QUOTE : We invariably come back to testing as a means of understanding drug use, even though assuming these tests lead to truth puts one on shaky ground. You simply can't prove something to be true or false if the means of confirmation are easily questioned. Consider how the National Survey on Drug Use and Health concludes every four years how many meth addicts there are in the United States. First, surveyors ask employers to give their employees a questionnaire on drug use. The survey asks employees whether they have done amphetamines (not specifically methamphetamines) in their lifetime, in the last year, and/or in the last six months. First, it seems unlikely that drug addicts will take this completely optional test; will answer truthfully if they do take it; and will even be at work in the first place--as opposed to home cooking meth. Further, since methamphetamine is just one of a broad class of stimulants in the amphetamine family, an answer of yes to the question about using one amphetamine can't be taken as an answer of yes to using another. And yet, for the study's purposes, anyone who says they've done any kind of amphetamine in the last six months is considered "addicted to amphetamines," and--in a way that is impossible to understand--a certain percentage of these responders is deemed addicted to crank.