# Also in the Article

Statistical analysis

Procedure

Significance of network-based measures were evaluated using controls and statistical tests. The statistical significance of the Louvain communities in each graph was calculated using non-parametric permutation testing (Park et al., 2009) by randomly permuting the community labels assigned to the nodes 5000 times. The final communities reported were Bonferroni corrected p < 0.05. The network measures in-degrees, out-degrees, and coreness of nodes were correlated with the power responses, the significant correlations, $probability p < 0.05$, FDR corrected were reported. All comparisons of classification accuracy among different feature spaces were reported with statistical significance for p <0.05, using t tests, also FDR corrected for multiple comparisons, among feature spaces.

Note: The content above has been extracted from a research article, so it may not display correctly.

Q&A