Evidence for measurement bias of the short form health survey based on sex and metropolitan influence zone in a secondary care population

Table 3 Measurement invariance tests regarding sex and Metropolitan Influence Zone

	Model	χ2	df	RMSEA [95% CI]	CFI
Sex	1. Configural	120.002	86	.043 [.022–.060]	.971
	2. Metric	118.349	101	.028 [.000–.047]	.985
	3. Scalar	139.229	111	.035 [.010–.051]	.976*
	3a. Partial Scalar (intercepts for items 10,11 free)	129.661	109	.030 [.000–.048]	.982
	4. Partial Strict (intercepts for items 10,11 free)	141.434	121	.028 [.000–.046]	.983
MIZ	1. Configural	118.580	86	.042 [.021–.060]	.973
	2. Metric	122.457	101	.032 [.000–.050]	.982
	3. Scalar	138.409	111	.034 [.008–.051]	.977*
	3a. Partial Scalar (Intercepts for item 2 free)	135.244	110	.033 [.000–.050]	.979
	4. Partial Strict (Intercepts for item 2 free)	163.871	122	.040 [.022–.055]*	.965*

*significant deterioration in model fit indicated by change in RMSEA > .01 and/or change in CFI < −.004 [40]

ISSN: 1477-7525