Active information-seeking during word learning and categorization: comparing infancy and adulthood

Active Grant Awarded

Award ID
F31HD122364
Recipient
Awarding agency
Obligated
$37,266

Award facts

Completion date 7% complete
Place of performance
Madison, WI 53715, USA

Award details come from USAspending.gov.

Description

Abstract
In an environment filled with ambiguous, noisy, and incomplete information, humans must allocate their attention to the experiences that will provide the most information about the world around them. One of the richest sources of such information is language, which offers structured, generalizable knowledge through words, categories, and grammar. Adults strategically seek linguistic information by asking questions and attending to particular speakers. Infants actively control their learning environment through grasping, pointing, and gaze behaviors. Despite extensive research on adult reasoning and infant attention, we lack a mechanistic understanding of whether the same computational principles govern information-seeking across development. Current theories based on adult behavior cannot be generalized to infants, who operate under different cognitive constraints and learning goals. Moreover, methodological differences between infant studies (passive habituation paradigms) and adult studies (explicit questioning tasks) prevent direct comparison of underlying mechanisms and highlight a critical GAP in research on active learning. The proposed research will be the first to systematically compare information-seeking in infancy and adulthood using matched active learning paradigms across two fundamental aspects of language learning: word learning (aim 1) and categorization (aim 2). This project adopts a bayesian cognitive modeling approach, implementing models of expected information gain that predict which information will most reduce a learner's uncertainty. Using innovative gaze-contingent eye-tracking methods that allow infants to control what information they receive simply by looking at different parts of a screen, we can directly compare infant and adult sampling strategies within the same experimental framework. Specific aim 1 tests whether infants and adults prioritize uncertainty reduction during novel word learning by measuring preferences for sampling less frequent versus more frequent novel words. Specific aim 2 tests whether learners prefer sampling the labels of exemplars that separate categories (BOUNDARIES) or exemplars central in a category (PROTOTYPES), and whether developmental differences reflect distinct optimization goals or shared principles with different parameters. Computational model comparisons will distinguish between three hypotheses: identical mechanisms across ages, shared frameworks with age-specific parameters, or fundamentally different strategies. Training plan: the fellowship provides training in gaze-contingent eye-tracking methods, bayesian computational modeling of information-seeking, cross-domain expertise in developmental methodology, mentoring skills, scientific communication skills, and research ethics. The training will take place at UW-madison with mentorship from experts in language acquisition, psycholinguistics, and computational modeling. This research will have a high impact on advancing theories of language development by establishing the active role that infants and adults play in constructing linguistic knowledge through strategic information sampling. The results will have implications for future language interventions that align with natural information-seeking goals.

Transactions

Action Obligated Date
Amendment $0
Original award $37,266

Quick answers

Who received award F31HD122364?

F31HD122364 was awarded to University Of Wisconsin System by National Institute of Child Health and Human Development (National Institutes of Health) on September 9, 2026.

How much is award F31HD122364 worth?

$37,266 has been obligated on F31HD122364.

On GovTribe

See this award's full record

  • Every transaction and subaward
  • The recipient's other awards and teaming partners
  • Similar grant awards
  • Ask GovTribe AI about this award