Evaluating the psychometric properties of the German adaptation of the client attachment to therapist scale

Table 7 Fit indices for the analysis of measurement invariance - Sample 2

Model	χ²	df	Δ χ²	Δdf	Δ p	CFI	ΔCFI	RMSEA	ΔRMSEA
Gender
Configural invariance	171.683	102				.974		.067
Loading invariance	195.790	111	24.107	3	<.001	.968	.006	.070	.003
Threshold invariance	227.103	145	31.313	34	.600	.969	.001	.061	.009
Intercept invariance	222.572	154	4.531	9	.873	.974	.005	.054	.007
Age groups (≤39, 40–59, ≥59)
Configural invariance	250.897	153				.967		.079
Loading invariance	287.178	171	36.281	18	.006	.962	.005	.082	.003
Threshold invariance	386.929	231	99.751	60	.001	.949	.013	.082	.000
Partial threshold invariance ^a	351.020	221	35.909	10	<.001	.957	.008	.076	.006
Intercept invariance	373.635	235	22.615	14	.067	.954	.003	.076	.000

CFI Comparative fit index, RMSEA Root mean square error of approximation
^a = The thresholds of Item 15 (“I think my counselor disapproves of me.”) and 18 (“I don’t like to share my feelings with my counselor.”) was freed to vary between groups

ISSN: 1471-2288