Quantifying Kitaoka's enhanced checkered illusion
16.513, Friday, May 10, 5:30 - 8:00 pm, Vista Ballroom
Dejan Todorovic1, Marija Milisavljevic1; 1Laboratory of Experimental Psychology, Department of Psychology, University of Belgrade, Serbia
The enhanced checkered illusion (Kitaoka, 2007) involves a pattern of checks that contain pairs of small squares (inductors) centered at one diagonal of the checks, near their opposite corners. The presence of inductors generates an apparent tilt in the orientation of the lines formed by the edges of the checks. The illusion is compelling (see http://www.michaelbach.de/ot/geom_KitaokaBulge/index.html), but little is known about its cause or relation to other tilt illusions. We studied how the illusion is affected by the position of the inductors, luminance polarity of checks and inductors, and distance of observation. The stimuli were circular insets of 18x18 enhanced checkerboards (sides of checks: 9.4 mm; sides of inductors: 1.9 mm), at five levels of rotation (-2, -1, 0, 1 and 2 degrees from vertical), each in two symmetrical mirror versions, observed from two distances (far: 100cm, near: 50cm). The task was to judge whether the orientation of the near-vertical edges of the checkerboard was tilted clockwise or counterclockwise. We obtained psychometric functions of percentage clockwise tilt judgments against degree of rotation. In Experiment 1 the check/inductor pattern was lightgray/black vs. darkgray/white. We varied the gap between the sides of the inductors and the sides of the checks, which was either zero, intermediate (0.9 mm), or large (1.8 mm). For the far observation distance the strongest illusion was found for the intermediate gap, it was weaker for the zero gap, and absent for the large gap. For the near observation distance the illusion was only present for the intermediate gap. In Experiment 2 we found illusory effects of similar strength for checkerboards with the black/white vs. white/black check/inductor pattern and checkerboards with the darkgray/lightgray vs. lightgray/darkgray pattern, but no effect for checkerboards with the darkgray/black vs. lightgray/white pattern. The results provide constraints for physiologically based theories of the illusion.