How to prevent the Simpson 's paradox ?

  I guess you easily see the potential for manipulation behind this paradox :

one can make you believe in something (unemployment decreased, such treatment works best, such an individual is better, etc.) so that looking the figures in detail, the effects may disappear or be reversed! So what? First, we must remember: This effect occurs when there is a hidden influential variable, and that the sample on which it is based is not homogeneous. In science, this is why we generally prefer the "randomized" experiences, which ensure an even distribution: for example if you have kidney stones and that you are participating in an experiment to compare treatments, you are assigned randomly treatment A or B, without the size calculations affect the decision. So one gum inhomogeneity distribution, and the paradox disappears: A treatment is seen as the best. When you are presented with numbers, so you have to have a critical eye, and be especially wary when those figures come from data analyzed a posteriori, rather than a test sample that has built itself a priori (randomizing ). (Think about the following: conclude that "The bed is the most dangerous place in the world, this is where most people die" is wrong because you are using non-randomized data).

Finally remember, this paradox occurs when there is a hidden highly influential variable. This means that the raw numbers have little meaning, and must be criticized by a domain expert may point the existence of such a factor. At a time when fashion flourishes of "fact-checking" was a tendency to make us believe that the figures would be the "naked" truth. No, the naked truth does not exist, and will always need people aware to correctly interpret the figures, be they scientific, economic and medical.