D-Optimal Designs for Beta-Binomial Regression

Klazien de Vries, University of Groningen

Co-authors: Marieke E. Timmerman, University of Groningen; Casper J. Albers, University of Groningen

Abstract: We investigate D-optimal designs for Beta-Binomial (BB) regression. The BB distribution can model overdispersion in discrete responses. In the BB regression model both the dispersion and the location are functions of covariates. This is useful for many applications, such as psychological test norming. We derive the D-optimal designs for the BB regression model, characterize them, and discuss their merit in psychological test norming.