A bifactor representation of the Center for Epidemiological Studies Depression Scale for children: gender and age invariance and implications for adolescents’ social and academic adjustment

Table 2 Model fit statistics for factor analyses and measurement invariance testing

Models	χ2	df	CFI	TLI	RMSEA	[90% CI]	ΔCFI	ΔRMSEA
Unidimensional Model	1374.495	170	0.834	0.814	0.118	[0.112-0.123]
CFA
Correlated traits	545.109	164	0.947	0.939	0.067	[0.061-0.074]	0.113	− 0.051
Higher-order CFA	538.294	166	0.949	0.941	0.066	[0.060-0.072]	0.002	− 0.001
Bi-CFA	399.587	151	0.966	0.957	0.057	[0.050-0.063]	0.016	− 0.009
ESEM
Correlated traits	332.135	116	0.970	0.951	0.060	[0.053-0.068]
Higher-order ESEM	328.117	118	0.971	0.953	0.059	[0.051-0.067]	0.002	− 0.001
Bi-ESEM	254.136	100	0.979	0.960	0.055	[0.047-0.063]	0.007	− 0.004
Models	χ2	df	CFI	TLI	RMSEA	[90% CI]	ΔCFI	ΔRMSEA
Invariance across gender
Configural	344.536	215	0.980	0.965	0.049	[0.039-0.058]
Scalar invariance	447.846	310	0.979	0.975	0.042	[0.033-0.050]	− 0.001	− 0.007
Strict invariance	462.385	330	0.980	0.977	0.040	[0.031-0.048]	0.001	− 0.002
Latent means invariance	584.676	335	0.962	0.957	0.054	[0.047-0.061]	0.018	0.014
Invariance across age (12–14, 15–16, 17–18)
Configural	462.981	330	0.981	0.967	0.049	[0.038-0.059]
Scalar invariance	661.315	520	0.979	0.978	0.040	[0.030-0.049]	0.002	− 0.009
Strict invariance	718.436	560	0.977	0.977	0.041	[0.031-0.049]	− 0.002	0.001
Latent means invariance	787.315	570	0.968	0.968	0.047	[0.039-0.055]	0.009	0.006

Note. χ² = chi square (weighted least square estimator was used); df = degrees of freedom; CFI = Comparative fit index; TLI = Tucker-Levis Index; RMSEA = root mean square error of approximation; 90% CI = 90% confidence interval for the RMSEA; CFA = confirmatory factor analysis; ESEM = exploratory structural equation model; Bi = bifactor

ISSN: 1753-2000