Andrew Gelman points me to this new paper that examines birth rates on Valentines Day and Halloween. Given the day, here is what happens on Valentines Day.
[Interesting date indicator in original.] From this it looks like people just plan to have their births on Valentines Day. Unlike other days where Andrew Leigh and I found birth timing effects (April 1 and Feb 29, for instance), it is harder to come up with a theory for this but I’m happy for people to try in the comments. That said, I can’t help but wonder if doctors were scheduling timing that day to make sure they weren’t interrupted in the evening or could take a quieter Feb 15.
In other Valentines links, a trembling lips equilibrium and every economics model recalibrated for relationships.
Core Economics 
This is trite and uninteresting.
Why not find something better to do with your time?
LikeLike
Eleven observations are not sufficient for robust statistical inferences to be made. What is the standard deviation and the t-test? Only when this pretty basic analysis is conducted may we begin to speculate as to the cause of any “blip” or indeed whether is is just random as compared with any other day of the year.
Were births planned for Feb 14 one would observe a statistically significant increase in births before and after Feb 14. Again, more data needed.
It would be useful to deliniate data from countries/cultures which observe (“Christian”?) Valentines Day from those which don’t. This would eliminate the assumption that St V day is somehow of some importance. North/South hemisphere data also. Holidays mid May? So many factors to consider, and then do other statistical tests (eg, F-test) to determine if they are different in any meaningful manner.
Disclaimer – the anniversary of my birthday is April 1, so I am a bit overly sensitive about these things, my apologies.
LikeLike