Figure2

Results from the experiment. An average fixation map was created for all the iHybrids. Maps were created by collapsing z-scored fixation durations across all correct trials of the experiment. This fixation map was then split into separate maps (b) showing fixation durations when responses were based on local face identifications and fixation durations when responses were based on global face identifications. Subtracting the fixation map for global identification responses from the fixation map for local identification responses resulted in a difference map. On the local and global maps shown here, white contours surround regions of significantly longer fixation durations than were observed for other areas; on the difference map, white contours indicated regions of significant differences between the local and global maps (i.e., fixation bias). The graph (c) shows the conditional probability of identification responses (local face or global face) as a function of the location of the first fixation (in the eye-mouth region or in the center of the face). Error bars show standard errors of the mean.

Done