Measures of Reliability in the Kinetics of British Army Marching Drill

Abstract

Marching drill is considered an essential component of basic military training. Assessing the magnitude, and understanding the implications of forces experienced when performing drill movements requires the determination of reliable measures for each of the drill movements. Little is currently known regarding relative effects of systematic bias (SB), within-subject (WS) variation, or test-retest reliability to the overall reliability of kinetic variables of marching drill. PURPOSE: Statistically analyse measures of reliability (SB, WS variation and test-retest reliability) in the kinetics of marching drill. METHODS: 15 male participants performed 10 trials of; stand at ease (SaE), stand at attention (SaA), quick march (QM), halt; normal walk was included for comparison. Vertical ground reaction forces (vGRF) were sampled at 1000Hz via Kistler force plate. Rate of force development (RFD) was determined from the vGRF data. SB was calculated from a one-way Anova. WS variation and test-retest reliability were calculated as the coefficient of variation of the typical error (CVte%) and intra-class correlation coefficients (ICC), using an iterative process; adding additional data points and recalculating until acceptable values were achieved. RESULTS: There was no SB (sig. difference between sessions) for Halt, QM and Walk (p> 0.05). Statistically sig. differences were observed between session (S) 1 and S2 (p=.044) and S1 and S3 (p=.004) for SaA and between S1 and S2 (p=.011) and S1 and S3 (p=.014) for SaE. Due to SB, the initial 10 trials of SaA and SaE were removed, with the remaining 20 trials used for further reliability calculations. The CVte% and ICC values were: vGRF - SaA (CVte%=7.6, ICC=0.75), SaE (CVte%=7.0, ICC=0.82), QM (CVte%=8.4, 0.91), halt, (CVte%=7.2, ICC=0.79), walk (CVte%=2.4, ICC=0.88); and RFD - SaA (CVte%=13.8, ICC=0.75), SaE (CVte%=10.4, ICC=0.79), QM (CVte%=41.4, ICC=0.79), halt (CVte%=14.9, ICC=0.65), and walk (CVte%=49.9, ICC=0.75). CONCLUSION: Differences between S1 and the remaining 2 sessions for SaA and SaE indicate that a familiarization session was required. Beyond 8 trials diminishing returns were observed in CVte% with the addition of extra data. At 8 trials ICC values of ≥0.75 were achieved for all tasks, except the Halt drill indicating reasonable/good levels of test-retest reliability. Therefore to accurately represent the kinetics of marching drill performance 8 trials are required.