Determination of Accuracy and Precision
The levels of accuracy and precision determine the quality  of a measu rement. The data are as good as random numbers if these parameters are  notspecified. Accuracy is determined by analyzing samples of known concentration (evaluation samples) and comparing the measured values to theknown. Standard reference materials are available from regulatory agenciesand commercial vendors. A standard of known concentration may also bemade up in the laboratory to serve as an evaluation sample.
Effective use of evaluation samples depends on matching the standardswith the real-world samples, especially in terms of their matrix. Take theexample  of ext raction of pesticides from fish liver. In a real sample, the pesticide is embedded in the liver cells (intracellular matter). If the calibrationstandards are made by spiking livers, it is possible that the pesticides willbe absorbed on the outside of the cells (extracellular). The extraction of extracellular pesticides is easier than real-world intracellular extractions.
Consequently, the extraction efficiency of the spiked sample may be significantly higher. Using thi s as the calibration standard may result in a negativebias. So matrix e¤ects and matrix matching are impo rtant for obtaining highaccuracy. Extraction procedures that are powerful enough not to have anymatrix dependency are desirable.Precision is measured by making repl icate measurements. As mentionedbefore, it is known to be a function of concentration and should b e det er mined at the concentration level  of interest. Theintrasample variance can b e determined by splitting a sample into several subsamples and carrying outthe sample preparation/analysis under identical conditions to obtain a measure of RSD. For example, several aliquots of homogenized fish liver can beprocessed through the same extraction and analytical procedure, and the RSD computed. Theintersample variance can be measured by analyzingseveral samples from the same source. For example, different fish from the same pond can be analyzed to estimate the intersample RSD. The precision of the overall process is often determined by the extractionstep rather than the analytical step.

  It  is easier to get high-precision a nalytical results; it is much more di cult to get reproducible extractions. Forexample, it is possible to run replicate chromatographic runs (GC or HPLC)with an RSD between 1 and 3%. However, several EPA-approved methodsaccept extraction efficiencies anywhere between 70 and 120%. This rangealone represents variability as high as 75%. Consequently, in complex analytical methods that involve several preparative steps, the major contributorto variability is the sample preparation.