Poster #62 - Are color categories helpful? Exploring acuity and target detection in infants and adults

• In Event: Poster Session 07

Abstract

Previous research has demonstrated that infants as young as 4 months responded categorically to color (Franklin et al., 2005; Ozturk et al., 2013; Skeleton et al., 2022). Specifically, infants were quicker to detect a target embedded in a background color that varied across a color boundary (e.g., blue target on a green background, between-category) than an equally spaced non-category crossing color pair (e.g., blue target on bluish-green background, within-category). The appearance of categorical perception so early in infancy suggests it may afford some sort of processing benefit. One possible advantage of categorical color perception might be increasingly sensitive color-based acuity, which may be critical for visual exploration of objects. If this is the case, then acuity gratings created using between-category pairs, should be more discriminable than equaluminant gratings made of between-category pairs. Thus, we predict that infants will both orient more quickly to between-category gratings and will demonstrate higher acuity for between-category gratings.

Data collection is ongoing, but the preliminary sample includes 8 7-month-olds, 14 11-month-olds, and 40 adults (for comparison). Infants were tested on an eye-tracker in an acuity procedure modeled on the Teller Acuity Cards (Dobsen & Teller, 1978), only rather than square-wave gratings, infants were shown Gabor patches presented against a within- or between-category background. Color pairs were created using 2 shades of blue, and 2 shades of green that were all equaluminant, and equally spaced in HSL color space (saturation = 81, lightness = 44.5 for all colors, Figure 1). These colors were then used to create 2 sets of within- and between-category pairs (blue and green) and participants each saw both a within- and between-category set of gratings varied in difficulty from 256 to 2 pixels per cycle (PPC; lower PPC indicates greater difficulty). Note, color set and order of category presentation was counterbalanced.

Primary measures included minimum PPC, and reaction time (RT) for target detection. Interestingly, adults seem to show a RT benefit for within-category color contrasts, but only for the green set, F(1,38)=5.401, p=.026, corrected follow-up (p=.065). A similar interaction was observed for the PPC measure, F(1,38)=4.119, p=.049, though follow-ups were not significant, likely due to a ceiling effect. Preliminary infant data reveal a significant main effect of category on min PPC, F(1,11)=6.261, p=.029, with significantly higher acuity ratings for within-category pairs (follow-up p=.047; Figure 2). Interestingly, this category effect was not observed in the RT data, though power was limited in this preliminary sample.

Results are counterintuitive and suggest that the effects of categorical color perception may be highly influenced by stimulus choice. Though we did not replicate basic advantages for categorical perception, we did demonstrate significant color category effects. It is important to note that our stimuli were fully equaluminant, which substantially constrained our color pair choice. Additional planned analyses for a complete infant sample include a close examination of “searching” prior to target detection, dwell time on target once found, and age effects, particularly as they pertain to detection of equaluminant targets.

Authors

• Micalee Anne Segers, University of Tennessee - Knoxville
  Presenting Author

• Alicia Hemmings

• Victoria Jones, University of Tennessee - Knoxville

• Esther Reynolds, University of Tennessee - Knoxville

• Shannon Ross-Sheehy, University of Tennessee - Knoxville