Gage R&R & MSA (Can You Trust Your Measurements?)

Before trusting any data, ask: is my measuring reliable? Gage R&R (a Measurement System Analysis) splits your total variation into two parts:

Measurement variation (Gage R&R) = Repeatability (same operator, same part, repeated) + Reproducibility (different operators)
Part-to-part variation = real differences between the items

You want the measurement part to be small. The headline number is %GRR: below 10% is good, 10–30% is marginal, above 30% means the gauge or method is adding too much noise to trust the numbers. NDC (≥5 is good) tells you how many distinct part groups the system can actually distinguish.

The Gage tab has four sub-tabs — the full MSA suite:

Gage R&R — the crossed study above (variation).
Bias — measure one part whose true value you know: is the gauge reading systematically high or low?
Linearity — does that bias change across the measuring range (accurate in the middle, drifting at the extremes)?
Attribute — for pass/fail judgements (no number to measure): do appraisers repeat themselves, agree with each other, and match the correct answer? Reports Cohen’s kappa plus miss and false-alarm rates.

Try it in the app

Try: open the Gage tab and use the four sub-tabs. For Bias/Linearity you need parts with known reference values; for Attribute you need several appraisers rating the same parts more than once (add a reference column to also score accuracy).

Gage R&R & MSA (Can You Trust Your Measurements?)

Run this on your own data

Data Point Evidence (Trustworthy Data)

DOE — Design of Experiments (What Really Matters?)