3 ms·
This data ignores the common preference each gender has for age. It's as common for a heterosexual man to be 10 years older than his partner, as it is for a man
by iratewizard 5y ago
This data ignores the common preference each gender has for age. It's as common for a heterosexual man to be 10 years older than his partner, as it is for a man to be two years younger. This would be more meaningful if the data was weighted by the probability of a marriage for each given age and gender.
- samatman 5y agoThis was well put, thank you. There are a lot of posters making this general point in terms of common sense and evolutionary biology, but the post we're responding to is about statistics. There is a statistical (and therefore actionable) way to weight the curves, in terms of likelihood of marriage between a man and a woman of a given age. Getting the relevant data might be difficult, but all marriages are recorded, and matching the names of the relevant parties to other records showing their age may at least be contemplated. If anyone has a few months to pull this together I'm sure we'd all enjoy it!
- zdkl 5y agoChallenge considered. Are there unified (federal) marriage & birth certificate online registries for the US or is that a matter for archives?