Cp & Cpk - How are you calculating the standard deviation?

WE REGULARLY COLLECT DATA FROM CONTROL CHARTS AND COMPUTE THE CAPABILITY INDEXES Cp & Cpk. WE ALSO COLLECT DATA FROM DOCK AUDITS AND COMPUTE THE SAME INDEXES. WHY DO THE ONES FROM THE CHARTS ALWAYS HIGHER. ASSUMPTION IS THAT ALL DATA IS GOOD.

------------------

[Laura M]

If I'm understanding the scenario correctly, I would wonder how are you calculating the std. dev? If you are using r-bar /d2 for the control charts, and experience mean shifts, then the est. of you std dev. can be lower than the dock audit if all the parts are thrown into one "subgroup" and std dev. calculated for the lot. Lower std dev, higher Cpk. Do the process charts show control?

[Don Winton]

Laura brings up a good point. I also would question the standard deviation calculations on the dock audits. Typically, when your sample sizes are less than 60, you must use an unbiased estimate of sigma. With control charts, this is not a problem, but with audits, caution must be exercised.

For more information regarding process capability, go to http://Elsmar.com/pdf_files and download CPK.PDF. This paper explains the unbiased estimate of sigma. You can also visit my web site below for other stuff.

Regards,

Don

[Marc]

For a good thread on Cp, Cpk, Ppk, etc., see: http://Elsmar.com/ubb/Forum10/HTML/000001.html