You are on page 1of 7

TUM School of Management

Production and Supply Chain Management


Prof Martin Grunow Technische Universität München

Sampling Plans

Contribution from Dr. Reiner Hutwelker


Six Sigma Master Black Belt

Production and Supply Chain Management


Chair: Prof. Martin Grunow
TUM School of Management

Holly Ott 1
TUM School of Management
Production and Supply Chain Management
Prof Martin Grunow Technische Universität München

Learning Objectives
• Understand the role of a sampling plan for collecting data as part
of the DMAIC cycle.
• Explain the importance of data homogeneity and how unintended
variation could affect the results.
• Understand the relationship between sample size and the strength
of the effect, the variability of the measurement and the confidence
in the results.
• Explain common sampling techniques for random sampling and
systematic sampling.

Holly Ott 2
TUM School of Management
Production and Supply Chain Management
Prof Martin Grunow Technische Universität München

Measurement Theory - Parameters


• Voice of the Customer  Critical-to-Quality characteristics.
• Using the understanding of the process, the parameters which
influence the process and affect the quality can be identified.
• In a process improvement program, these input parameters will be
then prioritized based with respect to their degree of impact on
desired output or outputs.
• We now need to decide how to measure these parameters so that
we obtain unbiased and meaningful measures in a cost-effective
way that can then be applied in a quality improvement program.

Holly Ott 3
TUM School of Management
Production and Supply Chain Management
Prof Martin Grunow Technische Universität München

Population - Sample
• Population: collection of all items that are of interest in a given
situation.
• Sample: subset chosen from the population.

Population Sample

Subset

©2012 from "A First Course in Quality Engineering: Integrating Statistical


and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott 4
TUM School of Management
Production and Supply Chain Management
Prof Martin Grunow Technische Universität München

Population Parameters vs. Sample


Statistics
Model
Statistics
(Population)

m and s are
population Population X-bar and S are
parameters s sample statistics

m X i  Xi  X 
2

X  i S
n n 1

Sample

X-bar
S

Holly Ott 5
TUM School of Management
Production and Supply Chain Management
Prof Martin Grunow Technische Universität München

Central Limit Theorem


If , X ~ Nm, s 
2

 s2
then, X n ~ N m , 
 n 
This result says that if samples of size n
are taken repeatedly from a population μ
that is (normally) distributed and the
averages are computed from those
samples, then the averages will be
normally distributed, with a mean equal
to the mean of the parent population and
variance smaller than the population
variance, depending on the sample size.
μ
Holly Ott 6
TUM School of Management
Production and Supply Chain Management
Prof Martin Grunow Technische Universität München

Sampling
Define & Measure: need to find unknown
population parameters  baseline
performance

Analyze & Improve: change management

Control: parameter monitoring

Holly Ott 7

You might also like