Validity of block start performance without arm forces or by kinematics-only methods.

The purpose of this study was to validate the calculation of reaction time (RT) and normalised power in block starts without considering arm ground reaction forces (GRFs) or using two kinematics-only methods. The RT and normalised power in the action phase were calculated using four different methods: using GRFs of arms and legs by force plates (whole F-based method), which can be regarded as the most valid method, using GRFs of legs captured by force plates (legs F-based method), using position of the centre of mass of the entire body captured by high-speed cameras (whole P-based method), and using only a partial subset of segment position (partial P-based method). Bland-Altman plots demonstrated that the RT of the legs F-based method was not similar to that of the whole F-based method: the mean difference was 7.4 ms and the 95% limits of agreement was-45.1 to 59.8 ms, and it was the least valid method for the calculation among the four methods. In contrast, the normalised power was more valid in the legs F-based method, followed by whole and partial P-based methods. This information will help researchers and practitioners to decide upon their analysis methods when analysing block start performance.