Ontogeny of early event memory: III. Attentional determinants of retrieval at 2 and 3 months

Abstract

Two-month-olds forget more rapidly than identically trained 3-month-olds, but the reason for this is not understood. In two studies, we tested the hypothesis that this retention difference reflects an ontogenetic difference in the pattern of visual attention that selectively reduces the number of potential retrieval cues likely to be sampled by the younger infants in visual learning tasks. In Experiment 1, we sought to alleviate the retention deficit of 2-month-olds by altering the spatial distribution of retrieval cues in the stimulus array and by providing serial access to environmental stimuli known to cue retrieval at 3 months. When the distribution of visual information in the stimulus array maximized the probability that potential retrieval cues would be detected without repeated gaze-shifting, 2-month-olds exhibited perfect 24-hour retention; otherwise, they exhibited no retention, whether their environmental context was highly distinctive or not. In Experiment 2, identically trained 3-month-olds exhibited excellent 24-hour retention irrespective of either the distribution of stimulus information or the distinctiveness of their training context. These data suggest that 2-month-olds' poorer retention results from a leaner memory representation of the training event. This stems not from slower encoding of potential retrieval cues that they have detected but from their failure to detect as many potential retrieval cues in a spatially distributed visual array as older infants within a limited period of time.