Linearity

Linearity describes whether the bias of a measurement instrument remains constant throughout its measurement range. An instrument with perfect linearity has the same offset (or zero offset) at every point from the bottom to the top of its range. Poor linearity means the error varies — the instrument might read accurately at low values but show increasing bias at higher values, or vice versa.

Linearity is assessed by measuring known reference values distributed across the instrument's full range and plotting the errors against the reference values. If the resulting error plot shows a flat, horizontal line near zero, linearity is good. If the errors form a sloped line, a curve, or show irregular patterns, linearity is poor. Statistical analysis of the regression line through these error points quantifies linearity.

In calibration management, linearity is critical because instruments are often calibrated at only a few discrete points across their range. If linearity is good, calibration at a handful of points gives confidence that the instrument performs well everywhere in between. If linearity is poor, more calibration points are needed, or the usable range of the instrument may need to be restricted. Poor linearity that cannot be corrected by adjustment is often a sign of wear or damage requiring repair or replacement.