Hacker News new | past | comments | ask | show | jobs | submit login

It's really really hard to figure out which correlations mean something. If the New York Times says tomorrow that wearing knit caps during pregnancy makes your baby smarter, then all the high-conscientious women who are high-income to be in the NYT's readership will wear knit caps. 20 years later when you measure, sure enough, there is a correlation between knit caps in pregnancy and educational outcomes!

You have to work hard to correct for all these effects. For example, is there a difference between : single mom because mom never married; single mom because parents divorced; single mom because dad died on a business trip?




All good points; which is why I was careful to call it correlation instead of causation. For example, distance from the Canadian border is also well-known to correlate with education outcomes (with larger distances correlating with worse outcomes).

That said, by far the strongest correlation is with dual parents. I don't know if anyone has studied the details you ask for, but it does stand to reason that a single parent -- of either sex and regardless of cause -- has a much tougher life and therefore less time to spend one on one with their child.


It's easier to figure out if correlations mean something if they make intuitive sense. We're not talking about knit caps, we're talking about the notion that kids are more likely to be successful if they have two parents raising them instead of one, all other things being equal. This is borderline common sense, just like it is also borderline common sense to suppose that kids having parents who are uneducated drug addicts are less likely to succeed than those who have successful educated parents.


The knit cap is simply to point out the ridiculousness of the problem. Breast-feeding is a common issue where there is the "intuitive sense" but it's really a "just-so story." There's "well, it's just common sense" theories that turned out to be useless when studied by smart people that weren't blinded by "well of course we will find some effect from this" bias.

Seriously, compare the life outcomes of single-parent households by breaking out the reasons for single-parenthood. http://www.clasp.org/resources-and-publications/states/0086.... There is a lot more going on than "no dad -> sucky results."

Zuck needs to figure out what the real causes are before pouring another $100,000,000 down the drain.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: