT. Body mass and exercise variable relationships to lactate derived from gravity-independent devices. Aviat Space Environ Med 78: 864-870, 2007.
7. Caruso, JF, Williams, JA, Hari, P, McCoy, JD, Coday, MA, and Ramsey, CA, and Capps, LB. Data reproducibility from the instrumentation of an inertial resistance exercise device suggested for use during space travel. Isokinet Exerc Sci 14: 371-382, 2006.
8. Charter, RA. Effect of measurement error on tests of statistical significance. J Clin Exp Neuropsychol 19: 458-462, 1997.
9. Coldwells, A, Atkinson, G, and Reilly, T. Sources of variation in back and leg dynamometry. Ergonomics 37: 79-86, 1994.
10. Davison, S. Inertial Exercise Clinical Training Instructional Manual. Newnan, GA: Engineering Marketing Association, 1997.
11. Eliasziw, M, Young, SL, Woodbury, MG, and Fryday-Field, K. Statistical methodology for the concurrent assessment of interrater and intrarater reliability: using goniometric measurements as an example. Phys Ther 74: 777-788, 1994.
12. Flansbjer, U-B, Holmback, AM, Downham, D, Patten, C, and Lexall, J. Reliability of gait performance tests in men and women with hemiparesis after stroke. J Rehabil Med 37: 75-82, 2005.
13. Heinonen, A, Sienanen, H, Viitasalo, J, Pasanen, M, Oja, P, and Vuori, I. Reproducibility of computer measurement of maximal isometric strength and electromyography in sedentary middle-aged men. Eur J Appl Physiol 68: 310-314, 1994.
14. Hopkins, WG. Measures of reliability in sports medicine and science. Sports Med 30: 1-15, 2000.
15. Hopkins, WG, Hawley, JA, and Burke, LM. Design and analysis of research on sport performance enhancement. Med Sci Sports Exerc 31: 472-485, 1999.
16. Jidovtseff, B, Croisier, JL, Lhermerout, C, Serre, L, Sac, D, and Crielaard, JM. The concept of iso-inertial assessment: reproducibility analysis and descriptive data. Isokinet Exerc Sci 14: 53-62, 2006.
17. Kovaleski, JE, Heitman, RJ, Gurichek, LR, Erdmann, JW, and Trundle, TL. Reliability and effects of leg dominance on lower extremity isokinetic force and work using the closed chain rider system. J Sport Rehabil 6: 319-326, 1997.
18. Kovaleski, JE, Ingersoll, CD, Knight, KL, and Mahar, CP. Relibility of the BTE dynatrack isotonic dynamometer. Isokinet Exerc Sci 6: 41-43, 1996.
19. Lewis, JE, Boyle, E, Magharious, L, and Meyers, MG. Evaluation of a community-based automated blood pressure measuring device. Can Med Assoc J 166: 1145-1148, 2002.
20. Liehr, P, Dedo, YL, Torres, S, and Meininger, JC. Assessing agreement between clinical measurement methods. Heart Lung 24: 240-245, 1995.
21. Madsen, OR. Torque, total work, power, torque acceleration energy and acceleration time assessed on a dynamometer: reliability of knee and elbow extensor and flexor strength measurements. Eur J Appl Physiol 74: 206-210, 1996.
22. Matheson, JW, Kernozek, TW, Denni, CW, and Davies, GJ. Electromyographic activity and applied load during seated quadriceps exercises. Med Sci Sports Exerc 33: 1713-1725, 2001.
23. 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.
24. Ottenbacher, KJ and Tomchek, SD. Measurement variation in method comparison studies: an empirical examination. Arch Phys Med Rehabil 75: 505-512, 1994.
25. Palmer, GS, Dennis, SC, Noakes, TD, and Hawley, JA. Assessment of the reproducibility of performance testing on an air-braked cycle ergometer. Int J Sports Med 17: 293-298, 1996.
26. Pearson, SJ, Harridge, SDR, and Grieve, DW. A variable inertial system for measuring the contractile properties of human muscle. Med Sci Sports Exerc 33: 2072-2076, 2001.
27. Pezzullo, D, Karas, S, and Irrgang, J. Functional plyometric exercises for the throwing athlete. J Athl Train 30: 22-26, 1995.
28. Roebroeck, ME, Harlaar, J, and Lankhorst, GJ. The application of generalizability theory to reliability assessment: an illustration using isometric force measurements. Phys Ther 73: 386-401, 1993.
29. Seger, JY, Westing, SH, Hanson, M, Karlson, E, and Ekblom, B. A new dynamometer measuring concentric and eccentric muscle strength in accelerated decelerated or isokinetic movements. Eur J Appl Physiol 57: 526-530, 1988.
30. Shrout, PE and Fleiss, JL. Intraclass correlation: uses in assessing rater reliability. Psychol Bull 86: 420-428, 1979.
31. Sleivert, GG and Wenger, HA. Reliability of measuring isometric and isokinetic peak torque development integrated electromyography, and tibial nerve conduction velocity. Arch Phys Med Rehabil 75: 1315-1321, 1994.
32. Stokes, M. Reliability and repeatability of methods for measuring muscle in physiotherapy. Physiother Pract 1: 71-76, 1985.
33. Tracy, J, Obuchi, S, and Johnson, B. Kinematics and electromyographic analysis of elbow flexion during inertial exercise. J Athl Train 30: 254-258, 1995.
34. Trappe, SW, Trappe, TA, Lee, GA, and Costill, DL. Calf muscle strength in humans. Int J Sports Med 22: 186-191, 2001.
35. Wannagetfast [Web site]. Available at: www.wannagetfast.com. Accessed November 1, 2007.
36. Wyse, JP, Mercer, TH, and Gleeson, NP. Time-of-day dependence of isokinetic leg strength and associated interday variability. Br J Sports Med 28: 167-170, 1994.
37. Zehr, EP and Sale, DG. Reproducibility of ballistic movement. Med Sci Sports Exerc 29: 1383-1388, 1997.
Keywords:

instrumentation; data reproducibility; force transducer

© 2008 National Strength and Conditioning Association