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

Statistical studies have long shown that (1) education outcomes strongly correlate with parenting; especially having two parents. (2) education outcomes have near-zero correlation with funding.

Given this, one would expect this exact outcome. Santayana and all that.

If you want to use money to improve education, I suspect the best use would be to provide parental education as soon as the child is born. It would be nice if someone provided a small amount of money to try that out.

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.

I'm currently in the process of investigating the literature on correlations between teacher pay and student outcomes.

Would you mind listing some studies showing point (2). I'd really appreciate anything you've got handy.

I don't agree with their proposed remedies, but the book "Schoolhouses, Courthouses, and Statehouses" by Eric Hanushek and Alfred Lindseth has thorough documentation of the disconnect between spending and outcomes.


I don't mean to be that jerk, but I'm going to anyway: citations are needed on this: "Statistical studies have long shown that..."

"We"—schools, society, etc.—can't really control parenting. But we can control schools, and per the articles I list here: http://jseliger.wordpress.com/2009/11/12/susan-engel-doesnt-... , it is probably possible to get substantially better outcomes than the ones we're getting now, chiefly through better teachers. At the moment, most public school teachers are paid in lockstep based on seniority—CS teachers and PE teachers get the same pay—and can't be fired after their second or third year of teaching, and that creates a lot of perverse incentives.

""We"—schools, society, etc.—can't really control parenting."

Sounds like you need a citation of your own. I'm sure you could come up with hundreds of ways to apply money towards the solution of improving parenting.

"We"—schools, society, etc.—can't really control parenting. But we can control schools,

I agree with this statement as it is. But I think it's one of the problems.

There have been complaints elsewhere in this thread about the strict work rules under which teachers operate. And I agree with them.

You know the story of the drunk looking for his keys under the streetlight instead of in the bushes where he lost them, saying "the light's better here"?

If we keep on expecting schools to be the magical institution that fixes society's problems, we are going to keep on being disappointed, all the while greatly bothering all the people working at the school in the meantime.

See aaronharnly's comment for a book full of cites https://news.ycombinator.com/item?id=7739206

Hanushek is an advocate for some kind of charter reform, per his Econtalk talk: http://www.econtalk.org/archives/2008/07/hanushek_on_edu_2.h... . I think he would loosely agree with my above points.

> "We"—schools, society, etc.—can't really control parenting.

Programs exist for this with some positive results. Here's one from a quick Google search: http://evidencebasedprograms.org/1366-2/nurse-family-partner...

There are innumerable small-scale programs that show limited positive results, but almost none of them scale up (see http://www.theatlantic.com/business/archive/2007/11/scale-ma... for more on that general problem; her book The Upside of Down is also good on this subject). Lots of small-scale Head Start programs show promise too, but the program's effects on the whole fade out after a couple years, and on a large scale it hasn't done anything except provide daycare and jobs.

Programs like "Nurse-Family Partnership – Top Tier" already operate. I know because I've written numerous Health Resources and Services Administration (HRSA) Healthy Start Initiative (HSI) proposals (see more about the program here: http://blog.seliger.com/2013/12/15/hrsas-healthy-start-initi...) that attempt to do just this. I do grant writing for nonprofit and public agencies, so I see citations like yours all the time. Next time I write an HSI or similar program, I'll cite "Nurse-Family Partnership – Top Tier." Doing so isn't going to make the program any better, because HSI has been operating for a couple decades, under different names, and hasn't accomplished much on a large scale, in part because of the scale-up problems described in the first paragraph.

Ideas like "Nurse-Family Partnership – Top Tier" sound good, but the gap between the real world and the proposal world is quite wide (http://blog.seliger.com/2010/04/11/the-real-world-and-the-pr...). Zuckerberg has evidently learned this.

Apologies for the length of this post, but the issue is a complex one that is hard to explain and harder to understand—hence all the links!

Many thanks for the info - this is a topic I care about deeply so I appreciate your time and the resources you mentioned.

I understand that scaling is very difficult. But what makes you think scaling these types of programs is harder than scaling those that directly target education? In fact, the first link you provide talks about the difficulty of scaling an educational program.

I'd love for there to be strong evidence backing a scalable program that addresses generational poverty, be it an educational approach or not, but I don't know of any. Until we find that, I really don't think it's helpful to limit public discourse to only one potential solution to such a complex problem.

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