Secondary Logo

Journal Logo

Original Research

Test-Retest Reliability, Criterion-Related Validity, and Minimal Detectable Change of Score on an Abbreviated Wingate Test for Field Sport Participants

Hachana, Younes1,2; Attia, Ahmed1,2; Nassib, Sabri2; Shephard, Roy J.3; Chelly, Mohamed Souhaiel1,2

Author Information
Journal of Strength and Conditioning Research: May 2012 - Volume 26 - Issue 5 - p 1324-1330
doi: 10.1519/JSC.0b013e3182305485
  • Free



The standard Wingate test (30-secT) is a 30-second cycle-ergometer test developed to evaluate an individual's power output (8,40). Performance indices derived from this test include the peak power output (PPO), mean power output (MPO), and fatigue index (FI); these reflect, respectively, the greatest average mechanical power developed over any 5-second period, the average power maintained over the six 5-second segments of the test, and the decline in power over 30 seconds, expressed as a percentage of the peak value (20). The test requires 30 seconds of maximal effort against a constant resistance, traditionally set at 7.5% of the subject's body mass (kilograms) (3,5,22,33). Because of its widespread use, the 30-secT has become a standard procedure to validate other measures of sprint performance such as the maximal anaerobic running test (29), and the tethered running test (24,42). It has also been used to evaluate the training status of sprinters (17), and it is considered a good predictor of short-distance running performance (27,31,34,41).

Nevertheless, there are both theoretical and practical objections to the 30-secT. The theoretical premises underlying the test can be traced back to the classic studies of Margaria et al. (26), Di Prampero (14), and Andersen and the WHO working party on exercise testing (1). These several authors envisaged that maximal effort proceeded in 3 phases, corresponding to the usage of phosphagen reserves (usually exhausted within 2–4 seconds), accumulation of lactate (completed over about 45 seconds), and a steady maximal intake of oxygen. The Wingate test was intended to provide information about the phosphagen component (“Anaerobic Power”) and the lactate component (“Anaerobic Capacity”). When performing a Wingate test, subjects typically exhibit a sharp rise of power output over the first few seconds, but they are subsequently unable to maintain this rate of working so that power output declines exponentially during the remainder of the test (5). Much of the total power output is derived from anaerobic metabolism, and there is an accumulation of [H+] from the accumulating by-products of anaerobic glycolysis (22), sometimes with associated hypoglycemia (37). Nevertheless, it is now recognized that there is a considerable overlap between the 3 phases previously described as anaerobic power, anaerobic capacity, and aerobic power. Perhaps more importantly from the viewpoint of athlete testing, participants in field sports and runners who are contesting distances of 50–100 m do not engage in all-out effort for as long as 30–45 seconds. When assessing the ability of such individuals to engage in sport-appropriate bursts of all-out activity, a test of a 15-second duration (15sec-T) has greater construct validity than the traditional 30-second Wingate Test (11,12,16).

A further objection to the traditional 30-second test is that it often induces unpleasant side effects such as nausea, vomiting, and light to severe headaches. Moreover, subject awareness of these side effects has a potential to inhibit maximal effort, with a high attrition rate if tests are repeated over the course of a training program (22).

We were thus interested in developing a shortened protocol that would minimize side effects but still provide reliable and valid results. There have been several previous suggestions for shortened protocols (4,22,33). Perhaps the most noteworthy proposition is that of Bar-Or et al. (4), originators of the 30-second test. They argued that a 15-second cycle-ergometer sprint provided a safe and reliable measure of both peak and mean power in men and women of advanced age. They found that the 15-secT was less fatiguing than the original protocol was, thus allowing them to make several evaluations of PPO and MPO within a given trial. However, all previous authors began from the standpoint of using a shortened test to predict 30-second values, rather than considering the 15-secT as having a greater construct validity in its own right. Furthermore, the reliability of the 15-secT has not been evaluated in young and athletic adults, nor has its criterion validity been assessed relative to the standard 30-secT. This study thus examined the reliability, validity, and value of the 15-secT in a substantial sample of physical education students; measures included the intraclass correlation coefficient (ICC), the SEM, the minimal detectable change (MDC), and a direct comparison of 15-secT with 30-secT values. The ICC is unfortunately influenced by intersubject variability and moreover cannot assess systematic bias (7,10). The SEM is not affected by intersubject variability (19); it quantifies the precision of individual scores and can be used to establish MDC (18), the amount by which an athlete's performance must change to be sure that an apparent alteration in score is not simply because of measurement error. The MDC must in turn be distinguished from the smallest worthwhile change (SWC), a subjective assessment of the smallest change in outcome that has practical importance for an athlete.


Experimental Approach to the Problem

The test-retest reliability and the criterion-related validity of the 15-secT were evaluated on physical education students. The outcome measures of absolute and relative PPO, MPO, and the FI were validated relative to the data obtained from standard 30-second Wingate tests. The findings justify the use of the abbreviated (and subjectively more pleasant) protocol in young athletes.


The subjects were 46 male and 23 female sport-science students, pursuing degrees in Exercise Science and Physical Education with an emphasis upon various team sports, including football, basketball, rugby, and handball (mean ± SD: age 20.7 ± 1.6 years, stature 1.71 ± 0.09 m, body mass 73.2 ± 9.4 kg). They had 6 ± 1.5 years of experience in their respective sports and were training in their clubs 5–6 times per week (i.e., 10–12 hours weekly). All the participants gave their written and informed consent to the study after receiving a thorough explanation of the protocol. This conformed to internationally accepted policy statements regarding the use of human subjects and was approved by the University Ethics Committee in accordance with the Helsinki declaration.


All the subjects were fully familiarized with the measurement protocols before data collection began. To avoid the effects of diurnal variation, all testing was conducted at the same time of the day. Laboratory ambient temperatures were controlled to 22 ± 1° C and a relative humidity of 62 ± 2%, monitored by a digital environmental station (Vaisala Oyj, Helsinki, Finland). All the participants had followed their normal diet, consuming a light meal with a caffeine-free beverage >3 hours before testing. They had normal sleep and avoided vigorous activity for 24 hours before testing. Data were first gathered on the absolute and relative reliability of the 15-secT. Each subject performed the 15-secT twice, with a maximum of 7 days between the tests. A PPO increase of >8% was considered a distorted response (e.g., because of test learning) (3); 11 subjects showing such a response were omitted from the study, leaving a final sample of 58 subjects. For test validation, 43 of the 58 subjects (21 men and 22 women) (mean ± SD: age 21.9 ± 1.9 years, stature 1.67 ± 0.1 m, mean body mass 71.3 ± 12.4 kg) performed the 30-secT and the 15-secT in random order, with a minimum of 2 days between assessments.

Both the 15-secT and 30-secT tests were conducted on a “Monark model 894E” cycle-ergometer (Monark, Vansbro, Sweden), fitted with toe clips and heel straps, and individually adjusted for saddle height. The warm-up (35) for each test consisted of alternating 30-second periods of active rest (zero-resistance pedaling at 60 rpm) and three 30-second bouts of exercise at increasing resistance (25, 50, 75% of the definitive test loading). The 15-secT and 30-secT sprints were begun from a standing start (i.e., with no preacceleration) and were performed against a constant resistance (7.5% of the subject's body mass [3,5,22,33]). The participants were instructed to accelerate as rapidly as possible and were strongly encouraged throughout. Outcomes (expressed as PPO, MPO, and FI) were calculated according to accepted procedures (5,7), PPO and MPO being expressed as absolute values (watts), relative to body mass (watts per kilogram) and relative to the 0.67th power of body mass (W·kg0.67) (13).

To quantify negative side effects, the subjects provided a yes or no answer with respect to the presence of nausea, light headedness, leg fatigue, and any other physical side effects immediately after the WAnT15 (22).

Statistical Analyses

Data analyses were performed using SPSS version 17.0 for Windows. Hypotheses were evaluated by parametric tests. Means and SDs were calculated after verifying the normality of distributions using Kolmogorov-Smirnov statistics. Systematic bias was investigated using a dependent t-test to evaluate the hypothesis that there was no significant mean difference between test and retest values. Estimates of effect size, mean differences, and 95% confidence intervals (CIs) protected against type 2 errors (15). The relative reliability of the 15-secT was determined by calculating ICC, and the absolute reliability was expressed in terms of the SEM and coefficients of variation (CV) (19,23,32). To complement the SEM, the SWC was reported; this is based on a rearrangement of Cohen's d effect size, where the smallest worthwhile effect (0.2) is multiplied by the between-subject SD (25). The sensitivity of the test was assessed by comparing the SWC and SEM, using the thresholds proposed by Liow and Hopkins (25). In brief, if the SEM is smaller than the SWC, the ability of the test to detect a change is “good”; if SEM equals SWC, then the test is “satisfactory,” but if the SEM is greater than the SWC, then the test is rated as “marginal.” Heteroscedasticity was assessed using a zero-order correlation between the absolute residuals and predicted scores for each participant. Knowledge of the SEM also allowed calculation of the MDC95; this reflects the 95% CI of the difference in score between paired observations, calculated as

(6,18,23). Pearson's correlation coefficients tested the validity of 15-secT data relative to the corresponding 30-secT values. Significance for all the statistical tests was accepted at p ≤ 0.05.


In all the subjects, side effects at the end of the 15-secT were limited to local muscle fatigue. Residual data for the between trials comparison were normally distributed (Kolmogorov-Smirnov p values for all outcomes ranging from 0.06 to 0.27). Respective heteroscedasticity coefficients for absolute PPO, relative PPO, derived PPO, absolute MPO, relative MPO, derived MPO, and FI were all small and statistically nonsignificant (r = 0.25 [p = 0.69], r = 0.07 [p = 0.59], r = 0.13 [p =0.32], r = 0.14 [p = 0.27], r = 0.07 [p = 0.57], r = 0.07 [p = 0.57], and r = 0.05 [p = 0.69], respectively). Dependent t-tests evaluating the equality of means showed no significant test-retest bias for absolute PPO (t = 0.04, p = 0.07, dz = 0.005), PPO relative to body mass (t = 0.01, p = 0.99, dz = 0.001), or PPO relative to body mass0.67 (t = 0.01, p = 0.99, dz = 0.001); absolute MPO (watts) (t = 1.52, p = 0.13, dz = 0.16), MPO relative to body mass MPO (t = 1.58, p = 0.11, dz = 0.16), or MPO relative to body mass0.67 (t = 1.55, p = 0.12, dz = 0.19) or FI (t = 1.37, p = 0.17, dz = 0.17) (Table 1). The estimated effect size (dz) was more than trivial for all outcomes. Test-retest reliability coefficients (ICC) ranged from 0.85 to 0.99 (Table 1), with the largest coefficients (0.989 and 0.993, respectively) for absolute PPO and MPO and the lowest ICC (0.854) for FI.

Table 1
Table 1:
Relative reliability of the 15-secT as seen in duplicate measurements with a 7-day intertest interval.*

The SEM % values for the 15-secT were all under 10%, the smallest (2.6%) being for absolute MPO and the largest (9.6%) for FI. When expressed in terms of percentage, the largest MDC95 was 26.6 for FI, whereas the smallest MDC95 was 7.3 for absolute MPO (Table 2).

Table 2
Table 2:
Absolute reliability indices and MDC95 of the 15-secT.*

Comparisons between the 15-secT and 30-secT data showed no significant bias for PPO, irrespective of their form of expression (dz ranged from 0.01 to 0.15) (Table 3). However, there were significant difference intertest differences in MPO whether expressed in absolute units (621 ± 152 and 523 ±124 W, respectively, t = 10.82; p < 0.0001; dz = 0.69), relative to body mass (8.5 ± 1.6 and 7.3 ± 1.2 W·kg1, respectively, t = 10.22; p < 0.0001; dz = 0.84), or relative to body mass0.67 (35 ± 7 and 29.7 ± 5.5 W·kg−0.67, respectively, t = 11.34; p < 0.0001; dz = 0.86) MPO. The FI values from the 30-secT (60.5 ± 9.6%) were also larger (t = 13.70; p < 0.0001; dz = 2.39) than for the 15-secT (39.0 ± 8.2%). Nevertheless, the 15-secT absolute, relative, and derived MPO, and FI were all significantly correlated with the corresponding 30-secT values (Table 4), with Pearson correlation coefficients significant at p < 0.001 (Table 3). The corresponding regression equations are shown in Table 4.

Table 3
Table 3:
Comparisons of peak power output, mean power output, and fatigue index between the 15-secT and 30-secT.*
Table 4
Table 4:
Regressions equations (±SE) to estimate scores on a 30-secT from MPO and FI as measured during a 15-secT.*


Although there are logical reasons to suggest that a 15-secT has greater construct validity than a 30-second effort when evaluating participants in field sports, the test-retest reliability of observations is also critical if scores are to be used in guiding such decisions as a return to sport or physical activity after injury. It might be thought that the reliability of testing would decrease with shortening of the period of observation, although pursuing a fatiguing test for a further 15 seconds could also make it unreliable. This is the first time that the reliability of an abbreviated Wingate anaerobic test has been evaluated in physical education students. The results suggest that despite the shorter period of observation, the 15-secT offers a highly reliable evaluation of both PPO and MPO. Reliability of the 30-secT has been studied by numerous investigators (5,21,36) but only in terms of Pearson product correlations (which ranked from 0.90 to 0.97); this index does not allow the estimation of SEM or SWC.

Quantification of the measurement error at a given time is important in interpreting individual test scores (38), for instance when examining the efficacy of an intervention such as a training regimen or use of a nutritional supplement. Our assessment of the 15-secT is based on 2 complementary indices—relative and absolute reliabilities. Relative reliability is indicated by the magnitude of the ICC (39), although scores depend also on the heterogeneity of the sample (10). Our values were excellent for relative (0.984), absolute (0.989), and derived (0.985) PPO and for relative (0.985), absolute (0.993), and derived (0.988) MPO but were only moderate for FI (0.854) (Table 1). Absolute reliability is indicated by the SEM, the CV, and the limits of agreement. The choice of index depends on the extent of heteroscedasticity within the data set; if data are heteroscedastic, CV is the appropriate measure of absolute reliability, but if data are homoscedastic, the absolute or relative SEM is the appropriate choice (9,39).

Estimation of the limits of agreement is recommended by some statisticians (9), but according to Hopkins (19), this measure of absolute reliability is too stringent to determine whether a change in an individual's score is real or is an artifact of measurement error. Given that our data were homoscedastic (Table 1), we elected the SEM to examine absolute reliability. The FI showed a larger SEM % (9.6%) than did MPO or PPO, independently of the mode of expression (Table 2). As might be expected, absolute, relative, and derived PPO (3.7, 3.5, and 3.7%, respectively). All showed a somewhat larger SEM % than the MPO scores; the most sensitive index of real change in the 15-secT outcome was the MPO, irrespective of how this was expressed (Table 2). We also calculated the likelihood that differences in 15-secT outcomes were substantial (i.e., SWC larger than the SEM); this was the case for both PPO and MPO, irrespective of how these variables were expressed (Table 2), indicating that such data have a good potential to detect real changes in the maximal power output of the legs.

In contrast, the SWC for the FI (1.8%) was smaller than its SEM (4.4%) (Table 2). Thus, one may question the use of performance decrement analyses when characterizing the maximal power of the legs. Oliver (30) has suggested that the mathematical procedures involved in calculating fatigue levels can influence their reliability. This was the case with our data; the relative reliability of FI (as shown by the ICC) was moderate (r = 0.854), but in terms of absolute reliability this measure was unsatisfactory. Atkinson and Nevill (2) have underlined that although relative reliability is conceptually useful, results are sufficiently affected by sample heterogeneity that a high correlation can coexist with a measurement error that is unacceptable for some analytical goals.

Assessment of an apparent change in performance depends on the magnitude of the change in score relative to error size (MDC95). The MDC95 for the absolute PPO value was 87.1 W; thus, a change in the PPO score exceeding 87.1 W can be accepted as a true response (2). In 95% of instances, the 15-secT will demonstrate a random variation of <87.1 W, 1.1 W·kg1, 4.7 W·kg0.67 and 46.8 W, 0.64 W·kg1, 2.6 W·kg0.67 for absolute, relative, and derived PPO and MPO, respectively, and 12.2 for FI. In addition to reliability, the 15-secT eliminated most of the unpleasant symptoms reported after a 30-secT, the only complaints being of “leg exhaustion.” This could encourage compliance with retesting and increase the test–retest reliability.

There are several points of caution to be emphasized when evaluating our findings. Perhaps most importantly, 11 of our 69 subjects were excluded from the analysis; if these individuals had been included, our estimates of reliabilities would have been poorer. Possibly, the effects of learning could have been overcome by greater familiarization, but insistence on repeated preliminary testing would limit the value of the test. Our subjects were also young athletes who were accustomed to undertaking maximal anaerobic effort. The reliability of test scores would likely have been smaller in older and less well-trained individuals. Moreover, in older subjects, the percentage effect from a given level of unreliability would be greater in terms of both PPO and MPO, because their absolute scores would be substantially smaller. Finally, the power outputs that we obtained should not be interpreted in any absolute sense, because there were inevitably substantial frictional losses of power in the chain and pedal bearings of the cycle ergometer.

The raw data in the validity tests showed some evidence of heteroscedasticity (Table 3). Natural log transformation of raw data of absolute, relative, and derived PPO of 30-secT and 15-secT reduced heteroscedasticity from r = 0.41 (p = 0.018) to r = −0.40 (p = 0.034) for absolute PPO, from r = 0.27 (p = 0.135) to r = −0.45 (p = 0.015) for relative PPO and from r = −0.032 (p = 0.87) to r = −0.46 (0.01) for derived PPO and gave a mean bias ± the 95% limits of agreement of 0.01 ± 0.05, 0.001 ± 0.07, and 0.01 ± 0.06 for absolute, relative, and derived PPO, respectively. The use of antilogs resulted in a mean bias on the ratio scale of 1.01, 1.00, and 1.01 and an agreement of ×/÷1.05, ×/÷1.07, and ×/÷1.06 for absolute, relative, and derived PPO, respectively. For any new subject, the biases of 0.05, 0.07, and 0.06% for absolute, relative, and derived PPO, respectively, would seem to be negligible; 95% of the ratios for 15-secT PPO should lie between the values 0.96 (1.01 ÷ 1.05) and 1.06 (1.01 × 1.05) for absolute PPO, between 0.93 (1 ÷ 1.07) and 1.07 (1 × 1.07) for relative PPO and between 0.95 (1.01 ÷ 1.06) and 1.07 (1.01 × 1.06) for derived PPO. In other words, the absolute, relative, and derived PPO as determined by the 15-secT would differ by not >1, 0.1, and 1%, respectively, from the PPO as measured by a 30-secT (28). To put these limits of agreement in a practical context, if a subject presented with an estimated 15-secT absolute PPO of 800 W, the worst case scenario is that on a 30-secT this athlete would develop a PPO as low as 800 × 0.96 = 768 W or as high as 800 × 1.06 = 848 W.

All correlations between 15-secT and 30-secT indices were statistically significant (Table 3). Somewhat surprisingly, although the duration of the 15-secT was only half that of the standard Wingate test, the MPO apparently still reflected the individual's anaerobic capacity. One limitation of our study was that we did not examine blood lactate values and acid-base responses.

In conclusion, this study suggests that motivation is easier for the 15-secT than for the 30-secT. The PPO and MPO as calculated from the 15-secT have sufficient reliability to detect small but practically significant changes in the peak performance of fit athletes, with the MPO score being most responsive to real change. Further, 15-secT outcomes are significantly correlated with traditional 30-secT indices.

Practical Applications

The 15-secT offers a reliable and valid tool for evaluating the maximal power output of the legs in young adults, and it seems to be the appropriate test to adopt when assessing participants in field sports and 50- or 100-m track events. Strength and conditioning coaches may benefit from the repeated use of the 15-secT as a simple practical means of detecting changes in an athlete's sprint performance. In particular, the determination of the MDC95 may help conditioning specialists when assessing responses to a particular training program or dietary regimen.


The authors would like to thank the subjects for their enthusiastic participation and the “Ministère de l'enseignement supérieur et de la Recherche Scientifique, Tunisia” for financial support.


1. Andersen KL, Shephard RJ, Denolin H, Varnauskas E, Masironi R. Fundamentals of Exercise Testing. Geneva, Switzerland: World Health Organization, 1971.
2. Atkinson G, Nevill AM. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med 26: 217–238, 1988.
3. Bar-Or O. The Wingate anaerobic test. An update on methodology, reliability and validity. Sports Med 4: 381–394, 1987.
4. Bar-Or O, Berman L, Salsberg A. An abbreviated Wingate anaerobic test for women and men of advanced age. Med Sci Sports Exerc 24(Suppl.):S22–1992.
5. Bar-Or O, Dotan R, Inbar O. A 30 s all-out ergometric test: Its reliability and validity for anaerobic capacity. Isr J Med Sci 13: 126, 1977.
6. Beckerman H, Roebroeck ME, Ankhorst GJ, Becher JG, Bezemer PD, Verbek AL. Smallest real difference, a link between reproducibility and responsiveness. Qual Life Res 10: 571–578, 2001.
7. Beekhuizen KS, Davis MD, Kolber MJ, Cheng MS. Test-retest reliability and minimal detectable change of the hexagon agility test. J Strength Cond Res 23: 2167–2171, 2009.
8. Beneke R, Pollmann C, Bleif I, Leithäuser RM, Hütler M. How anaerobic is the Wingate Anaerobic test for humans? Eur J Appl Physiol 87: 388–392, 2002.
9. Bland JM, Altman DA. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: 307–310, 1986.
10. Bland JM, Altman DG. Comparing two methods of clinical measurements: A personal history. Int J Epidemiol 24(Suppl.): 7–17, 1995.
11. Chelly MS, Cherif N, Ben Amar M, Hermassi S, Fathloun M, Bouhlel E, Tabka Z, Shephard R. Relationships of peak leg power, 1-RM half back squat and leg muscle volume to 5-m sprint performance of junior soccer players J Strength Cond Res 24: 266–271, 2010.
12. Chelly MS, Hermassi S, Shephard RJ. Relationships between power and strength of the upper and lower limb muscles and throwing velocity in male handball players. J Strength Cond Res 24: 1480–1487, 2010.
13. Cooper SM, Baker JS, Eaton ZE, Matthews N. A simple multistage field test for the prediction of anaerobic capacity in female games players. Br J Sports Med 38: 784–789, 2004.
14. Di Prampero PE. Anaerobic capacity and power. In: Frontiers of Fitness. R.J. Shephard, ed. Springfield, IL: C.C. Thomas, 1971. pp. 155–191.
15. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 39: 175–191, 2007.
16. Glaister M, Howatson G, Pattison JR, McInnes G. The reliability and validity of fatigue measures during multiple-sprint work: an issue revisited. J Strength Cond Res 22: 1597–1601, 2008.
17. Granier P, Mercier B, Mercier J, Anselme F, Préfaut C. Aerobic and anaerobic contribution to Wingate test performance in sprint and middle-distance runners. Eur J Appl Physiol Occup Physiol 70: 58–65, 1995.
18. Haley SM, Fragala-Pinkham MA. Interpreting change scores of tests and measures used in physical therapy. Phys Ther 86: 735–743, 2006.
19. Hopkins WG. Measures of reliability in sports medicine and science. Sports Med 30: 1–15, 2000.
20. Inbar O. Bar-Or, O, and Skinner, J. The Wingate Anaerobic Test. Champaign, IL:Human Kinetics, 1996.
21. Kaczkowksi W, Montgomery DL, Taylor AW, Klissouras V. The relationship between muscle fibre composition and maximal anaerobic power and capacity. J Sport Med Phys Fitness 22: 407–413, 1982.
22. Laurent CMJ, Meyers MC, Robinson CA, Green JM. Cross-validation of the 20- versus 30-s Wingate anaerobic test. Eur J Appl Physiol 100: 645–651, 2007.
23. Lexell JE, Downham DY. How to assess the reliability of measurements in rehabilitation. Am J Phys Med Rehabil 84: 719–723, 2005.
24. Lima M. Standardization of the tethered running in sprint athletes: Analysis of relation between power and performance in running. Master's thesis, Sao Paulo State University, Rio Claro, Brazil, 2007.
25. Liow DK, Hopkins WG. Velocity specificity of weight training for kayak sprint performance. Med Sci Sports Exerc 35: 1232–1237, 2003.
26. Margaria R, Cerretelli P, di Prampero PE, Massari C, Terelli G. Kinetics and mechanisms of oxygen debt contraction in man. J Appl Physiol 18: 371–377, 1963.
27. Meckel Y, Atterbom H, Grodjinovsky A, Ben-Sira D, Rotstein A. Physiological characteristics of female 100 metre sprinters of different performance levels. J Sports Med Phys Fitness 35: 169–175, 1995.
28. Nevill AM, 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.
29. Nummela A, Alberts M, Rijntjes RP, Luhtanen P, Rusko H. Reliability and validity of the maximal anaerobic running test. Int J Sports Med 17(Suppl.) S97–S102, 1996.
30. Oliver JL. Is a fatigue index a worthwhile measure of repeated sprint ability? J Sci Med Sport 12: 20–23, 2009.
31. Perez-Gomez J, Rodriguez GV, Ara I, Olmedillas H, Chavarren J, González-Henriquez JJ, Dorado C, Calbet JA. Role of muscle mass on sprint performance: Gender differences? Eur J Appl Physiol 102: 685–694, 2008.
32. Schreuders TA, Roebroeck ME, Goumans J, van Nieuwenhuijzen JF, Stijnen TH, Stam HJ. Measurement error in grip and pinch force measurements in patients with hand injuries. Phys Ther 83: 806–815, 2003.
33. Stickley CD, Hetzler RK, Kimura IF. Prediction of anaerobic power values from an abbreviated WAnT protocol. J Strength Cond Res 22: 958–965, 2008.
34. Tharp GD, Newhouse RK, Uffelman L, Thorland WG, Johnson GO. Comparison of sprint and run times with performance on the Wingate anaerobic test. Res Q Exer Sport 56: 73–76, 1985.
35. Vanderford ML, Meyers MC, Skelly WA, Stewart CC, Hamilton KL. Physiological and sport-specific skill response of Olympic youth soccer athletes. J Strength Cond Res 18: 334–342, 2004.
36. Vandewalle H, Heller J, Pérès G, Raveneau S, Monod H. Étude comparative entre le Wingate test et un test force-vitesse sur ergocycle. Sci Sport 2: 279–284, 1987.
37. Vincent S, Berthon P, Zouhal H, Moussa E, Catheline M, Bentué-Ferrer D, Gratas-Delamarche A. Plasma glucose, insulin and catecholamine responses to a Wingate test in physically active women and men. Eur J Appl Physiol 91: 15–21, 2004.
38. Wang CY, Sheu CF, Protas EJ. Test-retest reliability and measurement errors of six mobility tests in the community-dwelling elderly. Asian J Gerontol Geriatr 4: 8–13, 2009.
39. Weir JP. Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res 19: 231–240, 2005.
40. Wilson RWN, Snyder AC, Dorman JC. Analysis of seated and standing triple Wingate tests. J Strength Cond Res 23: 868–873, 2009.
41. Zagatto AM, Beck WR, Gobatto CA. Validity of the running anaerobic sprint test for assessing anaerobic power and predicting short-distance performances. J Strength Cond Res 23: 1820–1827, 2009.
42. Zemkova E, Hamar D. “All-out” tethered running as an alternative to Wingate anaerobic test. Kinesiology 36: 165–172, 2004.

maximal power; leg power; working capacity; cycle ergometer; relative reliability; absolute reliability; criterion validity

© 2012 National Strength and Conditioning Association