Interestingly, however, we found that participants consistently underestimated their intake of the candies on their desks yet overestimated how much they ate when the candies were farther away.