Table 1. The effects of age, sex, and IQ on global network properties.
Age effecta | Sex effectb | Age-sex interactionb | IQ effectc | |||||||
T-score | p-value | Model | T-score | p-value | Model | T-score | p-value | r | p-value | |
C | 1.491 | 0.142 | −1.970 | 0.055 | −0.484 | 0.630 | −0.126 | 0.380 | ||
L | 0.365 | 0.717 | −2.295 | 0.026 | F>M | −0.106 | 0.916 | −0.100 | 0.487 | |
NC | 2.254 | 0.029 | L+ | 0.037 | 0.971 | 1.259 | 0.214 | 0.064 | 0.657 | |
NL | 0.459 | 0.648 | −2.194 | 0.033 | F>M | 0.112 | 0.911 | −0.106 | 0.458 | |
SW | 2.170 | 0.035 | L+ | 0.180 | 0.858 | 1.230 | 0.225 | 0.069 | 0.630 | |
LE | 2.297 | 0.026 | L+ | −1.734 | 0.089 | −0.421 | 0.676 | −0.093 | 0.516 | |
GE | −0.355 | 0.724 | 2.285 | 0.027 | F<M | 0.110 | 0.913 | 0.075 | 0.602 |
Two multiply linear regressions that modeled age and age2 as predictors, along with sex as a covariate; the best model was determined by AIC.
A multiply linear regression that modeled age, sex, and age-sex interaction.
Pearson's correlation analysis between IQ and global network properties, each of which was regressed by a multiply linear regression that modeled age, sex, and age-sex interaction.
L+: Linear regression model showing significant positive correlation. F>M: female shows significantly higher values than male; F<M: female shows significantly lower values than male. Significances are set at p<0.05 and shown by bold characters.
C, clustering coefficient; L, characteristic path length; NC, normalized clustering coefficient; NL, normalized characteristic path length; SW, small-worldness; LE, local efficiency; GE, global efficiency.