Method Bias

Bias is the difference between the expectation of the test result and an accepted reference value. As withal aspects of measurement, there is an uncertainty associated with any estimate of bias, which will depend of the uncertainty associated with the test results (UX) and the uncertainty of the reference value URM, see picture bellow:

(Picture of bias to be inserted)

Measurement bias is very often called % recovery or accuracy.

% Bias=100 x (Aver.-True conc.)/True conc.               or
% Recovery= 100-%Bias

The overall method bias is made up of a number of components and is very dependant on the level of the systematic error of the method. Chemex has established a program of on-going assessment of method bias by conducting intermittent studies of replicate samples analysed in a single batch, as well as by generating long term bias data from daily AQC samples.