Table 3.

Poisson regression results

Step	Predictors	B	SE	t
1	(intercept)	1.56	0.02	83.57***
	Race (Black = 1)	−0.31	0.15	−2.08*
	Gender (Male = 1)	0.14	0.04	3.56***
2	(intercept)	−0.12	0.13	−0.94
	Race	0.10	0.11	0.92
	Gender	−0.20	0.03	−6.03***
	Alcohol use	0.71	0.05	14.36***
3	(intercept)	−0.25	0.14	−1.78
	Race	−0.26	0.25	−1.04
	Gender	0.13	0.10	1.41
	Alcohol use	0.77	0.06	13.64***
	Race × Gender	−0.03	0.18	−0.15
	Race × Alcohol Use	0.18	0.11	1.73
	Gender × Alcohol Use	−0.13	0.03	−3.76***
4	(intercept)	−0.26	0.14	−1.82
	Race	−0.10	0.25	−0.40
	Gender	0.15	0.10	1.53
	Alcohol use	0.77	0.06	13.72***
	Race × Gender	−0.52	0.48	−1.09
	Race × Alcohol Use	0.10	0.11	0.95
	Gender × Alcohol Use	−0.13	0.03	−3.90***
	Race × Gender × Alcohol Use	0.23	0.17	1.33

Notes: The dependent variable was the total sum scores of the 18 dichotomous Rutgers Alcohol Problem Index items. Two dummy variables to control for time frame in months and 14 dummy variables to control for study differences were included as control variables in all steps of the Poisson regression model. N slightly varies across the four models because of missing responses (maximum n = 14,716 and minimum n = 14,343).

^*p <.05;

^***p < .001.