Table 1.

Estimates of the overlap between states and state-differences, expressed in percentage of variance explained

Comparison	Method	Overlap			Difference			Difference/Overlap
		R-SM	R-M	M-SM	R-SM	R-M	M-SM	R-SM	R-M	M-SM
Average FC	1. C - SH	87	64	63	12	29	36	0.14	0.45	0.57
	2. C - SH - MR	87	64	63	12	29	36	0.14	0.45	0.57
	3. C - SH - MM	85	63	62	13	29	36	0.15	0.46	0.58
	4. C - PR	87	64	63	13	36	37	0.15	0.56	0.58
	5. P - SH	84	61	62	15	32	38	0.18	0.52	0.61
Age effects	1. C - SH	53	26	43	34	54	46	0.64	2.08	1.07
	2. C - SH - MR	42	17	34	41	52	49	0.98	3.06	1.44
	3. C - SH - MM	35	11	24	30	43	48	0.86	3.91	2.00
	4. C - PR	53	26	43	41	66	51	0.78	2.54	1.20
	5. P - SH	47	27	41	38	52	48	0.81	1.93	1.17
Individual differences	1. C - SH	21	9	19	21	32	23	1.00	3.56	1.21
	2. C - SH - MR	19	6	10	19	26	22	1.00	4.33	2.20
	3. C - SH - MM	17	11	15	20	27	22	1.18	2.45	1.47
	4. C - PR	21	9	19	48	50	46	2.25	5.32	2.42
	5. P - SH	20	9	22	22	33	21	1.10	3.67	0.95

For each comparison that is made in the main text, we show the results of four different analysis approaches. In order these are: (1) the standard split-half analysis as reported in the main text; (2) the results when regressing out motion at the group level; (3) the results using a subset of 147 participants across the lifespan who were matched on average motion; (4) the results using phase-randomization as a null model to estimate the size of state differences; and (5) the results using a different set of ROIs by Power et al. (2011). C, Using the set of ROIs by Craddock et al. (2012); P, using the set of ROIs by Power et al (2011); SH, split-half method; MR, results after regressing out effects of average motion for each connection; MM, using a subset of 147 participants across the lifespan who were motion-matched; PR, phase-randomization method.