Hacker News Reader

Comment by silvestrov | Hacker News Reader

The kids were a self-selected group because attending the school required that the parents had to do something active to get their kids into this school.

They try to control for this by comparing those kids to kids who lost in the lottery. But this only tells us about kids who have engaged parents. You cannot generalise to kids who have non-engaged parents.

Having engaged parents is the single most important condition for getting a good education.

Secondly, not having disruptive kids in class.

And only thirdly are quality of teachers.

If "free dental, medical, and mental health services" were all it took, then no school in Scandinavia should have any problem with poor kids. But we do have problems because some parents don't care at all about their kids education.

View on HN

> If "free dental, medical, and mental health services" were all it took, then no school in Scandinavia should have any problem with poor kids. But we do have problems because some parents don't care at all about their kids education.

It is probably true that those aren't sufficient, but things like getting enough to eat and not having serious, untreated dental or medical problems are basic prerequisites to a good education that aren't met for a lot of American children.

Also worth considering that the vast majority of parents who don't care simply can't afford to care. Working 2+ jobs, they may not have the time or emotional effort to invest in the same way that more privileged parents with elaborate PTO schemes have access to.

Free dental & medical (Medicare for All or similar) would go a long way to alleviating the pressure on parents and allow them the space and time to care. As would stronger labor laws to ensure all workers have access to a reasonably flexible working environment.

As De Boer puts it, if there's any possibility at all that an education result came from sample bias, it's definitely sample bias.

The standard narrative for "better schools" is that a trial program works great, everyone gets excited, and then it declines to mean the instant it's expanded. If they had lotteried people in entirely at random and still gotten better outcomes, I'd at least be curious.

https://fredrikdeboer.com/2017/03/29/why-selection-bias-is-t...

Essentially this.

It's not really the schools or teachers - they are generally decent. Some better than others ... but teaching is not rocket science. Basic education is not some magical, mysterious thing. It's mostly straightforward, and a little mundane.

If you have decent students, prepared to learn a bit - which mostly comes from a reasonably positive and stable home - then all you need is basic things to get them to a decent level of literacy.

Surely some systems are better than others but I think it's far more about homes and communities than it is dumping 'bad teachers' and certainly not about making teaching 'competitive'.

>>> but teaching is not rocket science

I actually think teaching children is a really difficult and important part of society but I think many people don't value the contribution incredible teaching can make to individual students lives.

A great school and teacher obviously can change the lives of children in poverty but it's astonishingly difficult and requires immense talent to achieve. Not commenting on the quality of this study however.

I taught high school for almost 10 years and I agree with both of you. You don't have to be a genius, some of the best Algebra 1 teachers I worked with couldn't teach calculus because they didn't understand it. Getting the students to understand the material is generally the easy part, their brains are highly receptive. If a student wants to learn the subject, you can basically just sit there and answer their questions. Engaging and motivating them to want to do it is the part that takes a lifetime to master.

I didn't say it was easy, or that it wasn't hard work, or that I didn't value it.

Teaching is 'hard work', it requires educated, dedicated people, and it's very important work.

But it's not rocket science. Doing it is not risky - especially if you're teaching something that's been taught a million times before. Most well educated people I think could teach at least one subject, though they may not have the wherewithal or stamina to stick it out, or may not have the empathy/social skills to do it well.

> Doing it is not risky - especially if you're teaching something that's been taught a million times before. Most well educated people I think could teach at least one subject

I think this is something that many well educated people think, right up until they are placed alone in a class-room of 30 children who aren't desperately interested in what they have to say.

The five year teacher attrition rate is under 20% based on this study: (https://edsource.org/2015/half-of-new-teachers-quit-professi...) That's rather low and includes many who don't like teaching even if they have the skills. Which suggests most people really can do it.

Granted, that's not the first time these people have been in a classroom, but collage Education degrees don't see high attrition either. Also, "97 percent of teachers who earned more than $40,000 their first year returned the next year, compared with 87 percent who earned less than $40,000."

You are correct. http://blog.mrbigler.com/2016/04/23/those-who-can-do-those-w...

Being a schoolteacher is risky - kids kill themselves over things that happen at school, or they metaphorically throw their life away with other stupid actions. A good, engaged teacher can help to steer children through that minefield. It's not just about teaching a subject.

The idea that one only needs to be good at a subject to teach it well is bunkum.

It's not rocket surgery, it's potentially far more important.

Most well educated people I think could teach at least one subject

And in essence you have hit on the problem. Everybody thinks they can be a teacher. And for some value of $teacher they can, but it's hard to be really good.

You might watch LeBron James play basketball and think "I can do that," go out on the court and take a few shots, and think you're pretty good. But you're likely not.

This is why assessing teachers is soooooo important. We need reliable, verifiable ways to separate the good from the bad.

And yet, assessing teachers reliably is a very difficult task.

As a teacher, the moment that a reliable assessment method is introduced there will be significant push back; however, it will also be the key to broad improvements in the education system and the further professionalisation of teaching (and, by extension, teacher retention/compensation).

I also believe that this starts by enabling classroom teachers to assess their students with the reliability of standardized tests but within their individualized curriculum. That's not an easily ask - but it is an exciting concept that I return to every few months.

I disagree a bit here, but I do have a limited understanding of the history of education. US high schools lag significantly behind international averages [1] across almost all metrics, and are used for socializing more than other other countries.

It's so easy for us to blame teachers, or to wishfully think they could turn it around on their own, but the approach has to extend beyond the classroom. Non-school factors are estimated to have 4-8 times the impact as teacher quality. [2]

We should work on removing vocational program stigma and drop the illusion that everyone should aspire to a 4 year degree, even if in disciplines with terrible job markets, to "succeed".

I really wish there was some way of incentivizing parents to be more involved, or moving to year round all day education to mitigate their impact.

[1] http://www.slate.com/articles/life/education/2014/02/high_sc...

[2] https://www.rand.org/education/projects/measuring-teacher-ef...

"US high schools lag significantly behind international averages"

This is a red herring - to the point where it is misleading.

A) Start with US 'students students' are behind maybe instead of schools. Because the comparison scores are PISA and they measure people not institutions.

But more importantly:

B) It's ethnicity. (And I'm not attributing anything to race here).

'European Americans' have been doing the same for the last 40 years. They are doing a little bit better than Europeans from Europe.

'Asian Americans' have been doing the same since we've measured them as a group, they do a little bit better than European Americans, and tend to do a little bit better than Asians from Asia.

African Americans fare the worst, Latino Americans fare better, but both groups have been doing a little bit better over time. They do a little bit better respectively than Latinos from South America and Africans from Africa.

Demographically, there are considerably more Latinos, and a few more Asian and African Americans in the country - which tilts the data.

So you see: any group from anywhere in the world actually fares better in America than they would in their place of origin. Which is really a different way of looking at it, no?

Because of the social and economic disparity between these groups, it lends well to the notion that it's about the students and not the schools.

Small but slightly less relevant note: the top American students tie the top UK students as being among the best in the world. The US/UK have a little bit of an elitist function going on.

So consider this: I'm Canadian, always proud that 'we fare better than the US' on PISA testing, but really, once you normalize for ethnicity (we don't have large groups of Latinos and African Americans) - then we do about the same.

And of course I should add I don't think this is racial issue at all, however I do think that the social + economic status of any group is tied to things like outlook on education, history of literacy in the family yada yada.

All of this is based on PISA scoring which is available on their site and other places.

I saw something like this years before, but have absolutely failed to come up with the source of the data. Do you know how I can get my hands on it? If it's true and honest then we need to focus on the real problems- at least the aspects that we can solve.

Other than the PISA scores comment - I agree with everything you've said. It's a multifaceted problem and, in many ways, the problems of schools are not strictly educational problems but the broad problems of society.

That said, educators are the tip of the spear - and I do believe that there is opportunity for additional technological mediation in improving the teacher's professional practice. It could also obsolete significant chunks of our multi-billion dollar standardized testing industry (a benefit to both the operating enterprise and the budgets of school districts).

The way we measure educational attainment across cultures is by PISA scores.

The whole basis of 'US is doing worse than OECD or Greece' or whatever is based on PISA scores.

It's very crude, but it's what we have.

My concern is that without some form of broad standardized testing, we won't have proper metrics to measure the success of programs and/or teachers against.

I agree (though I challenge the utility of current standardized testing to fulfill that job-to-be-done). My position is that we are already paying tens of thousands of teachers to assess the ability level of their students everyday while maintaining a secondary assessment industry that is able to exist because teacher's assessments are not able to be trusted. If we can make teachers' assessments trustable (for ranking/sorting/certification of skill & knowledge) then we won't need the very expensive secondary assessment industry to nearly the same degree. Likewise, many of the negative externalities of standardized testing could be mitigated by ubiquitously assessing student ability rather than relying on traditional "snapshots" as is the status quo.

I agree. If we ever get there then this will be the way that top teachers finally start making the money they deserve. The caveat, of course, is that they will be in limited supply so we'll need a way to extend their influence.

Yes - but we will actually have a tool upon which to make effective evaluations and offer individualized learning goals for teachers. Understanding my weaknesses is the first step in strengthening them (or leaving the profession if I can't).

The difficult part is being able to inspire the kids to be interested in the subject; the subject matter is fairly boilerplate. Wanting to learn a subject is much better than being forced to learn a subject.

In other words, it's much more about soft skills than hard skills. Think of a good speaker. Two people can talk about the same subject, but one leaves you with the inspiration and curiosity to learn more, the other makes you want to break for lunch.

Of course, you have to get rid of the disruptive kids. US schools are so scared of getting sued, the principles just bounce them back to the classrooms. When I was growing up, they had a school for bad kids so they wouldn't disrupt the ones that were there to learn. They closed that program down. My state is in the lower half of the school system.

Money also matters. You can't each in a classroom full of 30-40 kids at the high school level. Teachers can't be spending their money on school supplies the state won't budget for. Earning a college degree for a profession that doesn't pay very well isn't a great option, so fewer people do it.

For something so critical to our civilization, it seems many states treat education as a cost center.

Disclaimer: Both my parents were teachers as were many in the family. I am not, but if things were better, I certainly would consider it. I'm a programmer by trade.

The risk is turning away young minds.

> A great school and teacher obviously can change the lives of children in poverty but it's astonishingly difficult and requires immense talent to achieve.

Teaching children in poverty is a specialization. Teachers can excel at teaching children at upper middle class schools but fail miserably at an inner city school. People complain about privilege on teacher quality difference between inner city and good schools. The truth is that the "good" teachers at an upper middle class school wouldn't be much of an improvement over the "bad" teachers at inner city schools. The "good" teachers only know how to teach well-behaved, studious students!

That is why integration failed. The environment at good schools are good for children with familial support and low behavioral issues. It is not good for children with behavioral issues without familial support.

Is there empirical evidence for this?

There are many, many schools that are not decent. There are plenty of others that are 'decent' but have no resources to support any student who has even a minor learning deficient. In order to pull themselves up and compete on an even playing field with the rich/elite kids the majority of students need more than a 'basic education'.

Yup. Good teacher don't scale.

My son is in a charter school the controversial Success Academy which do really well (top 3% in new york state) even with kids who are homeless and living in shelters.

But this is because

1) Parents have to apply for the lottery and if they get in they have to be engaged 2) School is run almost like a Catholic school in how strict it is (but with a much more effective reward/punishment system) 3) Teachers aren't part of the union which means they can fire bad teachers. They are also fairly young.

This is a very different school than the one I attended in Denmark where I grew up but I have to say it's working remarkedly well.

My hope is that the school will provide my boys with an understanding of self-discipline and then I will take them to a more free-form school (like alt school) in the second part of their educational lives so that they can benefit from the freedom to explore and find their own perspectives using their basic skills (reading, writing, math & music)

Have no idea if this will work but that's how I am going to do it.

> They try to control for this by comparing those kids to kids who lost in the lottery. But this only tells us about kids who have engaged parents. You cannot generalise to kids who have non-engaged parents.

> Having engaged parents is the single most important condition for getting a good education.

If that was true then there would be no significant difference between the students in the study, since all of them had engaged parents.

Your claim is not true. If you factor out the most significant variable, some other variable becomes the most significant one.

Yeah, maybe more expensive teachers, for example.

Or more talented kids.

Richer or more connected parents maybe

> there would be no significant difference between the students in the study

I didn't say this. I say that you can't generalise to students NOT in the study, i.e. students with parents who didn't care to participate in the lottery.

Of course you can. You can control for the factor/variable you said was the "single most important", which is done through the design of the study, by your own admission. And then look at results from the other factors.

Whether your factor is important or not doesn't even enter the picture.

Just because something is the most important condition doesn't make it the only important condition.

But it makes it the most significant condition. It was in the stated initial premise.

And if they're all important (according to GPs list) then maybe more expensive teachers, better properties/real estate, supplies, etc is important enough anyways.

> > Having engaged parents is the single most important condition for getting a good education.

> If that was true then there would be no significant difference between the students in the study, since all of them had engaged parents.

This does not follow. Just because something is the most significant condition doesn't mean that two things meeting that condition will have no significant differences.

EDIT: For example, let's say breathing is the single most important thing to do in order to stay alive. That doesn't mean eating isn't important to do as well. If two people breathe, but only one of them eats, they will have significant differences in outcome even though they both met the single most important condition.

Sure, but OP ranked them specifically to downplay the effect of teachers (and other factors caused by wealth).

You can't have it both ways.

Either engaged parents is significantly more important, or the other factors are still very important.

And this discussion is entirely sidestepping the fact that OPs list is entirely arbitrary.

I think this is pretty simple and folks are talking over each other.

Involved parenting is a blocker issue, not just a critical one. There may be other critical issues in the queue, but those are blocked on that first one.

So the other issues being resolved may not matter at all if the first isn't completed first.

That's more or less how I see GPs argument. My rather diverse education background certainly lets me see that hypothesis as absolutely true - but I have my own personal biases due to experiences.

Man, when I listen to my mom tell me all the things she does for my sister's kids, who she's raising, it boggles my mind, and makes me a little jealous.

With all she's doing, there's no way those kids can't not succeed in school, yet it's a level of effort that really precludes having a career. Engaged, stay-at-home parents amazingly raise the quality of life, not just of kids, but of entire neighborhoods and it's one of those things that you can't fake or buy cheaply.

This comment is gold. Sample bias might be at play here and still that is encouraging.

Poor people do not love to be poor and are actively looking to get out of poverty. One of the ways they do this is by taking unusual risks and imitation. (Trust me I was poor). Public education system takes away both these tools from them. The school you will go to will be determined based on where you live and nothing else. Because of this stupid constraint you lose a very critical market signals that other schools could have used to filter out better kids and nurture them better.

I think poor children should have access to charter and private schools too but those schools should have 100% freedom to discriminate. Say for every certified poor kid that is admitted to private/charter school government gives a $10K voucher and you can admit only at max 25% of total capacity with such students. The poor kids and their parents line up outside the school. They have to prove that they are worthy of the seat by showing things that matter to the school. E.g. my kid is not a bully, parents of the kids have no criminal history, that parents are happily married, parent doing 2 jobs, parents have not done drugs etc. At least bare minimum parents willing to take efforts to enrol them into better schools and doing research of which schools are better.

Assuming it is the case that quality schools will only work for the subset of kids with engaged parents, that is still a huge net positive.

We should make every school high quality. I'm not sure why anyone is opposed to this idea.

>Secondly, not having disruptive kids in class.

If that turns out to be the 2nd order thing we can make that happen with sufficient effort too.

> We should make every school high quality. I'm not sure why anyone is opposed to this idea.

In the abstract I don't think anyone is; what's contentious is what measures make a school "high-quality."

And what the cost is. We can reduce traffic fatalities to zero by lowering speed limits to 25 and building cars like tanks, but we don't. Why not? It's not a reasonable expense to achieve the goal.

Everyone wants to do better. Just because people don't agree on means doesn't mean they are cold hearted bastards or something.

==We can reduce traffic fatalities to zero by lowering speed limits to 25 and building cars like tanks, but we don't. Why not? It's not a reasonable expense to achieve the goal.==

Except we have lowered speed limits and mandated security measures on cars, just not to the arbitrary level you stated. We followed scientific studies that showed positive results of lower speed limits in certain areas and implemented car safety requirements for manufacturers.

Right, but 40,000 people still died from traffic accidents in the U.S. in 2017. To get that number to zero, the U.S. might have to drastic things that probably wouldn't be a good tradeoff. The existence of a discussion about tradeoffs and costs doesn't mean that some people are anti-safe-driving or something.

Sure, but just because you took an unrelated example to the extreme to show that everything can't be done doesn't mean nothing should be done.

With car accidents, a problem was identified, scientific research was consulted to craft solutions, they were implemented, the numbers improved. This is likely because everyone, regardless of income level faces the risks of driving. Bad schools can be avoided by having more money and living in an area with good schools.

I appreciate the first point which I believe is a vital factor. Regarding the order of the second and third though, wouldn't the quality of teachers determine whether there are disruptive kids in the class in the first place?

Not likely. My sister teaches in a magnate school in a poor urban U.S. district.

The amount of bad behavior with which she must contend is shocking. It's entirely foreign to me from our childhood, and she has relatively little freedom in how she can handle chronically problematic students.

If it is theoretically possible for the teachers to keep the disruption level low enough for much learning to happen, then the formula is a secret.

It's very depressing, and it has left many teachers bitter, burned out, and I suspect more racist.

Yes, I see what you mean - it's easy for the collective level of class disruption to overwhelm the teacher, however good he/she may individually be.

> Not likely. My sister teaches in a magnate school in a poor urban U.S. district.

Couldn't we draw the conclusion then that wealth plays an important role? Which in turn also affects the pay of the teachers in the area, also affecting the quality of said teachers?

> It's very depressing, and it has left many teachers bitter, burned out, and I suspect more racist.

I'm not disagreeing with what you're saying, but what I think is depressing is that it leads to racism, instead of anger over systemic inequality.

The teachers who can't hack it, or don't care enough to stay at the bad schools leave. Good riddance, right? Most of the teachers at bad schools I've met are good teachers.

Wealth is highly predictive of educational attainment but I think there is more going on than just how much they can pay the teachers.

Disruptive behaviour is usually due to external factors I imagine - are you struggling with your parents' divorce, are you over tired from gaming all night, are you on drugs, suffering teen angst, ...

Disruption because the class is above or below your level is also, to some extent, uncontrollable by the teacher - they have 30 kids (typically in UK) to cater to.

Teachers can be entertaining, but that isn't always conducive to the subject matter.

There was a good blog post and discussion about your point here on HN 9 months ago.

https://news.ycombinator.com/item?id=15007942

Outside the US most public school systems are vastly more successful narrowing the education gap between rich and poor students.

Reply on news.ycombinator.com