Change blindness, subset segmentation, and the perceptual underestimation of subset numerosity

Area of Science:

• Visual Perception
• Cognitive Psychology
• Computational Neuroscience

Background:

• Humans can simultaneously assess numerosity of multiple color-defined subsets.
• Perceptual underestimation occurs when comparing similar colors, like light gray and white dots on a gray background.

Purpose of the Study:

• To test the hypothesis that figure-ground segmentation based on color similarity causes perceptual underestimation of foregrounded dots.
• To investigate the role of segmentation in accurate subset estimation.

Main Methods:

• Six experiments were conducted.
• A color-change detection task was used to assess segmentation accuracy.
• Attentional biasing was manipulated to test its effect on performance.

Main Results:

• Accurate subset estimation is achieved for sets that resist foreground selection.
• Foreground selection is incomplete when colors are highly similar (e.g., light gray among white).
• Attentional biasing improved performance for chromatic colors but not for similar gray shades.

Conclusions:

• Figure-ground segmentation based on color similarity can lead to incomplete segmentation and perceptual underestimation.
• Perceptual underestimation of mixed-color supersets may result from combining an underestimated foreground with an accurately estimated background.
• Segmentation difficulty, not accuracy, correlates with accurate subset estimation.