Table 7.

Weighted Parameter Estimates and 95% Confidence Intervals for a Logistic Regression Model of Past 30-day Cigarette Use at Wave 2 among 12-to-17-year-olds at Wave 1^a

Parameter	Estimate	Standard error	Value of t	Pr > ∣t∣	95% Confidence limits
Intercept	−1.02	0.07	−14.97	< .001	−1.15	−0.88
Wave 1: Cigarettes cause a lot of harmb	−0.12	0.06	−1.95	.054	−0.24	0.00
Wave 2: Cigarettes cause a lot of harmc	−0.50	0.06	−7.95	< .001	−0.62	−0.37
Race is White (alone)	0.24	0.05	4.69	< .001	0.14	0.35
Wave 2 household income above $50,000	−0.31	0.05	−6.14	< .001	−0.41	−0.21
Wave 1 past 30-day cigarette used	1.83	0.06	30.09	< .001	1.71	1.95

Note.

^aCases with a missing value for any of the variables in the model were excluded.

^bWave 1 value recoded from R01_YC1125 (How much do you think people harm themselves when they smoke cigarettes? A lot of harm vs some, a little, or no harm).

^cWave 2 value recoded from R02_YC1125 for youth (as done for R01_YC1125) and R02_AC9050 for those 18 or older at Wave 2 (How harmful do you think cigarettes are to health? Extremely or very harmful vs Somewhat, slightly, or not at all harmful).

^dWave 1 value recoded from R01R_Y_CUR_CIGS (smoked a cigarette in past 30 days). Wave 2 value (dependent variable) recoded from R02R_Y_CUR_CIGS for youth and R02R_A_P30D_CIGS for those 18 or older at Wave 2 (smoked a cigarette in past 30 days).