Skip to main content
. 2010 Sep 22;12(11):1099–1107. doi: 10.1093/ntr/ntq156

Table 2.

Results for the Mediated Moderation Model

Model 1a
Model 2
Model 3
DV: Wave 2 thirty-day smokingb
DV: Wave 2 perceived friend prevalencec
DV: Wave 2 thirty-day smoking
Independent variable β SE p β SE p β SE p
Intercept −2.91 0.40 <.0001 .001 0.26 .99 −3.08 0.42 <.0001
C: Wave 1 thirty-day smoking 1.68 0.21 <.0001 1.46 0.23 <.0001
C: Wave 1 perceived friend smoking 0.25 0.03 <.0001 0.08 0.05 .11
X: Program (Tx = 1 and control = 0) 1.63 0.75 .03 0.09 0.31 .77 1.98 0.79 .01
Z: Comorbidity (CoM) 0.96 0.53 .07 1.33 0.39 .0007 0.85 0.57 .14
XZ: Program × CoM −1.39 0.66 .04 −1.20 0.49 .01 −1.29 0.70 .07
M: Wave 2 perceived friend smoking prevalence 0.29 0.04 <.0001
MZ: Wave 2 perceived friend smoking prevalence × CoM −0.06 0.14 .64

Note. C = control variable; DV = dependent variable; M = mediator variable; X = independent variable; Z = moderator variable; MZ = interaction between mediator and moderator variable; XZ = interaction between independent and moderator variable.

a

All models controlled for attrition propensity score and school.

b

Thirty-day smoking represents the number of smokers who indicated they smoked in the past thirty days within the indicated wave (1 = yes and 0 = no).

c

Perceived friend prevalence represents the number of friends they believe smoked in the past thirty days (0 = none to 12 = 12 friends or more).