Table 2.

Results of logistic regression analysis predicting the probability of DSM-IV alcohol dependence during the first year of college.

Variable	Bivariate Models					Multivariate Model
	B	SE	p	OR	95% CI	B	SE	p	OR	95% CI
High school alcohol usea	1.59	.25	<.001	4.90	[3.00, 8.01]	1.20	.26	<.001	3.31	[1.98, 5.53]
Genderb	−.03	.16	.86	.97	[.71, 1.33]	−.08	.18	.67	.93	[.65, 1.31]
Racec	.38	.20	.05	1.46	[.99, 2.14]	.21	.22	.33	1.24	[.80, 1.91]
Mother’s educationd	−.03	.19	.86	.97	[.67, 1.40]	−.02	.20	.93	.98	[.66, 1.46]
Importance of religione	−.18	.16	.27	.84	[.61, 1.15]	−.01	.18	.95	.99	[.70, 1.40]
Sensation seeking	.22	.04	<.001	1.25	[1.15, 1.35]	.17	.04	<.001	1.18	[1.08, 1.29]
Parental monitoring	−.07	.013	<.001	.93	[.91, .96]	−.05	.02	.001	.95	[.93, .98]

Note. CI=confidence interval for odds ratio (OR). Sample sizes range (n=1,075 to n=1,239) due to item-level missing data.

^aAlcohol use=1, No alcohol use=0

^bMale=1, Female=0

^cWhite=1, non-White=0

^dBachelor’s/graduate degree=1, Some college/high school or GED/less than high school=0

^eModerately/Extremely important=1, Slightly/Not important=0