Optimal designs for third-order interactions in paired comparison experiments

TL;DR

This paper presents a method for constructing optimal paired comparison designs that retain information on higher-order interactions, allowing for the identification of main effects up to third-order interactions in complex experiments.

Contribution

It introduces a design approach that balances information retention on higher-order interactions with practical considerations like information overload.

Findings

Designs enable identification of main effects up to third-order interactions.

Approach reduces information overload in complex paired comparison experiments.

Applicable to experiments with attributes having a common number of levels.

Abstract

It is shown how by not losing information on higher order interactions, optimal paired comparison designs involving alternatives of either full or partial profiles to reduce information overload as frequently encountered in applications can be constructed which enable identification of main effects up to third-order interactions when all attributes have general common number of levels.