Table 4.
Goodness-of-fit indices for testing measurement invariance across samples
| First-order model | Second-order model | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| CFI | TLI | NFI | RMSEA | Δχ2 | CFI | TLI | NFI | RMSEA | Δχ2 | |
| Default model | .97 | .96 | .93 | .04 (.03, .04) | .97 | .96 | .93 | .04 (.03, .04) | ||
| + Factor loadings set equally across samples | .96 | .96 | .93 | .04 (.03, .04) | p = .212 | .96 | .96 | .93 | .04 (.03, .04) | p = .212 |
| + Factor loadings and inter-factor correlations (first-order model) or regression weights (second-order model) set equally across samples | .96 | .96 | .93 | .04 (.03, .04) | p = .363 | .96 | .96 | .93 | .04 (.03, .04) | p = .170 |
CFI Comparative fit index, TLI Tucker Lewis index, NFI Normed fit index, RMSEA Root mean square error of approximation