Table 5.
Accuracy by treatment controlling for the round.
| Label | 0.35*** |
| (6.30) | |
| Photo | 0.33*** |
| (5.96) | |
| Video | 0.23*** |
| (4.11) | |
| Second round | 0.65*** |
| (11.80) | |
| Label × second round | −0.33*** |
| (−4.13) | |
| Photo × second round | −0.19* |
| (−2.44) | |
| Video × second round | −0.16* |
| (−2.01) | |
| Constant | −0.11** |
| (−2.95) | |
| N | 844 |
| Raters | 422 |
| Log-likelihood | −434.94 |
| Akaike information criterion | 886 |
| Bayesian information criterion | 924 |
| Chi-squared (7 degrees of freedom) | 359.40*** |
Z-Value in parentheses. *** p < 0.001, ** p < 0.01, * p < 0.05. All general least squares regressions use measures constructed over the round where the rater is the panel and rounds are the trials. All results are robust using ordinary least squares regression.