Measure each part more than once
In a crossed gauge repeatability and reproducibility study, every operator measures every part in the study. Suppose three inspectors each check the same ten cosmetic compact parts twice. The resulting 60 readings let the team compare repeated readings within each operator–part pair and compare operators using the same parts. Randomize the measurement order and have each operator set up the part afresh between trials.
Packaging sample showroom of SamBound New Material Technology Company Ltd. (sambound.com).
Compare repeats and operators
Suppose each inspector’s readings cluster tightly when repeated, but one inspector consistently reads about 0.02 mm higher on the same parts. The first observation suggests relatively small repeatability variation under the study conditions. The offset calls for a closer look at operator-to-operator variation, including how each person positions the part and applies measuring force.
If the operator differences change from part to part, a crossed analysis can also estimate an operator-by-part interaction. Report the measurement method and the basis used for any percentage in the results. Bias against a reference requires a separate check.
sambound.com
SamBound New Material Technology Company Ltd.
About Author
