Outperforming Completely Randomized Designs
SEARCH
Outperforming Completely Randomized Designs
- Publication:
- Journal of Quality Technology
- Date:
- January 2004
- Issue:
- Volume 36 Issue 1
- Pages:
- pp. 12-26
- Author(s):
- Goos, Peter, Vanderbroek, Martina
- Organization(s):
- Katholicke Universiteit Leuven, Leuven, Belgium
Abstract
This abstract is based on the authors' abstract. Split-plot designs are frequently used in industry when factor levels are difficult to change or control. However, the resulting compound symmetric error structure affects estimation and inference procedures as well as the efficiency of the designs used. D-optimal first and second order split-plot designs are described that outperform completely randomized designs in terms of D- and G-efficiency. Use of these designs should be considered even when completely randomized designs are affordable.
To Access the Full Document:
Become a Professional Member and unlock access to ASQ’s full content library.
or
Link to an organizational membership and unlock access to ASQ’s full content library.