Ensemble perception in depth: Correct size-distance rescaling of multiple objects before averaging.

Area of Science:

• Visual perception
• Cognitive psychology

Background:

• Humans excel at estimating ensemble summary statistics, like average object size.
• The basis of these estimates (retinal vs. contextual size) remains debated.

Purpose of the Study:

• To investigate whether ensemble size estimates rely on raw retinal input or context-aware (distal) size representations.
• To determine if the visual system rescales individual object sizes based on perceived distance before averaging.

Main Methods:

• Experiments used a stereoscope to present ensembles of objects at varying apparent distances.
• Participants adjusted the mean size of object ensembles.
• Experiment 1 manipulated depth of all items; Experiment 2 varied correlations between angular size and apparent distance.

Main Results:

• Overestimation of mean size occurred when ensembles were presented in a remote plane, suggesting distance rescaling.
• Precision of size averaging varied with apparent size range, not angular size, indicating pre-averaging rescaling.
• Results support the hypothesis that individual sizes are rescaled to perceived distance before averaging.

Conclusions:

• Ensemble summaries of basic features like size are based on elaborated, context-informed representations.
• The visual system actively rescales object sizes based on perceived distance prior to ensemble averaging.
• Findings have significant implications for understanding size constancy and visual perception of object properties.