Table 2.
Model fit statistics for growth mixture models with increasing number of trajectory classes
| AIC | BIC | aBIC | Entropy | Smallest n | |
|---|---|---|---|---|---|
| 1 class | 13289.11 | 13331.36 | 13305.95 | 1453 | |
| 2 classes | 13009.12 | 13067.21 | 13032.27 | 0.82 | 255 |
| 3 classes | 12842.10 | 12916.04 | 12871.57 | 0.81 | 124 |
| 4 classes | 12770.70 | 12860.49 | 12806.48 | 0.77 | 59 |
| 5 classes | 12725.53 | 12831.16 | 12767.63 | 0.79 | 24 |
| 6 classes | 12672.87 | 12794.34 | 12721.28 | 0.72 | 30 |
| 7 classes | 12627.05 | 12764.36 | 12681.49 | 0.77 | 8 |
Note: For AIC (Akaike information criterion), BIC (Bayesian information criterion), and aBIC (sample-size adjusted BIC) lower values indicate better fitting models. For entropy higher values indicate better quality of class assignment. Smallest n refers to the class with the smallest number of participants.