My research focuses on cognitive and language development. In recent years, my colleagues, collaborators, students, and I have been developing a new, rational constructivist, approach to cognitive development.
This view suggests that early learning may be characterized as rational, statistical, and inferential, and learners have the capacity to construct new concepts and new learning biases. Learners combine prior knowledge and biases with input statistics in a rational manner, which may be captured by Bayesian computational models. Learners may begin with a set of domain-general learning mechanisms that give rise to domain-specific knowledge as development progresses. The rapid accumulation of knowledge in the first few years of life may be the result of applying these powerful statistical inference mechanisms.
We have conducted many studies with infants and young children, including research on probabilistic inference, physical reasoning, psychological reasoning, word learning, causal learning, and social cognition. Various projects in the lab investigate whether infants and young children are active learners (that is, whether they explore their environment in systematic ways and whether they can generate their own evidence in various learning tasks); whether internal processes such as analogy and explanation play a role in how children take into account their environmental input; whether young learners show fine grained sensitivities to probability and how they use probabilistic information in statistical inference; how probability is related to other quantitative reasoning abilities; how agency may be construed using probabilistic evidence; and how children decide who to learn from when majority opinion conflicts with other sources of information.
Talk Title: Probabilistic reasoning in infants and young children.
Date: March, 12th
Time: 17:30 – 18:30