Skip to main content
. Author manuscript; available in PMC: 2019 May 1.
Published in final edited form as: Psychol Addict Behav. 2018 May;32(3):338–349. doi: 10.1037/adb0000360

Table 4.

Results Testing Video Condition as a Moderator of Implicit and Explicit Alcohol Approach on Alcohol Consumption

Alcohol consumed (ml) B=ln eB SE(B) z p eB LL95% UL95%
Intercept   5.069 0.058 86.94 < .001 159.03 141.86   178.28  
Weekly Consumption   0.009 0.008 1.18    .24 1.01 0.99   1.03  
Gender −0.464 0.122 −3.81 < .001 0.63 0.50   0.80  
Video Condition −0.129 0.123 −1.04    .30 0.88 0.69   1.12  
Approach IAT −0.125 0.140 −0.89    .37 0.88 0.67   1.16  
Explicit Approach   0.041 0.035 1.17    .24 1.04 0.97   1.12  
Video × Approach IAT   0.107 0.277 0.39    .70 1.11 0.65   1.92  
Video × Explicit Approach   0.052 0.075 0.69    .49 1.05 0.91   1.22  
alpha (dispersion parameter)   0.473 0.064 0.362 0.617
Alcohol Percent of Total B SE(B) t p β UL95% UL95%
Intercept   0.546 0.012 46.72 < .001 – –   –  
Weekly Consumption   0.003 0.002 1.86    .07 0.16 −0.01   0.34  
Gender   0.004 0.025 0.16    .88 0.01 −0.16   0.19  
Video Condition −0.032 0.025 −1.27    .21 −0.11 −0.27   0.06  
Approach IAT   0.052 0.028 1.87    .06 0.16 −0.01   0.33  
Explicit Approach   0.007 0.007 1.02    .31 0.09 −0.08   0.25  
Video × Approach IAT −0.042 0.055 −0.77    .44 −0.06 −0.23   0.10  
Video × Explicit Approach −0.003 0.015 −0.21    .84 −0.02 −0.18   0.14  

Note. Bolded rows indicated significant effects; alpha = .034. B = unstandardized parameter estimate, log-linked for truncated negative binomial regression of alcohol consumed (ml); SE = standard error of the parameter estimate; eB = exponentiated coefficient also known as the Incident Rate Ratio; LL95% and UL95% = lower and upper limits of 95% confidence intervals. Weekly consumption equals total typical weekly alcohol consumption in US standard drinks as reported on the Daily Drinking Questionnaire. Gender was coded 0 = male, 1 = female. Video condition = exposure to either happy/neutral (Muppets or Denali) or sad (Schindler’s List) video clip; video condition was coded 0 = happy/neutral, 1 = sad. Predictors were mean centered prior to creation of product terms.