Skip to main content
. 2018 Jul 5;74(12):2187–2202. doi: 10.1002/jclp.22659

Table 2.

Robust maximum likelihood estimation of CFA models of the latent structure of the MHC‐SF items

Model 1 Model 2 Model 3
Single factor Two factors Three factors
Fit indices Well‐being Hedonic and eudaimonic well‐being Emotional, psychological and social well‐being
SB χ 2 281.2 207.7 176.6
df 77 76 74
NCP 204.2 131.7 102.6
AIC 337.2 265.7 238.6
RMSEA 0.075 0.061 0.054
CFI 0.978 0.986 0.989
SRMR 0.056 0.051 0.046
GFI/AGFI 0.86/0.81 0.89/0.85 0.91/0.87
Model 1 vs. 2 Model 2 vs. 3
Scaled Δχ 2 66.62*** 35.31***

Note. CFA, confirmatory factor analysis; MHC‐SF, Mental Health Continuum‐Short Form; SB χ 2, Satorra‐Bentler scaled chi‐square (smaller values indicate better fit); df, degrees of freedom; NCP, estimated noncentrality parameter (smaller values indicate better fit); AIC, Akaike's information criterion (smaller values indicate better fit); RMSEA, root mean square error of approximation (good fit ≤ 0.05; acceptable fit ≤ 0.08); CFI, comparative fit index (good fit ≥ 0.97; acceptable fit ≥ 0.95); SRMR, standardized root mean square residual (good fit ≤ 0.05; acceptable fit ≤ 0.10); GFI, goodness of fit index (good fit ≥ 0.95; acceptable fit ≥ 0.90); AGFI, adjusted goodness of fit index (good fit ≥ 0.90; acceptable fit ≥ 0.85).

***p < 0.001, two‐tailed.