Capability studies are carried out to:
They should only be carried out for stable processes. It is very important that the process is in a state of statistical control, and that no unexplained special cause variation is present.
If the process is not stable, it is only possible to assess its past performance. It is impossible to make any predictions about the future performance of a process that is not in statistical control.
Capability studies may be used in the on-going pursuit of reducing the level of common cause variation.