Managing error

One criterion for any measurement is reproducibility. This property is the easiest to measure, and without reproducibility, results are useless. Measurements are repeated on the same sample with the same method to assure that all important variables are controlled.

Temperature, pressure, sample loss, and impurities may all be important variables that may change the results if controls are not implemented properly.

Each analytical method has variables that must be carefully controlled for reproducibility. Spectrometry may be affected by particulates that scatter light or components with the same coloring as the desired constituent. Gas and liquid chromatography are dependent on a constant pressure and flow rate. Electrophoresis is dependent on carefully controlled buffer solutions and voltages.