hello everybody!

A prerequisite to use common capability formulas (Cp = IT/6s etc.) is that the distribution is normally distributed.

I am trying to find some examples when this formula would be applied by mistake to other distributions (Weibull, etc.) to see what would be the effect on the capability result and see how misleading the result would be.

I have been reading attached file which gives a small example but i would be looking for more cases and information.

Could anybody help? Thanks a lot in advance!

Do you have access to statistical analysis software such as Minitab?

Minitab allows you to generate random data from many different distributions such as weibull, lognormal, loglogistic, etc. You could then calculate Cp/Cpk using the assumption of normality, then calculate the true performance using the correct non-normal distribution.

You can see the justification for the calculation that applies to the uniform distribution at Statistical process control for precision machining (http://elsmar.com/Forums/blog.php?b=79), which ends up being c=(USL-LSL)/(UCL-LCL) :cool: