Figure 7:

(a) Correlation between the DFC estimate and nuisance norm versus the RMS correlation between the raw nuisance signal and the two seed time courses. Correlation thresholds are indicated by the black dashed lines such that points that lie above and below these lines correspond to cases in which the DFC estimates are considered to be ‘strongly correlated’ (|r| > 0.5 with empirical p ≤ 0.09) with the nuisance norms such that nuisance norms explain more than 25% of the total variance in the DFC estimates. The relative density of correlations (maximum density is normalized to 1.0) is indicated by the color map on the right-hand side. The relative density is computed by summing the total number of data points in smaller sub-grids and normalizing by the total number of points. (b) The blue bars show the percentage (%) of cases in which the nuisance norms explain more than 25% of the total variance in the DFC estimates as a function of the RMS correlation between the raw nuisance and the two seed time courses. The red dotted line (dots are located at bin centers) shows the cumulative sum of the percentage of strong correlations as a function of the RMS correlations. The green bars show the percentage (%) of strong correlations that are accounted for by the GS regressor.