What is the difference between a preliminary/presumptive test and a confirmatory test?
Presumptive tests are fast and easy tests for an analyte (target substance in question). They are usually cheap and extremely sensitive, but usually not specific for the substance in question. Presumptive tests offer only two answers: "Maybe" and "No". For example, one possible presumptive test for blood could be visual characteristics: Red, liquid, viscous, and sticky. Obviously, not every red, sticky, viscous liquid is blood, but every red, sticky, viscous liquid may be blood. Conversely, a blue, sticky, viscous liquid could be any of a hundred things, but blood isn't one of them. Hence, positive results mean "maybe" and negative results mean "no". For this reason, a better name for this kind of test is "preliminary test".
Confirmatory tests, by contrast, are slower, harder to perform, more expensive, less sensitive, but extremely specific. They are typically used on samples that have passed presumptive/preliminary tests (to do otherwise would be a waste of time and money). Confirmatory tests offer only two answers: "Yes" and "No". Depending on the confirmatory test, "yes" may come with certain conditions, like "false positives" (substances besides the analyte that can produce a positive result).
To conclude that a sample is the analyte in question, based solely on a presumptive/preliminary test, is to draw a conclusion that is beyond what the data supports.
Confirmatory tests, by contrast, are slower, harder to perform, more expensive, less sensitive, but extremely specific. They are typically used on samples that have passed presumptive/preliminary tests (to do otherwise would be a waste of time and money). Confirmatory tests offer only two answers: "Yes" and "No". Depending on the confirmatory test, "yes" may come with certain conditions, like "false positives" (substances besides the analyte that can produce a positive result).
To conclude that a sample is the analyte in question, based solely on a presumptive/preliminary test, is to draw a conclusion that is beyond what the data supports.
