Table 3 School friendship networks descriptive statistics.

From: Selection homophily and peer influence for adolescents’ smoking and vaping norms and outcomes in high and middle-income settings

Setting	School	Year group size (nodes)	Time-point^a	Participants with identifiable nominations	Edges	Mean (SD) nominations^b	Network density^c	Dyadic reciprocity^d	Edgewise reciprocity^d	Number of reciprocated ties^d	Total number of transitive ties^e	Mean (SD) number of transitive ties^e	Transitivity^f	Transitive triplets^f	Mean (SD) number of actors at distance 2^g	Mean (SD) proportion of network at distance 2^g	Three-cycles^h	Jaccard indexⁱ
NI	1	83	Base	73	537	7.36 (2.40)	0.079	0.45	0.62	167	536	6.00 (3.15)	0.45	1513	14 (8)	0.17 (0.10)	430	0.4435
	1	83	FU	70	524	7.49 (2.64)	0.077	0.40	0.58	151	524	5.82 (3.62)	0.40	1300	15 (9)	0.18 (0.11)	318	0.4435
	2	120	Base	105	849	8.09 (2.53)	0.059	0.40	0.57	243	838	6.41 (3.60)	0.42	2656	17 (11)	0.15 (0.09)	722	0.4033
	2	120	FU	95	755	7.95 (2.49)	0.053	0.38	0.55	208	745	5.51 (3.75)	0.43	2269	15 (10)	0.13 (0.09)	575	0.4033
	3	115	Base	101	851	8.43 (2.32)	0.065	0.41	0.58	247	848	6.63 (3.54)	0.44	2825	18 (11)	0.16 (0.10)	748	0.4556
	3	115	FU	88	756	8.59 (1.88)	0.058	0.41	0.58	218	750	5.97 (3.77)	0.45	2316	15 (11)	0.13 (0.10)	599	0.4556
	4	125	Base	102	787	7.72 (2.57)	0.051	0.35	0.52	205	779	5.58 (3.77)	0.37	2015	17 (10)	0.13 (0.08)	494	0.3996
	4	125	FU	98	740	7.55 (2.60)	0.048	0.31	0.47	173	732	5.10 (3.69)	0.34	1575	16 (12)	0.13 (0.10)	322	0.3996
	5	97	Base	85	706	8.31 (2.41)	0.076	0.41	0.58	204	702	6.72 (3.45)	0.44	2192	18 (10)	0.19 (0.11)	550	0.5552
	5	97	FU	93	787	8.46 (2.20)	0.085	0.45	0.62	243	782	7.67 (2.79)	0.45	2876	20 (10)	0.21 (0.10)	796	0.5552
	6	178	Base	157	1230	7.83 (2.42)	0.039	0.41	0.58	355	1224	6.09 (3.33)	0.38	3376	20 (12)	0.11 (0.07)	939	0.4173
	6	178	FU	140	1015	7.25 (2.63)	0.032	0.32	0.49	247	998	4.84 (3.57)	0.41	2293	13 (10)	0.08 (0.06)	470	0.4173
Bogotá	7	148	Base	135	957	7.09 (2.71)	0.044	0.32	0.48	232	950	5.56 (3.27)	0.38	2366	17 (9)	0.11 (0.06)	559	0.3433
	7	148	FU	133	796	5.98 (2.62)	0.037	0.41	0.58	232	787	4.51 (3.04)	0.40	1770	13 (8)	0.09 (0.05)	504	0.3433
	8	107	Base	103	703	6.83 (2.68)	0.062	0.38	0.55	193	698	5.76 (3.01)	0.39	1757	16 (7)	0.15 (0.06)	465	0.4238
	8	107	FU	100	651	6.51 (2.57)	0.057	0.41	0.58	190	645	5.33 (3.03)	0.37	1510	15 (7)	0.14 (0.07)	423	0.4238
	9	106	Base	97	743	7.66 (2.71)	0.067	0.34	0.51	190	741	6.09 (3.28)	0.31	1624	23 (10)	0.22 (0.10)	420	0.4108
	9	106	FU	95	641	6.75 (2.84)	0.058	0.34	0.51	163	639	5.09 (3.26)	0.33	1289	17 (10)	0.16 (0.10)	341	0.4108
	10	133	Base	120	828	6.90 (2.50)	0.047	0.38	0.55	227	821	5.20 (2.91)	0.33	1820	19 (12)	0.14 (0.09)	498	0.327
	10	133	FU	119	775	6.51 (2.50)	0.044	0.40	0.57	221	768	4.98 (3.02)	0.37	1746	16 (10)	0.12 (0.08)	473	0.327
	11	92	Base	75	473	6.31 (2.80)	0.056	0.35	0.52	122	463	4.32 (3.17)	0.36	991	13 (9)	0.14 (0.10)	255	0.2623
	11	92	FU	63	374	5.94 (2.80)	0.045	0.40	0.57	107	367	3.36 (3.12)	0.46	859	8 (8)	0.09 (0.09)	218	0.2623
	12	140	Base	108	739	6.84 (2.69)	0.038	0.31	0.48	177	731	4.22 (3.44)	0.32	1416	16 (12)	0.12 (0.08)	348	0.3772
	12	140	FU	103	590	5.73 (2.42)	0.030	0.33	0.49	146	584	3.34 (2.95)	0.33	939	11 (9)	0.08 (0.06)	233	0.3772

NI Northern Ireland, SD standard deviation.
^aBase: baseline; FU: follow-up.
^bMean (SD) friendship nominations based on pupils who provided at least one identifiable friend nomination. There were 1266 participants who nominated at least one friend at baseline (5 had no identifiable nominations), and 1200 who nominated at least one friend at follow-up (3 had no identifiable nominations). There were 9538 friendship nominations at baseline (9403 [98.6%] were identifiable) and 8476 friendship nominations at follow-up (8404 [99.2%] were identifiable). At baseline 6473/9403 (68.8%) nominations were of friends from the same school class. At follow-up 5692/8404 (67.7%) nominations were of friends from the same school class. At baseline, 5124/9403 (54.5%) identifiable friend nominations were reciprocated (follow-up: n = 4598/8404 [54.7%]). On average, participants made 7.5 identifiable friend nominations at baseline, and 7.0 at follow-up (based on participants providing at least one identifiable friend nomination). The average school class size was 28.3 with a standard deviation (SD) of 6.8 (NI: mean = 24.5, SD = 5.0; Bogotá: mean = 32.1, SD = 6.2). The average school year group size was 125.7 with an SD of 26.7 (NI: mean = 127.1, SD = 32.0; Bogotá: mean = 124.4, SD = 20.0). Each network had one component.
^cThe ratio of observed edges to the total number of possible edges.
^dIn directed networks, reciprocity occurs when an actor receives a tie from another actor, and sends a tie back to the same actor. Dyadic reciprocity refers to the proportion of dyads that are symmetric (i.e., the ratio of mutuals to non-null dyads). Edgewise reciprocity refers to the proportion of edges in the network that are reciprocated. The number of reciprocated ties in the network is calculated as 0.5*(Edgewise reciprocity)*(Number of edges in the network).
^eThe total number of transitive ties refers to the number of edges where there is a direct (friendship nomination) as well as an indirect connection (through a mutual friend). The mean (SD) number of transitive ties is based on the total number of nodes in the network.
^fTransitivity is the ratio of transitive triads to (transitive plus intransitive triads). Transitive triplets refer to the total number of transitive triads in the network. A transitive triad occurs when if actor ‘a’ is connected to actor ‘b’ and actor ‘b’ is connected to actor ‘c’, then actor ‘a’ is connected to actor ‘c’. The Transitivity and Transitive triplets statistics account for ‘weak-form’ transitivity (i.e., actor ‘a’ nominates actor ‘b’, actor ‘b’ nominates actor ‘c’, and actor ‘a’ nominates actor ‘c’).
^gMean (SD) number of actors tied to each node through at least one intermediary (i.e., number of actors at geodesic distance 2), and mean (SD) proportion of the network tied to each node through at least one intermediary. Calculated based on directed networks.
^hTotal number of three-cycles in the network (i.e., closed triads in which actor ‘a’ nominates actor ‘b’, actor ‘b’ nominates actor ‘c’, and actor ‘c’ nominates actor ‘a’).
ⁱJaccard similarity index measures the degree of similarity between the friendship networks at baseline and follow-up in each school. It is computed for each school as: (the number of edges that are common to the baseline and follow-up network)/(the total number of edges in at least one of the networks [baseline or follow-up]). Range: 0 (completely dissimilar) to 1 (completely similar).

Back to article page

Table 3 School friendship networks descriptive statistics.

Search

Quick links