Table 6.
Study 2: unadjusted and adjusted multi‐level regression model with serving size predicting observed alcohol consumption (UK units); participants are clustered in teams (level 2) and quiz nights (level 3).
| Unadjusted (n = 164) | Adjusted (n = 148) | |||||
|---|---|---|---|---|---|---|
| B (SE) | (BCa 95% CI) | P | B (SE) | (BCa 95% CI) | P | |
| Fixed components | ||||||
| Intercept | 2.88 (0.23) | (2.41, 3.36) | 0.001 | 1.59 (0.53) | (0.43, 3.07) | 0.004 |
| Serving size condition (reference: standard) | −1.14 (0.28) | (−1.68, −0.60) | 0.001 | −0.73 (0.52) | (−1.78, 0.27) | 0.14 |
| Gender (reference: male) | −0.54 (0.57) | (−1.59, 0.42) | 0.35 | |||
| Serving size × gender | −0.31 (0.69) | (−1.68, 1.05) | 0.66 | |||
| AUDIT‐C | 0.30 (0.08) | (0.14, 0.44) | 0.001 | |||
| Consumption before quiz | 0.03 (0.10) | (−0.12, 0.17) | 0.73 | |||
| Random components | ||||||
| Level 3 × 2 variance (quiz night × teams) | 1.38 (0.37) | 1.17 (0.37) | ||||
| Level 1 variance (participants) | 2.84 (0.35) | 2.23 (0.28) | ||||
AUDIT = Alcohol Use Disorders Identification Test; AUDIT‐C scores range between 0 and 12. CI = confidence interval; SE = standard error; BCa = bias‐corrected and accelerated.