Which color was used for the larger circles was counterbalanced across trials in the actual experiment. In this example, the blue circles were larger than the red circles on average, and the reported sizes of the tested circle are indicative of bias. We measured whether observers reported different sizes for the tested circle when it was red versus when it was blue (the circle was in fact the same size in both presentations).
The illustration in (a) shows how bias was calculated from a pair of matched displays.
Bias in size estimates in the three experiments.
0 kommentar(er)
