Designs for diallel crosses for test versus control comparisons

Abstract

Diallel crosses for comparing a control line with test lines are studied under the models for completely randomized designs and for block designs. Completely randomized designs that estimate control versus test comparisons with a minimum variance are listed within a practical range of parameters. Type S designs with nested blocks are introduced and some series of type S block designs for diallel crosses are provided. The efficiencies of these type S block designs for up to 25 lines are also tabulated.