Skip to main content

Table 3 Multilevel logistic regression models with odds ratios and 95% confidence intervals predicting lifetime drunkenness

From: Adolescent substance use and peer use: a multilevel analysis of cross-sectional population data

Model

Empty

Full

Individual level fixed effects †

 

OR

95% CI for ORs

Academic grades

 

0.952***

0.940-0.964

Time spent with parents

 

0.864***

0.837-0.892

Peer drunkenness

 

2.423***

2.245-2.619

Peer respect for alcohol use

 

3.538***

2.272-5.509

Daily smoking

 

7.466***

4.465-11.227

Lifetime cannabis use

 

8.032***

5.864-11.003

Gender

 

1.396***

1.217-1.601

Family structure

 

1.399***

1.225-1.598

Parental Education

 

0.981

0.956-1.006

Attends school in home neighborhood

 

0.942

0.771-1.152

Family financial status

 

1.038

0.969-1.113

Resid. moving in the last 12 months

 

0.984

0.736-1.316

Changed schools in the last 12 months

 

0.973

0.688-1.378

Resides in the capital area

 

0.808*

0.683-0.955

School level fixed effects ††

   

Peer drunkenness

 

1.518**

1.129-2.042

Cross-level interaction effect

   

Peer respect for alcohol use*School level peer drunkenness

 

0.702***

0.565-0.872

Random effects

 

Var. comp.

SD

df

χ 2

Family financial status u5

 

0.012*

0.01

135

163.76

Intercept u0

 

0.064

0.25

134

188.72

Model fit statistics

     

MOR

1.47

 

ICC

0.047

 

R2dicho

 

0.87

−2 log likelihood

−9903.62

−10329.97

  1. † Based on Wald T and 6,513 df.
  2. †† Based on Wald T and 139 df.
  3. * p < .05, ** p < .01, *** p < .001.
Back to article page

Substance Abuse Treatment, Prevention, and Policy

ISSN: 1747-597X