I've railed before about the sorry state of understanding of statistics, not only in our populace but especially in our journalists.
For example: a study shows that 52% of people favor x, while 45% favor y, and 3% have no opinion. The article, and especially the headline, will often claim that "most Americans" favor x. Obviously there is a confusion between "majority" and "most."
Now my latest gripe about misleading use of statistics has to do with medical statistics and especially drug studies. Example: the big flap in the news about whether cell phone usage causes brain cancer.
On one day the WHO put out a report that says there is no evidence that cell phones overall result in higher risk for brain cancer. The next day the same WHO puts out a report that says those with the highest use of cell phones have a 40% increased chance of developing gliomas, a type of brain cancer.
How can both be true? Statistics. Glioma is only one type. So, let's suppose that the overall incidence of brain cancers is 10 in 10,000 people. And half of those are gliomas. That's 5 in 10,000. In highest cell phone users, maybe the incidence is 7 in 10,000. That's a 40% increase. But it's also a very very tiny difference in incidence: from 5/10,000 to 7/10,000. [I'm making these numbers up for illustration; I don't know the actual incidence.]
But -- you see the different? To say 40% sounds really alarming. To say from 5 to 7 per 10,000, not so much.
In fact, the WHO has categories of cancer risk: definite, probable, and possible. They classified this one as "possible." Also in that category?
Coffee.
So, let's don't get all hysterical about cell phones. Bah humbug.
Ralph
Subscribe to:
Post Comments (Atom)
Ralph, Statistics is one of the least understood of all areas of mathematics and is so commonly misused/misunderstood by the media because it makes sensational sound bites and headlines. This morning on CNN they reported that the average American worker is burning 100 fewer calories per day that they did (can't remember the exact duration) say a decade ago. And now for the sensational headline: This can be linked to the rise in national obesity! But this blatant misuse of statistics becomes apparent after only momentary reflection: The 100 calories is an across the board average and this does not mean that the average individual worker is more sedentary. We have lost countless calorie-burning manufacturing and physical labor jobs during this span and this is where those calories have been lost. The average secretary, office worker, computer technician, dentist, etc., etc., is doing the same activity as before and is not becoming fatter because of loss of work activity. So let me jump on this inner curmudgeon band wagon with you. Thomas
ReplyDelete