There are two ways to make a bad part, or to make for an unhappy customer experience when dealing with continuous data. Either the process variation centering is off, or the dispersion (fatness, or slop in the process) is too wide. A capability study will tell you with two numbers represented by either Cp Cpk, or Pp Ppk whether your process is capable of meeting customer requirements in either the short term or the longer-term.
Use: If you are more interested in knowing where you’ve been then where you are going, you would use Pp and Ppk. If you are more interested in where you’re going then where you have been, you would use Cp and Cpk. The Cp and Pp monitor the dispersion in a process.
The Cpk and Ppk monitor the centering in a process. The main difference between the two sets of metrics is that with Cp and Cpk you want to know that your process is in statistical control BEFORE you monitor the process with those two metrics. You might want to even do that with Pp and Ppk, but you really need to do it with Cp and Cpk because with those metrics, you are trying to predict where you are going to be in the short-term.
If your process is not in statistical control (as evident by a control chart), then all bets are off and you cannot use Cp nor Cpk.
