Table 2. Hierarchical logistic regression.

Drinking Category			β	S.E	Wald	p.	OR	95% C.I for OR
Step 1	Drinker vs. nondrinker	Intercept	-.68	1.04	.43	.51
		Age	.07	.05	1.76	.19	1.07	[.97, 1.18]
		Gender (f = 0, m = 1)	.28	.18	2.61	.11	1.33	[.94, 1.88]
		Home (urban = 0, rural = 1)	.07	.16	.16	.69	1.07	[.78, 1.47]
	Recent drinker vs. nondrinker	Intercept	-2.87	1.06	7.30	.01
		Age	.13	.05	6.41	.01**	1.14	[1.03, 1.26]
		Gender (f = 0, m = 1)	1.36	.18	58.23	< .01**	3.88	[2.74, 5.49]
		Home (urban = 0, rural = 1)	.47	.18	7.17	.01**	1.60	[1.13, 2.25]
Step 2	Drinker vs. nondrinker	Intercept	-.65	1.04	.39	.53
		Age	.07	.05	1.74	.19	1.07	[.97, 1.18]
		Gender (f = 0, m = 1)	.29	.18	2.73	.10	1.34	[.95, 1.89]
		Home (urban = 0, rural = 1)	.04	.17	.06	.80	1.04	[.75, 1.44]
		Chinese CO	.23	.29	.62	.43	1.26	[.71, 2.21]
		Western CO	.32	.27	1.36	.24	1.37	[.81, 2.33]
	Recent drinker vs. nondrinker	Intercept	-2.68	1.07	6.26	.01
		Age	.13	.05	5.89	.02*	1.14	[1.03, 1.26]
		Gender (f = 0, m = 1)	1.34	.18	55.15	< .01**	3.80	[2.67, 5.41]
		Home (urban = 0, rural = 1)	.33	.18	3.40	.07	1.39	[.98, 1.98]
		Chinese CO	.10	.31	.10	.75	1.10	[.61, 2.01]
		Western CO	1.31	.29	20.52	< .01**	3.71	[2.10, 6.54]

Note: OR: odds ratio. Step 1 χ² = 114.58**, Step 2 χ² = 143.13**

*significant at .05 level

** significant at .01 level.