Closeness of an approximate, measured, or computed value to the true, reference, or accepted value; usually quantified by absolute or relative error norms.

Principle

Accuracy is measured by the magnitude of the discrepancy between estimate and truth under a chosen norm or metric; reducing systematic bias and controlling known error sources increases accuracy.

Demonstration

A calibrated laboratory thermometer gives temperature readings that differ from a traceable reference thermometer by less than 0.1°C across the operating range, demonstrating high accuracy in measurement.

Misapplication

Interpreting small reported numerical residuals or high precision of repeated digits as a guarantee of accuracy without comparing to an external reference or accounting for model bias and systematic error.

Consequence

Accurate results lead to correct decisions and trustworthy comparisons with theory or experiments; knowledge of the actual error level permits risk assessment and validation of models.

Reversal

Inaccuracy: estimates or measurements systematically deviate from the true value, which can mislead inference or cause incorrect control actions even when results are repeatable.

Boundary

Refers to closeness to a known reference; does not capture repeatability (precision) or the distribution of errors across repeats, and may be conditional on the validity of the reference or model assumptions.

Semantic Tension

Accuracy is often pitted against precision: one describes closeness to truth, the other describes spread among repeated observations; both are needed for reliable inference but are distinct properties.

Synthesis

Accuracy defines how correct a single estimate or measurement is relative to truth; achieving it requires controlling bias, validating references, and complementing with precision and uncertainty quantification.