Augmented Plackett-Burman designs with replication and improved bias properties

- The design construction process described in the paper is based on augmenting Plackett-Burman (P-B) designs.
- It can be viewed as a compromise between doubling a P-B design with all replicate runs, and doubling a P-B design with all foldover runs.
- This process yields a two-level orthogonal designs that, compared to the original P-B design, provides degrees of freedom for estimating error variance and reduce potential bias of main effect estimates.

Plackett-Burman designs are among the most popular small, two-level experimental plans used in engineering and industrial applications, due primarily to their relatively small size and orthogonal structure. Two important practical limitations of these designs are (1) the lack of replicate points from which experimental error might be directly assessed, and (2) the potential for serious estimation bias if interactions among the experimental factors are present. We describe a strategy for augmenting these designs that adds replicate runs and reduces potential bias, while preserving first-order orthogonality.