Table 5 

Maximum likelihood goodnessoffit indices for confirmatory factor models 

Original SixItem ASC 



Invariance 
χ^{2} 
df 
p 
Δχ^{2} 
df 
p 
RMSEA 
CFI 
NNFI 


Form 
32.0 
16 
.01 
.055 
.992 
.985 

Slope 
33.7 
22 
.05 
1.7 
6 
.95 
.038 
.994 
.992 
Intercept 
39.9 
28 
.07 
6.2 
6 
.40 
.035 
.994 
.994 
Item Error 
46.1 
34 
.08 
6.2 
6 
.40 
.033 
.994 
.995 
Covariance 
46.5 
35 
.09 
0.4 
1 
.53 
.033 
.994 
.995 


Revised FiveItem ASC (children's health item excluded) 



Invariance 
χ^{2} 
df 
p 
Δχ^{2} 
df 
p 
RMSEA 
CFI 
NNFI 


Form 
11.2 
8 
.19 
.033 
.998 
.996 

Slope 
15.8 
13 
.26 
4.6 
5 
.47 
.024 
.999 
.998 
Intercept 
21.4 
18 
.26 
5.6 
5 
.35 
.015 
.999 
.998 
Item Error 
30.5 
23 
.14 
9.1 
5 
.11 
.030 
.996 
.997 
Covariance 
30.9 
24 
.16 
0.4 
1 
.53 
.029 
.997 
.997 


χ^{2 }is the chisquare value for the given model with degrees of freedom, df, and probability, p. Δχ^{2 }is the difference in chisquare between the given model and the previous one. RMSEA is the Root Mean Square Error of Approximation, CFI is the Comparative Fit Index, and NNFI is the NonNormed Fit Index. Listwise deletion of missing values was used for all of the models. 

Gotay and Pagano Health and Quality of Life Outcomes 2007 5:15 doi:10.1186/14777525515 