Table 3.

Regression Analyses with Linear and Nonlinear Cumulative Risk (Ranging from 0 to 4) Predicting Psychosocial Outcomes

Predictors	Perceived Academic Difficulties β	Internalizing Symptoms β	Externalizing Symptoms β	Cigarette Use β	Heavy Drinking β	Illicit Drug Use OR	Criminal Behavior OR
Model 1
Gender^a	.00	−.72^***	−.27^***	−.11^***	−.01	.78^*	3.89^***
Number of Missing CR Variables	.05^***	.00	.04^**	.05^***	.06^***	1.10^**	1.04
CR Linear	.13^***	.09^**	.15^***	.15^***	.21^***	1.62^***	1.32^**
CR Quadratic	.00	−.05	−.04	.02	−.08^*	.93	.96
Model 2
Gender^a	.02	−.72^***	−.29^***	−.11^***	−.02	.72^*	3.65^***
Number of Missing CR Variables	.05^***	.00	.04^**	.05^***	.05^***	1.10^**	1.04
CR Linear	.15^***	.09^*	.11^**	.15^**	.18^***	1.43^*	1.17
CR Quadratic	−.02	−.06	−.01	.01	−.05	.97	.99
CR Linear X Gender	−.04	−.01	.06	.00	.04	1.35	1.19
CR Quadratic X Gender	.02	.01	−.04	.01	−.04	.91	.96

Note. N = 6963. CR = Cumulative Risk. β = standardized regression coefficients. OR = Odds Ratio. Girls = 0 and Boys = 1.

p < .05.

^**

p < .01.

^***

p < .001.

The CR Index that ranged from 0 to 4 was used in these analyses.

Because Gender is a binary independent variable, the STDY Standardization values are reported for all continuous outcome variables (Muthén & Muthén, 1998 – 2012).