L. Armstrong et Le. Marks, DIFFERENTIAL-EFFECTS OF STIMULUS CONTEXT ON PERCEIVED LENGTH - IMPLICATIONS FOR THE HORIZONTAL-VERTICAL ILLUSION, Perception & psychophysics, 59(8), 1997, pp. 1200-1213
Six experiments examined orientation-specific effects of stimulus cont
ext on the visual perception of horizontal and vertical lengths: Using
a paired-comparison method, Experiments 1-5 showed that the probabili
ty of judging a given vertical Line to be longer than a given horizont
al line was relatively great when the stimulus set comprised relativel
y long horizontals and short verticals, and relatively small when the
stimulus set comprised short horizontals with long verticals. To the e
xtent that stimulus context exerts orientation-specific effects on per
ceived length, it thereby modulates the degree to which verticals appe
ar longer than physically equivalent horizontals: the horizontal-verti
cal illusion (HVI). Under various contextual conditions, the HVI was a
s small as 3% (horizontals had to be 3% greater than verticals to be p
erceived as equally long) and as great as 15%, equaling about: 12% in
a ''neutral'' context. In Experiment 6, subjects judged the absolute p
hysical length of each stimulus, and the results indicated that stimul
us context acted largely by decreasing perceived lengths. The results
are consistent with the hypothesis that differential effects of contex
t reflect a process of stimulus-specific perceptual attenuation.