Abstract
The visual system has higher sensitivity to expansion than to contraction (e.g. Ptito et al., 2002; Takeuchi, 1997). Early development of the anisotropy for expansion/contraction detection have investigated by Shirai et al. (2004a). They tested 2–8 month olds' sensitivity to expansion/contraction by using visual search displays and demonstrated that the anisotropy for expansion/contraction detection emerges at about 3 months of age.
In the present study, we re-examined development of the anisotropy by using dynamic random dot patterns (RDPs). RDP is a popular stimulus to measure young infants' motion sensitivity (e.g., Banton & Bertethal, 1996; Shirai et al., 2004b; Wattam-Bell, 1996).
A total of 115 2–4 month-old infants participated. We presented two RDPs to the infants side by side. One was a target composed of radially moved signal dots and randomly moved noise dots. The other was a distractor composed of noise dots only. There were 4 experimental blocks: 2 target motions (expansion/contraction) X 2 signal strengths (a ratio of the number of signal dots to that of signal + noise dots in a target; 1 or 0.5). Each experimental block consisted of 4 presentations of a pair of RDPs. Each infant participated in only one experimental block. We calculated the infant's preference score (PS) for target based on a videotaped infant's looking behavior. The PS was defined as the ratio of total looking time for target to that for target + distractor in an experimental block.
The results indicated that the infants aged over 3 months showed significant preference for the expansion target only when the target had strong (=1) signal strength. None of the infants showed significant preference for the contraction target even the target has strong signal strength. These results suggest that the anisotropy for expansion/contraction detection emerges at about 3 months of age independent of stimulus types.