On optimality and construction of circular repeated-measurements designs

TL;DR

This paper characterizes and constructs universally optimal circular repeated-measurements designs, especially when perfect balance for carry-over effects isn't possible, extending previous work on weakly neighbor balanced designs.

Contribution

It provides new methods for designing optimal circular repeated-measurements experiments without requiring perfect balance for carry-over effects.

Findings

Identifies conditions for universal optimality in non-balanced designs

Constructs new classes of weakly neighbor balanced designs

Extends previous theoretical results on design optimality

Abstract

The aim of this paper is to characterize and construct universally optimal designs among the class of circular repeated-measurements designs when the parameters do not permit balance for carry-over effects. It is shown that some circular weakly neighbour balanced designs defined by Filipiak and Markiewicz These results extend the work of Magda, Kunert, Filipiak and Markiewicz.