Table 3.
Tests of longitudinal measurement invariance between the urban and rural groups for preference of solitude.
| Model | χ 2 | df | RMSEA | CFI | TLI | χ 2 | Δχ2 | Δdf | p |
|---|---|---|---|---|---|---|---|---|---|
| Model fit before allowing residuals of items to covary | |||||||||
| M1 | 887.865 | 372 | 0.060 | 0.904 | 0.891 | – | – | – | – |
| M2 | 915.866 | 390 | 0.059 | 0.902 | 0.894 | M2 vs. M1 | 26.377 | 18 | 0.091 |
| M3 | 950.390 | 411 | 0.058 | 0.899 | 0.897 | M3 vs. M2 | 31.890 | 21 | 0.061 |
| Model fit after allowing residuals of two items to covary a | |||||||||
| M1 | 741.279 | 366 | 0.051 | 0.930 | 0.919 | ||||
| M2 | 767.888 | 384 | 0.051 | 0.928 | 0.921 | M2 vs. M1 | 25.138 | 18 | 0.121 |
| M3 | 803.364 | 405 | 0.050 | 0.926 | 0.923 | M3 vs. M2 | 34.331 | 21 | 0.033 |
M1, Configural invariance; M2, Metric invariance; M3, Scalar invariance.
Residuals of the two items “I enjoy being by myself” and “I like spending time alone in my room” were allowed to covary according to model modification index.