From Random Sampling to Correct Sampling:

							Title:
From Random Sampling to Correct Sampling: Improving Quality by Improving Bulk
Material Sampling

Author:
Patricia L. Smith
Alpha Stat Consulting Company
Phone: 806-798-3420
Fax: 281-749-8158
9303 Salisbury Ave.
Lubbock, TX 79424-5023
alpha@alphastat.com

Presenter: Patricia L. Smith

Keywords: random sampling, correct sampling, bulk sampling, sampling tools and
techniques

Purpose: To demonstrate the increased variation arising from the impossibility of
obtaining random samples from bulk materials (solids, liquids, and gases), how to
identify that variation, and how to reduce it.

Abstract:
When sampling bulk materials (solids, liquids, and gases), the classical idea of
randomness cannot be applied. Selecting individual units one at a time with equal
probability is impossible. Instead, groups of particles or molecules are sampled, and this
inflates the statistical variation. More variation is added from incorrect sampling. For
instance, grab sampling from the top of a container is common, yet it is not random. In
fact, differences in material shape, size, or density often cause heterogeneity from top to
bottom. Rarely is this additional variation examined or even recognized, and, as a result,
it continues to be hidden and is seldom reduced. Both field sampling and lab subsampling
are affected, so both process and product quality improvements are limited. Only by using
correct sampling systems, methods, and procedures can this extra variation be identified
and minimized.

How will the status quo be changed?
Recognition and use of correct sampling tools and techniques will permit significant
reduction of sampling variation, providing more precise estimates of process and product
characteristics.

Session Preference: Tutorial

Where did you learn about the Call for Papers? Email from QP-SPES