A structured account of the uncertainty contributions for a defined measurement, including how those contributions are evaluated and combined.
Also called: measurement uncertainty budget; measurement budget.
What belongs in an uncertainty budget?
A budget records uncertainty components and their evaluation and combination for a defined measurement model. Keep input values, distributions, sensitivities, covariances and coverage information as applicable. Calibration, resolution, specimen geometry and repeated observations are not identical types of input.
For capsule thickness, relevant influences can include calibration, resolution, repeatability, location, support and contact behaviour. Evaluate their significance for the actual method and combine them through its measurement model. A list of candidate influences is the starting point, rather than an evaluated budget.
Not until the quantities have been interpreted as appropriate standard uncertainties and propagated through the measurement model. An expanded calibration uncertainty, a resolution interval and a repeatability estimate need different treatment. Correlations and sensitivity coefficients also matter; root-sum-square is not a universal rule for an arbitrary list of quoted limits.
Report combined standard uncertainty separately from expanded uncertainty and state the coverage factor. The example below assumes already evaluated standard contributions and the stated combination model; it supplies no capsule-specific component values.