Skip to main content
. 2019 Feb;19(1):e116–e126. doi: 10.1016/j.clbc.2018.10.008

Supplemental Table 3.

Model Fit Statistics for BMQ and PSM Variables (n = 406)

Class
1 2 3 4 5
Parameters 10 16 22 28 34
LL −2090.764 −1997.354 −1964.527 −1942.876 −1931.403
AIC 4201.528 4026.709 3973.054 3941.752 3930.807
BIC 4241.592 4090.810 4061.194 4053.930 4067.023
Entropy 0.666 0.759 0.777 0.730
Sample Size per Class, % (n) Class 1 = 62 (252)
Class 2 = 38 (154)
Class 1 = 49.3 (200)
Class 2 = 3.4 (14)
Class 3 = 47.3 (192)
Class 1 = 9.9 (40)
Class 2 = 33 (134)
Class 3 = 2.7 (11)
Class 4 = 54.4 (221)
Class 1 = 17 (69)
Class 2 = 3.0 (12)
Class 3 = 32.2 (131)
Class 4 = 43.1 (175)
Class 5 = 4.7 (19)
VLMR-LRT P Value .0001 .2951 .1590 .7835
LMR-LRT P Value .0001 .3023 .1652 .7856

Two participants had missing data for all 5 variables and were excluded from the analysis.

Abbreviations: AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion; BMQ = Beliefs about Medicines Questionnaire; LL = log-likelihood; LMR-LRT = Lo–Mendell–Rubin adjusted likelihood ratio test; PSM = Perceived Sensitivity to Medicines Scale; VLMR-LRT = Vuong–Lo–Mendell–Rubin likelihood ratio test.