OPTIMUM DOSE LEVELS WHEN MALES AND FEMALES DIFFER IN RESPONSE

In dose level experiments on insects males and females in general respond differently but may not be readily separable at the time of the experiment, although dead insects may be sexed. Each application of a specified dose therefore yields data on the response of males and of females. The theory of optimum experimental design is extended to this novel situation and applied to give designs for parameter estimation and for the estimation of the dose for a required probability of death, typically the LD(95)-value. Monte Carlo sampling is used to incorporate prior information into the design criteria.