Research Article
Montreal Cognitive Assessment: Seeking a Single Cutoff Score May Not Be Optimal
Table 2
Information for models of 2–4 classes.
| Model | Classes | AIC | BIC | ABICa | Entropy | LRT | Smallest class % |
| (1) Mixture modeling of ADL, MoCA, and CDR | 2 | 19770 | 20319 | 19932 | 0.87 | <0.00 | 36.6 | 3 | 19365 | 19964 | 19541 | 0.83 | <0.01 | 13.1 | 4 | 19249 | 19897 | 19400 | 0.79 | 0.68 | 2.5 | (2) Mixture modeling of only MoCA | 2 | 12642 | 12867 | 12708 | 0.83 | <0.00 | 37.2 | 3 | 12362 | 12628 | 12441 | 0.78 | <0.01 | 18.1 | 4 | 12286 | 12592 | 12376 | 0.80 | <0.01 | 3.6 | (3) Latent class analysis of only MoCA | 2 | 12635 | 12941 | 12725 | 0.84 | <0.00 | 37.7 | 3 | 12390 | 12867 | 12531 | 0.82 | 0.76 | 14.8 | 4 | 12348 | 12996 | 12539 | 0.78 | 0.79 | 12.0 |
|
|
Note. LRT = likelihood ratio test of k vs. k − 1 number of classes.
|