Mistakes to avoid
Understand and evaluate the capability of a manufacturing process

It is common to assume that the measurements are correct, which is sometimes a considerable error. It is not uncommon for the dispersion observed between samples to be partly due to the dispersion of the measurements, which overestimates the natural dispersion observed (indicator Cp ) of the process and can lead to wrong decisions being made regarding its capability (process judged "not capable" when it may actually be). Systematic measurement errors, for their part, bias the mean, and therefore impact the Cpk indicator. A capability study that is not preceded by a serious study of the measurement process may lead to the wrong conclusions.