Abstract
Motion pop-out occurs when there is a single moving target among multiple static distractors. Motion pop-out occurs in human infants, but the extent of pop-out is modulated by the characteristics and spatial distribution of the static elements in the visual field (Dannemiller, 2000). When equal numbers of low and high contrast static elements are distributed unevenly across the visual field, infants are less likely to orient toward the moving target when it appears contralaterally to the hemifield with more of the high contrast static elements. The static elements compete with the moving target to determine the likelihood of motion pop-out. A signal detection model with a maximum-response decision rule captures these results with infants reasonably well. This model predicts a set-size effect as the number of static elements in the visual field increases unlike the case with adults in which no set size effects are predicted when the moving singleton is well above detection threshold. In two separate samples (n = 97 and n = 98) with infants from 2 to 5 months using Teller's (1979) Forced-Choice Preferential Looking procedure, the predicted set size effect did not occur with a range from 1 to 14 distractors. Instead, infants oriented toward the moving target at levels above chance but below 100%, and set size had no effect. We now show with two additional experiments (N = 64) that both the competition effect with heterogeneous distractors and the lack of a set size effect with homogeneous distractors are explained by a high-threshold-like model in which motion pop-out occurs on some percentage of trials. On the complementary percentage of trials pop-out fails to occur but orienting is not completely random; instead, it is determined by the properties of the static elements. This is analogous to a less-than-ideal observer whose orienting decision is based on the target on some percentage of the trials and on the distractors on the remaining trials.