Abstract
Recent studies reported that one can accurately extract the mean size of similar objects. However, few studies tested that calculation of the mean size was based on either perceived size or physical size. To investigate this, we used Ebbinghaus illusion. We either simultaneously (Experiment 1) or sequentially (Experiment 2) presented four Ebbinghaus illusion figures per each visual field. Participants' task was to decide which visual field had the larger mean size. They had to calculate the mean size based only on center circles and ignore illusion-inducers. The center circles and the inducers had a different color to help participants to focus on the center circles. In both experiments, participants consistently chose the visual field that contained inducers smaller than center circles, although the sizes of center circles were always same across two visual fields. In Experiment 3, we investigated how much calculation of the mean size was influenced by Ebbinghaus illusion. We measured PSE using the staircase method. We found that the average difference between PSE and the actual mean size was about 13 min. These results suggest that participants' calculation of the mean size was based on perceived size. Moreover, they imply that judgment of mean size is achieved in a later stage of visual processing, at least after resolving size constancy.
This research is supported by Korea Research Foundation Grant # H00039.