TABLE 4.
Model | Fit indices |
||||||
S-B χ2 | df | Scaling correction | CFI | TLI | RMSEA [90% CI] | SRMR | |
Bifactor men | 27.4 | 28 | 0.988 | 1.000 | 1.003 | 0.000 [0.00 –0.095] | 0.044 |
Bifactor women | 49.6* | 28 | 1.261 | 0.974 | 0.959 | 0.046 [0.061–0.088] | 0.050 |
(1) Configural | 79.7* | 56 | 1.124 | 0.981 | 0.970 | 0.056 [0.023–0.082] | 0.049 |
(2) Metric | 86.7 | 73 | 1.190 | 0.989 | 0.986 | 0.037 [0.000–0.064] | 0.067 |
(3) Scalar | 115.2* | 83 | 1.170 | 0.974 | 0.972 | 0.053 [0.026–0.075] | 0.102 |
| |||||||
Model comparison | Δχ2 | df | P | ΔCFI | Conclusion | ||
| |||||||
1 vs. 2 | 9.7 | 17 | 0.917 | 0.008 | Equivalent | ||
2 vs. 3 | 30.9 | 10 | <0.001 | 0.015 | Not equivalent |
The conclusion in the model comparison section is based on a joint consideration of Δχ2 and ΔCFI. Chi-square statistics were estimated with the robust maximum likelihood (MLR) estimator and Δχ2 was calculated with the scale-adjusted chi-square difference test (http://www.thestatisticalmind.com/calculators/SBChiSquareDifferenceTest.htm, accessed 2021-09-06). df, degrees of freedom; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; TLI, Tucker–Lewis Index; CFI, comparative fit index; CI, confidence interval. *p < 0.05.