Fig. 6.

Predictive power of individual infant/toddler functional networks against null models. The distribution of partial correlation coefficients between true and predicted age (adjusted for nuisance variables) is shown light blue and corresponding null model distributions are shown in grey. *Indicates p < .05 for each network (comparison of true model’s median of partial r with the null distribution from 500 bootstraps) with Holm-Bonferroni correction for multiple comparisons. Networks are ordered by size, with the largest network (aFPN) on the left. The dashed blue line shows the median partial r of full within-network model from the analyses in Fig. 5 for comparison.