Numerosity and density judgments: biases for area but not for volume

Area of Science:

• Visual Perception
• Psychophysics
• Computational Neuroscience

Background:

• Human observers possess a remarkable ability for rapid numerical estimation in visual scenes.
• This numerosity perception is thought to rely on specialized neural mechanisms encoding quantity or spatial density.
• Previous research indicates that perceived density, but not necessarily number, can be influenced by changes in stimulus area.

Purpose of the Study:

• To investigate whether judgments of visual number and density are susceptible to biases introduced by changes in volume.
• To compare the effects of area versus volume on the precision and bias of numerosity and density perception.
• To determine if biases observed in two-dimensional (2-D) perception extend to three-dimensional (3-D) representations.

Main Methods:

• Stimuli consisted of non-overlapping black and white luminance-defined dots presented in 2-D and 3-D using a stereoscope.
• A temporal two-alternative forced-choice (2AFC) task combined with the Method of Constant Stimuli (MOCS) was employed.
• Precision and bias (point of subjective equality shift) of numerosity and density judgments were measured for varying areas and volumes.

Main Results:

• For 2-D stimuli, perceived density was significantly biased by increasing area, aligning with prior findings; perceived number remained unaffected.
• In 3-D stimuli, despite a strong subjective impression of volume, no bias in perceived density or number was observed with increasing volume.
• A control experiment confirmed observers' ability to accurately perceive stereoscopic depth (disparity) in the 3-D stimuli.

Conclusions:

• Number and density judgments, which are demonstrably biased by area manipulations in 2-D, are not similarly affected by volume changes in 3-D.
• The mechanisms underlying numerosity and density perception appear robust to volumetric scaling, suggesting distinct processing pathways from area-based biases.
• These findings challenge the generalization of area-based perceptual biases to volumetric contexts in visual number and density estimation.