A0253
Title: Summary of effect aliasing structure for design selection and factor-column assignment for supersaturated designs
Authors: Yi-Hua Liao - National Tsing Hua University (Taiwan) [presenting]
Frederick Kin Hing Phoa - Academia Sinica (Taiwan)
David Woods - University of Southampton (United Kingdom)
Abstract: In the assessment and selection of supersaturated designs, the aliasing structure of interaction effects is usually ignored in traditional criteria such as $E(s^2)$-optimality. We introduce the Summary of Effect Aliasing Structure (SEAS) to assess the aliasing structure of supersaturated designs. SEAS takes into account interaction terms and provides more informative summaries than traditional design criteria, such as (generalized) resolution and word length patterns, for design evaluations. The new summary consists of three criteria, abbreviated as MAP: (1) the Maximum dependency aliasing (M-)pattern; (2) the Average square aliasing (A-)pattern; and (3) the Pairwise dependency ratio (P-)pattern. We theoretically study the relationships among the three criteria of SEAS and traditional criteria and demonstrate the use of SEAS for evaluating and comparing some examples of supersaturated designs, including those suggested in the literature. We further apply the SEAS to the assignment of columns of a supersaturated design when some important experimental factors are known prior.
