Sociodemographic characteristics associated with alcohol consumption and alcohol-related consequences, a latent class analysis of The Norwegian WIRUS screening study

Table 1 Comparison of model fit from 1 to 5 classes

Model	log-likelihood	resid. df	BIC	aBIC	cAIC	Likelihood-ratio	Entropy	VLMR-LRT	LMR-A-LRT
1 class	−24,274.68	4278	48,825.54	48,720.68	48,858.54	10,949.552	–	–	–
2 classes	−21,761.46	4244	44,083.64	43,870.74	44,150.64	5923.110	0.769	p < .0001	p < .0001
3 classes	−21,050.71	4210	42,946.68	42,625.74	43,047.68	4501.606	0.825	p < .0001	p < .0001
4 classes	−20,855.28	4176	42,840.36	42,411.39	42,975.36	4110.743	0.755	p = .7537	p = .7537
5 classes	−20,681.74	4142	42,777.83	42,240.82	42,946.83	3763.674	0.758	p = .5294	p = .5306

Bold indicates statistically significant differences
BIC Bayesian Information Criteria
AIC Aikaike Information Criteria
VLMR-LRT Vuong-Lo-Mendell-Rubin likelihood-ratio test for k-1 versus k classes
LMR-A-LRT Lo-Mendell-Rubin adjusted likelihood-ratio test for k-1 versus k classes

ISSN: 1471-2458