Table 2.

Optimal number of trajectories: assessment statistics

TrajectoriesLog likelihoodBICSSABICFree parametersEntropyLRT
1−4,643.019,320.409,304.525NANA
2−3,549.327,153.657,128.2480.93P < 0.0001
3−3,332.386,740.386,705.44110.88P < 0.0001
4−3,187.816,471.886,427.41140.89P < 0.0001
5−3,153.706,424.286,370.29170.88P = 0.0689
Model with covariates−2,698.465,717.115,567.84470.91P < 0.0001
  • NOTE: BIC (−2logL + r lnn), where r is the number of free model parameters; SSABIC (n in BIC formula is replaced by n* = (n + 2)/24); Vuong-Lo-Mendell-Rubin likelihood ratio test for k (H0) versus k − 1 classes.