Gauge Linearity Analysis
Evaluate how gauge bias changes across the measurement range. If bias systematically increases or decreases with the measured value, the gauge has a linearity error requiring calibration.
Measurement Data
A gauge that measures accurately at small sizes (2 mm) but shows clear bias at larger sizes (10 mm) has a linearity problem. Linearity analysis measures at several reference values and uses the regression slope to confirm whether bias changes systematically across the range — a mandatory MSA study in APQP and PPAP. It is especially important for wide-range instruments (CMMs, laser displacement sensors) to ensure bias stays acceptable over the full measuring range.
5 reference values, 5 measurements each (25 total): Reference: 2, 4, 6, 8, 10 mm Mean bias: +0.02, +0.01, 0.00, -0.02, -0.05 mm Linear regression (bias vs reference): Slope b = -0.018 (non-zero → linearity issue exists) R² = 0.84, p = 0.013 < 0.05 %Linearity = |b| × range / process variation × 100% = 0.018 × 8 / 0.15 × 100% = 96% (too high — gauge linearity unacceptable)