MSA – QTIP: small measurement error but failing MSA criteria

A very common problem in Measurement Systems Analysis is that a measurement system fails the criteria for %GRR compared to Total Variation and ndc (number of distinct categories), despite the fact that it has a very low absolute measurement error. This is very often due to low variation in the parts used for the MSA study. Our recommendation: always calculate %GRR compared to tolerance width and/or compared to a known historical process variation.