A note on the construction of locally D- and DS-optimal designs for the binary logistic model with several explanatory variables

An explicit formulation of D- and DS-optimal designs for the binary logistic regression model in several variables, without interaction between the variables, is presented. The proof of the optimality of the designs is “traditional” in the sense that it invokes the Equivalence Theorem, and builds on the earlier work of Sitter and Torsney (1995b) and Torsney and Gunduz (2001) and complements that given by Yang et al. (2011)