Fig. 2.

The hierarchical clustering analysis using Ward’s method. Each row is a participant, with autistic participants being marked as gold in the small column on the left, while typically developing participants are blue. The four main columns show normalized GFP in each loudness condition. Each column also depicts changes in normalized GFP over time, with earlier time-points being on the left of each column and later time-points on the right of each column. As shown in the scale in the upper left corner, smaller values, (reflecting a weaker normalized GFP in the loudness condition) are redder, while larger values (reflecting a stronger GFP) are yellower. A histogram showing the distribution of individual data points (individual data points represent a participant’s normalized GFP value in a given condition and at a given time-point) is superimposed over the scale. The dendrogram on the far left shows hierarchical clusters of similar participants. The horizontal lengths of the dendrogram branches represent the distance between clusters. The clusters selected for the purposes of this analysis are separated by blank space, and group numbering proceeds from top to bottom (i.e., C1 is at the top, C4 is at the bottom)