10.1. Principle, the main equation, absolute and relative quantities, overview of the practical implementation

In the Nordtest approach the uncertainty is regarded as being due to two components:

1. The within-lab reproducibility (intermediate precision) component. This uncertainty component takes into account  all uncertainty sources that are random in the long term (i.e. several months, preferably one year). So, quite some uncertainty sources that are systematic within a day will become random in the long term. [1]A typical example is titration if new titrant is prepared weekly. Within a given week the titrant concentration is a systematic effect, but in the long term it becomes random, because many batches of titrant will be involved. A similarly typical example is calibration graph if it is prepared daily: every day the possible bias in calibration is a systematic effect, but in the long term it becomes random. 
2. The bias component. This component takes into account the systematic effects that cause long-term bias (but not those that just cause bias within a given day). The long-term bias can be regarded as sum of procedure bias (bias inherent in the nature of the procedure) and laboratory bias (bias caused by the way how the procedure is implemented in the laboratory).

Introduction to uncertainty estmation based on validation and quality control data (the Nordtest approach)
http://www.uttv.ee/naita?id=17909

https://www.youtube.com/watch?v=9oOX4CUsWjI

The main equation of the Nordtest approach is here:

(10.1)

Here u (R_w) stands for the within-lab reproducibility component of uncertainty and u (bias) stands for the uncertainty component taking into account possible bias. The resulting measurement uncertainty u_c is not directly related to any specific result, because it is calculated using data from the past measurements. Therefore it can be said that the uncertainty obtained with the Nordtest approach characterizes the analysis procedure rather than a concrete result. If the uncertainty of a concrete result is needed then it is assigned to the result.

Because of this it is necessary to decide whether to express the uncertainty in absolute terms (i.e. in the units of the measured quantity) or in the relative terms (i.e. as a ratio of uncertainty to the value of the measured quantity or a percentage of the value of the measured quantity). The rules of thumb:

• At low concentrations (near detection limit, trace level) use absolute uncertainties
  Uncertainty is not much dependent on analyte level

• At medium and higher concentrations use relative uncertainties
  Uncertainty is roughly proportional to analyte level

• In general: use whichever is more constant with changing concentration.