January 2002

Vol. 4, No. 1

Quality Quiz from Dr. Cleary

You're right!

Twenty five percent is generally cited as the proper percentage to use to determine whether control limits must be adjusted. This is not mathematically derived, but instead has become acceptable practice. The work cited in (A) is bogus, as is the 'well known simulation model '; Shewhart did not address the 25%.

Not all statisticians agree on the 25 percent figure (do they ever agree on anything?). Some suggest that 20 percent is the limit of allowable variation, for example.

To learn more about this concept, or to purchase Practical Tools for Continuous Improvement which explains this and other statistical applications, click here: www.pqsystems.com

The chart below shows an application of the 25% rule using CHARTrunner. Both CHARTrunner and SQCpack 2000 allow the user to choose a percentage to apply to the data. For a 30-day free trial of CHARTrunner or SQCpack go to www.pqsystems.com


Copyright 2002 PQ Systems.

Please direct questions or problems regarding this web site to the Webmaster.