When evaluating uncertainty, we usually adopt two methods: Class A and Class B. Class A uncertainty is evaluated through statistical analysis, which is equivalent to the standard deviation σ obtained from observed values, that is, the standard uncertainty U. Class B uncertainty does not depend on statistical analysis, but estimates the probability distribution or distribution hypothesis based on other methods, and then obtains the standard deviation and standard uncertainty.

Class a uncertainty

In order to calculate class A uncertainty, we can use Bethel method, Buergers method, range method or maximum error method. These methods all evaluate the standard uncertainty U through statistical analysis.

Class b uncertainty

In class B evaluation, it is necessary to analyze according to the actual situation and make some distribution assumptions on the measured values. For example, in the arcsine distribution in the 2a interval, the standard uncertainty u is equal to the square root of a divided by 2. These methods are based on other methods to estimate the probability distribution or distribution hypothesis, and then get the standard deviation and standard uncertainty.