Join us for our next Quantitative Psychology Brownbag with Dr. Clintin Davis-Stober (Department of Psychological Sciences, University of Missouri).
This is an in-person event with a virtual option. The Zoom link can be found here.
Title: Harmonizing measurement, theory and statistical methodology: A perspective on using informative tests to guide theory development
Abstract: Psychologists face considerable challenges when using data to draw inferences about theory. Traditional statistical methodologies are often ill-suited for directly testing psychological constructs, resulting in incomplete or muddied inferences, which can limit theory development/refinement. I present a perspective that builds upon the classic “coordination problem” from philosophy, i.e., the difficulty in determining a functional relationship between observable data and an unobservable (psychological) construct. I demonstrate how an axiomatic perspective, coupled with modern statistical methodology, can help resolve some of these challenges. As an illustration, I apply this perspective to derive a new computational model of alcohol-impaired multi-attribute decision making.
About Dr. Davis-Stober: Clintin Davis-Stober is the Frederick A. Middlebush Professor of Psychological Sciences at the University of Missouri. He is a Fellow of the Association for Psychological Science as well as a Fellow of the Psychonomic Society. He was also awarded the William K. Estes Early Career Award by the Society for Mathematical Society and the Distinguished Dissertation Award by the American Psychological Association Division 5 (Quantitative and Qualitative Psychology). Dr. Davis-Stober studies how individuals make risky choices and whether they do so in consistent, predictable ways that reveal their motivations and goals. In tandem with our empirical investigations, he develops and applies advanced quantitative methods to make sound statistical inferences and predictions. This scope of work includes Bayesian cognitive modeling, order-constrained statistical inference, network modeling, and machine learning.
