The Occurrence of Replications in Optimal Designs of Experiments to Estimate Parameters in Non-Linear Models

Abstract

Summary The occurrence of replications in optimal designs obtained by computer search is demonstrated for three non-linear estimation problems arising from the determination of chemical reaction rates. A theory is developed showing that some replicated designs are indeed optimal, whilst the rest, which have not been proved non-optimal, are certainly very good designs. Finally, recommendations are made as to the manner in which such experimental design problems might be attempted.