Skip to main content
. 2013 Sep;74(5):736–745. doi: 10.15288/jsad.2013.74.736

Table 3.

Regression analyses for drinking frequency outcomes

Variable Estimate SE p
Outcome: Drinking days
 Time 1 model
Overall mental health MHI-5 -.06 .12 .636
Perceived norms .43 .26 .104
MHI-5 × Perceived Norms .00 .08 .995
 Time 2 model
Overall mental health MHI-5 -.01 .13 .914
Perceived norms .38 .27 .165
MHI-5 × Perceived Norms -.03 .08 .706
Outcome: Heavy drinking days
 Time 1 model
Overall mental health MHI-5 .07 .12 .545
Perceived norms .60 .25 .020
MHI-5 × Perceived Norms -.07 .08 .373
 Time 2 model
Overall mental health MHI-5 .04 .12 .705
Perceived norms .43 .25 .085
MHI-5 × Perceived Norms -.04 .07 .559

Notes: MHI-5 = Mental Health Inventory score. SE = standard error. Significant effect is bolded. We did not run follow-up tests for the number of drinking days or the number of heavy drinking days because interactions between the full MHI-5 composite score and the perceived norms interaction terms were nonsignificant for either outcome variable.