Research Project

 

Modeling Project:


I will be working to build a computational model of a neural net. This will be based on word learning in young children. The focus will be on the relation between nouns and verbs in a young child’s vocabulary and how that appears when mapped in a graph where words are nodes and edges are their relations. We are using data from adult free - response of the first noun that comes to mind when they see a given Bates - MacArthur Communicative Developmental Inventory (MCDI) verb. We filtered out the nouns to include only MCDI nouns and then compressed them into counts of the number of times adults paired the noun and verb. We will be comparing the graphs produced by different children in different stages of development.



Behavioral Project:


We will be extending previous research that has been conducted in this lab and others. We wish to see what defines a name for an object in a child’s mind. We will use line drawings or pictures to train each child in the names of new objects, which are word - like sounds (made up words), related sounds (animal sounds for animal objects, vehicle sounds for vehicle objects), and unrelated sounds (vehicle sounds for animal objects, animal sounds for vehicle objects). We will then see which ones they treat as valid words to name objects based on the choices they make in regards to their selection of images and novel objects related to the image.