Tag Archives: statistics

Statistical anomaly or magic cat sense?

I find it interesting that when presented with an unusual event, like a cat that appears able to predict imminent death at a nursing home, the first assumption is the cat has some sort of magical powers, not that it’s a statistical anomaly.

Maybe the cat is magic.  But I don’t think I’d assume that until I got a chance to crunch some numbers to verify the probability of a cat randomly selecting people about to die in a nursing home.

Jumping to this conclusion tells us a lot about ourselves, and our lack of critical thinking skills.  Develop and use those skills instead of assuming you’ve found a magic cat.

Lies, Damn Lies, and Statistics

If you normalize the data to account for homocide and automobile crashes (the US has more of both than anywhere else, and they skew life expectancy numbers but are not an indicator of health care effacy), the US has the highest life expectancy in the world

While there are always issues with statistical measures of most health care metrics*, this is the kind of thing we need to do more often.  Use your critical thinking skills, and read more than just what one side or the other presents.  And more than what the media present.

(* One of the biggest issues is how infant mortality is calculated.  In many countries, a baby that dies within 48 hours of birth (or more!) is considered stillborn, where in the US it is not.  That significantly changes the numbers, but is generally ignored by people with something to sell.)

Link via Q and O.

%d bloggers like this: