Skip to main content
. 2013 Feb 24;43(1):70–80. doi: 10.1007/s10964-013-9929-1

Table 2.

Competing model fits for tests of measurement invariance across gender and time

Model χ df Corr RMSEA CFI Model ΔCFI Δchi
1. Configural model 699.908 448 1.163 0.032 0.971
2. Weak invariance across gender 714.370 468 1.171 0.031 0.972 M2-M1 0.001 16.690 (20)
3. Weak invariance across gender and time 799.309 483 1.175 0.035 0.964 M3-M2 0.008 78.982 (15)**
4. Partial weak invariance across gender and timea 725.455 479 1.173 0.031 0.972 M4-M2 0.000 11.471 (11)
5. Strong invariance across gender 757.982 499 1.166 0.031 0.970 M5-M4 0.002 32.903 (20)*
6. Partial strong invariance across genderb 752.311 498 1.167 0.031 0.971 M6-M4 0.001 26.570 (19)
7. Strong invariance across gender and time 1,079.069 516 1.162 0.045 0.936 M7-M6 0.035 367.240 (18)**
8. Partial strong invariance across gender and timec 806.534 499 1.168 0.034 0.965 M8-M6 0.006 38.466 (1)**

p < 0.05, ** p < 0.01

aInvariance was relaxed at item 4 at age 23, item 5 at age 13, and item 6 at age 15 and 23

bInvariance was relaxed across gender at item 2 at age 13

cInvariance was relaxed at item 2 at age 15 and 18, item 3 at age 13 and 23, item 4 at age 13 and 15, item 5 at age 13 and 15, and item 6 at age 15 and 18. In addition,at age 13, invariance at item 2, 3, 4, 5 and 6 was relaxed across gender, and at age 15, invariance at item 2 and 4 was relaxed across gender, and at age 23, item 3 invariance was relaxed across gender