A cross-cultural re-evaluation of the Exercise Addiction Inventory (EAI) in five countries

Sports Medicine - Open

Table 3 Degree of model fit of the exercise addiction inventory in six samples from five different countries and tests of measurement invariance

	χ ²	df	p -Value	RMSEA	Cfit of RMSEA	CFI	TLI	SRMR	Δ χ ²	Δ df	p -Value	ΔRMSEA	ΔCFI
Confirmatory factor analysis in each country separately
Spain	6.1	9	0.727	<0.001	0.946	1.000	1.000	0.022
UK	32.6	9	<0.001	0.094	0.017	0.942	0.903	0.042
US	58.4	9	<0.001	0.065	0.051	0.920	0.867	0.033
Denmark	14.3	9	0.113	0.032	0.828	0.985	0.975	0.024
Hungary	21.4	9	0.011	0.049	0.491	0.976	0.961	0.027
Hungary_2	80.5	9	<0.001	0.054	0.266	0.949	0.915	0.027
Multigroup analyses to test the measurement invariance
Configural invariance	211.5	54	<0.001	0.055		0.955	0.925	0.029
Configural vs. metric invariance									114.2	25	<0.001	0.002	0.025
Metric invariance	325.4	79	<0.001	0.057		0.930	0.920	0.051
Metric vs. scalar invariance									2,140.0	25	<0.001	0.093	0.571
Scalar invariance	2,346.2	104	<0.001	0.150		0.361	0.447	0.136

The latent variables were identified by fixing one factor loading being equal to 1.
χ ² Chi-square value of model fit of each model, df degree of freedom, RMSEA root mean squared error of approximation, Cfit of RMSEA is a statistical test that evaluates the statistical deviation of RMSEA from the value 0.05, and non-significant probability values (p > 0.05) indicate acceptable model fit; CFI comparative fit index, TLI Tucker–Lewis index, SRMR the standardized root mean square residual, Δχ ² Satorra–Bentler scaled (S–B scaled) χ ² difference test, Δdf the difference of df in two models compared, ΔRMSEA the difference of RMSEA values in two models compared, ΔCFI the difference of CFI values in two models compared.