Table 3.
Fit Indices for Measurement Invariance (MI) Models
| Model | χ2 (df) | Δχ2 | CFI | ΔCFIa | NCI | ΔNCI | ΔNCI Criteriaa |
|---|---|---|---|---|---|---|---|
| Heavy drinking | .0072 | ||||||
| Unconstrained | 255.20 (24) | - | .982 | - | 1.0840 | - | |
| Weak MI | 278.95 (28) | 23.75* | .980 | .002 | 1.0912 | .0072 | |
| Strong MI | 459.19 (33) | 180.24* | .966 | .014 | 1.1603 | .0688 | |
| Strict MI | 546.71 (39) | 87.52* | .959 | .007 | 1.1938 | .0334 | |
| Sensation Seeking | .0062 | ||||||
| Unconstrained | 175.33 (8) | - | .943 | - | 1.0601 | - | |
| Weak MI | 176.30 (10) | 0.97 | .943 | .000 | 1.0597 | .0004 | |
| Strong MI | 183.43 (12) | 7.13* | .941 | .002 | 1.0616 | .0019 | |
| Strict MI | 188.20 (15) | 4.77 | .941 | .000 | 1.0622 | .0006 | |
| Impulsivity | .0062 | ||||||
| Unconstrained | 74.23 (8) | - | .973 | - | 1.0233 | - | |
| Weak MI | 74.71 (10) | 0.48 | .974 | .001 | 1.0228 | .0005 | |
| Strong MI | 75.95 (12) | 1.24 | .974 | .000 | 1.0225 | .0002 | |
| Strict MI | 85.25 (15) | 9.30* | .971 | .003 | 1.0248 | .002 | |
| Autonomy | .0062 | ||||||
| Unconstrained | 200.93 (8) | - | .955 | - | 1.0696 | - | |
| Weak MI | 203.64 (10) | 2.71 | .955 | .000 | 1.0698 | .0002 | |
| Strong MI | 206.84 (12) | 3.20 | .955 | .000 | 1.0703 | .0005 | |
| Strict MI | 219.00 (15) | 12.16* | .953 | .002 | 1.0737 | .0034 |
Note. Rows in bold indicate the selected measurement invariance models for each construct. CFI = Comparative fit index; NCI = McDonald’s noncentrality index.
a
The minimum criteria for change in model fit indicated by the CFI (.002) and NCI were taken from Meade and colleagues (2008).
*
p < .05.