Skip to main content
. 2019 May 6;14(5):e0216031. doi: 10.1371/journal.pone.0216031

Table 2. Fit indices for the long-term DC profiles with different numbers of latent classes (Latent Profile Analyses).

No. of classes Log L BIC Entropy VLMR
(p)
LMR
(p)
Latent class proportions
n (%)
1 -587.69 1231.56 - - - 108 (100)
2 -533.10 1155.16 0.82 0.006 0.007 69 (64) / 39 (36)
3 -492.96 1107.66 0.81 0.341 0.352 37 (34) / 28 (26) / 43 (40)
4 -465.71 1085.93 0.84 0.170 0.177 30 (28) / 22 (20) / 16 (15) / 40 (37)
5 -449.31 1085.91 0.88 0.267 0.273 9 (8) / 8 (7) / 35 (33) / 23 (21) / 33 (31)
6 -435.99 1092.04 0.88 0.406 0.416 9 (8) / 7 (7) / 25 (23) / 19 (18) / 13 (12) / 35 (32)
7 -419.29 1091.41 0.88 0.620 0.627 20 (19) / 6 (5) / 7 (6) / 14 (13) / 19 (18) / 30 (28) / 12 (11)

Note: BIC = Bayesian information criterion; VLMR = Vuong-Lo-Mendell-Rubin test; LMR = Lo-Mendell-Rubin test.