Skip to main content
[Preprint]. 2023 Nov 18:2023.05.11.540387. Originally published 2023 May 12. [Version 2] doi: 10.1101/2023.05.11.540387

Table 3.

Comparison of prediction accuracies between states

Domain of predicted performance State A RMSE mean (SD) State B RMSE mean (SD) t p-value
Significant main effect of state
WM REST 1.10 (0.06) WM 1.02 (0.05) 5.10 < 0.001
REST 1.10 (0.06) SOCIAL 1.05 (0.06) 3.49 0.013
REST 1.10 (0.06) EMOTION 1.05 (0.06) 3.68 0.007
Significant post-hoc tests for specific state—task—network combinations
WM REST (n-back, Power nodes) 1.06 (0.09) WM (n-back, Power nodes) 0.95 (0.09) 4.83 < 0.001
REST (n-back, EMO network) 1.13 (0.10) WM (n-back, EMO network) 1.00 (0.10) 3.51 0.016
REST (List Sorting, EMO network) 1.19 (0.09) WM (List Sorting, EMO network) 1.05 (0.09) 4.10 0.002
REST (n-back task, Power nodes) 1.06 (0.09) SOCIAL (n-back task, Power nodes) 0.98 (0.09) 3.21 0.043
REST (n-back task, Power nodes) 1.06 (0.09) EMOTION (n-back task, Power nodes) 0.97 (0.09) 3.72 0.008
REST (n-back, EMO network) 1.13 (0.10) EMOTION (n-back, EMO network) 1.00 (0.10) 3.40 0.023
REST (List Sorting, EMO network) 1.19 (0.09) EMOTION (List Sorting, EMO network) 1.10 (0.09) 3.32 0.030

Note. Machine-learning-adjusted t-test to assess state specificity using the averaged 100 RMSE values obtained from 100-fold cross-validation within the state listed in column “State A” versus the state listed in column “State B”. p-values are Bonferroni corrected for multiple comparisons. Post-hoc t-tests between individual predictions of the task in the network (both noted in brackets) and the state listed in column “State A” versus the state listed in column “State B”.