Poster E90, Monday, March 26, 2:30-4:30 pm, Exhibit Hall C
Statistical learning of categorical regularities in adults and children
Yaelan Jung1, Dirk B. Walther1, Amy S. Finn1; 1University of Toronto
Humans are able to learn statistical regularities that are present in the environment with very little exposure (Saffran, Aslin, & Newport, 1996). This learning can even occur rather abstractly; adults have been shown to learn statistical regularities that occur across categories and not specific exemplars (Brady & Oliva, 2008). Do younger children learn these abstract regularities like adults? Given evidence that children are more sensitive to features of individual items than adults (Sloutsky & Fisher, 2004) they may or may not show rapid learning of statistical structure at the category level. To address this, we performed statistical learning experiments on adults (18-22 y old) and children (6-9 y old). Observers were exposed to a 5-minute stream of animal images which had a statistical structure – 4 triplets, in which 3 animals appeared in the same order, were randomly distributed in the stream. Critically, the images were always different exemplars in the stream (i.e., different monkey picture each time). Observers were tested using two alternative forced-choice and reaction time based target detection tasks. We observed that both adults and children were able to learn the statistical regularities when the patterns were at the category level. However, is the exposure to multiple exemplars necessary for category level learning or is an exposure to a single item (i.e., same monkey picture each time) sufficient to extract the category level knowledge? We found that both adults and children could still learn the statistical regularities at the category level even with single item exposure.
Topic Area: PERCEPTION & ACTION: Development & aging