Shape constancy and theory of mind: is there a link?

In three experiments, children aged between 4 and 7 years viewed a circular disc oriented at a slant. The disc was made of luminous material and situated in a darkened chamber. Children of all ages exaggerated the circularity of the disc when they knew that the object was really a circle (the circle task), and the effect was greatest in the younger members of the sample. Crucially, however, a group of children in Experiment 3 who viewed an identical shape that they knew emanated from an actual ellipse did not exaggerate circularity. In the second experiment, children tackled three standard theory of mind tasks in addition to the circle task mentioned above. A significant correlation emerged (even with age partialed) between the extent of exaggeration made by those who knew that the shape was a circle and ability to pass the theory of mind tests. It seems knowledge of reality contaminates judgements of appearance in the circle task. This might be the same bias that features in realist errors in theory of mind tasks.