Table 1.
Fit Indices for the Latent Profile Analysis Determining the Typology of Pediatric Disinhibited Eating Behaviors.
| Clusters | Parameters | BIC | aBIC | cAIC | LL | Entropy | Largest Residual |
|---|---|---|---|---|---|---|---|
| 1 | 7 | 4581.81 | 4559.60 | 4588.81 | −2269.86 | 1.00 | 263.49 |
| 2 | 15 | 4161.55 | 4113.95 | 4176.55 | −2035.65 | 0.99 | 33.26 |
| 3 | 23 | 4097.75 | 4024.76 | 4120.75 | −1979.69 | 0.83 | 33.30 |
| 4 | 31 | 4076.05 | 3977.69 | 4107.05 | −1944.78 | 0.85 | 0.83 |
| 5 | 39 | 4073.17 | 3949.41 | 4112.17 | −1919.27 | 0.79 | 1.80 |
| 6 | 47 | 4085.51 | 3936.37 | 4132.51 | −1901.38 | 0.77 | 0.50 |
| 7 | 55 | 4118.58 | 3944.05 | 4173.58 | −1893.84 | 0.78 | 0.41 |
| 8 | 63 | 4147.32 | 3947.41 | 4210.32 | −1884.15 | 0.75 | 0.22 |
Note: BIC = Bayesian Information Criterion; cAIC = Consistent Akaike Information Criterion; aBIC = sample size adjusted BIC; LL = Log-likelihood; Lower BIC, aBIC, and cAIC values indicate better model fit. Entropy is a measure of classification accuracy, with high values indicating better accuracy. The high residual tests the conditional independence assumption (should be less than 3.0).