Fig. 8.

Path lengths and clustering coefficients for surrogate data sets. A: path lengths (left) and clustering coefficients (right) for surrogate data sets fit to randomly rewired correlation matrices compared with Erdos-Renyi random networks. Randomly rewired networks match the path lengths and clustering coefficients of Erdos-Renyi networks. B: path lengths (left) and clustering coefficients (right) for surrogate data sets fit to small-world correlation matrices compared with Erdos-Renyi random networks. Small-world networks have similar path lengths and greater clustering compared with Erdos-Renyi random networks. C: path lengths (left) and clustering coefficients (right) for surrogate data sets fit to maximally clustered correlation matrices compared with Erdos-Renyi random networks. Clustered networks have greater path lengths and clustering than Erdos-Renyi random networks. ***P < 0.001 by Wilcoxon rank sum test.