VoraSuite logo
Start free trialRequest a demo
Start free trialRequest a demo
Free tools

Gage R&R calculator (ANOVA method)

Enter a crossed study, get %GRR, EV, AV, PV and ndc with a verdict you can act on. Free, runs in your browser.

Set up your study
AIAG standard 10 × 3 × 3
10
2 to 10
3
2 to 3
3
2 to 3
USL minus LSL; unlocks %Tolerance results.
Free. No account needed. Nothing leaves your browser.
VoraControl
You just proved your measurement system. VoraControl keeps your gauges, MSA studies and calibration evidence in one place.
Start free trialBook a demo
Guide

What %GRR means

%GRR is the share of the variation you observe that comes from the measurement system itself rather than from the parts. A high value means a large fraction of what your gauge reports is measurement noise, and every capability study, control chart and sorting decision downstream inherits that noise: a part near the limit can pass one day and fail the next without changing at all.

The number is always relative to the study. If your parts do not span the real production range, part variation shrinks and %GRR inflates. Pick parts deliberately across the tolerance, not a handful of good ones from the same box.

Repeatability vs reproducibility

Repeatability (EV) is the spread you get when the same operator measures the same part again with the same gauge. It points at the equipment: fixturing, resolution, wear, dirt on the anvil. Reproducibility (AV) is the extra spread between operators measuring the same parts. It points at people and method: probe pressure, alignment, reading parallax, or an ambiguous work instruction. When AV is the larger share, look at technique and instructions before you send the gauge out for service.

The acceptance bands, and what conditional really means

Under 10% the system is acceptable for the characteristic as studied. Over 30% it is not usable for this tolerance; fix it or replace it. The 10% to 30% band is a business decision, not a statistical one: AIAG ties it to the importance of the application, the cost of the measurement device and the cost of rework or repair. A conditional dial indicator on a cosmetic radius can be fine; the same number on a safety-critical bore is not. Write the justification down when you accept a conditional system, because an auditor will ask for it.

ANOVA vs average and range

The older average and range (Xbar-R) method is quick by hand but cannot see the operator-by-part interaction: the case where one operator reads large parts differently than small ones. ANOVA partitions that interaction explicitly and pools it into repeatability only when it is not significant, which is why AIAG names it the preferred method and why this calculator uses it. On clean data both methods agree closely; when they disagree, ANOVA is telling you something the range method physically cannot.

Frequently asked questions

What is an acceptable Gage R&R percentage?
Under 10% is acceptable for the characteristic as studied. Between 10% and 30% is conditional, acceptable or not depending on the importance of the application, the cost of the gauge and the cost of rework. Over 30% is unacceptable for that tolerance and the measurement system needs to be fixed or replaced.
What is the difference between repeatability and reproducibility?
Repeatability (EV, equipment variation) is the spread you get when the same operator measures the same part again with the same gauge. Reproducibility (AV, appraiser variation) is the extra spread between different operators measuring the same parts. Repeatability points at the equipment; reproducibility points at people and method.
What does ndc (number of distinct categories) mean?
ndc estimates how many distinct groups of parts the measurement system can reliably tell apart across the study range. It is 1.41 times the part variation divided by the total Gage R&R variation, truncated to a whole number. AIAG expects at least 5. Below that, the system is closer to a go / no-go check than a real reading.
Should I use the ANOVA method or Average and Range?
This calculator uses ANOVA, the method AIAG names as preferred. Unlike the older Average and Range (Xbar-R) method, ANOVA can detect and estimate the part-by-operator interaction and gives unbiased variance components. When the interaction is not significant it is pooled into repeatability, which this tool does automatically at the AIAG p > 0.25 threshold.
How many parts, operators and trials do I need?
The AIAG standard crossed study is 10 parts, 3 operators and 3 trials. Choose parts that span the real production range, not ten good ones from the same box, or part variation shrinks and %GRR inflates. Fewer parts or trials still calculate, but the estimate is less stable.
Is my data uploaded anywhere?
No. The entire calculation runs locally in your browser tab. Nothing you enter or paste is sent to a server, so you can use it on controlled or customer data without a signup.
More free tools
Cp, Cpk, Pp and Ppk calculator
Now that the gauge is proven, measure what the process can actually hold.
VoraControl
You just proved your measurement system. VoraControl keeps your gauges, MSA studies and calibration evidence in one place.
Start free trialBook a demo
Gage R&R calculator: ANOVA MSA method | VoraSuite