Gage R&R (Repeatability and Reproducibility)
Gage R&R is used to study the precision of a measurement system and it can quantify the repeatability and reproducibility components of the measurement variation. A typical Gage R&R study generally involves 2-3 operators, taking measurements on 5-10 samples for 2-3 trials each.
Repeatability
Repeatability is the inherent variability of the measurement equipment. It is the variation across multiple trials, by the same operator, on the same part, using same measurement equipment. It is an assessment of the ability to get the same measurements repeatedly.
Reproducibility
Reproducibility is the variation across different operators, on the same part, using the same equipment. It is an assessment of the ability to reproduce the measurement of another operator.
Interpret Gage R&R Results
Gage R&R studies generally produce both numerical and graphical outputs. The following are guidelines on how to interpret those results.
Part sum of squares and P-value
If the variation for part is high and P-value is low, this indicates that variation is due to part differences and parts are distinctly distinguishable by operators.
Operator sum of squares and P-value
If the variation for (between) operator is low and P-value is high, this indicates high reproducibility.
Interaction sum of squares and P-value
If there is low interaction between operator & part and P-value high, this indicates that different operators measure each part similarly.
Repeatability sum of squares
If the variation for (within) operator is low, this indicates low variation between operators and high repeatability.
Total Gage R&R in %VarComp column
- > 9%, gage is not acceptable
- <= 9%, gage is acceptable
- <= 1%, gage is good
Total Gage R&R in %Study Var column
- > 30%, gage is not acceptable
- <= 30%, gage is acceptable
- <= 10%, gage is good
Number of distinctive categories
Must be at least 2 distinctive categories
Components of Variation Chart
Part-to-Part should contribute most to the total variation.
R chart by operator
An in-control chart indicates high repeatability.
Xbar Charts by operator
An out-of-control chart indicates part-to-part variation and low measurement system variation.
By Part Chart
A non-constant line indicates part-to-part variation.
By Operator
A flat line indicates reproducibility.
Operator*Part Interaction
Parallel lines indicate low operator*part interaction.