Proceedings of the Annual Meeting of the Cognitive Science Society

Young children tend to generalize novel names by shape; when asked to match a novel object to one of two objectsthey often choose the one that matches in shape. This shape bias has been shown in laboratory tasks to be connected tovocabulary learning: children who know less than 50 words do not show this bias and training using object categorieswell-organized by shape improves children’s word-learning. An open question is whether experience with real (3D)objects is necessary or children can transfer from practice matching 2D objects. In this project, we used a week-long athome intervention with an iPad game. Compared to a version of the game that asks children to establish identity matches,children who played with 2D shape matches for a week have a more robust shape bias with real-world objects at posttestas well as a modest effect in vocabulary growth 2 months later.