Vertical asymmetry in central color evaluation: How spatial position and scene structure modulate the perceived size of colored regions

Humans exhibit vertical asymmetry in color perception, with the lower visual field generally showing superior performance. We investigated whether this asymmetry extends to central vision by having participants judge the relative size of colored and grayscale regions within small, centrally presented stimuli (1.5° × 2.0°). Natural-scene images were partially desaturated so that either the upper or lower portion remained in color, and participants reported which region (colored or grayscale) appeared to be larger. To examine the role of scene structure, we also used pixel-shuffled versions of the same images, which preserved the vertical chromaticity distribution but removed scene recognizability. Two presentation durations were tested: 50 ms and 250 ms. With a brief presentation duration (50 ms), the colored region was overestimated in size relative to the grayscale region more strongly in the lower colored than in the upper colored condition for natural-scene images, t(19) = 5.34, p < 0.001, d = 1.20 in Experiment 1; \(F( {1,14} )\; = \;23.54,\;p\; < \;0.001,\;\hat \eta _p^2\; = \;0.627\) in Experiment 2-1, but this asymmetry was not observed for pixel-shuffled images, \(F( {1,14} ) = \;0.28,\; p\; = \;0.602, \hat \eta _p^2\; = \;0.020\). With a longer presentation duration (250 ms), the lower-half effect was significant across both natural-scene and pixel-shuffled conditions, \(F( {1,15} )\; = \;9.85,\;p\; = \;0.007,\;\hat \eta _p^2\; = \;0.396\), with no significant interaction between stimulus type and color position, F(1, 15) = 1.14, p = 0.303. These findings suggest that the lower visual space biases color evaluation even in central vision and that this bias emerges more readily when a natural-scene structure is available.