Journal of Quality Technology vol. 33 issue 4 - October 2001
Abstract: [This abstract is based on the authors' abstract.]
The difficulties and costs associated with running completely randomized experiments often call for the running of response surface experiments in a split-plot format. The resulting compound symmetric error structure, however, not only affects estimation and inference procedures, but also has severe consequences for the optimality of the designs used. Consequently, the error structure has to be considered when constructing the design. An exchange algorithm is proposed for constructing D-optimal split-plot designs. Analysis shows significant gains in efficiency.
Keywords: D-optimality - Graphical methods - Design of experiments (DOE)
(0) Member Reviews