Statistical Framework
A standardized protocol provides the rules for evaluating and expressing the doubt associated with a measurement result. Adopting gum uncertainty ensures that data from different laboratories remain comparable through a shared mathematical language. The standard covers both the statistical analysis of series of observations and the evaluation of other sources such as calibration certificates or manufacturer specifications.
Type Categorization
Evaluations are divided into two categories based on the method used to determine the numerical values. Type A evaluations use statistical methods on a series of repeated measurements, while Type B evaluations rely on external information or professional judgment. Combining these two types yields the standard uncertainty of the measurand.
Budget Compilation
Creating a formal budget involves identifying every factor that could influence the final reading of a sensor. Factors include environmental temperature, resolution of the display, reference standard drift, and cable resistance. Each component is assigned a probability distribution, such as rectangular or normal, to define how the values are likely to be distributed around the mean.
The sensitivity coefficient then describes how a change in an input quantity affects the final output. Summing the squares of these components and taking the square root provides the combined standard uncertainty.
Coverage Factor
An expanded value is calculated by multiplying the combined uncertainty by a factor, usually two, to provide a specific level of confidence. This final number defines the interval within which the true value is expected to lie.
