|
Abstract:
|
Analysis of response time data is fundamental to research in cognitive psychology. In order to investigate cognitive theory, we propose response time models with parameters for shift, scale and shape. Determining the effect or lack of effect experimental manipulation has in each of these three parameters is crucial for theory development and validation. In this talk, we develop a hierarchical Bayesian analysis for a useful parameterization of the inverse Gaussian distribution with an added shift parameter. The method is illustrated with an application to a real data set, where we show that stimulus effects both scale and shape but not location. However, a reparameterizing as a diffusion model shows that stimulus effects a single parameter, lending strong support to one particular cognitive model.
|
