Table 2.

Regression Parameters (B) and Odds Ratios (OR) for Trait Predictors in Elementary School for Cigarette Smoking Class Comparisons in High School

	Contrast 1		Contrast 2		Contrast 3
	All smokers (n = 249) vs. Nonsmokers (n = 684)		Stable High Smokers (n = 56) vs. Rapid Escalators and Experimenters (n = 223)		Rapid Escalators (n = 82) vs. Experimenters (n = 141)
Predictors	B	OR [95% CI]	B	OR [95% CI]	B	OR [95% CI]
Gender	0.36*	1.44 [1.05, 1.96]	0.76*	2.13 [1.08, 4.21]	0.65*	1.91 [1.04, 3.50]
Tried cigarettes at grade 4	0.44	1.55 [0.79, 3.06]	1.56**	4.78 [1.65, 13.88]	0.65	1.92 [0.44, 8.39]
Free or reduced lunch	0.38*	1.46 [1.09, 1.96]	0.71**	2.04 [1.05, 3.72]	0.86**	2.36 [1.33, 4.18]
Sensation seeking-I	3.09***	21.96 [9.40, 51.35]	1.06	2.90 [0.46, 18.43]	0.34	1.40 [0.27, 7.23]
Sensation seeking-S	3.63***	37.78 [7.37, 193.59]	−2.71	0.07 [0.00, 2.27]	0.25	1.28 [0.05, 31.7]

Note. For all contrasts, the first-named group was coded 1. A positive B indicates that variable predicts greater odds of being in the first class rather than the second. Gender coded: male = 0, female = 1. I = intercept, S = slope. Results are for models with the nonsignificant interactions between sensation seeking and gender removed.

*p < .05. **p < .01. ***p < .001.