FIG. 3.
Variability of walking as a function of AP and ML visual perturbations. Step variability data (filled circles), defined as root-mean-square (RMS) deviations of step width and length, are plotted against perturbation amplitude. Data were fit well by a linear relationship (solid lines for linear regression fits); we defined perturbation sensitivity as the change in step variability per unit change in perturbation amplitude (i.e., slope of linear fit). Step width generally exhibited greater variability than that of step length. Only ML perturbations produced significant changes in step variability. Error bars denote SD. Asterisks (*) denote significant sensitivity (P < 0.05).