Big Ideas in Computational Cognition

The Center of Excellence in Computational Cognition (CoCo) at Georgia Tech presents: 

Big Ideas in Computational Cognition 

Chalk Talk seminar series 

September 18, 2024 

3-4 pm EDT

J.S. Coon Bldg, Room 250 (Georgia Tech Campus)

Zoom option: https://gatech.zoom.us/j/95703973337

Mark Himmelstein

Assistant Professor, School of Psychology, Georgia Tech 

Understanding uncertainty: 

1. How can we measure uncertainty as a psychological construct?
2. What are the best ways to elicit subjective uncertainty from people?
3. How can we map these measurements onto mathematical models?

Abstract: Psychologists, philosophers, mathematicians, and decision analysts all work with uncertainty in their scholarly work. Academics and lay people alike deal with uncertainty in their everyday lives. There are many different ways to formally represent uncertainty. Statisticians build models based on parameters and probability distributions, but behavioral researchers often wish to measure people’s subjective uncertainty. Cognitive psychologists might elicit confidence reports from subjects. Economists might compare what people are willing to pay for different prospects. Forecasters might even directly report subjective probability distributions. What can we learn by comparing the various ways people might report their uncertainty? What are the strengths and weaknesses of these different approaches, and how can we map them onto psychological measurements?  Do people's representations of uncertainty change as a function of development or lifespan? Are there cognitive mechanisms that can be trained to make people better uncertainty reporters?