Evaluating Müller-Lyer effects using single fin-set configurations.

Four experiments were conducted to study the bias of perceived length for Müller-Lyer configurations that contained a single set of fins (i.e., two segments that join to form a vertex). The experiments manipulated several factors that have been shown to be critical to the effect: (1) version (which way the apex pointed), (2) length of the stimulus span, (3) presence or absence of a line segment in the span being judged, (4) fin length, (5) fin angle, and (6) the zone in which the response was rendered. Using percent error as the index of perceptual distortion, the major finding was that the two versions show an opposite slope for strength of effect as a function of span. When stimulus spans were plotted against response means (not converting to percent error), an almost perfect linear relation was found. These results indicate that the perceptual effects can be modeled as a linear system having two parameters through which the treatments exert their influence. The results are discussed in relation to major theories of mechanism for the Müller-Lyer illusion.