Table 1.
Model | ln(L) | k | BIC |
---|---|---|---|
Latent class model | |||
2 classes | −3,702.914 | 11 | 3,746.638 |
3 classes | −3,684.760 | 17 | 3,752.333 |
4 classes | −3,677.175 | 23 | 3,768.598 |
5 classes | −3,677.050 | 29 | 3,792.322 |
Latent trait model | |||
2 values | −3,702.914 | 11 | 3,746.638 |
3 values | −3,696.191 | 12 | 3,743.890 |
4 values | −3,694.466 | 13 | 3,746.140 |
5 values | −3,694.549 | 14 | 3,750.198 |
Normal distribution | −3,695.429 | 10 | 3,735.178 |
Note. N = 2,835. Values shown in tables are log-likelihood, ln(L); number of parameters, k; and Bayesian information criterion, BIC. Models with smaller values of BIC provide a better fit to the data.