Skip to main content
. 2014 Nov;75(6):982–992. doi: 10.15288/jsad.2014.75.982

Table 4.

Model parameters for mid-term drinking outcomes (6-month follow-up)

graphic file with name jsad982tbl4.jpg

Predictors Coeff. IRR or Ba SE [95% CI] z p
Drinking frequency
 Baseline drinking frequency 1.07 0.005 [1.06, 1.08] 15.57 <.001
 PNF 0.93 0.06 [0.83, 1.05] -1.16 .25
 DBF 0.97 0.06 [0.86, 1.10] -0.44 .66
Total alcohol quantity
 Negative binomial process
  Baseline alcohol quantity 1.02 0.001 [1.01, 1.02] 16.64 <.001
  PNF 0.89 0.06 [0.77, 1.02] -1.65 .10
  DBF 0.82 0.06 [0.71, 0.94] -2.81 .01
 Zero-inflated process
  Baseline alcohol quantity -0.02 0.01 [-0.04, -0.005] -2.53 .01
Alcohol problems
 Negative binomial process
  Baseline alcohol problems 1.08 0.01 [1.06, 1.09] 10.73 <.001
  PNF 0.93 0.10 [0.75, 1.14] -0.70 .48
  DBF 0.75 0.08 [0.61, 0.93] -2.59 .01
 Zero-inflated process
  Baseline alcohol problems -0.71 0.19 [-1.09,-0.33] -3.66 <.001

Notes: Coeff. = coefficients; IRR = incident rate ratio; CI = 95% confidence interval; PNF = personalized normative feedback; DBF = decisional balance feedback.

a

For the negative binomial process, coefficients are represented as IRRs. For the zero-inflated process, raw coefficients (B) are presented.