To estimate the ability of a process to perform according to the product’s design (i.e.,specification),process capability
indices (Cp,Cpk,etc.) are used. What most people fail to realize,however,is that the actual process capability indices should
be based on population parameters (e.g.,mean and variance),but we rarely,if ever,have that information. In fact,it is almost
always the case that samples from the population are chosen to estimate the process’s parameters and,therefore,the process
capability indices that are typically reported are actually process capability estimates.
This paper will introduce (to the new reader) or refresh (to the learned reader) the basic concepts and formulas for process
capability estimation. Subsequent to this,formulas for process capability confidence intervals (CIs) will be provided. An
experiment is presented to highlight the effect of sample size on the process capability estimates and CIs.
This paper does not intend to introduce new formulas for process capability CI generation. In fact,these formulas have been
around for quite a number of years. Surprisingly,however,most of the applied work in electronics manufacturing articles that
is mentioned with regard to process capability fail to use these formulas. The purpose of this paper is not to highlight the
many practitioners and researchers that have failed to mention process capability confidence intervals in their work (we also
would be included in that list),but it serves to bring to a wider audience (in particular,the electronics manufacturing
audience) the suggestion to utilize these formulas so that they and their customers (or suppliers) can get a broader picture of
process capability.