Distracting effects in length matching.

In psychophysical experiments, subjects matched two spatial intervals of a three-spot stimulus into what appeared to them to be equal. The stimulus was flanked by stripes. The length matching errors increased in proportion to the referent interval of the stimulus and approached 6-12 percent of its length. Also, the error increased with an increase of the width of the gaps between the spots and the distracting stripes. Error reached a maximum at gaps equal to 10-15 percent of the length of the referent interval of the stimulus. When the luminance of the stripes increased or decreased, in comparison to the luminance of the background, length matching errors grew symmetrically and became approximately constant at higher contrasts. The experimental findings show the presence of local positional averaging which may be described quantitatively by means of spatial filtering procedures.