Table 7.
Estimated marginal means and 95% CIs for the categorical variables within the hierarchical mixed-model regression framework calculated for each specialty.
| Factor | Level | Layperson | Orthodontist | Prostho-Resto | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SE | 95% CI | Mean | SE | 95% CI | Mean | SE | 95% CI | |||||
| Judge’s Sex | Female | 4.964 | 0.125 | 4.719 | 5.209 | 4.543 | 0.278 | 3.986 | 5.100 | 4.226 | 0.258 | 3.708 | 4.745 |
| Male | 4.801 | 0.205 | 4.397 | 5.205 | 5.587 | 0.331 | 4.922 | 6.251 | 4.999 | 0.429 | 4.128 | 5.870 | |
| Facial index | 80% (Euryprosopic) | 4.622 | 0.127 | 4.372 | 4.871 | 4.834 | 0.227 | 4.382 | 5.285 | 4.341 | 0.259 | 3.821 | 4.861 |
| 86% (Mesoprosopic) | 5.101* | 0.127 | 4.852 | 5.351 | 5.243* | 0.227 | 4.792 | 5.695 | 4.846* | 0.259 | 4.325 | 5.366 | |
| 94% (Leptoprosopic) | 4.925 | 0.127 | 4.676 | 5.174 | 5.118* | 0.226 | 4.667 | 5.569 | 4.651* | 0.259 | 4.132 | 5.171 | |
| Midline deviation | 4 mm to the left | 3.854 | 0.142 | 3.576 | 4.133 | 3.576 | 0.253 | 3.076 | 4.076 | 3.101 | 0.290 | 2.523 | 3.678 |
| 3 mm to the left | 4.363 | 0.142 | 4.085 | 4.642 | 4.127 | 0.253 | 3.627 | 4.627 | 3.777 | 0.290 | 3.199 | 4.355 | |
| 2 mm to the left | 4.468 | 0.138 | 4.196 | 4.739 | 4.780 | 0.246 | 4.292 | 5.269 | 4.110 | 0.283 | 3.546 | 4.674 | |
| 1 mm to the left | 5.580* | 0.142 | 5.301 | 5.859 | 6.039* | 0.253 | 5.538 | 6.539 | 5.605* | 0.290 | 5.028 | 6.183 | |
| No midline deviation | 5.767* | 0.113 | 5.544 | 5.989 | 6.385* | 0.203 | 5.977 | 6.794 | 6.129* | 0.231 | 5.659 | 6.599 | |
| 1 mm to the right | 5.871* | 0.142 | 5.592 | 6.149 | 6.290* | 0.253 | 5.790 | 6.791 | 5.863* | 0.290 | 5.285 | 6.440 | |
| 2 mm to the right | 5.166 | 0.142 | 4.887 | 5.444 | 5.739 | 0.253 | 5.239 | 6.240 | 5.158 | 0.290 | 4.580 | 5.736 | |
| 3 mm to the right | 4.701 | 0.142 | 4.422 | 4.979 | 4.719 | 0.253 | 4.218 | 5.219 | 4.272 | 0.290 | 3.694 | 4.850 | |
| 4 mm to the right | 4.175 | 0.142 | 3.897 | 4.454 | 3.930 | 0.253 | 3.429 | 4.430 | 3.501 | 0.290 | 2.923 | 4.078 | |
| Midline roll | 15° to the left | 4.429 | 0.142 | 4.150 | 4.709 | 4.565 | 0.253 | 4.063 | 5.066 | 4.071 | 0.291 | 3.492 | 4.651 |
| 10° to the left | 4.325 | 0.142 | 4.046 | 4.605 | 4.401 | 0.253 | 3.899 | 4.903 | 3.747 | 0.291 | 3.168 | 4.327 | |
| 5° to the left | 4.534 | 0.142 | 4.254 | 4.813 | 5.102 | 0.253 | 4.600 | 5.604 | 4.367 | 0.291 | 3.787 | 4.946 | |
| No midline roll | 5.567* | 0.112 | 5.345 | 5.788 | 5.979* | 0.202 | 5.573 | 6.386 | 5.652* | 0.230 | 5.184 | 6.120 | |
| 5° to the right | 5.132 | 0.142 | 4.852 | 5.411 | 5.292 | 0.253 | 4.791 | 5.794 | 4.805 | 0.291 | 4.225 | 5.384 | |
| 10° to the right | 4.948 | 0.142 | 4.668 | 5.227 | 5.041 | 0.253 | 4.539 | 5.543 | 4.738 | 0.291 | 4.158 | 5.317 | |
| 15° to the right | 5.244 | 0.142 | 4.965 | 5.524 | 5.075 | 0.253 | 4.573 | 5.577 | 4.909 | 0.291 | 4.330 | 5.489 | |
SE, standard error; CI, confidence interval. Within each factor (categorical predictor), asterisks highlight the highest beauty score as well as features that are not significantly different from the top score according to the Bonferroni post hoc test at a 0.008 level of significance; in other words, a number of asterisks in a predictor section mark the zone of acceptability. A lack of asterisks within the ‘sex’ variable means non-significant main effects estimated using the mixed-model regression analyses.