Table 3.

Goodness-of-fit indices for the hierarchical CFA

	Model°	likelihood ratio χ2*	df	RMSEA (90% C.I.)	CFI	SRMR
INITIAL MODEL: 52 variables	0F	2382	1253	0.085 (0.079; 0.090)	0.90	0.10
	1F	2459	1267	0.086 (0.081; 0.092)	0.89	0.11
	2F	2436	1266	0.086 (0.080; 0.091)	0.90	0.10
Model phase 1: LONG selection	0F	1225	719	0.075 (0.068; 0.082)	0.94	0.08
	1F	1297	733	0.078 (0.071; 0.085)	0.93	0.10
	2F	1262	732	0.076 (0.069; 0.083)	0.94	0.09
Rerun with non-imputed data	2F	1059	732	0.096 (0.083; 0.11)	0.87	0.12
Model phase 2: SHORT selection	0F	642	384	0.073 (0.063; 0.083)	0.95	0.07
	1F	717	398	0.080 (0.070; 0.089)	0.95	0.10
	2F	681	397	0.075 (0.066; 0.085)	0.95	0.08
Rerun with non-imputed data	2F	537	397	0.084 (0.065; 0.10)	0.93	0.11

* For all likelihood ratio χ²: p < .00001 ° See methods section for the description of the model structures. 0F = basic model with seven factors; 1F = seven factors and one hierarchical latent factor; 2F = proposed structure of seven factors and two hierarchical latent factors, see also Figure 1.