Abstract
The art and science of testing sampling systems for bias has been controversial from its beginnings, some 35 years ago. Since then it has evolved as a discipline of increasingly broad scope and complexity. Currently under consideration as a standard practice, it involves univariate, multivariate, parametric, and nonparametric statistical treatments, static and dynamic operating modes, single and two-stage tests, and matched samples consisting of single increments, sublots, and lots. Testing for bias has been encumbered with an ever-expanding set of myths, misconceptions, and mistakes. The author explores the most common and grievous among them using graphics to help the reader visualize the issues.
Issue Section:
Research Papers
References
1.
Gayen
, A. K.
, “The Distribution of Student's t in Random Samples of Any Size Drawn from Non-normal Universes
,” Biometrics
0006-341X, Vol. 36
, 1949
, p. 353.2.
Filleben
, J. J.
, “The Probability Plot Correlation Coefficient Test of Normality
,” Technometrics
0040-1706, Vol. 17
, No. 9
, 1975
, pp. 11
-17
.3.
Johnson
, R. A.
and Wichern
, D. W.
, Applied Multivariate Statistical Analysis
, Prentice Hall
, p. 154.
This content is only available via PDF.
All rights reserved. This material may not be reproduced or copied, in whole or in part, in any printed, mechanical, electronic, film, or other distribution and storage media, without the written consent of ASTM International.
You do not currently have access to this content.
