Table 3.

Fit indices and class distributions of the six best growth mixture models for depression.

Rank	Random	$N_C$	AIC	BIC	SABIC	Entropy	%.C1	%.C2	%.C3	%.C4	%.C5	%.C6	%.C7
Depression
1	I	6	21,662.86	21,793.95	21,714.53	0.75	9.4	1.1	2.8	77.4	5.4	3.9
2	I	4	21,736.78	21,825.92	21,771.91	0.77	1.5	7.9	79.6	11.1
3	I	7	21,675.50	21,827.56	21,735.44	0.75	3.6	2.6	1.0	69.8	11.4	0.0	11.7
4	ISQ	7	21,661.97	21,840.25	21,732.25	0.52	1.0	2.5	16.7	9.4	0.0	66.6	3.7
5	I	5	21,763.66	21,873.78	21,807.07	0.75	78.9	4.6	2.6	2.7	11.2
6	ISQ	3	21,779.88	21,874.26	21,817.08	0.76	81.7	7.9	10.4
Anxiety
1	IS	3	22,587.22	22,665.87	22,618.22	0.52	13.2	59.0	27.8
2	IS	4	22,569.39	22,669.01	22,608.66	0.61	56.3	3.4	32.9	7.4
3	I	6	22,543.54	22,674.63	22,595.21	0.53	37.0	31.4	3.2	17.5	6.8	4.1
4	I	5	22,565.88	22,675.99	22,609.28	0.57	26.2	4.6	58.0	7.4	3.8
5	IS	5	22,562.23	22,682.83	22,609.77	0.61	56.6	2.1	30.5	2.4	8.4
6	ISQ	3	22,591.20	22,685.59	22,628.41	0.53	13.2	59.0	27.8

The models are sorted by increasing BIC. The ‘Random’ column indicates the combination of random parameters in the model (‘I’ stands for intercept, ‘S’ for linear slope, and ‘Q’ for quadratic slope). The $N_C$ column gives the number of classes estimated in the model.