Fig. 2.

Correlations between resting state network amplitude, between subject motion and behavioural modes. For between subject amplitude-vs-motion correlations, root mean square relative (timepoint-to-timepoint) motion was averaged across each subject's 4 runs before calculating the between-subject correlation against network amplitude. For the between-subject correlations with the CCA modes of covariation, the subject-weight variables for the first two modes obtained from the CCA analysis were correlated against amplitudes. Dotted lines indicate the minimum correlation required to pass two-tailed significance testing (p < 0.05, Bonferroni corrected for multiple comparisons across 25 networks; |r| = 0.11).