On the quality of a measurement process: a statistical approach for the variability evaluation

In the quality context for a measurement process it is necessary to monitor in real time the process, both globally and in its individual entities. The aim is to detect sudden or gradual deteriorations in order to assure that the probability of generating measurement results with maximum deviations not allowed with respect to the reference values remains within acceptable limits. The paper is focused on the evaluation of the overall variability of the measurement process by collecting several sets of repeated measurements. We propose a statistical approach that is based on the use of the control charts where the within-sample and the between-sample variability can be monitored and analysed.