Abstract
Distances between points within the surface of an object are illusorily expanded compared to equivalent distances outside of an object, an effect we have termed “object-based warping” (Vickery & Chun, 2010, Psychological Science). Further insights into the causes of this distortion would benefit from examining variation of object-based warping over the entire object's surface, and across different enclosing shapes. To measure warping, we asked participants to adjust the spacing between two adjustment dots to match the spacing between two fixed reference dots. We compared judgments when the reference dots were placed in and around objects to when they were presented in the absence of objects. Prior studies used points symmetrically distributed around the object mid-line. Here, we examined variation over a vertical rectangular object's surface by considering seven equally-spaced points along its vertical midline from the top edge to the bottom edge. On each trial, dots were presented at two of these points. Every possible position pair was probed. Warping effects were observed for all interior segments tested, such that estimated spacing was 5–12% greater with than without an object. Segments with at least one point at the object's boundary showed no significant warping, and significantly weaker warping than non-boundary segments for every spacing, implying that object-based warping depends on surface enclosure. For distorted interior segments, there was no systematic variation due to segment size, other than a tendency to report the smallest distortions for small segments relative to baseline (5%, relative to 9–13% distortions for larger segments). Across multiple additional experiments, we observed object-based warping on variable shapes (rectangles, squares, circles, triangles), as well as on the surface of three-dimensional objects. Object-based warping is uniform across the interior of an object and extends to different shapes and volumes, suggesting that warping is a generalizable consequence of object-hood.
This research was supported by research grants NIH R01-EY014193 and P30-EY000785 to M.C.