Table 2. Parameter Estimates for Four-Class Model of Past Two-Week Alcohol Use from Ages 18 through 25/26.

Modal Age	Overall Proportion	Non-Drinkers	Legal Non-HI Drinkers	Persistent Non-HI Drinkers	HI Drinkers
			Latent Class Membership Probabilities
		0.21	0.23	0.40	0.16
			Item Response Probabilities
	No past 30-day alcohol use
18	0.51	0.81	0.85	0.28	0.19
19/20	0.44	0.77	0.89	0.15	0.07
21/22	0.28	0.80	0.42	0.01	0.03
23/24	0.25	1.00	0.07	0.08	0.03
25/26	0.22	0.69	0.11	0.15	0.03
	Past 30-day alcohol use without 10+ drinking
18	0.39	0.14	0.10	0.66	0.45
19/20	0.45	0.23	0.10	0.78	0.43
21/22	0.58	0.20	0.58	0.89	0.32
23/24	0.63	0.00	0.90	0.88	0.40
25/26	0.66	0.27	0.87	0.82	0.38
	Any 10+ drinking in past 2 weeks
18	0.10	0.04	0.04	0.05	0.36
19/20	0.11	0.00	0.01	0.07	0.50
21/22	0.15	0.00	0.00	0.10	0.64
23/24	0.12	0.00	0.04	0.04	0.57
25/26	0.11	0.04	0.02	0.03	0.59

Notes: Unweighted n = 1,078. HI=high-intensity. Bold font indicates item-response probabilities of ≥.50. Latent class membership probabilities sum to 1 horizontally. Item response probabilities (reflecting the probabilities of each level of use for the sample overall, as well as within latent classes), sum to 1 vertically within age. For example, at age 18, the overall probability of no alcohol use in the past 30 days was 0.51, the probability of having 1-9 drinks per occasion during the past two weeks was 0.39, and was 0.10 for having 10+ drinks.