Abstract
The study aimed to quantify the measurement error / reliability of a musculoskeletal profiling test battery administered in young, elite academy soccer players, and to examine if the order in which the test battery was administered, and who it was administered by, influenced reliability. Players (n = 75; age 12-20 years; stature 1.47-1.95 m; body mass 36-89 kg) from U-12 to U-23 age groups were assigned to either: 1) intra-rater-fixed order; 2) intra-rater-non-fixed order; 3) inter-rater-fixed order; or, 4) inter-rater-non-fixed order groups. On two separate occasions separated by 3 to 7 days, 12 raters conducted a musculoskeletal profiling test battery comprising 10 tests (Supine Medial Hip Rotation, Supine Lateral Hip Rotation, Hamstring 90/90, Prone Medial Hip Rotation [degrees]; Combined Elevation, Thoracic Rotation, Weight-Bearing Dorsiflexion, Y-Balance [centimetres]; Beighton, Lumbar Quadrant [categorical]). The measurement error / reliability for tests measured in degrees and centimetres was evaluated using the intraclass correlation (relative reliability), coefficient of variation and ratio limits of agreement (absolute reliability). Intraclass correlations varied from 0.04 ("poor") to 0.95 ("excellent"), coefficient of variation from 2.9 to 43.4%, and the ratio limits of agreement from 1.058 (*/÷ 1.020) to 2.026 (*/÷ 1.319) for the tests measured in degrees and centimetres. The intraclass correlation, coefficient of variation and ratio limits of agreement were smallest for five out of eight tests measured in degrees and centimetres when the tests were administered in an intra-rater-fixed test order. These findings emphasise that different testing methods, and the administration of a musculoskeletal profiling test battery using a less than optimal design, will influence measurement error and hence test reliability. These observations need to be considered when investigating musculoskeletal function and age, injury, training or asymmetry in young, elite academy soccer players.