Right. There's more than 20 western countries. If P is around 5% we would expect the result just from random fluctuation. If you start looking at other factors, such as number of children, when you have children, age gaps between children, etc., let's say 5 different potential scenarios, you would expect to see things that only have a 1% chance of happening. If 5 other groups are doing similar analysis, one of them is likely to see trends that only have a 0.2% chance of happening. I would take this result with a huge grain of salt.