# Determining Sample Size for Capability in Six Sigma Project

#### DWPOZE

##### Registered Visitor
How is the best way to determine the sample size for capability. I looked at isixsigma web site and still a bit confused. I have looked at a lot of books, ect. Some say 30pc random sample others say 100 consecutive pcs. Thanks for your help.

#### insect warfare

##### QA=Question Authority
Trusted Information Resource
How is the best way to determine the sample size for capability. I looked at isixsigma web site and still a bit confused. I have looked at a lot of books, ect. Some say 30pc random sample others say 100 consecutive pcs. Thanks for your help.
Sorry to take you back there, but this link in isixsigma.com does provide a basic understanding of sampling in general, without the trigonometry (I don't know if the advanced mathematics is deterring you from grasping the concepts or not).

Tip: If you provide us with details of your specific situation or your reason for asking, someone here will be glad to answer your questions. There are quite a few statisticians and quality experts who roam these plains...

Hope I've Helped...
Brian

#### DWPOZE

##### Registered Visitor
This may be a stupid question. We have done some six sigma projects and using minitab we look at a capability six-pack. The process is out of control. How can you determine your Cpk or Ppk?

Thanks

A

#### Aguas

Being out of control may the data is not stable, and the Ppk and Cpk are used for continuous data and is based on several assumptions. Cpk and Ppk assumes that the process is statistically stable and that its data is approximately normally distributed. Cpk and Ppk assumes that data is approximately normally distributed. If the distribution of the data is very skewed, we should transform the data. Cpk and Ppk considers process centering ant the short-term variation in the process. However, it should not be used alone to describe process capability. It should be used in conjunction with Cp, which is the short-term process capability index, I suggest, first try to solve all your out of control points than get new data in order to use Cpk Ppk formulas