Triadic comparisons reveal representational motifs in human color perception.
eScholarship (California Digital Library) · 18.06.2025
DOI: 10.48448/4czb-bq31 ↗Kurz gefasst
Fachartikel in der Zeitschrift eScholarship (California Digital Library), erschienen am 18.06.2025. Verfasst von Cognitive Science Society 2025 und 2 weiteren Autor:innen.
Abstract (Originalsprache)
While individual differences have been found in many aspects of color perception, it is unclear whether mental representations of color vary between individuals in a structured way. To test for structured variability, we collected perceptual similarity judgments for 58 colors in a triadic-comparisons procedure, and from each participant's judgments, embedded the colors into a personalized three-dimensional space. The personalized embedding predicted a participant's similarity judgments on held-out items significantly better than did (a) embeddings from other participants, suggesting reliable individual differences in perceived color similarity (p < 0.0001), and (b) color coordinates in a standardized perceptual color space (CIELAB; p < 0.0005). Across individuals, embedding structure did not vary randomly but fell into two clusters, one encompassing more distinct color categories and the other a more continuous perceptual space. The results suggest the existence of previously unrecognized “motifs” in how people represent colors.
Schlagwörter
Zitieren
APA
2025, C. S. S., Rogers, T. T., & Zimnicki, C. (2025). Triadic comparisons reveal representational motifs in human color perception. eScholarship (California Digital Library). https://doi.org/10.48448/4czb-bq31
Vancouver
2025 CSS, Rogers TT, Zimnicki C. Triadic comparisons reveal representational motifs in human color perception. eScholarship (California Digital Library). 2025. doi:10.48448/4czb-bq31
BibTeX
@article{2025,
author = {2025, Cognitive Science Society and Rogers, Timothy Thomas and Zimnicki, Clementine},
title = {Triadic comparisons reveal representational motifs in human color perception},
journal = {eScholarship (California Digital Library)},
year = {2025},
doi = {10.48448/4czb-bq31}
}
EUVIMED