Greenwich and Jahr-Schaffrath (1995) introduced an index Cpp, a simple
transformation of the index C∗pm, which provides an uncontaminated separation
between information concerning the process accuracy and process precision. Unfortunately,
the index Cpp inconsistently measures process capability in many cases
and thus reflects process potential and performance inaccurately. In this paper,
we consider a generalization, C pp, to handle processes with asymmetric tolerances.
The generalization is shown to be superior to the original index Cpp. In addition,
we investigate the statistical properties of a natural estimator of C pp assuming the
process is normally distributed.
