From: Preference heterogeneity in health valuation: a latent class analysis of the Peru EQ-5D-5L values
Model | Log-likelihood | Parameters | Bayesian information criterion (BIC) | Akaike information criterion (AIC) | Consistent Akaike information criterion (AIC) |
---|---|---|---|---|---|
Conditional logit | − 19,377.1 | 22 | 38,981.67 | 38,798.12 | 39,003.67 |
Heteroskedastic logit | − 19,041.8 | 29 | 38,383.63 | 38,141.68 | 38,412.63 |
Latent class logit (2 classes) | − 17,185.7 | 81 | 35,209.13 | 34,533.34 | 35,290.13 |
Scale adjusted latent class logits | |||||
1 Scale and 2 taste classes | − 17,359.2 | 66 | 35,400.94 | 34,850.3 | 35,466.94 |
1 Scale and 3 taste classes | − 17,068.9 | 103 | 35,203.06 | 34,343.72 | 35,306.06 |
2 Scale and 2 taste classes | − 17,208.5 | 87 | 35,316.91 | 34,591.06 | 35,403.91 |
2 Scale and 3 taste classes | − 16,914.7 | 124 | 35,112.00 | 34,077.46 | 35,236 |