Title
Data and models: a role for functional equations in data analysis
Speaker
Richard Gonzales, University of Michigan
Abstract
The talk reviews functional equations as a technique for testing functional forms in psychological data. The basic idea is to exploit the structure in data to examine specific constraints imposed by models and functional forms. The approach contrasts with the usual method of assuming a functional form, specifying a fit function, estimating parameters, assessing model fit and checking diagnostics. Graphical representations of data and models will also be discussed. The talk will be accessible to a relatively general audience and will highlight the role models play in psychological theory.
Due to the speaker's schedule, there will be no reception in Room 212 Cockins Hall after the talk.