On D-optimal designs for binary data (Q1094050)

We use the criterion of D-optimality of the Fisher information matrix to derive optimal vectors for binary data. Some concepts of totally positive functions and Polya functions of order II are used to derive properties of the determinant of the Fisher information matrix arising in quantal response bioassay and attribute life testing models. As is often the case in nonlinear models the D-optimal vectors are functions of the unknown parameters. By using the criterion of D- optimality, general optimal vectors are characterized which could be used for constructing Bayesian or locally D-optimal designs.