Table 4.

Bivariate and multivariate relationships between predictors and susceptibility to smoking^a

Predictor	OR (95% CI) for susceptibility to smoking, bivariate	OR (95% CI) for susceptibility to smoking, multivariate
Demographic characteristic
Age	1.00 (.90, 1.11)	.91 (.79, 1.05)
Gender	.90 (.70, 1.17)	1.12 (.78, 1.61)
African-American race	1.11 (.58, 2.11)	.66 (.26, 1.68)
Other race	1.41 (.70, 2.84)	1.59 (.60, 4.18)
Parental education^b	1.02 (.88, 1.17)	1.10 (.91, 1.33)
Environmental factor
Parent smoking	1.36 (1.03, 1.79)^c	.92 (.64, 1.33)
Sibling smoking	1.94 (1.39, 2.72)^c	1.14 (.73, 1.79)
Friend smoking	4.28 (3.19, 5.73)^c	3.28 (2.30, 4.68)^c
Responsive parenting	.61 (.48, .77)^c	1.15 (.79, 1.68)
Demanding parenting	.50 (.40, .64)^c	.72 (.51, 1.03)
Electronic media use	1.04 (1.01, 1.06)	1.00 (.96, 1.03)
Stress	1.07 (.92, 1.25)	1.11 (.88, 1.40)
Intrinsic factor
Grades	.57 (.46, .71)^c	.91 (.67, 1.23)
Depression	1.05 (.87, 1.27)	.72 (.54, .96)^c
Self-esteem	.63 (.50, .80)^c	.54 (.39, .76)^c
Rebelliousness	2.72 (2.09, 3.53)^c	1.60 (1.09, 2.33)^c
Sensation seeking	2.25 (1.80, 2.81)^c	1.66 (1.22, 2.27)^c
Knowledge of harm and addictiveness of smoking	1.03 (.90, 1.16)	1.00 (.85, 1.18)
SML level^d
SML as a continuous variable (1 point on a 10-point scale)	.64(.57, .71)^c	.68 (.58, .79)^c
SML above median (vs below median)	.42 (.32, .55)^c	.49 (.35, .68)^c

OR = odds ratio; CI = confidence interval; SML = smoking media literacy. For this analysis, the sample includes only those individuals who were not current smokers.

Parental education was used as a surrogate for socioeconomic status.

Result is statistically significant (p < .05).

SML was measured in two different ways: as a continuous variable and as a dichotomous variable.