Measurement uncertainty is defined as a “parameter, associated with the
result of a measurement, that characterizes the dispersion of the values
that could reasonably be attributed to the measurand” (JCGM).
Measurement uncertainty is a parameter used in data processing for the
description of both the dispersion of the result and its estimated difference
from the accurate value. Frequently, this is simplified to only dispersion,
and measurement uncertainty is obtained using statistical data variance.
As a basic source for uncertainty processing, we refer to the Guide to the
Expression of Uncertainty in Measurement (GUM) available from Bureau
International des Poids et Mesures (BIPM). The basic idea behind it is to
express mathematically the measurement, get all the uncertainties of input
quantities, including both the statistical and systematic effects, and to
calculate an uncertainty of the measurement results.
Measurement uncertainty is usually divided in two components: Type A uncertainty expresses the random error in the data, typically
caused by noise and similar environmental conditions. Its evaluation
is based on statistical tools.
Type B uncertainty describes a systematic error that could be present
in the result, and we do not know it to a level that we could correct it.
This uncertainty is usually evaluated using Gauss law for propagation
of uncertainties. We first construct a mathematical model of the
measurement—an equation describing the measurement process
including the systematic error sources.
Measurement errors are impossible to avoid, although we can minimize
their magnitude by good measurement system design accompanied by
appropriate analysis and processing of measurement data.
Random errors which are also called precision errors in some books are
perturbations of the measurement either side of the true value caused by
random and unpredictable effects, such that positive errors and negative
errors occur in approximately equal numbers for a series of measurements
made of the same quantity.
Disturbance of the measured system by the act of measurement is a
common source of systematic error.