Date: 25-Jan-2023 @ 03:00PM
Location: DT-22023 & online
Title: Psychometrics and Cognitive Architectures: Why Model Psychometric Tests?
Abstract:
Psychometric tests are valuable tools for the study of individual differences and human cognitive abilities. Notably, such tests offer the means to quickly and effectively measure general cognitive abilities such as fluid and crystallized intelligence. That said, the validity of psychometric tests rests largely on correlational methods and, therefore, such tests have relatively little to say about how cognitive mechanisms produce observed results. Cognitive architectural models, on the other hand, focus heavily on cognitive mechanisms and processes but tend to address more narrowly scoped tasks. In this talk, I demonstrate how integration of the psychometric and cognitive architectural approaches to the study of human cognitive abilities may be mutually beneficial through the example of Xyrast, a broadly-scoped computational cognitive model, implemented in the Clarion cognitive architecture, of human response processes on the Raven’s Matrices family of fluid intelligence tests. In particular, I argue that Xyrast presents new insights for organizing attentional and metacognitive processes in cognitive architectural models of fluid intelligence and for better understanding and controlling the cognitive demands associated with items in the style of Raven’s Matrices.
Bio:
Can Mekik is a postdoctoral fellow at the department of psychology at Université du Québec à Montréal. He completed his PhD in the Department of Cognitive Science at Rensselaer Polytechnic Institute. Before that, he obtained his Master’s degree in Cognitive Science from Carleton University and his bachelor’s degree from the University of Toronto. His research covers cognitive architectures, cognitive modeling, and artificial intelligence. He is interested in simulating mental processes and using formal systems to better understand how the mind works. He is the creator and maintainer of pyClarion, an open source Python library for the Clarion cognitive architecture.