FIGURE 2.

The diagram shows the entire data analysis process from the step-by-step gait features extraction to the ANCOVA analysis, considering a single exemplifying gait feature (the range of motion of the ankle: “ROM ankle”). Light pink panel shows the single subject analysis (run subject per subject). (a) At first the gait features are extracted for each step, both for the right and left side. Then, (b) normality is checked considering the values of all the M steps and, if verified, (c) the mean value over all the steps is computed, for both the right and left side. Light blue panel depicts group analysis (run across all subject of TD and ASD group). Left and right (mean) values for each subject are collected and (d) normality is tested; (e) t-test or Wilcoxon test, depending on the data distribution, is performed to check for differences between left and right side. (f) When difference are not significant the mean value between left and right sides is computed for each subject. (g) Shapiro–Wilk test is run to check for normality within groups and, since verified, (h) ANCOVA analysis with QI as covariate is performed.