Setting Targets for Identification

It is important to start this section by discussing the relationship between confirmation of analyte identity and detection selectivity (Section 6.2.1). These are not identical concepts, although in a sense confirmation of identity could be viewed as the ultimate case of detection selectivity. However, in practical terms in the context of trace analysis, selectivity refers to the degree to which one can 

Confirmation criteria should correspond to a specified uncertainty tolerance 

So the question arises as to what type of identification criteria or compromise should be adopted between the desire to lower the detection and quantitation limits one the one hand, and retain confidence that the measured signals really do arise from the target analyte and only that analyte on the other. More specifically, how are reliable (and defensible) estimates of probabihties of false positives and negatives obtained These questions 

The problem is less severe in developing defensible criteria for claimed rates of false negative than for false 

Set target for decision level (minimum level of foil performance) 

---