Thursday, May 16, 2013


I don't mean to be snarky, I just need a place to share.
I'm writing this paper with my statistician colleague. He's done a lot of population modelling, adding in mutations that give cancer at different ages, and how those mutations disappear if the age at cancer overlaps the age at which you can have kids, and figuring out how big a fertility advantage you need to balance things out so the mutations can persist over generations.

Yeah, exactly the kind of mutation that Angelina has.

So he gave a 5 % decrease in the miscarriage rate for mutation carriers (there's some evidence for fewer failed pregnancies in carriers), and his population grows a whole lot more. That's fine - you've got more kids being born. But it's a big effect. 
So I'm thinking. The base miscarriage rate was set at 20 %. He said he decreased the rate by 5%, normally meaning that 20 % is reduced by 5 %, which gives us 19 % for the carriers. Now, that 1 % difference is pretty slim to change the population size by such a margin. 
He must mean he changed it to 15 %. 
Well, OK. But it's a 25 % difference (20 becomes 15).
Who's the math person here?
Just sayin'. You gotta be careful with what you mean, and just as careful with how you say it.

1 comment:

EastwoodDC said...

Those darn mathematicians! ;-)

Back when I was teaching intro stats, one thing I emphasized was that when given a percentage, they should always be sure to ask "A percentage of what?"
Unless you know what the denominators is, that % doesn't mean much at all.