Abstract
The cognitive processes that mediate visual perception are dissociable from those of visually-guided action. For example, pictorial illusions bias perceptual judgments, but not the calibration of manual grasps. We wondered whether pictorial illusions elicit perceptual biases similar to those of real-world three-dimensional stimuli (that afford, but do not require, manual interaction). To address this question, we conducted a psychophysical experiment to measure the effect of display format on the tilt illusion. In the tilt illusion, the perceived orientation of a central test grating is altered by the presence of a simultaneously presented surround grating with a different orientation. On each trial, observers adjusted the orientation of a central test grating so that it appeared to be vertical in orientation. The surround gratings were oriented 12○ to the left or right of vertical. Critically, in 'real' blocks, participants performed their perceptual judgments on a tilt illusion stimulus comprised of solid white wooden rectangular prisms mounted against a black background; the orientation of the test grating was varied around the vertical axis using a computer-controlled servomotor. In 'image' blocks, participants performed their perceptual judgments on high-resolution two-dimensional computerized images of the real-world display. The images were matched closely to their real-world counterparts for pictorial depth cues, illumination, size and distance. We found that although the tilt illusion was present in both stimulus formats, the magnitude of the illusion was markedly reduced for the real-world displays.
Meeting abstract presented at VSS 2018