Skip to main content

Table 3 Model fit indices for one- to four-latent profile solutions

From: Portals to frailty? Data-driven analyses detect early frailty profiles

Profile

(−)2LL

npar

AIC

BIC

SABIC

LMR

VLMR

Entropy

1

− 5029.43

42

− 4945.43

− 4757.47

− 4890.81

–

–

–

2

− 5584.96

51

− 5482.96

− 5254.72

− 5416.64

< .001

< .001

0.99

3

− 5880.55

60

− 5760.55

− 5492.02

− 5682.52

< .001

< .001

0.99

4b

− 6235.52

69

–

–

–

–

–

–

  1. (−2)LL − 2 log-likelihood; npar number of parameters free; AIC Akaike information criterion; BIC Bayesian information criterion; SABIC sample size adjusted BIC; LMR adjusted Lo-Mendell-Rubin likelihood ratio test; VLMR adjusted Vuong-Lo-Mendell-Rubin likelihood ratio test
  2. bThis model was not considered due to non-replicated log-likelihood