Bayesian average or truncation at boundaries? The mechanisms underlying categorical bias in spatial memory

Area of Science:

• Cognitive Psychology
• Neuroscience
• Spatial Cognition

Background:

• Spatial memory is subject to biases influenced by implicit or explicit category boundaries.
• The dot-localization task reveals systematic biases toward category centers.
• Two models, weighted-average and truncation, explain these biases but are empirically indistinguishable.

Purpose of the Study:

• To differentiate between the weighted-average and truncation models of spatial memory bias.
• To investigate the impact of explicit category boundaries on spatial localization errors.
• To provide experimental evidence for distinct mechanisms underlying categorical biases in spatial memory.

Main Methods:

• Utilized a novel distribution analysis approach in a dot-localization task.
• Compared error distributions under standard conditions versus when an explicit V-shaped category boundary was presented.
• Analyzed spatial recall errors in three controlled experiments.

Main Results:

• Symmetrical error distributions shifted toward category centers when default categories were used, supporting the weighted-average model.
• Highly skewed error distributions were observed when an alternative V-shaped category was used, consistent with the truncation model.
• The study successfully differentiated the two proposed mechanisms based on experimental conditions.

Conclusions:

• The findings provide the first experimental evidence distinguishing between weighted-average and truncation mechanisms in spatial memory.
• The results suggest that different cognitive processes are engaged depending on whether default or explicitly presented category boundaries are salient.
• This research advances our understanding of how spatial information is encoded and recalled under varying contextual influences.