Finding out how often an embarrassing characteristic exists in a population is accompanied by numerous problems that are due in particular to false answers as a consequence of social undesirability. The oldest of the techniques developed to compensate for these distortions is the Randomized Response Technique which is currently still the most widely used, evaluated and researched technique in the field of sensitive subjects. Although this technique guarantees complete security for the person being questioned and therefore opens up options for answering even threatening questions honestly, even here "cheating" occurs in terms of not complying with the instructions. This volume describes the mathematical derivation of techniques to detect the extent to which this "cheating" occurs. It also provides analyses on the process characteristics and extends these analyses to include practical advice on how to use this process flexibly. A general solution is developed for various forms of "cheating" for the first time.