Table 3.
Latent class model fit indices.
| 2 Classes | 3 Classes | 4 Classes | 5 Classes | 6 Classes | |
|---|---|---|---|---|---|
| Pearson’s chi-squared | 250.423 | 174.469 | 125.068 | 90.822 | 64.962 |
| LR chi-squared | 242.742 | 160.678 | 116.547 | 84.817 | 70.056 |
| Chi-squared df | 112 | 104 | 96 | 88 | 80 |
| Log-likelihood | −7513.337 | −7472.305 | −7450.240 | −7434.375 | −7426.994 |
| AIC | 15,056.674 | 14,990.610 | 14,962.480 | 14,946.750 | 14,947.989 |
| BIC | 15,141.441 | 15,120.586 | 15,137.665 | 15,167.143 | 15,213.592 |
| Adj-BIC | 15,093.785 | 15,047.513 | 15,039.174 | 15,043.236 | 15,064.268 |
| Entropy | 0.639 | 0.551 | 0.643 | 0.742 | 0.734 |
| Vong–Lo–Mendell–Rubin likelihood ratio test (LMR) | 1 Versus 2 Classes | 2 Versus 3 Classes | 3 Versus 4 Classes | 4 Versus 5 Classes | 5 Versus 6 Classes |
| LMR p-value | 0.0000 | 0.0000 | 0.0000 | 0.0000 | 0.2069 |
Note 1: Smaller AIC or adjusted BIC values and larger entropy values indicate a more favorable model fit. Note 2: LMR test compares the improvement in model fit between sequential classes, p < 0.0001 is considered to have a good model fit.