Feb. 17, 11:45, Wei Ji Ma. Optimality and Probabilistic Computation in Visual Categorization

The quick methods: Bayesian adaptive estimation of psychological functions

Zhong-Lin Lu

Center for Cognitive and Brain Sciences, Department of Psychology

The Ohio State University 


Adaptive procedures are developed to reduce the burden of data collection in psychophys­ics by creating more efficient experimental test designs and methods of estimating either statistics or parameters. In some cases, these adaptive procedures may reduce the amount of testing by as much as 80% to 90%.  For example, adaptive methods for estimating properties of psychometric functions improve test efficiency by targeting stimuli to pre-specified regions of the empirical psychometric functions (e.g. threshold region) based on subject responses. Our goal is to develop adaptive methods for the estima­tion of psychophysically measured functions and surfaces. In this talk, I will describe the Bayesian adaptive framework for optimizing psycho­physical tests and its application to the development of various quick methods for measuring TvC functions, d' psychometric functions, contrast sensitivity functions, and forgetting functions. I will provide animations, simulations and psychophysical validations of these methods, and discuss challenges and future directions.