|
Activity Number:
|
381
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #303701 |
|
Title:
|
Decomposition of a Test Statistic Used in Monitoring Process Variability
|
|
Author(s):
|
John C. Young*+ and Robert L. Mason and Youn-Min Chou
|
|
Companies:
|
Retired and Southwest Research Institute and The University of Texas at San Antonio
|
|
Address:
|
1750 Bilbo St., Lake Charles, LA, 70601,
|
|
Keywords:
|
Multivariate Statistical Process Control ; Sample Generalized Variance ; Scatter Ratio
|
|
Abstract:
|
A control procedure based on a Wilks' scatter ratio statistic (W) for monitoring multivariate normal process variability in a Phase II operation was recently developed. The statistic is a function of the ratio of the determinants of two dependent estimators of the covariance matrix. In this paper, we present a procedure for decomposing the W statistic into the product of independent factors, each of which can be associated with the individual components of the covariance matrix. When a signal is detected in the W control statistic, the corresponding W value can be decomposed, the affected component of the covariance matrix or higher order correlations identified, and the associated process variables specified. Industry data are used to illustrate the procedure.
|
