Table 3
Fit statistics for the models. Comparisons of conditions using measurement invariance procedures.
Goodness-of-fit indices and model comparisons for tested models | |||||||
Model | CMIN | DF | CMIN/DF | CFI | RMSEA | ECVI | AIC |
0 | 463.269 | 240 | 1.930 | .959 | .050 | 1.594 | 583.269 |
0a | 463.095 | 240 | 1.930 | .954 | .053 | 1.762 | 583.095 |
0b | 299.156 | 240 | 1.246 | .958 | .054 | 4.990 | 419.156 |
1 | 806.332 | 480 | 1.680 | .953 | .038 | 2.250 | 1046.332 |
2 | 834.803 | 501 | 1.666 | .952 | .038 | 2.221 | 1032.803 |
3 | 844.630 | 507 | 1.666 | .952 | .038 | 2.216 | 1030.630 |
4 | 848.113 | 511 | 1.660 | .952 | .038 | 2.207 | 1026.113 |
5 | 900.382 | 540 | 1.667 | .948 | .038 | 2.194 | 1020.382 |
Comparison of conditions using measurement invariance procedures | |||||||
Model | Dif. DF | Dif. CMIN | p | Dif. CFI | Dif. RMSEA | ||
Assuming that model 1 is correct | 2 | 21 | 28.471 | .127 | .001 | .000 | |
3 | 27 | 38.298 | .073 | .001 | .000 | ||
4 | 31 | 41.781 | .094 | .001 | .000 | ||
5 | 60 | 94.050 | .003 | .004 | .000 | ||
Assuming that model 2 is correct | 3 | 6 | 9.827 | .019 | .000 | .000 | |
4 | 10 | 13.310 | .000 | .000 | .000 | ||
5 | 39 | 65.579 | .000 | .006 | .000 | ||
Assuming that model 3 is correct | 4 | 4 | 3.483 | .001 | .000 | .000 | |
5 | 33 | 55.752 | .000 | .004 | .000 | ||
Assuming that model 4 is correct | 5 | 29 | 52.269 | .005 | .004 | .000 | |
Note: Model 0, total runners; model 0a male runners; model 0b, female runners; model 1, indicates that no parameters are restricted to being equal in all groups; model 2, factorial loads restricted to being equal; model 3, observed strctural weights and factorial loads restricted to being equal; model 4, observed structural covariances, structural weights and factorial loads restricted to being equal; model 5, observed structural residues, structural covariances, structural weights and factorial loads restrictied to being equal. Dif. CMIN, difference between the model and the other models;, Dif. DF. difference between the model and the other models; p, level of significance across models. |