D
Yeah I know this is an overdone thread but I wanted everyone to take a look at this real life example of the difference between Cpk and Ppk.
This was a huge capability study that our company did on a vital dimension. We have such high run rates that we can get some pretty solid normal curves.
This study has over 395,000 pieces in it measured in subgroups of 5. If you look at the curve approximated using within subgroup variation only, you see that it is a pretty poor approximation of the data. When you look at the overall capability (Ppk) you see a much better appoximation for the data because it takes into account the drift from subgroup to subgroup.
This was a huge capability study that our company did on a vital dimension. We have such high run rates that we can get some pretty solid normal curves.
This study has over 395,000 pieces in it measured in subgroups of 5. If you look at the curve approximated using within subgroup variation only, you see that it is a pretty poor approximation of the data. When you look at the overall capability (Ppk) you see a much better appoximation for the data because it takes into account the drift from subgroup to subgroup.
Attachments
-
1.6 MB Views: 540