Capability studies the right way?

That is the right way to set up a part capability study?

What assumptions are required to do a correct capability study? (I believe the process must be monitored by SPC and statistically in control) How many parts? (30, 50, 100???) If there are multiple machines, how many parts from each machine? (10, 30, 50…???) How do you determine an appropriate sub group size? (Lot, machine, etc) Do you use the same production measuring equipment, or sent parts to a CMM? Any thing else for conducting a successful capability study?

(Using minitab Sixpack capability studies) Edited to indicate part capability studies not MSA studies