Table 3.

The Results of LCA with Covariates ^a

	Model 1 (N = 7350) ^b			Model 2 (N = 7315)^b

	Substance Users ^c	Bullies ^c	Substance-using bullies ^c	Substance Users ^c	Bullies ^c	Substance-using bullies ^c

	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]
Gender
Female	1.35 [1.07–1.70]	0.75 [0.60–0.92]	0.58 [0.43–0.79]	1.49 [1.14–1.94]	0.77 [0.62–0.94]	0.68 [0.48–0.95]
Male (ref)
Age	1.90 [1.72–2.10]	0.90 [0.82–0.98]	1.42 [1.30–1.55]	1.87 [1.69–2.08]	0.88 [0.81–0.96]	1.38 [1.27–1.50]
Race/Ethnicity
African-American	0.79 [0.53–1.17]	1.86 [1.28–2.71]	1.40 [0.99–1.97]	0.53 [0.34–0.83]	1.50 [1.02–2.20]	0.84 [0.58–1.20]
Hispanic	1.36 [1.07–1.72]	1.00 [0.74–1.42]	1.76 [1.25–2.48]	1.10 [0.84–1.43]	0.90 [0.64–1.26]	1.30 [0.91–1.85]
Other race/ethnicity Caucasian (ref)	1.00 [0.67–1.49]	0.76 [0.53–1.09]	0.92 [0.61–1.37]	0.79 [0.54–1.16]	0.66 [0.45–0.98]	0.69 [0.43–1.21]
Maternal knowledge (lower vs. higher)				2.81 [2.32–3.41]	1.87 [1.36–2.56]	5.38 [3.73–7.74]
Paternal knowledge (lower vs. higher)				2.39 [1.90–3.00]	1.61 [1.33–1.94]	2.05 [1.47–2.88]
Spending evenings with friends (3–7 vs 0–2 evenings/week)				2.62 [2.12–3.26]	1.41 [1.13–1.76]	3.00 [2.19–4.43]

Note.

The analysis is analogous to a multinomial logistic regression, with the four-category latent class membership as the outcome variable. Due to missing data on the covariate variables, 230 adolescents (3.0%) were excluded in the analysis of Model 1 and 265 adolescents (3.5%) were excluded in the analysis of Model 2.

The purpose of Model 1 was to examine socio-demographic differences, whereas the main purpose of Model 2 was to examine the influence of peer affiliation with socio-demographic variables as covariates.

Class 1 (the noninvolved) was the referent for the latent class membership.