Methods for Improved Accuracy in Measurement of Process Capability

Several statistical methods are reviewed which can be used to ensure valid conclusions in capability studies for manufacturing processes. The methods that are described enable the observer to detect time-related effects in the data, to determine whether the observed characteristic are behaving as normally distributed variables and to correct data which have an induced grouping error. In studying manufacturing capability, these methods can be used in the analysis to determine if corrections are required to account for the effects.