Table 2. Results of the analysis of variance from Study 2.
| Source | df | SS | η2 | F | p |
|---|---|---|---|---|---|
| Identity (I) | 44 | 58.08 | 0.43 | ||
| Observer Sex | 1 | 0.37 | 0.00 | 2.79 | .102 |
| Observer Sex × I | 44 | 5.85 | 0.04 | ||
| Cosmetics | 1 | 33.35 | 0.25 | 47.89 | < .001 |
| Cosmetics × I | 44 | 30.64 | 0.23 | ||
| Observer Sex × Cosmetics | 1 | 1.73 | 0.01 | 20.66 | < .001 |
| Observer Sex × Cosmetics × I | 44 | 3.68 | 0.03 | ||
| Total | 179 | 133.70 |
There is no error term for conducting an F test on differences between models, and as such, no ratio is calculated. df = degrees of freedom, SS = sums of squares. Interactions with the Identity measure can be interpreted as an error term for that variable [35].