Table 2.
Goodness-of-fit statistics for different PSS-10 models and multigroup confirmatory factor analyses in nonclinical vs. clinical individuals.
| χ2 (df) | CFI | TLI | RMSEA (90%CI) | SRMR | AIC | BIC | ΔCFI | ΔTLI | ΔRMSEA | ΔSRMR | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Nonclinical sample | |||||||||||
| 1-factor | 544.851 (35) |
.907 | .881 | .108 (.100-.116) |
.057 | ||||||
| 2-factor | 275.559 (34) |
.956 | .942 | .075 (.067-.084) |
.040 | 317.55 | 425.27 | ||||
| 2-factor finala |
167.316 (32) |
.975 | .965 | .058 (.050-.067) |
.031 | 213.31 | 331.29 | ||||
| Clinical sample | |||||||||||
| 1-factor | 946.493 (35) |
.589 | .472 | .213 (.201-.225) |
.180 | ||||||
| 2-factor | 189.822 (34) |
.940 | .920 | .089 (.077-.102) |
.050 | 231.82 | 323.26 | ||||
| 2-factor finala |
151.644 (32) |
.954 | .935 | .081 (.068-.094) |
.046 | 197.64 | 198.54 | ||||
| Clinical vs. Nonclinical sample | |||||||||||
| Configural Invariance | 345.776 (64) |
.965 | .951 | .049 (.044-.054) |
.035 | ||||||
| Metric Invariance |
376.719 (72) |
.962 | .953 | .048 (.043-.053) |
.038 | -.003 | .002 | -.001 | .003 | ||
| Structural Invariance | 586.620 (82) |
.938 | .931 | .058 (.054-.063) |
.039 | -.024 | -.022 | .010 | .001 | ||
Note. The final 2-factor model was compared with the unmodified 2-factor model; a in the final model, errors of items 1 and 3 and items 1 and 9 were freed to correlate; df = degrees of freedom; CFI = Comparative fit index; TLI = Tucker-Lewis index; RMSEA = root mean square error of approximation (90% confidence interval shown in brackets); SRMR = standardized root mean square residual; Δ = difference; configural invariance = equal form; metric invariance = equal loadings; structural invariance = equal intercepts.