The impact of data reduction on the intra-trial reliability of a typical measure of lower limb musculoskeletal stiffness.

Double-leg repeated jumping tasks are commonly used as measures of lower limb stiffness in exercise science research. Within similar stiffness calculations, variation in data-reduction criteria exists. The impact of these varied data-reduction methods on stiffness measures is unknown. Sixteen adolescent female participants from varied physical activity backgrounds performed 15 self-paced, bent-knee continuous jumps (CJb) on two force plates. Leg stiffness was calculated as the ratio of the peak force and the centre of mass displacement for each contact. Using combinations of criteria based on previous literature, 83 data-reduction methods were applied to the raw data. Data reduction suitability was assessed based on intra-trial reliability, the number of participants excluded and the average contacts excluded. Four data-reduction methods were deemed suitable for use with adolescent female populations, with three consecutive contacts within 1 SD of the average jump frequency considered optimal. The average individual stiffness values were not greatly influenced by the data-reduction method; however, for a single participant, a stiffness change of up to 6 kN center dot m(-1) (30%) was observed. The role and potential impact of data-reduction methods used to evaluate measures of lower limb stiffness during repeated jumping tasks warrants consideration.