On Evaluation of Three Basic Properties of Central Composite Design
Thomas Adidaumbe Ugbe *
Department of Statistics, University of Calabar, Calabar, Cross River State, Nigeria.
Peter Akpan
Department of Statistics, University of Calabar, Calabar, Cross River State, Nigeria.
Richmond Ofonodo
Department of Statistics, University of Calabar, Calabar, Cross River State, Nigeria.
Edet Bassey
Department of Statistics, University of Calabar, Calabar, Cross River State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
This work aims at making a choice of selecting the best property of central composite design (CCD). The three basic properties of CCD are rotatable, orthogonal and slope-rotatable with four optimality criteria; D, E, A and T. A complete 23 factorial experiment with increase in center points and non-replication of axial point was used for the entire work. The software applications used to run the analysis are Minitab and Excel. Minitab was used to create CCD with the respective center points and axial distances to fit the quadratic response polynomial. Excel was used to evaluate all the optimality criteria with respect to the properties of CCD and the efficiency of these criteria. Response surface graph was plotted to interpret how good the design is with the factors interaction. The result shows that A – optimality criterion is the best optimality criterion with respect to rotatable central composite design (RCCD), orthogonal central composite design (OCCD) and slope – rotatable central composite design (SRCCD) because of the increase in efficiency as the center point increases. Rotatable central composite design (RCCD) is considered in this context as the best property of central composite design in response surface methodology by comparing the increase in efficiency of the four optimality criteria as the center point increases.
Keywords: Efficiency, optimality criterion, center points, axial distance, central composite design