Skip to main content
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: Res Child Adolesc Psychopathol. 2021 Jan 6;49(4):519–531. doi: 10.1007/s10802-020-00734-4

Table 4.

Model fit statistics for latent class models estimated up to 5 classes

Class LL AIC BIC aBIC BLRT Entropy
2 −2507.26 4587.05 4664.78 4595.05 0.000 0.848
3 −2271.52 4438.50 4544.51 4449.40 0.000 0.851
4 −2189.25 4379.09 4513.36 4392.90 0.000 0.898
5 −2147.35 4304.21 4466.75 4320.92 0.000 0.971

Note. LL = Log Likelihood; AIC = Akaike Information Criteria; BIC = Bayesian Information Criterion; aBIC = sample size adjusted BIC; BLRT = Bootstrap Likelihood Ratio Test. Note that for the 5-class solution, the number of random starts was increased to 200 to allow for model convergence (replication of the best loglikelihood value); convergence difficulties are sometimes an indication that the data do not support this number of classes. One participant was missing data on all 7 items, and thus the total sample size for latent profile analyses was n = 253.