Skip to main content
. Author manuscript; available in PMC: 2013 Jul 10.
Published in final edited form as: Prev Med. 2005 Mar;40(3):355–362. doi: 10.1016/j.ypmed.2004.06.014

Table 3.

Results from multi-level modeling 11th grade: frequency drinking at school last 30 days

Fixed effect Coefficient se t Ratio
Model for Individuals (N = 3,318)

Commercial sources 0.108 0.007 13.70**
Friends >21 source 0.012 0.009 1.34
Friends <21 source 0.004 0.013 0.32
Parent source 0.018 0.022 0.82
Stole from home source 0.113 0.020 5.46**

Model for Communities (N=93)
Commercial access rate (CAR) 0.058 0.025 2.28**
Minor in possession enforcement (MIP) −0.023 0.022 −1.05

Cross-Level Effects
CAR → Commercial source −0.043 0.008 −5.03**
CAR → Friend > 21 source 0.009 0.008 1.12
CAR → Friend < 21 source −0.000 0.014 −0.06
CAR → Parent source 0.059 0.023 2.53**
CAR → Home source 0.036 0.021 1.74+
MIP → Commercial source −0.019 0.008 −2.29**
MIP → Friend > 21 source −0.049 0.010 −4.83**
MIP → Friend <21 source 0.021 0.015 1.45
MIP → Parent source −0.021 0.021 −1.00
MIP → Home source −0.021 0.023 −0.94
Variance
Random effect Component se t-value
Community means, u0j 0.010 0.005 1.89*
Commercial source slopes, u1j 0.014 0.004 3.06**
Friends >21 source slopes, u2j 0.031 0.006 4.74**
Friends <21 source slopes, u3j 0.015 0.005 2.63**
Parent source slopes, u4j 0.030 0.015 1.95*
Stole from home source slopes, u5j 0.086 0.033 2.55**
Individual students, rij 0.725 0.018 38.65**
+

p < .10,

*

p < .05,

**

p < .01