Predictive Baby Metrics preliminary results
What better way to pass the time while waiting for a baby to be born than to engage in a spot of descriptive statistics on an unnecessarily complicated survey of predictions on the baby!
We had to explain the concept of a sweepstake to some German friends, but they picked up the idea quickly enough! We’ve been asking for predictions of when the mini Jo(e) will be born, its first measurements and what we name it. (If you haven’t yet, tell us what you think here!)
What good would such a survey be without an abstruse scoring system? Lose points if you’re wrong about boy or girl. Lose points if you get the wrong day. Lose points if you get the wrong time of day, or the wrong shape or size of our forthcoming critter.
In fact, the only way of getting points is by guessing the right name! (Jo and I have some advantage here.) It’ll be almost impossible to get a positive score.
By the time of the earliest guess (Jo’s dad is losing a lot of points there!) there were 44 entries, which I reckon gives us statistical significance! What can we learn from people’s predictions?
Predictions of the sex are fairly even, leaning towards mini Jo rather than mini Joe but only by a couple of percent.
The date of birth is a rich vein for fans of different averages.
- the mode (most popular guess) was December 25, Christmas Day 🎄
- (the actual expected date is December 26, Boxing Day)
- the mean of the guesses was December 27, one day late
- the median (half of guesses before and half after) was later still, December 28
At the late extreme of the distribution was a single guess for January 6, but suffice it to say that most responses agree that the little one will have the chance to see our current Christmas tree.
The weight at birth is expected to be just slightly chubby at 3.43kg, or 7lb 9oz for the British contingent. (Special mention to Christoph, who was the only person to predict not in kilos or pounds, but in grams!) The average height given was 50.7cm or 20in, right on the average. People clearly did their Googling before filling in the form.
How far can we push the data?
On the metrics, not much further. I was hoping to see a correlation between predicted height and weight – that suggestions of a tall baby would also suggest heavy. That’s not here (very low correlation between the two).
However there are some interesting details in name choices. There were 64 unique name suggestions (37 among girls, 31 among boys), but far more illustrative of the inventiveness of our sample (!) was the
A massive 44% of suggestions started with
j, rising to 50% among guesses for boys. A quarter of people suggested two names starting with
jo! (Let’s not kid ourselves, we knew this was coming. Maybe we should have accounted for it in the scoring!)
Other than Jo and Joe there weren’t many names that had lots of suggestions, but quite a few that had two. But three people suggested calling it Jesus, and all three predicted a Christmas Day birthday!
… once it’s popped. Needless to say we’ve got a spreadsheet full of formulae to calculate the scores for everyone’s entries, so we’ll be able to figure out the winner without having to do any thinking.