Table 3.

Regressions for people under the statutory curfew age who lived in zip codes that were 100% within the city boundaries with no interaction terms between curfews and gender, age, or race/ethnicity

Variable	Coef.	Robust SE	T	P > |t|
Model 1—Annual drinking [F(15, 16) = 274.3; Prob > F = .001]
Curfew	-.034	.023	-1.46	n.s.
Year	-.009	.002	-3.88	.001
Age	-.050	.005	-9.86	.001
Gender	-.027	.013	-2.04	.058
Black	-.139	.047	-2.95	.010
Hispanic	-.031	.028	-1.08	n.s.
Grades
A’s	-.281	.022	-12.93	.001
B’s	-.149	.027	-5.59	.001
C’s	-.059	.022	-2.74	.015
Having 1 parent in the house	-.057	.027	-2.12	.050
Having 2 parents in the house	-.002	.027	-0.06	n.s.
Model 2—Monthly drinking [F(15, 16) = 301.33; Prob > F = .001]
Curfew	-.063	.025	-2.53	.022
Year	-.007	.002	-4.59	.001
Age	-.040	.007	-5.79	.001
Gender	-.008	.015	-0.52	n.s.
Black	-.160	.037	-4.28	.001
Hispanic	-.021	.024	-0.87	n.s.
Grades
A’s	-.278	.025	-11.13	.001
B’s	-.176	.021	-8.21	.001
C’s	-.070	.020	-3.58	.002
Having 1 parent in the house	-.008	.027	-0.29	n.s.
Having 2 parents in the house	-.038	.028	-1.34	n.s.
Model 3—Heavy episodic drinking [F(15, 16) = 3290.6; Prob > F = .001]
Curfew	-.046	.019	-2.42	.028
Year	-.002	.001	-1.13	n.s.
Age	-.021	.007	-3.24	.005
Gender	-.032	.014	-2.28	.037
Black	-.125	.031	-4.05	.001
Hispanic	-.011	.022	-0.48	n.s.
Grades
A’s	-.215	.025	-8.51	.001
B’s	-.144	.025	-5.78	.001
C’s	-.056	.018	-3.06	.007
Having 1 parent in the house	-.012	.019	-0.62	n.s.
Having 2 parents in the house	-.049	.020	-2.44	.027

Notes: Coef. = coefficient; prob = probability; n.s. = not significant.