Skip to main content
. Author manuscript; available in PMC: 2021 Nov 1.
Published in final edited form as: J Psychiatr Res. 2020 Aug 1;130:194–200. doi: 10.1016/j.jpsychires.2020.07.032

Table 2.

Fit Indices for Latent Class Models

Class type BIC aBIC Entropy Lo-Mendel Rubin testa BLRTa Replicated p value
Two-class 29834.876 29726.911 0.747 p < 0.001 Yes p < 0.001
Three-class 29623.350 29464.578 0.817 p = 0.255 Yes p < 0.001
Four-classb 29499.492 29289.913 0.844 p = 0.132 No NA
Five-Classb 29477.808 29217.422 0.784 p = 0.698 No NA

Notes. AIC: Akaike information Criterion, BIC: Bayesian information criterion, aBIC: sample-size adjusted Bayesian information criterion. BLRT: Bootstrapped Ratio Likelihood Test.

a.

a significant result indicates that the solution with the higher number of classes is a better fit

b.

For BLRT testing, attempts were made to increase the LRTSTARTS, default is 20 and 5, successive runs were tried up to 1000 and 400.