Table 2. Summary of Latent Growth Model Fit Statistics.
| Models | -2 Log Likelihood | Free parameters | AIC | aBIC |
|---|---|---|---|---|
| Michigan longitudinal sample | ||||
| - Linear | 7071.40 | 6 | 14154.81 | 14158.94 |
| - Quadratic | 6961.99 | 10 | 13943.97 | 13950.86 |
| - Modified logistic | 7024.69 | 8 | 14065.38 | 14070.88 |
| - Latent basis | 7184.79 | 19 | 14407.57 | 14420.65 |
| - Exponential | 6934.36 | 10 | 13888.71 | 13895.59 |
| MLSELD preschool sample | ||||
| - Linear | 6688.04 | 6 | 13588.08 | 13595.82 |
| - Quadratic | 6760.81 | 10 | 13541.63 | 13554.52 |
| - Modified Logistic | 6783.97 | 8 | 13583.95 | 13594.26 |
| - Latent Basis | 6779.69 | 13 | 13585.38 | 13602.14 |
| - Exponential | 6760.21 | 10 | 13540.42 | 13553.32 |
| Oregon sample | ||||
| - Linear | 5216.08 | 6 | 10444.16 | 10448.97 |
| - Quadratic | 5192.96 | 10 | 10405.91 | 10413.92 |
| - Modified logistic | 5208.42 | 8 | 10432.85 | 10439.25 |
| - Latent basis | 5202.44 | 13 | 10430.88 | 10441.29 |
| - Exponential | 5189.12 | 10 | 10398.23 | 10406.24 |
Note. AIC refers to the Akaike information criterion, aBIC refers to the adjusted Bayesian information criterion. Bolded values indicate best fit.