Test-retest Reliability Of Selected Measures Of Soldier Performance

Abstract

PURPOSE: To determine the test-retest reliability (TRR) of measures of Soldier performance. METHODS: 7 Soldier performance tests (30-m grenade throw for distance [GT-D] and accuracy [GT-A], maximum box lift [1RM], running long jump with 20.5-kg load [LJ], 2-mile march with 33-kg load [LM], 80-kg dummy drag [DD], repetitive lift and carry of a 20.5 kg box [RLC], and 5-30-m rushes with 20.5 kg load [Rush],) were performed 4 times over an 8 week period, with adequate rest between tests. Repeated measures analyses of variance (RMANOVA) were used to examine changes over time. Intraclass correlation coefficients (ICC) and 95% confidence intervals (CI) were calculated using data from all Soldiers with 4 trials for a test. Models and forms for reliability coefficients were ICC (2,1), except for GT (ICC(2,5)) and LJ (ICC(2,3)). RESULTS: RMANOVA revealed no differences over time for GT-D, LJ, LM, DD or Rush. RLC and 1RM stabilized by the second trial and GT-A by the third trial. Table 1 lists the ICC, CI and sample size for each test. Three tests (GT, 1RM and LJ) had significantly higher overall TRR compared to 3 tests with lower TRR (CD, RLC, and Rush). By virtue of a CI that overlapped both high- and low-TRR groups, the TRR for LM was not different from either group. Table 1. Test-Retest results.Table 1: Test-Retest Reliability result.CONCLUSIONS: The measures of soldier readiness with the greatest TRR were GT, 1RM and LJ. Based on CI overlap, the LM was not different from the higher or lower TRR group. These 4 tests could be used to provide a reliable measure of soldier readiness.