Peer selection and influence effects on adolescent alcohol use: a stochastic actor-based model

BMC Pediatrics

Table 3 Stochastic Actor-Based Model Results for Network Selection

	β	SE(β)	p-value^a	Between school std. dev.	Fisher’s combination 1-side test
					Right-side (+)	Left-side (-)
Density	-3.42	0.38	<.001	1.91	1.000	<.001
Reciprocity	2.53	0.20	<.001	0.95	<.001	1.000
Transitive triplets	0.84	0.09	<.001	0.31	<.001	1.000
3-cycles	-0.44	0.14	.012	0.56	.997	<.001
In-degree popularity	0.09	0.01	<.001	0.08	<.001	1.000
Out-degree popularity	-0.16	0.03	<.001	0.18	1.000	<.001
Out-degree activity	-0.04	0.03	.197	0.16	.658	<.001
Adolescent age	-0.10	0.05	.097	0.31	.895	<.001
Friend age	0.05	0.03	.120	0.15	.012	.796
Age similarity	1.48	0.40	.002	1.81	<.001	1.000
Adolescent male	0.04	0.09	.664	0.84	.031	.423
Friend male	0.11	0.04	.016	0.45	.005	.995
Gender same	0.36	0.05	<.001	0.33	<.001	1.000
Adolescent minority	0.06	0.07	.408	0.30	.585	.725
Friend minority	-0.14	0.25	.608	0.99	.733	.145
Same race	0.58	0.43	.24.1	1.33	<.001	.991
Adolescent alcohol use	0.03	0.03	.361	0.25	.212	.767
Friend alcohol use	0.08	0.02	.001	0.18	.006	1.000
Alcohol use similarity	1.28	0.21	<.001	0.82	<.001	1.000

^aSnijders-Baerveldt two-sided test [39].
Parameter estimates β and standard error for stochastic actor-based model of the evolution of school friendships in the Add Health study. Coefficients correspond to the change in log-odds of a friendship nomination being present. Characteristics are bolded when the Fisher’s combination test yields a p-value less than 0.025.

ISSN: 1471-2431