Skip to main content
. 2021 Dec 21;2(1):39–49. doi: 10.1016/j.jointm.2021.11.001

Table 1.

Fit statistics for different number of trajectory groups.

Trajectory groups Best BIC Best AIC 2ΔBIC Likelihood AvePP (per trajectory group) Percent of patients with posterior probability <70% (per trajectory group) Percent of sample size (per trajectory group) Relative entropy
1 −443,433.2 −443,413.9 NA −443,408.9 1 0 100 NA
2 −393,822.8 −393,784.2 99,220.9 −393,774.2 0.99/0.98 0.3/0.7 69/31 0.951
3 −373,887.9 −373,830.0 39,869.7 −373,815.0 0.97/0.96/0.98 0.5/0.6/0.9 49/37/14 0.924
4 −365,471.5 −365,394.2 16,832.8 −365,374.2 0.96/0.94/0.95/0.97 0.6/0.7/0.8/0.9 37/35/20/8 0.916
5 −360,994.9 −360,898.4 8953.2 −360,873.4 0.95/0.91/0.91/0.95/0.97 0.7/0.7/0.8/0.9/0.9 33/30/18/14/6 0.884
6 −356,096.4 −355,980.5 9797.0 −355,950.5 0.94/0.89/0.90/0.92/0.95/0.97 0.7/0.9/0.8/0.9/0.9/0.9 28/12/28/16/12/4 0.897
7 −353,099.9 −352,964.7 5993.0 −352,929.7 0.94/0.86/0.90/0.88/0.91/0.95/0.98 0.8/0.8/0.9/0.9/0.9/0.9/1.0 24/20/12/18/12/10/4 0.886

AIC: Akaike Information Criterion; AvePP: Average post-test grouping probability; BIC: Bayesian information criterion; NA: Not avalible; 2ΔBIC: Twice the difference between the BIC values.