Jessica Horst

This series of studies is the preliminary work for my disseration. I have been working on this project with Larissa Samuelson in the Language and Category Development Lab.

Children were presented with several forced choice trials with two familiar objects and one novel object in which they determined the referent for either a familiar or novel word. After a five-minute delay children were asked follow-up questions to determine what they retained from the fast mapping trials and whether they would extend the newly learned words to novel instances of the categories. Overall, children were highly accurate on the fast mapping trials, and showed evidence of being able to retain one but not eight words.

Connectionist Model of Fast Mapping and Word Learning: (2005 - present):

This research has been a nice compliment to my dissertation research. I have been working on this project with Bob McMurray in the Mechanisms of Audiovisual Cognition Lab.

This model contains two layers of input units (an auditory "word" layer and a visual "object" layer). These layers are not connected to each other, but each connected to a third, hidden layer. Through unsupervised Hebbian learning the model learns the associations between words and objects. Simulations were conducted by presenting networks with same task presented to children in the fast mapping studies discussed above. Overall, the same pattern of results were found. Further analyses of weight changes provided new insights into how children solve fast mapping.

This work is currently being extended to investigate how and when a novel name becomes a known name.

Form & Function Categorization in Infants (2002 - 2004):

This series of studies examined the types of categories 10-month-old infants formed during a habituation task as a function of the organization of the category to which they were familiarized. Specifically, infants formed an exemplar-based category when familiarized to objects that shared a common appearance, but formed a summary representation when familiarized to objects that shared a common function.

This research was in collaboration with Lisa Oakes and Kelly Madole in the Infant Cognition Lab.

Categorization of Cats and Dogs in Infants With and Without House Pets (2003 - present):

This research tests the role of long-term exposure to a category outside the laboratory on infants' categorization in the laboratory. Specifically, in a paired comparison procedure infants were familiarized with pictures of dogs and cats. We investigated the types of categories infants formed as a function of what they were shown during the laboratory task (i.e., dog or cat condition) and what they were exposed to before coming into the laboratory (e.g., whether they had a dog or cat at home)

This research was in collaboration with Kristine Kovack-Lesh and Lisa Oakes in the Infant Cognition Lab.

Children's Behavior Towards Solid Objects and Nonsolid Substances (2003 - present):

This research specifically explores children's behavior toward solids and nonsolids when they are allowed to freely explore novel stimuli. A new coding scheme and procedure were devised to examine the specific actions children perform on solid objects and nonsolid substances (e.g., grabbing at, throwing, bringing to their mouths) as a function of the children's vocabulary level and age.

This research is in collaboration with Larissa Samuelson in the Language and Category Development Lab.

Children's Novel Noun Generalization of Solid Objects and Nonsolid Substances (2002 - present):

This work investigated how children extend names for solids (e.g., wood shapes) and nonsolids (e.g., lotion). Studies were conducted with both 16-18-month-old and 24-month-old children. For the 24-month-old children novel noun generalizations were related not only to the type of stimulus they were extending the name for (e.g., solid or nonsolid) and the training they had received, but also what stimuli they had seen on previous trials. Thus, these children's novel noun generalizations were a function not only of what they had brought to the task (an understanding of how solid and nonsolid things are named), but also what they had learned during the task and had seen up to any given moment. The studies with 16-18-month-old children also incorporate vocabulary data and a no-name version of the task

This research is in collaboration with Larissa Samuelson in the Language and Category Development Lab.

Task And Stimulus Factors Affect The Shape And Material Biases (2001 - 2004):

This series of studies explored how children extend names for rigid and deformable objects as a function of vocabulary. Specifically, two curvilinear patterns of overextension were found. In a yes/no task children overgeneralize the shape bias to the naming of deformable objects at age 3 but not at ages 2 and 4. In a forced-choice task children overgeneralize the shape bias between 19- and 21-months-of age depending on the number of count nouns in their productive vocabularies

This research is in collaboration with Larissa Samuelson, Anne Schutte and Brandi Dobbertin in the Language and Category Development Lab.

Form & Function Categorization in Toddlerhood (2001 - 2004):

This series of studies investigated toddlers' abilities to flexibly shift categorizing objects based on different dimensions: what the objects looked like and what they could do. Toddlers were presented with a sequential touching task and a demonstration that highlighted the features of some of the objects. Children's abilities to flexibly categorize the same objects into different categories varied as a function of the children's vocabularies and the particular demonstration.

This research is in collaboration with Larissa Samuelson and Lisa Oakes in the Language and Category Development Lab and the Infant Cognition Lab.