Table 2.

Cross-sectional multivariate regression analyses showed that lower friendship quality was uniquely associated with higher levels of CU traits at each time point during adolescence.

Outcome variable: Friendship Quality -Time 1
Predictors - Time 1	B	SE	β	P
CU traits	−.01	.003	−.10	.01
Grandiose-Manipulative traits	−.001	.002	−.03	.46
Impulsive-Irresponsible traits	−.001	.003	−.02	.64
Aggression	−.004	.02	−.01	.83
Gender	−.21	.04	−.15	<.001
Age	.004	.01	.01	.73
Race	−.13	.04	−.11	<001
Interview location	.06	.03	.06	.05
Outcome variable: Friendship Quality -Time 2
Predictors - Time 2	B	SE	β	P
CU traits	−.02	.003	−.24	<001
Grandiose-Manipulative traits	.001	.002	.01	.91
Impulsive-Irresponsible traits	.003	.002	.05	.21
Aggression	.01	.02	.02	.62
Gender	−.19	.04	−.13	<001
Age	.004	.01	.01	.75
Race	−.12	.04	−.10	.001
Interview location	.07	.03	.07	.02
Outcome variable: Friendship Quality - Time 3
Predictors - Time 3	B	SE	β	P
CU traits	−.01	.003	−.10	.01
Grandiose-Manipulative traits	−.001	.002	−.01	.79
Impulsive-Irresponsible traits	−.002	.003	−.04	.38
Aggression	.02	.02	.03	.35
Gender	−.17	.04	−.11	<.001
Age	.002	.01	.004	.90
Race	−.09	.04	−.07	.01
Interview location	.02	.03	.02	.61

Note. We also modeled covariances between all predictor variables in the models (results available on request).