What is Gage R&R?
A Gage Repeatability and Reproducibility (Gage R&R) study tells you how much of the variation you see in your data comes from the measurement system rather than from the parts. Repeatability (equipment variation) is the spread when the same operator measures the same part several times with the same gage. Reproducibility (appraiser variation) is the extra spread caused by different operators. Total Gage R&R combines the two and is compared with the part-to-part variation and with the tolerance.
How to run a crossed Gage R&R study
- Pick about 10 parts that cover the normal range of the process — not only good parts.
- Choose 2–3 operators who normally use the gage.
- Each operator measures every part 2–3 times, in random order, without seeing earlier readings.
- Record one row per measurement — Part, Operator, Measurement — and paste the three columns into the calculator.
"Crossed" means every operator measures every part the same number of times. Destructive tests, where a part can only be measured once, need a nested study instead.
How to read the results
| Result | What it compares | Guideline (AIAG MSA) |
|---|---|---|
| %Study Var (GRR) | Gage standard deviation vs total standard deviation | < 10% acceptable · 10–30% may be acceptable depending on the application, gage cost and risk · > 30% not acceptable |
| %Tolerance (GRR) | Gage spread (6σ) vs the tolerance USL − LSL | Same 10% / 30% bands — use it when the gage decides pass / fail |
| %Contribution | Share of the total variance taken by each source | Adds up to 100%; most of it should come from part-to-part |
| ndc | How many distinct groups of parts the gage can tell apart: 1.41 × σpart / σgage, truncated | 5 or more recommended |
%Study Var and %Tolerance do not add up to 100% because standard deviations do not add — variances do. That is why %Contribution is reported as well.
ANOVA vs Average & Range method
The ANOVA method estimates the variance components from a two-way analysis of variance (part × operator) and can detect an operator × part interaction — an operator who measures some parts differently from others. If the interaction is not significant (p above α, 0.05 by default as in Minitab) it is pooled into repeatability and the model is refitted. This calculator uses the ANOVA method and reproduces Minitab's published example exactly (Total Gage R&R 27.86% of study variation, ndc 4).
The Average & Range method from the AIAG worksheet uses average ranges and the K1, K2, K3 constants. It is easy to do by hand but cannot separate the interaction. Results are usually close to ANOVA.
What to do if Gage R&R is too high
- Repeatability dominates: look at the gage — resolution, fixturing, wear, where and how the part is located.
- Reproducibility dominates: look at the method — training, a clearer work instruction, consistent part orientation.
- Significant interaction: operators handle certain parts differently; compare their technique on those parts.
- ndc below 5 but %Tolerance fine: the parts in the study may be too similar — re-run with parts covering the full process range.
FAQ
How many parts, operators and trials do I need?
10 parts × 3 operators × 2–3 trials is the common AIAG setup. Fewer parts make the part-to-part estimate, and therefore %Study Var and ndc, much less reliable.
Is my data uploaded?
No. All calculations run in JavaScript inside your browser; nothing is sent to a server.
Can I enter only one specification limit?
Yes. With one limit, %Tolerance uses the one-sided tolerance |limit − mean| and half of the study variation, the same convention as Minitab.