Fig 4 -

Panel A compares shifts in True Positive Rates (TPR) across Study 1 and 2 conditions. Panel B compares shifts in False Positive Rate (FPR) across Study 1 and 2 conditions. Panel C compares Criterion shifts across Study 1 and 2 conditions. The Mask and No Mask bars are the experimental conditions from Study 1. The Mask+65 and Mask+95 correspond to the 65- and 95-algorithm conditions in Study 2, respectively. Significant (p p < 0.05) shifts across conditions and studies are indicated by horizontal lines and asterisks (*).