Abstract
Purpose: Previous studies indicate that infants use color as a motion correspondence cue surprisingly well (Dobkins & Anderson, 2002). Here, we investigated whether color influences motion integration, thought to occur at a later stage in processing. We compared coherent motion thresholds for a “homochromatic” (signal and noise dots both red or both green) versus a “heterochromatic” (signal and noise dots different colors) stimulus. If color acts as a segmentation cue for motion, thresholds should be lower for the “hetero” condition.
Methods: For 3-month-olds (n=16), data were obtained using a Directional (left/right) Eye Movement (DEM) technique. (Stimulus parameters: 43×34 deg display; 603 0.42 deg dots; 25 deg/sec; cone contrast between “red” and “green” = 27.7%; unlimited duration). Adults (n=12) were tested with the same stimuli (yet limited duration), using both psychophysical and DEM techniques. For each subject, a “color benefit ratio” was calculated: THR-homo/THR-hetero. Ratios > 1.0 indicate that color acts as a segmentation cue for motion.
Results: In line with previous studies, the mean adult ratio obtained psychophysically was 2.1. However, the adult ratio decreased to 1.3 with DEM, suggesting that adults use a higher-level strategy when tested psychophysically. The mean infant ratio was 1.0, that is, infants did not use color as a segmentation cue for motion. This null result is unlikely to be accounted for by poor red/green contrast sensitivity.
Conclusions: Possibly, infants use color as a correspondence, but not a segmentation, cue for motion. Alternatively, the use of color as a segmentation cue may require higher-level strategies, which infants lack.
NIH/NEI R01-EY12153-06 (KRD)