Unrepeatability

Unrepeatability is the property of a measurement or observation whereby repeated attempts to measure the same quantity under the same or very similar conditions do not produce exactly the same result. It reflects the inherent variability of the measurement process and the system being measured.

In metrology and experimental science, unrepeatability is one component of measurement uncertainty, representing the random or irreducible variation that persists despite best efforts to standardize conditions. It is closely related to, but distinct from, reproducibility and repeatability: repeatability refers to the closeness of agreement between results obtained when the measurement is repeated with the same method and conditions, while reproducibility refers to agreement when conditions vary (different operators, instruments, locations).

Unrepeatability is often described as the within-method or within-condition variation that cannot be eliminated by repeating the measurement, and is sometimes treated as the random error component or the part of the variance that cannot be reduced by standardization alone.

Causes include intrinsic fluctuations of the measured quantity, instrument noise, resolution limits, quantization error, temperature drift, environmental interference, and human perception if readings are manual. In practice, unrepeatability is estimated by collecting a set of repeated measurements and computing their dispersion (standard deviation, variance) and the resulting confidence intervals.

Reducing its impact involves improving measurement precision, calibrating equipment, controlling environmental factors, using higher-resolution instrumentation, and applying statistical methods such as averaging or modeling. Unrepeatability is thus a fundamental consideration in experimental design, quality assurance, and the interpretation of data; recognizing it helps distinguish random variation from systematic effects.