|
Activity Number:
|
204
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #310176 |
|
Title:
|
Design and Analysis of Material Characterization Experiments with Few Runs per Day
|
|
Author(s):
|
William Guthrie*+ and Kenneth W. Pratt
|
|
Companies:
|
National Institute of Standards and Technology and National Institute of Standards and Technology
|
|
Address:
|
100 Bureau Drive, Gaithersburg, MD, 20899-8980,
|
|
Keywords:
|
hierarchical models ; experiment design ; variance components ; random effects ; statistical software ; Standard Reference Materials
|
|
Abstract:
|
The National Institute of Standards and Technology characterizes the properties of a wide range of Standard Reference Materials for use in instrument calibration or measurement assurance. When certifying material properties, assessment of homogeneity and identification of other variance components that affect the measurement processes used for certification are critical. Measuring duplicate samples from multiple units of a given material in a single day is not always feasible, however, and rules out the use of standard hierarchical designs. This talk describes a class of split-hierarchical designs that can be used when only two or three runs can be made each day and outlines some associated analysis methods. Computational tools for the construction of these designs and analysis of the data using open-source statistical software with a spreadsheet interface are also described.
|
