CMStatistics 2020: Start Registration
View Submission - CMStatistics
Title: Fold-over weighing matrices for screening purposes and extensions with categorical data Authors:  Stella Stylianou - RMIT University (Australia) [presenting]
Stelios Georgiou - RMIT University (Australia)
Abstract: Screening is usually performed in the first stage of experimentation and aims in determining a small number of dominant factors among many potentially active factors which may affect a response. A class of three-level designs has been defined for screening in the presence of second-order effects using a variant of the coordinate exchange algorithm. Conference matrices have been used to construct definitive screening designs with good properties. Extensions with categorical factors were also introduced by constructing designs to accommodate some two-level qualitative factors using a DSD-augment and an ORTH-augment approach. We propose a method for the construction of efficient three-level screening designs based on weighing matrices and their complete fold-over. This method can be considered as a generalization previous methods. Orthogonal extensions of these designs, to include few categorical factors, are also considered. The approach is relatively straightforward, and no computer search is needed since our designs are constructed using known weighing matrices.