Table 4.
Risk Factors for Underestimating and Overestimating Wines and Spirits from Multivariate Multinomial Logistic Regression
| Wine pouring accuracy (n = 248) | Spirit pouring accuracy(n = 192) | |||
|---|---|---|---|---|
| RRR | 95% CI | RRR | 95% CI | |
| Within 0.5 units | 1.00 (reference) | 1.00 (reference) | ||
| Underestimated by >0.5 units | ||||
| Sex (female) | 1.68 | 0.76 to 3.69 | 2.03 | 0.72 to 5.71 |
| Age 35 to 54 (vs. 16 to 34) | 1.07 | 0.44 to 2.63 | 0.59 | 0.18 to 1.93 |
| Age 55+ (vs. 16 to 34) | 2.26 | 0.60 to 8.53 | 0.18 | 0.01 to 2.76 |
| Volume poured (ml) | 1.02 | 1.01 to 1.02** | 1.04 | 1.01 to 1.06* |
| No degree (vs. degree or equivalent) | 0.53 | 0.20 to 1.37 | 1.77 | 0.59 to 5.35 |
| Non-white (vs. white) | 3.88 | 1.65 to 9.16* | – | – |
| Unemployed or retired (vs. employed) | 4.30 | 1.08 to 17.07* | – | – |
| Full-time student (vs. employed) | 0.71 | 0.24 to 2.11 | – | – |
| Drank >4/3 on heaviest day in last week | – | – | 1.00 | 0.28 to 3.59 |
| Drank >8/6 on heaviest day in last week | – | – | 0.52 | 0.14 to 1.95 |
| Overestimated by >0.5 units | ||||
| Sex (female) | 1.08 | 0.58 to 1.99 | 0.92 | 0.44 to 1.92 |
| Age 35 to 54 (vs. 16 to 34) | 1.30 | 0.64 to 2.65 | 0.74 | 0.32 to 1.71 |
| Age 55+ (vs. 16 to 34) | 1.55 | 0.51 to 4.72 | 0.64 | 0.16 to 2.55 |
| Volume poured (ml) | 1.00 | 1.00 to 1.01 | 0.99 | 0.97 to 1.02 |
| No degree (vs. degree or equivalent) | 1.89 | 0.99 to 3.59* | 2.78 | 1.28 to 6.07* |
| Non-white (vs. white) | 1.44 | 0.68 to 3.02 | – | – |
| Unemployed or retired (vs. employed) | 2.40 | 0.70 to 8.20 | – | – |
| Full-time student (vs. employed) | 1.10 | 0.50 to 2.44 | – | – |
| Drank >4/3 on heaviest day in last week | – | – | 1.15 | 0.40 to 3.31 |
| Drank >8/6 on heaviest day in last week | – | – | 2.52 | 0.95 to 6.63 |
RRR, relative risk ratio; 95% CI, 95% confidence interval.
*p > 0.05, **p < 0.001.
Mutually adjusted. Slightly different models were constructed for wine and spirits based on the results of likelihood ratio tests. Covariates not included in the model are designated with dashes.