Table 3.
Fit statistics for all models
| Model | χ 2 | df | CFI | TLI | SRMR | RMSEA | Δ χ 2 | Δdf | p |
|---|---|---|---|---|---|---|---|---|---|
|
|
|||||||||
| CFA (full sample) | 332.235 | 111 | 0.958 | 0.949 | 0.039 | 0.053 | -- | -- | -- |
| CFA (male subsample) | 323.789 | 111 | 0.955 | 0.945 | 0.041 | 0.055 | -- | -- | -- |
| Model 1 (full sample) | 738.025 | 175 | .899 | .883 | 0.090 | 0.068 | -- | -- | -- |
| Model 2 (full sample) | 585.250 | 173 | 0.926 | 0.913 | 0.064 | 0.058 | 152.775 | 2 | <0.001 |
| Model 1 (male subsample) | 588.304 | 158 | 0.913 | 0.897 | 0.085 | 0.065 | -- | -- | -- |
| Model 2 (male subsample) | 431.361 | 156 | 0.944 | 0.933 | 0.057 | 0.052 | 156.943 | 2 | <0.001 |
Note: CFA = confirmatory factor analysis, χ2 = chi-square, df = degrees of freedom, CFI = comparative fit index, TLI = Tucker-Lewis index, SRMR = standardized root mean square residual, RMSEA = root mean squared error of approximation, Δχ2 = change in chi-square, Δdf = change in degrees of freedom. Model 1 did not include direct paths from posttraumatic stress disorder to eating disorder symptoms and food addiction symptoms, respectively. Model 2 included direct paths from posttraumatic stress disorder to eating disorder symptoms and food addiction symptoms, respectively.