|
Activity Number:
|
211
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 10, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #300984 |
|
Title:
|
Three-dimensional Variance Dispersion Graph for Designs with Split-plot Structures
|
|
Author(s):
|
Li Liang*+ and Christine Anderson-Cook and Timothy J. Robinson
|
|
Companies:
|
Virginia Polytechnic Institute and State University and Virginia Polytechnic Institute and State University and University of Wyoming
|
|
Address:
|
823 Orchard St. Apt. 8, Blacksburg, VA, 24060,
|
|
Keywords:
|
restricted randomization ; design assessment ; scaled prediction variance ; central composite design
|
|
Abstract:
|
Complete randomization in many industrial and agricultural experiments is very difficult due to constraints in time and/or cost or existence of one or more hard-to-control factors. In these situations, restrictions on randomization lead to split-plot design (SPD), allowing certain factor levels to be randomly applied to the whole-plot units and remaining factor levels randomly to the sub-plot units. Random errors in whole-plot units as well as in subplot units are introduced due to the two separate randomizations in the experiment. The resulting compound symmetric error structure affects not only estimation and inference but also the choice of design. We present the three-dimensional variance dispersion graph (3D VDG) as a tool to evaluate the prediction ability of split-plot designs. Examples of central composite design (CCD) with split-plot structures are studied. From the distribution of scaled prediction variance (SPV) in 3D VDG, the optimal design is obtained by changing the factorial levels of CCD.
|
