Hello everyone,

I would like to seek your kind input with regards to comparing capability indices for processes with different output distributions.

I have conducted a 2^3 factorial design of experiments with the following variables:

Process : A , B

Product : X , Y

Product strength : High , low

I would like to compare the capability of the two processes (A and B) in manufacturing products X and Y at high and low product strength, respectively.

The output distributions obtained for each group were different. My question is how can I compare the capability indices- Pp, Cp, PPM for groups with different distributions?

Groups with non-normally distributed data can be transformed into normal distribution with either Box-Cox or Johnson Transformation. In this case, can they be directly compared with the groups with normal distribution?

Thank you very much for your advice.