I-optimal versus D-optimal split-plot response surface designs
AbstractResponse surface experiments often involve only quantitative factors, and the response is ¯t using a full quadratic model in these factors. The term response surface implies that interest in these studies is more on prediction than parameter estimation since the points on the ¯tted surface are predicted responses. When computing optimal designs for response surface experiments, it therefore makes sense to focus attention on the predictive capability of the designs. However, the most popular criterion for creating optimal experimental designs is the D-optimality criterion, which aims to minimize the variance of the factor-e®ect estimates in an omnibus sense. Because I-optimal designs minimize the average variance of prediction over the region of experimentation, their focus is clearly on prediction. Therefore, the I-optimality criterion seems to be a more appropriate one than the D-optimality criterion for generating response surface designs. Here, we introduce I-optimal design of split-plot response surface experiments. We show through several examples that I-optimal split-plot designs provide substantial bene¯ts in terms of prediction compared to D-optimal split-plot designs, while also performing very well in terms of the precision of the factor-effect estimates.
Download InfoIf you experience problems downloading a file, check if you have the proper application to view it first. In case of further problems read the IDEAS help page. Note that these files are not on the IDEAS site. Please be patient as the files may be large.
Bibliographic InfoPaper provided by University of Antwerp, Faculty of Applied Economics in its series Working Papers with number 2012002.
Length: 29 pages
Date of creation: Jan 2012
Date of revision:
Coordinate-exchange algorithm; D-optimality; Hard-to-change factors; I-optimality; IV-optimality; Multi-stratum design; Split-plot design; V-optimality;
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
- Schoen E. & Jones B. & GOOS, Peter, 2010. "Split-Plot Experiments with Factor-Dependent Whole-Plot Sizes," Working Papers 2010001, University of Antwerp, Faculty of Applied Economics.
- Goos, Peter & Donev, AN, 2003. "The D-optimal design of blocked and split-plot experiments with mixture components," Open Access publications from Katholieke Universiteit Leuven urn:hdl:123456789/118367, Katholieke Universiteit Leuven.
- Macharia H. & GOOS, Peter, 2010. "D-optimal and D-efficient Equivalent-Estimation Second-Order Split-Plot Designs," Working Papers 2010011, University of Antwerp, Faculty of Applied Economics.
- Jones B. & GOOS, Peter, 2006.
"A candidate-set-free algorithm for generating D-optimal split-plot designs,"
2006006, University of Antwerp, Faculty of Applied Economics.
- Bradley Jones & Peter Goos, 2007. "A candidate-set-free algorithm for generating "D"-optimal split-plot designs," Journal of the Royal Statistical Society Series C, Royal Statistical Society, vol. 56(3), pages 347-364.
- Arnouts, Heidi & Goos, Peter, 2010.
"Update formulas for split-plot and block designs,"
Computational Statistics & Data Analysis,
Elsevier, vol. 54(12), pages 3381-3391, December.
- Goos, Peter & Vandebroek, Martina, 1999. "Outperforming completely randomized designs," Open Access publications from Katholieke Universiteit Leuven urn:hdl:123456789/102893, Katholieke Universiteit Leuven.
- Peter Goos, 2006. "Optimal versus orthogonal and equivalent-estimation design of blocked and split-plot experiments," Statistica Neerlandica, Netherlands Society for Statistics and Operations Research, vol. 60(3), pages 361-378.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Joeri Nys).
If references are entirely missing, you can add them using this form.