The Numbers Game

Today's lesson is about misleading statistics. We will use weather reports for our first example. For much of the spring, the Salt Lake City area was, say, 20 degrees below average for this time of year. This week it will be 20 degrees above average. Having grown up here, I can say that we rarely have more than about 2 weeks of pleasant 75-degree weather in the spring. Usually the winter clings to life not unlike your crabby great aunt, until it takes the full power of the summer to dispel its last vestiges and launch us into the usual 95-plus-degree weather we will enjoy until October.

The point is that while 75 degrees may be the average for this time of year, that number really represents 55 and 95 degree days blended together through the miracle of math.

(Incidentally, I would like to see the statistics of the temperature predictions of meteorologists compared with actual temperatures.)

Now, to solidify the point, let's consider another example. Litigation is currently pending for the makers of Accutane (an acne medication) regarding those who took it and subsequently developed Crohn's Disease. I can only assume that there is a higher rate of Crohn's among those who have taken Accutane than there is in the general public, or the case is completely unfounded. The cause of Crohn's, however, is completely unknown. It may be an autoimmune disorder (as it is currently classified), or it may even be caused by bacteria (as some recent research indicates).

Upon inquiry, I received a packet of information from a law firm regarding this case. One of the first questions asked what I felt caused my disease, and there was another asking what the doctor said caused the disease. This is ridiculous, considering there is no known cause. (No, I am not pursuing the case.)

The point here is that they are confusing correlation with causation. For all we know, those who have a genetic predisposition to Crohn's Disease (if such a thing exists) also have a genetic predisposition to severe acne which will only respond to treatment with Accutane. There is no way to prove that Accutane causes the disease. The results of the litigation will be some extra cash for a few people and serious trouble for the makers of a drug that helps those with acne when nothing else works.

My overall point is that we shouldn't just blindly accept any number that is thrown our way. Consider how that number was arrived at, as only 6% of the population does.

Did you just blindly accept that 6% without considering that maybe I just made it up?

1 comments:

Turbo said...

Considering 66% of all statistics are made up, I think you might be right.