Table 2. Criteria to choose the best number of classes.

Number of classes	BIC	Entropy	Number of individuals per class						Bootstrapped likelihood ratio test
			1	2	3	4	5	6	p Value
2	21481.114	0.78	374	569					0.001
3	21383.724	0.88	108	732	103				0.001
4	21414.326	0.72	228	499	107	109			0.078
5	21407.395	0.75	175	562	86	19	101		0.067
6	21308.773	0.76	179	552	84	19	103	6	0.079

Notes:

p Values were comparing with model with one fewer class.

BIC, Bayesian information criterion.