*2.3. Data Analysis*

From the acquired and calibrated dynamic data (i.e., Fz, Mx and My), the UCOP coordinates, XUCOP and YUCOP, with respect to the chair origin were calculated for each phase according to Equation (1), removed of their average value for each phase, and low-pass filtered through a third-order zero phase Butterworth digital filter with a 20-Hz cut-o ff frequency. This produced, for each phase, a 2D time series, {XUCOPA,YUCOPA}, {XUCOPB,YUCOPB} and {XUCOPC,YUCOPC}, respectively. Then, the corresponding resulting distances RDA, RDB and RDC (i.e., the magnitude of the displacement of the instantaneous position of UCOP from its mean value) were calculated. From these time series, in agreemen<sup>t</sup> with the definitions given in [22], mean distance values (MD, MDX and MDY), RMS distance values (RDIST, RDISTX and RDISTY), range values (RANGE, RANGEX and RANGEY), Mean Velocity values (MVELO, MVELOX and MVELOY), the 95% confidence circle area value (AREA-CC), the 95% confidence ellipse area value (AREA-CE) and the sway area value (SWAREA) were calculated for every subject and for each phase. Moreover, for each parameter and for each phase, the mean value and the standard deviation were calculated for all subjects. Pairwise t-tests were conducted with phases as a grouping factor for all the calculated parameters. In order to verify whether there was a significant di fference in parameters among the three phases, a p-value < 0.05 was chosen, but also di fferent acceptable values of p (i.e., p < 0.01, p < 0.001 and p < 0.0001) were selected to indicate di fferent levels of statistical e ffect.
