Table 3. The percentage preference for the different dish pairs in Experiment 7, sorted according to preference magnitude.
Note how the column defining whether the dishes were odd or even seems relatively random in terms of order, which implies no relationship between this and strength of Preference.
| Items | Dish with more elements, by % of participants, is | ||||
|---|---|---|---|---|---|
| Dish1 | Dish2 | Difference | Dish1, Dish2, odd or even items | Preferred (95% CI) | Bigger portion (95% CI) |
| 1 | 3 | 2 | OO | 73 (63.2, 81.39) *** | 47 (36.94, 57.24) |
| 2 | 4 | 2 | EE | 70 (60.02, 78.76) ** | 40 (30.33, 50.28) |
| 4 | 5 | 1 | EO | 66 (55.85, 75.18) * | 42 (47.71, 67.8) |
| 2 | 3 | 1 | EO | 62 (51.75, 71.52) | 42 (32.2, 52.29) |
| 1 | 2 | 1 | OE | 62 (51.75, 71.52) | 44 (34.08, 54.28) |
| 3 | 5 | 1 | OO | 59 (48.71, 68.74) | 50 (39.83, 60.17) |
| 5 | 6 | 2 | OE | 57 (46.71, 66.86) | 48 (37.9, 58.22) |
| 4 | 6 | 2 | EE | 57 (46.71, 66.86) | 60 (30.33, 50.28) |
| 3 | 4 | 1 | OE | 53 (42.76, 63.06) | 47 (36.94, 57.24) |
*
p < .05, as derived from Bonferroni-corrected exact Fisher’s tests; 95% CI in brackets.
**
p < .01.
***
p < .001.