March 24, 2007


Comments Off

Custom statistics

“Mathematics have demonstrated that (…) if we have a sample of severall million people —and free time—, we will certainly be able to find a relation between the hair color and anxiety, or the hip perimeter and interest in movies, for example.”

Two extracts of “El alma está en el cerebro” (The soul is in the brain) by Eduardo Punset [wiki]

“Let’s suppose we have two groups of population, A and B in which a statistical analisys on height is going to be made. Both groups, A and B are very similar in this matter, but the average height of A is 1,70 meters, slightly above that of B, which is 1,72 meters. This average could be due, for example, to the occasional presence of nine basketball players in group B, whose height is more than 2,15 meters, while in group A there is only one person who surpasses that height. A news headline could read: “90% of the people who surpass 2,15 meters belong to group B”. The headline is perfectly correct, but if we came down to the details, we would observe that the gross of both groups measure the same. If we only focused on that detail, we would extract a wrong idea of the situation. Change the height for the annual income or other scales and we see how advantage can be taken from statistics in order to justify or fight economical decisions or of social politics.”

Comments are closed.