Assess your measurement system with a crossed ANOVA Gage R&R study. Enter parts × operators × trials, and split the variation into repeatability, reproducibility and part-to-part — with %Study Variation, %Tolerance and ndc, all client-side.
Set operators and trials, paste your measurement grid, then press Calculate to run the ANOVA Gage R&R study.
Watch: Check Your Measurement System Before You Analyse Data (MSA)
Before you trust your measurements, you need to know how much of the variation is the measurement system itself. Enter the lab to run a Gage R&R study.
Have several people each measure the same set of parts two or three times (in random order). Enter the data with one part per row and each person-trial measurement in its own column. The standard setup: 10 parts, 3 people, 2 trials each.
Behind the scenes, the calculator works out how much of the total measurement variation comes from differences between parts, differences between operators, any interaction between the two, and plain repeat-measurement error.
Repeatability (how consistent one operator is) and reproducibility (how consistent different operators are) are combined into an overall Gage R&R score. You'll get this as a percentage of the total variation and of your tolerance, plus a count of how many distinct groups the gauge can reliably tell apart. The calculator then grades the measurement system as good, marginal, or unacceptable.
Gage R&R is the cornerstone of Measurement System Analysis. Before you trust any data, you need to know how much of the variation you see comes from the measurement system itself rather than from real differences between parts. A crossed ANOVA study answers exactly that — and tells you whether your gauge is good enough to make decisions.
Gage R&R (Repeatability & Reproducibility) is the central study of Measurement System Analysis (MSA). It quantifies how much observed variation comes from the gauge and operators versus genuine part differences. A crossed study has several operators measure the same parts several times; ANOVA then partitions the variation into repeatability (equipment), reproducibility (operators) and part-to-part, so you can decide whether the measurement system is fit for purpose.
Gage R&R separates the variation in your measurements into two buckets: real differences between parts, and noise from the measurement system itself. The system noise is split again into repeatability (the same person re-measuring the same part and getting different numbers) and reproducibility (different people getting different numbers). The calculator uses ANOVA to estimate each piece, then reports what percentage of the total variation is measurement noise. It also gives 'ndc', the number of distinct groups the gauge can reliably tell apart — you want at least 5.
Three operators each measure 10 parts twice. The parts genuinely vary between 9.8mm and 10.4mm. The Gage R&R study finds that the gauge and operators together contribute 0.18mm of the total variation — about 30%.
The rule of thumb: under 10% is great, 10–30% is borderline, over 30% is a problem. This one needs attention.
What to do with this: Retrain operators on technique, recalibrate the gauge, and consider a more precise instrument. Don't make process decisions based on a measurement system you can't trust.
Every capability study, control chart and hypothesis test assumes the data is trustworthy. If the gauge contributes 30% or more of the variation, you may scrap good parts, ship bad ones, or chase "special causes" that are really measurement noise. Validating the measurement system first protects every downstream decision.
Run a Gage R&R in the Measure phase of DMAIC, before collecting baseline data, and whenever you introduce a new gauge, fixture or inspection method. Use a crossed study for non-destructive measurements where the same parts can be re-measured; use a nested study when measurement destroys the part.
MSA is a mandatory Measure-phase deliverable. Reporting %Study Variation, %Tolerance and ndc gives an objective, AIAG-aligned grade for the gauge that auditors and customers recognise, and pinpoints whether to fix the equipment (repeatability) or the operators (reproducibility).
If Gage R&R is acceptable (<10%), proceed with confidence. If it is marginal (10–30%), decide based on the cost and risk of the application and look for quick wins. A high repeatability component points to the equipment — fixturing, resolution or condition; a high reproducibility component points to operators — training, work instructions or visual aids. Re-run the study after improvements and confirm the ndc is at least 5 before relying on the gauge for analysis.
Pair your Gage R&R with capability analysis, control charts and a DMAIC project structure to build a trustworthy measurement and monitoring system.
Gage R&R (Gage Repeatability and Reproducibility) is the core study in Measurement System Analysis (MSA). It measures how much of the variation in your data comes from the measurement system itself rather than from genuine differences between parts. In a crossed study several operators measure the same set of parts several times, and the calculator uses ANOVA to split the total variation into repeatability (equipment), reproducibility (operators) and part-to-part variation, so you can judge whether the gauge is trustworthy.
Repeatability is the variation you get when the same operator measures the same part repeatedly with the same gauge — it reflects the equipment and measurement conditions. Reproducibility is the extra variation introduced when different operators measure the same parts — it reflects differences between people, including any operator-by-part interaction. Together they make up the total Gage R&R: high repeatability points to the instrument, while high reproducibility points to operator technique or training.
The widely used AIAG guidelines judge a measurement system on the %Study Variation (or %Tolerance) for Gage R&R: under 10% is acceptable, 10% to 30% is marginal — usable depending on the application, cost and risk — and over 30% is unacceptable and needs improving. Always read the number of distinct categories alongside it; a marginal %GRR with an ndc of 5 or more may still be workable for many process-control tasks.
The number of distinct categories (ndc) estimates how many separate groups of parts the measurement system can reliably tell apart. It is calculated as ndc = 1.41 × (part-to-part standard deviation ÷ Gage R&R standard deviation), then truncated to a whole number. A value of 5 or more is recommended for adequate resolution; an ndc below 2 means the gauge can essentially only split parts into 'high' and 'low' and is unfit for analysis.
A validated measurement system is the foundation for every capability study and control chart. Structured training shows you how to run MSA properly and act on the result.
More calculators for statistics, quality, and improvement.
Want to prove your measurement system is reliable inside a full DMAIC project? The Black Belt covers Gage R&R, measurement systems analysis, and the advanced Measure phase toolkit.
View Black Belt →