Table 2.
Regression models assessing the impact of wine glass size on volume of wine solda
| Bar & Restaurant Estimate (95 % CI) [p-value] |
Bar Estimate (95 % CI) [p-value] |
Restaurant Estimate (95 % CI) [p-value] |
|
|---|---|---|---|
| Modelling of the mean (identity link): | |||
| (Intercept) | 4.027 (2.644; 5.411) [<0.001]*** | 3.869 (2.295; 5.443) [<0.001]*** | 5.027 (3.402; 6.652) [<0.001]*** |
| Busyness level (log) | 0.815 (0.623; 1.008) [<0.001]*** | 0.774 (0.539; 1.009) [<0.001]*** | 0.677 (0.422; 0.931) [<0.001]*** |
| Monday | −0.044 (−0.143; 0.056) [0.389] | −0.025 (−0.188; 0.139) [0.767] | −0.149 (−0.298; 0.000) [0.052]. |
| Tuesday | −0.113 (−0.199;-0.028) [0.011]* | −0.202 (−0.337;-0.066) [0.005]** | −0.102 (−0.219; 0.015) [0.092]. |
| Wednesday | 0.014 (−0.082; 0.110) [0.777] | 0.040 (−0.070; 0.150) [0.481] | −0.063 (−0.218; 0.091) [0.423] |
| Thursday | −0.002 (−0.114; 0.110) [0.973] | 0.003 (−0.147; 0.154) [0.965] | −0.059 (−0.238; 0.120) [0.522] |
| Friday | 0.134 (0.024; 0.244) [0.019]* | 0.071 (−0.128; 0.271) [0.486] | 0.283 (0.167; 0.398) [<0.001]*** |
| Saturday | 0.177 (0.049; 0.304) [0.008]** | 0.231 (0.049; 0.414) [0.015]* | 0.309 (0.171; 0.448) [<0.001]*** |
| Sunday | −0.166 (−0.246;-.085) [<0.001]*** | −0.119 (−0.234;-0.004) [0.045]* | −0.219 (−0.345;-0.093) [0.001]*** |
| Small glass | −0.021 (−0.108; 0.066) [0.632] | 0.058 (−0.068; 0.184) [0.367] | −0.087 (−0.201; 0.028) [0.143] |
| Large glass | 0.090 (0.019; 0.161) [0.015]* | 0.134 (0.033; 0.236) [0.011]* | 0.079 (−0.025; 0.183) [0.140] |
| Temperature | −0.009 (−0.017;-0.001) [0.023]* | −0.017 (−0.028;-0.007) [0.002]** | −0.005 (−0.015; 0.005) [0.310] |
| Modelling of the variance (log link): | |||
| (Intercept) | −1.560 (−1.753;-.367) [<0.001]*** | −1.123 (−1.328;-.919) [<0.001]*** | −1.222 (−1.415;-.029) [<0.001]*** |
| Small glass | −0.053 (−0.408; 0.301) [0.769] | −0.111 (−0.466; 0.243) [0.540] | −0.179 (−0.527; 0.168) [0.315] |
| Large glass | −0.638 (−0.996;-0.279) [0.001]*** | −0.745 (−1.156;-0.333) [0.001]*** | −0.451 (−0.809;-0.093) [0.015]* |
| Monday | −0.358 (−0.705;-0.010) [0.047]* | −0.038 (−0.375; 0.298) [0.823] | 0.051 (−0.287; 0.390) [0.767] |
| Tuesday | −0.071 (−0.395; 0.253) [0.670] | 0.160 (−0.178; 0.497) [0.356] | −0.193 (−0.525; 0.139) [0.258] |
| Wednesday | 0.145 (−0.181; 0.471) [0.386] | −0.257 (−0.614; 0.099) [0.160] | 0.325 (−0.003; 0.652) [0.055]. |
| Thursday | 0.431 (0.102; 0.760) [0.012]* | 0.325 (−0.008; 0.658) [0.059]. | 0.510 (0.186; 0.834) [0.003]** |
| Friday | 0.099 (−0.246; 0.443) [0.576] | 0.464 (0.116; 0.811) [0.010]* | −0.172 (−0.516; 0.172) [0.329] |
| Saturday | −0.319 (−0.656; 0.018) [0.067]. | −0.613 (−0.956;-0.270) [0.001]*** | −0.491 (−0.818;-0.164) [0.004]** |
| Sunday | 0.073 (−0.302; 0.447) [0.705] | −0.039 (−0.415; 0.336) [0.838] | −0.029 (−0.392; 0.334) [0.875] |
Significance key: ‘.’ for p-value < 0.1, ‘*’ for p-value < 0.05, ‘**’ for p-value < 0.01, ‘***’ for p-value < 0.001
aThe outcome is the daily volume of wine sold on log scale. ‘Treatment’ contrasts with standard glass as reference were used for the glass size predictor. ‘Sum’ contrasts were used for the days of week. Parameter 95 % confidence intervals and p-values respectively appear in parenthesis and in square bracket