Choice of performance measure affects the evaluation of print start performance

Previous sprint start analyses have evaluated performance using numerous diverse measures. Mendoza and Scho¨ llhorn (1993: Journal of Sports Sciences, 11, 25 – 29) found different performance measures (block velocity and time to 10 m) to affect the perceived success of experimental alterations to block phase technique. The selection of performance measure could therefore potentially influence the conclusions drawn when studying the sprint start. The aim of this study was to determine whether the choice of performance measure influences the performance-based ranking of a series of sprint starts. Following ethical approval, one elite male sprinter (age¼19 years, 100 m PB¼10.22 s) completed 15 maximal sprints to 30 m during four training sessions. Two-dimensional video data were recorded (200 Hz) in order to obtain full-body kinematics and to identify movement onset and block exit timings. Performance was assessed using six discrete measures (block velocity, average block acceleration, average block power, time to 2 m, velocity at 2 m, velocity at the end of the first step). Trials were ranked from best to worst for each performance measure. Using these ordinal data, Spearman’s rank order correlation coefficients (r) were calculated between each pair of performance measures. Of the 15 pairs of performance measures, none were perfectly correlated and only three produced significantly (P50.01) similar rankings. Average block acceleration was correlated with both block velocity (r¼0.71) and average block power (r¼0.78). Average block power was also correlated with time to 2 m (r¼0.83). None of the remaining 12 pairs were significantly correlated (all P40.05), and seven pairs showed negative r values, indicating a tendency for rank order to be reversed. The choice of performance measure was found to influence the performance-based rank order of a series of sprint starts. The conclusions obtained from research attempting to associate certain aspects of sprint start technique with improvements in performance are therefore influenced by the choice of performance measure. These data do not indicate whether one performance measure is superior to others. However, from a theoretical point of view, it is suggested that average block power may provide the most valid overall measure of sprint start performance, because it incorporates the changes in displacement, velocity, and time.