Latent class analysis of sexual health markers among men and women participating in a British probability sample survey

Table 2 Model fit statistics for different numbers of sexual health groups identified using latent class analysis

	Number of classes	AIC	BIC	Entropy	Smallest class (%)	lowest class probability	LMR p-value
Men	1	38,361	38,465		100.0	1.00
	2	36,783	36,998	0.70	18.5	0.86	<.001
	3	35,911	36,238	0.75	10.0	0.85	<.001
	4	35,565	36,003	0.89	4.3	0.88	<.001
	5	35,408	35,957	0.71	3.7	0.82	0.330
	6	35,352	36,013	0.70	3.6	0.67	0.182
Women	1	68,151	68,275		100.0	1.00
	2	63,649	63,903	0.78	16.9	0.90
	3	62,259	62,643	0.78	11.8	0.89	<.001
	4	61,900	62,414	0.85	2.2	0.86	<.001
	5	61,544	62,188	0.79	2.2	0.72	<.001
	6	61,364	62,139	0.73	2.0	0.80	0.011
	7	61,212	62,118	0.72	1.9	0.69	0.385
	8	61,107	62,143	0.73	1.9	0.68	0.403

Figures in bold show the fit of the four-class model selected for men, and the six-class model selected for women
AIC Akaike Information Criterion, BIC Bayesian Information Criterion, LMR Lo-Mendell-Rubin likelihood ratio test

ISSN: 1471-2458