Table 3.
Median scores of the 13 traits according to the category of anesthesiologist and AI text-to-image generator (N = 100 images in each cell, total 1,200 images).
| Trait | Anesthesiologist | |||||
|---|---|---|---|---|---|---|
| General | Cardiac | Pediatric | ||||
| AI1 | AI2 | AI1 | AI2 | AI1 | AI2 | |
| Threatening | 1.3 | 2.0 | 2.3 | 2.7 | 1.0 | 2.0 |
| Masculine | 4.7 | 5.0 | 5.3 | 5.0 | 2.3 | 3.3 |
| Feminine | 1.3 | 1.3 | 1.3 | 1.0 | 5.0 | 2.7 |
| Baby-faced | 1.7 | 1.0 | 1.3 | 1.3 | 3.7 | 6.0 |
| Attractive | 6.0 | 4.7 | 5.7 | 4.3 | 6.0 | 3.0 |
| Trustworthy | 5.0 | 5.0 | 5.0 | 5.0 | 6.0 | 2.0 |
| Happy | 1.7 | 1.3 | 2.3 | 1.3 | 5.3 | 2.3 |
| Angry | 1.0 | 2.0 | 1.3 | 1.7 | 1.0 | 2.0 |
| Sad | 1.3 | 2.3 | 1.3 | 1.7 | 1.0 | 2.7 |
| Disgusted | 1.0 | 1.3 | 1.0 | 1.0 | 1.0 | 2.0 |
| Surprised | 1.3 | 1.3 | 1.3 | 1.3 | 1.0 | 2.3 |
| Fearful/afraid | 1.3 | 1.7 | 1.3 | 1.3 | 1.0 | 2.7 |
| Unusual | 2.3 | 1.0 | 1.5 | 1.0 | 1.7 | 6.0 |
| Obstetric | Regional | HoD | ||||
|---|---|---|---|---|---|---|
| AI1 | AI2 | AI1 | AI2 | AI1 | AI2 | |
| Threatening | 1.3 | 1.7 | 1.0 | 2.7 | 1.3 | 1.3 |
| Masculine | 2.7 | 1.3 | 6.0 | 5.0 | 5.7 | 5.0 |
| Feminine | 4.3 | 5.7 | 1.0 | 1.0 | 1.0 | 1.0 |
| Baby-faced | 1.3 | 1.3 | 1.3 | 1.0 | 1.0 | 1.0 |
| Attractive | 5.3 | 5.7 | 6.3 | 5.0 | 5.7 | 5.0 |
| Trustworthy | 5.7 | 5.0 | 5.7 | 5.0 | 5.3 | 5.7 |
| Happy | 3.7 | 2.7 | 4.3 | 2.7 | 4.0 | 4.0 |
| Angry | 1.3 | 1.3 | 1.3 | 1.7 | 1.3 | 1.3 |
| Sad | 1.7 | 2.0 | 1.3 | 2.0 | 1.3 | 1.3 |
| Disgusted | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| Surprised | 1.0 | 1.3 | 1.3 | 1.3 | 1.0 | 1.0 |
| Fearful/afraid | 1.3 | 1.7 | 1.0 | 1.7 | 1.0 | 1.3 |
| Unusual | 1.7 | 1.7 | 2.7 | 1.0 | 3.0 | 1.0 |
Entries in bold font indicate traits whose median assessed score was at least 4 (the middle of the trait 1-7 scoring scale), thus highlighting the areas where the AI models exhibited the most pronounced biases.
AI1, ChatGPT DALL-E2; AI2, Midjourney; HoD, Head of Department.