Table 6.
Information criteria to select the number of classes for SAFA and SANA LC models *.
| SAFA LC Models | SANA LC Models | |||||||
|---|---|---|---|---|---|---|---|---|
| Number of classes | Npar | LL | BIC | CAIC | Npar | LL | BIC | CAIC |
| 1 | 10 | −1289.44 | 2631.70 | 2641.70 | 10 | −3190.41 | 6440.84 | 6450.84 |
| 2 | 23 | −1190.75 | 2503.02 | 2526.02 | 28 | −2848.49 | 5865.01 | 5893.01 |
| 3 | 36 | −1153.23 | 2496.65 | 2532.65 | 46 | −2671.54 | 5619.14 | 5665.14 |
| 4 | 49 | −1116.25 | 2491.38 | 2540.38 | 64 | −2560.59 | 5505.26 | 5569.26 |
| 5 | 62 | −1086.41 | 2500.38 | 2562.38 | 82 | −2490.60 | 5473.32 | 5555.32 |
| 6 | 75 | −1067.68 | 2531.60 | 2606.60 | 100 | −2439.32 | 5478.78 | 5578.78 |
* Npar = Number of estimated parameters.; LC = Latent class; LL = Log-likelihood; BIC (LL) = Bayesian information criterion based on the log-likelihood; CAIC (LL) = Consistent Akaike information criterion, based on the log-likelihood; SAFA = stated attribute full attendance; SANA = stated non-attendance.