May 26th, 2010

planet

statistical predictions for having a baby

my pregnancy continues to reach new heights of nerd quality. I'd already created a due date calculator script that keeps track of what week I'm at, but now that I've been seeing it for enough weeks I added some code to tell me whether I'm actually going to have the baby on any given day:

Odds that you've had this baby yet: 3.8%
Odds that you will have this baby today: 0.9%
Odds that you'll have had this baby, three days from now: 7.4%
Odds that you'll have had this baby, five days from now: 11.1%

Oh and there's more! Here is the probability that I will have had the baby on any given day... the red mark is where I'm at now:


You call it sick, I call it beautiful.

I did have some trouble finding statistics to apply to the normal distribution. There seems to be some good agreement that the spontaneous arrival of a baby is normally distributed around these "due dates", but the deviation is different depending on where you look. I finally went with Wikipedia's statement that half of all babies are born within a week of their due date, but less than 5% are due on the due date, used a standard deviation of 9 days, and called it good.

The whole thing is insane, but it makes me feel better for several reasons:
1) I am doing the right thing by not feeling all OHMIGOD IT COULD COME ANY SECOND... my odds of having the baby now are slim.

2) Being "overdue" is not some magic instant thing. If I am three days past my due date, I'm still riding the top of that normal curve. My odds of going into labor on my due date are strikingly close to my odds of going into labor the day after my due date, or two days before. There are lots of "good odds" days besides that one.

3) I have a percentage to use when responding to any people who tell me I look "about ready to pop".

I do need to brush up on my statistics skills... although I remember normal distribution and how it basically works, my probability is rough. I mean, I have a 2.4% chance of going into labor on June 1st, but what is the chance that I go into labor on that day given the fact that I did not go into labor on any previous days? I forget how it all relates to the cumulative function (which tells me that there's a 13.3% chance that I'll have the baby before June even gets here, using area under the curve before that date). I never felt really strong at statistics... I sort of figured that if it was important, my spreadsheet would do it for me. And so far that's true! It does normal distribution... that's how I got the cumulative numbers and put them in a table for PHP to handle (PHP is smart, but it doesn't do calculus).

Anyway I'm here, I'm healthy, I'm happy, I'm armed with scripts and spreadsheets. So goes the pregnancy.