IDENTIFICATION AND CORRELATION OF HUMAN FOOTFALL LOAD PARAMETERS USING MULTIVARIATE-ANALYSIS

This research had two main objectives: to identify and quantify the mu ltiple reaction parameters of human footfall load histories for 24 sub jects; and to seek statistical correlations of the reaction parameters with two gaits: fitness walking and running, with two footfall surfac es: rigid and mat, and with two subject attributes: gender and arch in dex. These reaction parameters, measured,vith a force plate, include t he subjects' foot reaction forces in the three orthogonal directions a nd the particular features of these forces such as their duration, ave rage values, peak values, and rates of loading. An automated data retr ieval-software system evaluated these reaction parameters. The statist ical correlations were made using principal component analysis (PCA), a method that projected the 13 identified footfall reaction parameters onto subsets of three or four parameters called principal components that contained most of the variance of the original thirteen. The resu lts, among others, show couplings between the vertical and the peak me dial lend, but an uncoupling of the posterior-anterior loads with the lends in the other two directions.