. 2020 Feb 21;8(1):20. doi: 10.3390/dj8010020

Table 5.

Fit indices in the one-group analysis and the multi-group analysis across dental and mental health patients for the first revision of the hypothetical model (R1).

Fit Indices		One-Group	Multi-Group Analysis
Fit Indices		One-Group	UC	MW	SW	SC	SR	MR
χ²		6.451	18.106	18.254	25.073	42.347	44.398	51.933
df		8	16	18	22	23	26	29
p		0.597	0.318	0.439	0.294	0.008	0.014	0.006
χ²/df		0.806	1.132	1.014	1.140	1.841	1.708	1.791
BS p		0.614	0.331	0.455	0.310	0.010	0.026	0.011
GFI		0.962	0.915	0.914	0.880	0.799	0.801	0.781
IFI		1	0.971	0.996	0.954	0.708	0.709	0.619
CFI		1	0.964	0.996	0.948	0.673	0.689	0.612
RMSEA (90% CI)	Point value	0	0.048	0.016	0.050	0.123	0.112	0.119
	LB	0	0	0	0	0.061	0.051	0.064
	UB	0.134	0.137	0.120	0.126	0.180	0.168	0.170
	p-close	0.689	0.462	0.593	0.461	0.030	0.048	0.025
SRMR		0.060	0.073	0.072	0.132	0.126	0.130	0.122

Note. Nested models in constrains for multi-group analysis: UC = Unconstrained model, MW = with constraints on the measurement weights, SW = with constraints on the structural weights, SC = with constraints on the structural covariances, SR = with constrains on the structural residuals, and MR = with constraints on measurement residuals. Fit indices: χ² = likelihood ratio chi-square statistic, df = degrees of freedom for the likelihood ratio chi-square test, p = probability value under null hypothesis of goodness of fit for the likelihood ratio chi-square statistic, χ²/df = relative chi-square, BS p = Bollen–Stine bootstrap probability value (with the simulation of 1000 random samples), GFI = Jöreskog-Sörbom goodness-of-fit index, IFI = incremental fit index through Bollen’s coefficient delta-2, CFI = Bentler’s comparative fit index, RMSEA (90% CI) = point estimation and interval estimate with 90% confidence level for Steiger–Lind root mean square error of approximation, LB = lower boundary and UB = upper boundary of a two-sided 90% confidence interval for population RMSEA, p-close = probability value under null hypothesis of close fit: H₀ = RMSEA ≤ 0.05, and SRMR = standardized root mean square residuals.