This study examined the intra-session reliability of sprint performance on a non-motorized treadmill amongst healthy, active male and female adults. One hundred and twenty participants (males n = 77; females n = 45) completed two familiarization sessions, followed by a third session that consisted of three trials (T1, T2, T3) of maximal sprints (4-s), interspersed by three minutes of recovery. Combining males and females exhibited moderate-to-excellent test-retest reliability (intra-class correlation coefficient, ICC), minimal measurement error (coefficient of variation, CV) and trivial differences between trials (effect size, ES) for speed, power, total work and acceleration (ICC = 0.82-0.98, CV = 1.31-8.45%, ES = 0.01-0.22). The measurement error was improved between comparisons of T1 vs. T2 (CV = 1.62-8.45%, ES = 0.12-0.22) to T2 vs. T3 (CV = 1.31-6.56%, ES = 0.01-0.07) and better for females (CV = 1.26-7.94%, ES = 0.001-0.26) than males (CV = 1.33-8.53%, ES = 0.06-0.31). The current study demonstrated moderate-to-excellent reliability and good-moderate measurement error during a 4-s sprint on a non-motorized treadmill. However, sex had a substantial impact with females exhibiting better values. Practitioners should employ at least two separate trials within a session, in addition to multiple familiarization sessions, to achieve reliable non-motorized treadmill sprint performances.
Keywords: measurement error; minimal detectable change; power; running; sex; speed.
Copyright: © Academy of Physical Education in Katowice.