Table 5 Results of ordinary least-squares linear regressions showing peer influence effects for focal participant responses to experimentally measured smoking and vaping norms, and other smoking outcomes, at follow-up.

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

Outcomes	Time-point	Outcome variable: Focal participant responses to outcomes at follow-up
		(1) −i = Average of nominated friends^a				(1) −i = Average of nominated friends^b			(2) -i = Average of school class^a				(3) −i = Average of school year group^a
		n	Response_{−i, t}			Response_{−i, t}			n	Response_{-i, t}			n	Response_{−i, t}
		n	b (SE)^c	p-value	β^d	b (SE)^c	p-value	β^d	n	b (SE)^c	p-value	β^d	n	b (SE)^c	p-value	β^d
Experiment Part 2: Injunctive norms for smoking/vaping (−1 = “extremely socially inappropriate” to +1 = “extremely socially appropriate”).
P2S2	t = base	1073	0.34 (0.09)	<0.001^h	0.1396	0.27 (0.09)	0.003	0.1096	1087	0.44 (0.12)	<0.001^h	0.1176	1087	0.78 (0.18)	<0.001^h	0.1510
P2S2	t = fu	1018	0.15 (0.06)	0.02	0.0760	0.08 (0.06)	0.22	0.0399	1087	0.31 (0.10)	0.003	0.0933	1087	0.57 (0.16)	<0.001	0.1199
P2S3	t = base	1072	0.11 (0.07)	0.11	0.0489	0.12 (0.07)	0.09	0.0508	1086	0.17 (0.16)	0.28	0.0357	1086	0.01 (0.28)	0.96	0.0017
P2S3	t = fu	1017	0.12 (0.07)	0.09	0.0535	0.11 (0.07)	0.09	0.0534	1086	0.21 (0.14)	0.12	0.0523	1086	0.18 (0.25)	0.48	0.0213
P2S4	t = base	1070	0.13 (0.08)	0.09	0.0605	0.13 (0.08)	0.09	0.0601	1084	0.17 (0.12)	0.15	0.0424	1084	0.46 (0.32)	0.15	0.0582
P2S4	t = fu	1016	0.12 (0.07)	0.08	0.0568	0.09 (0.07)	0.19	0.0424	1084	0.29 (0.11)	0.01	0.0780	1084	0.44 (0.20)	0.03	0.0748
P2S5	t = base	1073	0.19 (0.06)	0.002	0.0918	0.19 (0.06)	0.003	0.0897	1087	0.32 (0.10)	0.001	0.0981	1087	0.48 (0.15)	0.001	0.1028
P2S5	t = fu	1018	0.17 (0.06)	0.003	0.0902	0.17 (0.06)	0.005	0.0885	1087	0.45 (0.09)	<0.001^h	0.1629	1087	0.45 (0.14)	0.001	0.1098
P2S6	t = base	1070	0.16 (0.07)	0.02	0.0762	0.16 (0.07)	0.02	0.0753	1084	0.36 (0.11)	0.001	0.1025	1084	0.40 (0.19)	0.03	0.0774
P2S6	t = fu	1015	0.16 (0.06)	0.007	0.0803	0.15 (0.06)	0.007	0.0797	1084	0.45 (0.10)	<0.001^h	0.1417	1084	0.51 (0.15)	0.001	0.0977
P2S7	t = base	1072	0.16 (0.05)	0.003	0.0870	0.16 (0.05)	0.003	0.0861	1086	0.29 (0.09)	0.001	0.1006	1086	0.28 (0.14)	0.04	0.0622
P2S7	t = fu	1018	0.27 (0.06)	<0.001^h	0.1474	0.27 (0.06)	<0.001ⁱ	0.1471	1086	0.41 (0.09)	<0.001^h	0.1358	1086	0.31 (0.17)	0.07	0.0630
P2S8	t = base	1073	0.30 (0.05)	<0.001^h	0.1578	0.30 (0.05)	<0.001ⁱ	0.1589	1087	0.31 (0.09)	<0.001	0.1037	1087	0.45 (0.14)	0.001	0.0940
P2S8	t = fu	1018	0.23 (0.05)	<0.001^h	0.1235	0.23 (0.05)	<0.001ⁱ	0.1228	1087	0.47 (0.08)	<0.001^h	0.1703	1087	0.48 (0.13)	<0.001	0.1083
P2S9	t = base	1072	0.16 (0.07)	0.02	0.0711	0.15 (0.07)	0.02	0.0697	1086	0.51 (0.12)	<0.001^h	0.1230	1086	0.56 (0.21)	0.009	0.0767
P2S9	t = fu	1017	0.24 (0.07)	<0.001	0.1183	0.24 (0.07)	<0.001	0.1176	1086	0.41 (0.11)	<0.001^h	0.1226	1086	0.21 (0.20)	0.29	0.0330
Experimental injunctive norms scale (average P2S2–P2S9)	t = base	1064	0.22 (0.06)	<0.001^h	0.1098	0.23 (0.06)	<0.001ⁱ	0.1111	1078	0.35 (0.10)	<0.001	0.1054	1078	0.41 (0.15)	0.005	0.0766
Experimental injunctive norms scale (average P2S2–P2S9)	t = fu	1011	0.25 (0.05)	<0.001^h	0.1341	0.25 (0.05)	<0.001ⁱ	0.1352	1078	0.47 (0.08)	<0.001^h	0.1707	1078	0.30 (0.14)	0.03	0.0620
Experiment Part 3: Descriptive norms for smoking/vaping (−1 = “none of my peers” to +1 = “all of my peers”).
P3Q1	t = base	1073	0.09 (0.06)	0.15	0.0451	0.10 (0.06)	0.11	0.0503	1087	0.24 (0.11)	0.03	0.0723	1087	0.15 (0.15)	0.33	0.0323
P3Q1	t = fu	1018	0.09 (0.06)	0.14	0.0466	0.09 (0.06)	0.14	0.0462	1087	0.30 (0.10)	0.004	0.0902	1087	0.20 (0.17)	0.25	0.0354
P3Q2	t = base	1073	0.19 (0.05)	<0.001	0.1017	0.20 (0.05)	<0.001	0.1033	1087	0.31 (0.08)	<0.001^h	0.1187	1087	0.32 (0.11)	0.003	0.0927
P3Q2	t = fu	1018	0.18 (0.06)	0.001	0.0942	0.18 (0.06)	0.002	0.0926	1087	0.30 (0.08)	<0.001^h	0.1105	1087	0.28 (0.12)	0.02	0.0695
Experimental descriptive norms scale (average P3Q1–P3Q2)	t = base	1073	0.15 (0.06)	0.01	0.0764	0.16 (0.06)	0.007	0.0813	1087	0.26 (0.09)	0.003	0.0923	1087	0.20 (0.11)	0.08	0.0544
Experimental descriptive norms scale (average P3Q1–P3Q2)	t = fu	1018	0.13 (0.06)	0.03	0.0642	0.13 (0.06)	0.03	0.0646	1087	0.28 (0.09)	0.002	0.0938	1087	0.17 (0.13)	0.19	0.0383
Experiment Part 4: Willingness to pay to support anti-smoking norms (Donations to ASSIST/Dead Cool; 0 = “0 tokens donated” to 10 = “10 tokens donated”).
Donation to ASSIST/Dead Cool	t = base	1071	0.14 (0.06)	0.02	0.0740	0.12 (0.06)	0.03	0.0658	1085	0.15 (0.09)	0.12	0.0505	1085	0.26 (0.20)	0.20	0.0437
Donation to ASSIST/Dead Cool	t = fu	1016	0.34 (0.06)	<0.001^h	0.1877	0.32 (0.06)	<0.001ⁱ	0.1754	1085	0.48 (0.09)	<0.001^h	0.1604	1085	0.66 (0.14)	<0.001^h	0.1494
Self-report injunctive smoking norms (−2 = “think(s) that I definitely should smoke” to +2 = “think(s) that I definitely should not smoke”).
IN1	t = base	1073	−0.04 (0.06)	0.51	−0.0173	−0.04 (0.07)	0.60	−0.0153	1082	0.05 (0.13)	0.72	0.0105	1082	0.36 (0.22)	0.09	0.0463
IN1	t = fu	1070	0.07 (0.06)	0.27	0.0311	0.06 (0.06)	0.31	0.0288	1082	−0.17 (0.12)	0.16	−0.0378	1082	0.23 (0.23)	0.32	0.0326
IN2	t = base	1075	−0.02 (0.08)	0.80	−0.0061	−0.02 (0.08)	0.76	−0.0075	1084	0.20 (0.19)	0.28	0.0422	1084	0.23 (0.56)	0.68	0.0200
IN2	t = fu	1072	0.14 (0.08)	0.07	0.0627	0.14 (0.08)	0.07	0.0602	1084	0.12 (0.11)	0.25	0.0308	1084	0.08 (0.27)	0.77	0.0095
IN3	t = base	1073	0.11 (0.07)	0.09	0.0513	0.11 (0.07)	0.11	0.0496	1082	0.18 (0.11)	0.13	0.0425	1082	0.26 (0.19)	0.16	0.0473
IN3	t = fu	1070	0.08 (0.05)	0.13	0.0387	0.08 (0.05)	0.17	0.0358	1082	0.003 (0.10)	0.98	0.0007	1082	0.12 (0.20)	0.54	0.0184
IN4	t = base	1071	0.03 (0.06)	0.60	0.0134	0.03 (0.06)	0.55	0.0151	1081	−0.02 (0.13)	0.85	−0.0050	1081	0.69 (0.32)	0.03	0.0730
IN4	t = fu	1069	0.06 (0.05)	0.28	0.0277	0.04 (0.05)	0.45	0.0194	1081	0.18 (0.12)	0.13	0.0396	1081	0.41 (0.22)	0.06	0.0567
IN5	t = base	1073	−0.01 (0.06)	0.83	−0.0055	0.005 (0.06)	0.93	0.0023	1082	−0.05 (0.11)	0.66	−0.0123	1082	−0.01 (0.23)	0.95	−0.0017
IN5	t = fu	1070	0.03 (0.06)	0.58	0.0142	0.03 (0.06)	0.59	0.0140	1082	0.04 (0.11)	0.74	0.0088	1082	0.09 (0.20)	0.66	0.0120
IN6	t = base	1074	0.21 (0.06)	<0.001	0.1013	0.19 (0.06)	0.001	0.0909	1083	0.20 (0.10)	0.05	0.0570	1083	0.44 (0.16)	0.007	0.0824
IN6	t = fu	1071	0.25 (0.06)	<0.001^h	0.1256	0.22 (0.06)	<0.001ⁱ	0.1129	1083	0.27 (0.09)	0.003	0.0794	1083	0.41 (0.16)	0.01	0.0801
IN7	t = base	1075	0.11 (0.06)	0.06	0.0537	0.09 (0.06)	0.14	0.0431	1084	0.21 (0.12)	0.07	0.0556	1084	0.57 (0.18)	0.002	0.0996
IN7	t = fu	1072	0.19 (0.06)	0.003	0.0944	0.17 (0.06)	0.007	0.0866	1084	0.21 (0.11)	0.06	0.0572	1084	0.34 (0.18)	0.06	0.0644
Self-report injunctive norms scale (average IN1–IN7)	t = base	1068	0.07 (0.06)	0.23	0.0310	0.07 (0.06)	0.25	0.0288	1078	0.06 (0.12)	0.63	0.0145	1078	0.38 (0.18)	0.04	0.0634
Self-report injunctive norms scale (average IN1–IN7)	t = fu	1066	0.19 (0.05)	<0.001	0.0936	0.19 (0.05)	<0.001	0.0923	1078	0.20 (0.09)	0.03	0.0534	1078	0.24 (0.16)	0.13	0.0426
Self-report descriptive smoking norms (DN1: 1 = “smoke(s) very often” to 5 = “never smoke(s)”/“don’t know”; DN2: 1 = “almost all of them smoke” to 5 = “almost none of them smoke”/“don’t know”).
DN1.1	t = base	1075	0.30 (0.10)	0.005	0.1340	0.28 (0.10)	0.008	0.1274	1084	0.30 (0.14)	0.04	0.0842	1084	1.06 (0.30)	0.001	0.1793
DN1.1	t = fu	1072	0.33 (0.09)	<0.001^h	0.1702	0.31 (0.09)	<0.001	0.1632	1084	0.24 (0.13)	0.06	0.0703	1084	0.62 (0.18)	0.001	0.1384
DN1.2	t = base	1076	0.13 (0.05)	0.007	0.0665	0.12 (0.05)	0.01	0.0636	1085	0.08 (0.07)	0.24	0.0279	1085	0.16 (0.09)	0.06	0.0434
DN1.2	t = fu	1073	−0.005 (0.04)	0.91	−0.0028	−0.02 (0.05)	0.71	−0.0092	1085	0.02 (0.08)	0.76	0.0073	1085	0.11 (0.09)	0.23	0.0266
DN1.3	t = base	1075	0.08 (0.04)	0.06	0.0399	0.08 (0.04)	0.07	0.0386	1084	0.06 (0.08)	0.45	0.0172	1084	0.13 (0.13)	0.30	0.0273
DN1.3	t = fu	1072	0.02 (0.05)	0.64	0.0113	0.02 (0.05)	0.71	0.0089	1084	0.03 (0.08)	0.74	0.0072	1084	0.08 (0.14)	0.59	0.0142
DN1.4	t = base	1075	0.01 (0.04)	0.81	0.0047	0.01 (0.04)	0.79	0.0053	1084	0.07 (0.13)	0.59	0.0165	1084	0.11 (0.14)	0.43	0.0173
DN1.4	t = fu	1072	−0.09 (0.04)	0.03	−0.0417	−0.09 (0.04)	0.04	−0.0414	1084	0.10 (0.15)	0.50	0.0253	1084	−0.10 (0.17)	0.56	−0.0134
DN1.5	t = base	1075	0.08 (0.05)	0.13	0.0300	0.07 (0.05)	0.15	0.0294	1084	0.16 (0.12)	0.16	0.0357	1084	0.29 (0.19)	0.13	0.0291
DN1.5	t = fu	1072	−0.01 (0.05)	0.81	−0.0048	−0.01 (0.05)	0.81	−0.0049	1084	0.18 (0.11)	0.10	0.0453	1084	0.09 (0.16)	0.58	0.0116
Self-report descriptive norms scale 1 (average DN1.1–DN1.5)	t = base	1073	0.09 (0.05)	0.06	0.0460	0.09 (0.05)	0.07	0.0438	1082	0.05 (0.07)	0.47	0.0181	1082	0.18 (0.10)	0.09	0.0474
Self-report descriptive norms scale 1 (average DN1.1–DN1.5)	t = fu	1070	0.04 (0.05)	0.41	0.0201	0.03 (0.05)	0.49	0.0168	1082	0.02 (0.07)	0.80	0.0062	1082	0.14 (0.10)	0.16	0.0380
DN2.1	t = base	1076	0.17 (0.07)	0.02	0.0774	0.16 (0.07)	0.03	0.0729	1085	0.17 (0.11)	0.11	0.0515	1085	0.21 (0.15)	0.14	0.0464
DN2.1	t = fu	1073	0.19 (0.07)	0.005	0.1012	0.18 (0.07)	0.007	0.0961	1085	0.19 (0.11)	0.10	0.0584	1085	0.12 (0.16)	0.46	0.0246
DN2.2	t = base	1076	0.17 (0.06)	0.006	0.0784	0.14 (0.06)	0.03	0.0656	1085	0.30 (0.10)	0.002	0.0955	1085	0.48 (0.16)	0.003	0.0994
DN2.2	t = fu	1073	0.11 (0.06)	0.07	0.0556	0.08 (0.06)	0.17	0.0431	1085	0.17 (0.09)	0.07	0.0526	1085	0.38 (0.15)	0.01	0.0827
DN2.3	t = base	1076	0.26 (0.07)	<0.001	0.1278	0.25 (0.08)	0.001	0.1189	1085	0.57 (0.10)	<0.001^h	0.1910	1085	0.65 (0.13)	<0.001^h	0.1703
DN2.3	t = fu	1073	0.18 (0.06)	0.002	0.1076	0.16 (0.06)	0.01	0.0930	1085	0.39 (0.09)	<0.001^h	0.1525	1085	0.53 (0.11)	<0.001^h	0.1633
Self-report descriptive norms scale 2 (average DN2.1-DN2.3)	t = base	1076	0.22 (0.07)	0.001	0.1025	0.19 (0.07)	0.005	0.0903	1085	0.37 (0.10)	<0.001^h	0.1250	1085	0.45 (0.13)	0.001	0.1151
Self-report descriptive norms scale 2 (average DN2.1-DN2.3)	t = fu	1073	0.11 (0.06)	0.06	0.0624	0.08 (0.06)	0.18	0.0455	1085	0.25 (0.08)	0.003	0.0936	1085	0.36 (0.12)	0.003	0.1014
Survey questions about smoking behavior, intentions, knowledge, attitudes, and psycho-social antecedents, and objectively measured smoking behavior^e
Self-report smoking behavior (1–4)	t = base	1083	0.16 (0.07)	0.03	0.0726	0.16 (0.07)	0.03	0.0719	1093	0.22 (0.11)	0.04	0.0557	1093	0.55 (0.16)	0.001	0.0920
Self-report smoking behavior (1–4)	t = fu	1081	0.24 (0.06)	<0.001^h	0.1342	0.24 (0.07)	<0.001ⁱ	0.1344	1093	0.22 (0.08)	0.01	0.0691	1093	0.38 (0.10)	<0.001^h	0.0885
Intentions (1–6)	t = base	1078	0.17 (0.08)	0.04	0.0630	0.15 (0.08)	0.07	0.0561	1087	0.48 (0.14)	<0.001	0.1078	1087	0.96 (0.24)	<0.001^h	0.1184
Intentions (1–6)	t = fu	1075	0.21 (0.08)	0.005	0.1160	0.20 (0.08)	0.01	0.1095	1087	0.11 (0.11)	0.32	0.0310	1087	0.38 (0.14)	0.007	0.0811
Knowledge (0–6)	t = base	1080	0.20 (0.05)	<0.001	0.1001	0.12 (0.06)	0.04	0.0621	1089	0.41 (0.08)	<0.001^h	0.1616	1089	0.45 (0.10)	<0.001^h	0.1366
Knowledge (0–6)	t = fu	1077	0.29 (0.06)	<0.001^h	0.1568	0.24 (0.06)	<0.001ⁱ	0.1329	1089	0.41 (0.08)	<0.001^h	0.1773	1089	0.42 (0.11)	<0.001^h	0.1362
Attitudes (1–5)	t = base	1067	0.16 (0.06)	0.007	0.0763	0.16 (0.06)	0.007	0.0763	1076	−0.04 (0.11)	0.73	−0.0095	1076	0.34 (0.21)	0.10	0.0501
Attitudes (1–5)	t = fu	1064	0.17 (0.06)	0.009	0.0844	0.17 (0.07)	0.009	0.0847	1076	−0.10 (0.12)	0.41	−0.0246	1076	0.23 (0.17)	0.17	0.0417
Self-efficacy (Emotional; 1–6)	t = base	1072	0.11 (0.07)	0.13	0.0484	0.10 (0.07)	0.19	0.0414	1081	0.30 (0.12)	0.01	0.0735	1081	0.46 (0.17)	0.005	0.0797
Self-efficacy (Emotional; 1–6)	t = fu	1068	0.29 (0.07)	<0.001^h	0.1505	0.26 (0.07)	<0.001	0.1355	1081	0.22 (0.09)	0.02	0.0629	1081	0.42 (0.13)	0.001	0.0913
Self-efficacy (Friends; 1–6)	t = base	1078	0.06 (0.06)	0.33	0.0272	0.06 (0.06)	0.35	0.0256	1087	0.20 (0.11)	0.08	0.0495	1087	0.18 (0.17)	0.29	0.0294
Self-efficacy (Friends; 1–6)	t = fu	1075	0.20 (0.07)	0.003	0.0992	0.18 (0.07)	0.006	0.0919	1087	0.11 (0.10)	0.27	0.0286	1087	0.27 (0.18)	0.13	0.0455
Self-efficacy (Opportunity; 1–6)	t = base	1079	0.13 (0.07)	0.07	0.0506	0.11 (0.07)	0.10	0.0441	1088	0.38 (0.14)	0.006	0.0847	1088	0.42 (0.20)	0.04	0.0646
Self-efficacy (Opportunity; 1–6)	t = fu	1076	0.17 (0.07)	0.02	0.0846	0.14 (0.07)	0.04	0.0738	1088	0.21 (0.10)	0.04	0.0588	1088	0.29 (0.17)	0.08	0.0539
Perceived physical risks (0–100%)	t = base	1074	0.07 (0.05)	0.18	0.0387	0.07 (0.05)	0.21	0.0361	1082	0.15 (0.09)	0.09	0.0470	1082	0.07 (0.14)	0.61	0.0141
Perceived physical risks (0–100%)	t = fu	1070	0.12 (0.05)	0.03	0.0604	0.11 (0.05)	0.04	0.0571	1082	0.11 (0.11)	0.32	0.0312	1082	−0.15 (0.21)	0.48	−0.0238
Perceived social risks (0–100%)	t = base	1076	0.15 (0.05)	0.003	0.0865	0.08 (0.06)	0.18	0.0421	1085	0.27 (0.07)	<0.001^h	0.1070	1085	0.40 (0.09)	<0.001^h	0.1248
Perceived social risks (0–100%)	t = fu	1073	0.13 (0.06)	0.02	0.0711	0.06 (0.06)	0.36	0.0318	1085	0.24 (0.08)	0.005	0.0831	1085	0.40 (0.12)	0.001	0.1113
Perceived addiction risks (0–100%)	t = base	1016	0.23 (0.06)	<0.001^h	0.1188	0.05 (0.07)	0.45	0.0260	1025	0.51 (0.09)	<0.001^h	0.1870	1025	0.71 (0.10)	<0.001^h	0.2253
Perceived addiction risks (0–100%)	t = fu	1009	0.14 (0.05)	0.004	0.0833	−0.06 (0.06)	0.33	−0.0328	1025	0.39 (0.08)	<0.001^h	0.1594	1025	0.58 (0.09)	<0.001^h	0.2073
Perceived benefits (0–100%)	t = base	1029	−0.04 (0.06)	0.46	−0.0208	−0.04 (0.06)	0.47	−0.0207	1038	0.08 (0.12)	0.48	0.0198	1038	0.34 (0.20)	0.10	0.0474
Perceived benefits (0–100%)	t = fu	1025	0.03 (0.06)	0.62	0.0137	0.03 (0.06)	0.62	0.0137	1038	−0.08 (0.12)	0.50	−0.0192	1038	−0.15 (0.25)	0.55	−0.0175
PBC (easy to quit; 1–5)	t = base	1076	0.27 (0.06)	<0.001^h	0.1457	0.03 (0.07)	0.63	0.0185	1085	0.50 (0.08)	<0.001^h	0.1987	1085	0.68 (0.09)	<0.001^h	0.2347
PBC (easy to quit; 1–5)	t = fu	1073	0.17 (0.06)	0.003	0.0970	−0.10 (0.07)	0.14	−0.0580	1085	0.50 (0.08)	<0.001^h	0.2046	1085	0.60 (0.09)	<0.001^h	0.2177
PBC (to avoid smoking; 1–5)	t = base	1080	0.06 (0.07)	0.37	0.0283	0.01 (0.07)	0.92	0.0034	1089	0.36 (0.13)	0.008	0.0875	1089	0.77 (0.22)	<0.001	0.1259
PBC (to avoid smoking; 1–5)	t = fu	1077	0.09 (0.07)	0.20	0.0426	0.01 (0.07)	0.84	0.0071	1089	0.12 (0.13)	0.39	0.0304	1089	0.53 (0.18)	0.003	0.1036
Objectively measured smoking behavior (0–30)	t = base	1041	0.37 (0.05)	<0.001^h	0.2707	0.07 (0.07)^f	0.32	0.0484	1048	0.38 (0.05)	<0.001^h	0.2634	1048	0.59 (0.06)	<0.001^h	0.3702
Objectively measured smoking behavior (0–30)	t = fu	1022	0.48 (0.09)	<0.001^h	0.3852	0.39 (0.09)	<0.001ⁱ	0.3103	1048	0.84 (0.06)	<0.001^h	0.5470	1048	0.87 (0.06)	<0.001^h	0.5078
Smoking susceptibility (binary)^g		n	OR (SE)	p-value	OR (SE)	p-value	n	OR (SE)		p-value		n	OR (SE)		p-value
	t = base	1078	1.14 (0.04)	<0.001^h	1.12 (0.04)	<0.001	1087	1.10 (0.07)		0.12		1087	1.21 (0.12)		0.05
	t = fu	1075	1.14 (0.03)	<0.001^h	1.11 (0.03)	0.001	1087	1.17 (0.06)		0.004		1087	1.31 (0.10)		<0.001

^aIn each model the outcome variable is the focal participant’s (i) response to the relevant item at follow-up. The predictor variable is the average of the relevant peer group’s (-i) responses to the equivalent item at baseline (t = base) or follow-up (t = fu), where –i = (1) focal participant’s nominated friends; (2) focal participant’s school class; (3) focal participant’s school year group. All models include robust (Huber–White) standard errors specified using Stata’s ‘vce(robust)’ option. The following baseline variables are included as covariates in all models: Gender (0 = boy; 1 = girl/prefer not to say), Age (1 = 12 years or less; 2 = 13 years; 3 = 14 years or more), Intervention (1 = ASSIST; 2 = Dead Cool), Ethnicity (0 = no ethnic minority; 1 = ethnic minority), Socio-economic status (NI: 1 = NIMDM2017 ≤ 296.6; 2 = 296.6 < NIMDM2017 ≤ 593.2; 3 = NIMDM2017 > 593.2; Bogotá: 1 = Informal settlement/Lowest/Low; 2 = Middle-Low/Middle; 3 = Middle-High/High), and baseline values of the outcome variable.
^bResults of models adjusted for Setting (0 = Northern Ireland; 1 = Bogotá). Baseline variables included as covariates in all models are: Gender, Age, Intervention, Ethnicity, Socio-economic status, baseline values of the outcome variable, and Setting. Results are reported for –i = (1) focal participant’s nominated friends.
^cUnstandardized regression coefficients representing the average change in i’s response category to the relevant question at follow-up for a one-unit increase in the average of –i’s responses to the equivalent item at baseline (t = base) or follow-up (t = fu), holding other variables constant.
^dStandardized regression coefficients representing the standard deviation change in the outcome variable, for a one standard deviation increase in the predictor variable.
^eHigher numerical values represent higher self-report anti-smoking outcomes, apart from objectively measured smoking behavior (higher numerical values represent higher levels of smoking behavior).
^fFor the model including focal participants’ carbon monoxide readings at follow-up as the outcome variable, and average carbon monoxide readings for nominated friends at baseline as the predictor variable, variance inflation factors were 3.00 for Setting and 3.02 for the predictor variable when the model was adjusted for Setting.
^gLogistic regressions were run for models including focal participants’ smoking susceptibility as the outcome variable, with robust (Huber White) standard errors specified using Stata’s ‘vce(robust)’ option. Results are odds ratios, standard errors, and p-values. Odds ratios represent the multiplicative change in odds of being susceptible to commencing smoking for a 10% increase in the number of nominated friends/pupils in the same school class/pupils in the same school year group classified as being susceptible to commencing smoking (1 out of 10 nominated friends/pupils in the same school class/pupils in the same school year group).
^hRetained statistical significance at the 5% level after using the Holm–Bonferroni procedure to correct the p-values for multiple testing (p ≤ 0.05; based on all tests reported in Table 5, excluding results of models which were repeated to adjust for Setting).
ⁱIndicates where the p-value in models examining peer influence effects from the average responses of nominated friends still met the Holm–Bonferroni threshold when the model was additionally adjusted for Setting (p ≤ 0.05).
Bold values show the results meeting the significance level of p ≤ 0.01.

Back to article page

Table 5 Results of ordinary least-squares linear regressions showing peer influence effects for focal participant responses to experimentally measured smoking and vaping norms, and other smoking outcomes, at follow-up.

Search

Quick links