A Mixed-Methods Approach to Evaluating the Internal Validity of the Reactive Strength Index.

Abstract

The reactive capacity of the muscle-tendon complex is commonly assessed using the reactive strength index (RSI). Conventionally, the RSI is a ratio of rebound jump height to ground contact time in depth jumping. Several assumptions regarding the linear mechanics acting through the whole-body center of gravity may threaten the internal validity of computation and interpretation of RSI scores. First, it is common for rebound jump height to be predicted from rebound jump flight time. This assumes that the angular positioning of body segments is equivalent at the time instances of rebound jump take-off and landing. Prior literature supports a mixed-methods approach for computing the RSI that is void of this assumption. The mixed-methods approach gives a more valid estimation of rebound jump height. In this approach, rebound jump height is estimated from rebound jump take-off velocity of the whole-body center of mass. This is accomplished by subtracting an estimate of impact velocity, acquired using videography, from change in whole-body center of mass velocity estimated from integrated vertical ground reaction force data. Second, it is often assumed that vertical displacement of the whole-body center of mass during the drop phase of the depth jump is predicted perfectly from the height of the platform used to perform the drop. This assumption may affect the internal validity of comparing RSI scores across individuals and within individuals performing depth jumps from varied heights. The purpose of the present study was to investigate the internal validity of RSI scores computed using the conventional approach and impact velocity variability, which may affect the interpretation of RSI scores. Seventy physically active young adults performed depth jumps from drop heights of 0.51, 0.66, and 0.81 m. RSI was computed using the conventional approach and a mixed-methods approach featuring the use of 2-dimensional videography, body segment parameters, and force platform dynamometry. The two computational methods were compared using linear regression performed on data from each drop height. In addition, a 2 (computational method) by 3 (drop height) Analysis of Variance (ANOVA) was performed to evaluate for main effects and interactions in RSI data. Multiple one sample t-tests were performed to compare estimated and theoretical impact velocities. The ANOVA revealed no main effect or interactions between computational approaches (p = 0.467–0.938). Linear regression revealed moderately strong associations between RSI scores computed using the conventional and mixed-methods approaches (R2 = 0.685–0.741). Moreover, linear regressions revealed that the conventional approach tends to overestimate the mixed methods approach for RSI scores below 1.0 and underestimate the mixed methods approach for RSI scores above 1.0. Lastly, estimated impact velocities were observed to be as much as 13% lower versus theoretical (p < 0.001). Researchers with access to motion capture and force platform technology may consider using a mixed-methods approach for computing the RSI, which likely maximizes the internal validity of scores. In addition, results suggest for practitioners to practice caution when comparing conventional RSI scores across individuals.