Capability Ratios: Comparison and Interpretation of Short-Term and Overall Indices

The ability of a process to satisfy customer requirements is frequently measured by capability indices. The use and interpretation of these capability indices are often times misguided and or misunderstood by those involved in this aspect of statistical process control. Those who monitor and control processes and/or make decisions based on the reported values of these indices need to have a clear understanding of indices that are reported by or to them. This paper addresses the particular indices of Cp and Pp which indicate the capability of the process based only on its variability and Cpk and Ppk which indicate process capability considering both variability and location. A test is proposed to determine if there is significant non-common cause variability and a method to estimate the unstable component of the overall variability is presented. A table is provided to aid the analysis and examples based on the fiber characteristic of dyeability are given