A quantitative confirmation of visual capture of curvature.

An investigation was performed to quantify the experience of contour curvature formed under sensory discrepancy inspection conditions. N = 80 male and female undergraduates were used. Experimental Ss finger-tracked a horizontal straight edge while viewing limb movements through a curve inducing lens. Control Ss inspected the edge unimodally, either through vision (distorted) or proprioception. Rather than relying on verbal reports all Ss were required to match their impressions of contour shape with an adjustable metal curve. Findings indicated that (a) the mean impression of contour shape derived from discrepant visual and proprioceptive information was curved and did not differ from that derived from distorted visual information, (b) proprioceptive inspection alone resulted in accurate judgements of physical straightness, (c) response modality effects did not emerge, and (d) a manipulation designed to direct Ss attention to the left contour shape and during sensory discrepancy inspection did not affect mean curvature matches but created sizeable variability among matches. A sensory organization versus a selective processing interpretation of visual capture was discussed.