TABLE 1.

Summary of the two‐factor confirmatory factor analysis in the discovery and replication samples

Sample	Model	Number of factors	Included RSNs	Chi‐square statistic	Degrees of freedom	AIC	CFI	SRMR
Discovery sample	EFA‐based model	2	18	897	133	973	0.81	0.12
	Model for multivariate GWAS	2	17	575	118	645	0.84	0.12
Replication sample	Model for multivariate GWAS of the discovery sample	2	17	295	118	365	0.76	0.19

Note: For each sample listed in the first column, the second column distinguishes the stages composing the CFA approach. For the discovery sample, CFA consisted of two stages: the first stage, that is, EFA‐based model, tested the model design for the two‐factor model indicated by our EFA approach; the second stage, that is, modeling for multivariate GWAS, only kept RSNs that showed Bonferroni‐corrected factor loadings. The CFA conducted for the replication sample consisted of a single stage, which used the model employed in the multivariate GWAS of the discovery sample. In each stage, a model with a given number of factors was tested (third column), with a given number of RSNs (fourth column). For each model, we display the chi‐square statistic, degrees of freedom, AIC, CFI, and SRMR, from the fifth to the ninth columns.

Abbreviations: AIC, Akaike information criterion; CFA, confirmatory factor analysis; CFI, comparative fit index; EFA, exploratory factor analysis; GWAS, genome‐wide association study; RSN, resting state network; SRMR, standardized root mean square residual.