Fig. 3.

The similarity matrix S(TBI, AD) displays statistically significant equivalences of functional correlation ρ between the TBI and AD groups. Each cell S_ij encodes the result of testing the null hypothesis of equivalence in ρ_ij across these two groups. This null hypothesis is only tested if both TBI and AD differ significantly from HC. In other words, if there is no significant mean difference in ρ between/either (a) HC and TBI and/or (b) HC and AD, S_ij is drawn in white. If both study groups differ from the HC group but no significant statistical equivalence is found across TBI and AD, S_ij is also drawn in white. Elsewhere, the color-coded quantity is the value of the TOST procedure’s t statistic with the smallest magnitude. The color-coded quantity is a t statistic with 63 df. If a statistical equivalence is associated with a relatively stronger correlation in both TBI and AD relative to HC, t is positive and S_ij is drawn in red; if the correlation is weaker relative to HC, t is negative and S_ij is drawn in blue. Boundaries between M_α and M_β are delineated by thick black lines; boundaries between the submodules of M_β are delineated by thinner lines. See the caption of Fig. 1 for abbreviations. (B) Graph representation of S(TBI, AD). Nodes are color-coded and grouped by module. Edge colors encode t score values, according to the color bar in (A)