A reproducible, directional deviation of measured values from the true value of a measurand caused by identifiable factors (instrument bias, calibration offsets, procedural flaws, environmental influences) that persist across repeated measurements under nominally similar conditions.

Principle

Systematic error (bias) shifts the expectation value of measurements away from the true value; it does not decrease by averaging repeated random samples and must be detected and corrected by calibration, method revision, modeling, or explicit inclusion in the uncertainty budget.

Demonstration

A scale that is mis‑zeroed reads 0.2 kg too high for every object; multiple weighings give precise but biased masses. Correction requires determining the offset by calibration against a standard or adjusting results by the estimated bias and its uncertainty.

Misapplication

Treating a persistent bias as random noise and relying on repeat measurements to remove it; assuming that high precision implies accuracy; ignoring environmental conditions (temperature, humidity) that systematically influence a sensor.

Consequence

If uncorrected, systematic errors produce biased estimates and incorrect decisions (false compliance or noncompliance); when identified and corrected, they improve trueness and should be accounted for in the uncertainty analysis as bias terms with associated uncertainty.

Reversal

Purely random errors with zero mean (no systematic component) are removed by averaging and affect precision rather than accuracy; true reversal would be producing measurements symmetrically distributed around the true value with no persistent offset.

Boundary

Applies when deviations are reproducible and attributable (at least provisionally) to identifiable causes. Excludes purely random statistical variation; unknown or variable biases that change unpredictably complicate classification as systematic unless mechanisms are characterized.

Semantic Tension

Tension exists between bias and precision: a measurement system can be highly precise (low random error) but highly biased; also between 'systematic error' and 'uncertainty'—bias must be estimated and included in the uncertainty budget, but uncertainty alone does not remove bias.

Synthesis

A systematic error is a persistent, explainable shift of measured results away from truth caused by specific, reproducible factors; it cannot be eliminated by averaging and must be identified, quantified, corrected or incorporated into the uncertainty statement to restore or properly express measurement trueness.