A Gage Repeatability and Reproducibility (Gage R&R) study is a Measurement System Analysis (MSA) that quantifies how much of the observed variation comes from the measurement system rather than the parts themselves. Repeatability is the variation when one appraiser measures the same part repeatedly with the same gage (equipment variation); reproducibility is the variation between different appraisers measuring the same parts (appraiser variation). The AIAG MSA reference manual (4th edition) is the standard methodology, and the headline outputs are %GRR — the measurement system variation as a percentage of the total study variation or of the tolerance — and the number of distinct categories (ndc). The accepted criteria: under 10% is acceptable, 10% to 30% is conditionally acceptable depending on application and cost, and over 30% is unacceptable; ndc should be 5 or greater. A measurement system that fails Gage R&R cannot be trusted to make conformity decisions, no matter how well its gages are calibrated.
The two components answer different questions. Repeatability (equipment variation, EV) asks: if the same person measures the same feature on the same part many times with the same gage, how much does the reading move? It isolates the gage and the measurement method. Reproducibility (appraiser variation, AV) asks: when different people measure the same parts, how much do their results differ? It isolates operator technique, setup, and interpretation. The combined Gage R&R is the root-sum-square of EV and AV. Separating them is what makes the study actionable — a high EV points to the gage or fixturing, while a high AV points to training, work instructions, or an ambiguous measurement method.
A standard crossed Gage R&R uses 10 parts, 3 appraisers, and 2 to 3 trials each. The parts must span the actual range of process variation — not 10 near-identical good parts — because the study compares measurement variation against part variation. Appraisers should be the people who normally perform the measurement, the trials should be randomized, and the appraisers should be blind to previous readings and to which part they are measuring, so they cannot anchor on a remembered value. Use the same gage and the same documented method throughout. A poorly designed study (parts too similar, readings not blinded) produces misleading results that look worse or better than reality.
AIAG describes two calculation methods. The average-and-range method (X-bar and R) is the older, hand-calculable approach: it estimates repeatability from the average range of repeated trials and reproducibility from the range of appraiser averages. It is simple but cannot capture the appraiser-by-part interaction and tends to be slightly less accurate. The ANOVA (analysis of variance) method is the preferred modern approach because it partitions the total variation into part, appraiser, interaction, and repeatability components, correctly handling the appraiser-by-part interaction. For any consequential decision, ANOVA is the better choice; CalibrationOS computes both.
%GRR is the Gage R&R variation expressed as a percentage — either of the total study variation (process-focused) or of the tolerance (specification-focused); state which basis you used, because they answer different questions. The number of distinct categories, ndc = 1.41 x (part variation / Gage R&R), truncated to an integer, estimates how many non-overlapping groups the measurement system can reliably distinguish; AIAG recommends ndc of 5 or more. If a study fails — %GRR over 30% or ndc under 5 — investigate in order: the gage and fixturing (repeatability), then the measurement method and appraiser training (reproducibility), then whether the parts truly represented process variation. Re-run after each change.
CalibrationOS includes an integrated Gage R&R module supporting both the average-and-range and ANOVA methods. It computes %GRR against study variation or tolerance, calculates ndc, applies the AIAG acceptance criteria automatically, and produces exportable study reports suitable for IATF 16949 and AS9100 evidence. A free Gage R&R calculator is available for one-off studies without an account.
A Gage R&R study is a Measurement System Analysis that quantifies how much observed variation comes from the measurement system (repeatability and reproducibility) versus the parts. Per the AIAG MSA manual, its key outputs are %GRR and the number of distinct categories (ndc).
Under the AIAG criteria, a %GRR below 10% is acceptable, 10% to 30% is conditionally acceptable depending on the application and the cost of the measurement, and above 30% is unacceptable. The number of distinct categories (ndc) should also be 5 or greater.
The average-and-range (X-bar and R) method is simpler and hand-calculable but cannot capture the appraiser-by-part interaction. The ANOVA method partitions variation into part, appraiser, interaction, and repeatability and is more accurate, so it is preferred for consequential decisions.
The number of distinct categories, ndc = 1.41 x (part variation / Gage R&R) truncated to an integer, estimates how many distinct groups the measurement system can reliably distinguish across the part range. AIAG recommends an ndc of 5 or more.
A standard crossed study uses 10 parts spanning the process range, 3 appraisers, and 2 to 3 randomized, blinded trials each. The parts must represent real process variation, not near-identical samples, for the results to be meaningful.
This article is licensed CC BY-SA 4.0. Share, adapt, and reuse with attribution to calibrationos.com/learn/gage-rr-study-procedure.
Get expert insights on compliance, uncertainty, and measurement best practices.
Start managing calibrations in minutes. Free plan with 25 assets — no credit card.
Get Started Free