Hacker News Reader

Comment by yummyfajitas | Hacker News Reader

There is considerable evidence showing this is more or less true. "Intrinsically unschoolable" is a strong and incorrect statement, but "less schoolable" yields similar conclusions and is well supported by research.

Consider Raj Chetty's work; he shows that the primary determinant of student performance is intrinsic student qualities, but that teachers also matter a fair bit.

http://www.rajchetty.com/chettyfiles/value_added.pdf

http://www.rajchetty.com/chettyfiles/w19423.pdf

Why would you consider this factual claim to be an "incoherent position"?

View on HN

The papers you've cited make essentially the opposite case as your comment does. Chetty's work suggests that a single year of teaching under a high quality teacher raises the lifetime earnings expectation of a student by $50,000.

The papers show large and significant effects from student quality as well, a position you described as "incoherent". The fact that teachers also have an effect doesn't change this.

If you were actually correct, and teacher quality was the primary determinant, then we could skip VAM and just use raw test scores to measure teacher quality.

I.e., if teacher A had a class of underprivileged black students while teacher B had a class of rich Asian students, and teacher B got higher test scores, that would be evidence teacher B was a better teacher. Is this actually your belief?

Can you point out the section of the major Chetty VA study that distinguishes student circumstances from student intrinsic quality?

The context of the thread you've jumped into is what the best intervention to circumstance might be to improve outcomes for students. You appear to be begging the question.

I don't know what you mean by "intrinsic". Kalium said:

The educational experience of students is not dictated wholly by teachers, curriculum, and facilities.

You seemed to disagree with his claim. Due to this apparent disagreement, I interpreted "intrinsic" to mean something other than teachers, curriculum and facilities. The Chetty paper, and the entire need for VAM (as opposed to just measuring raw test scores), shows this to be true. Do you believe this to be false?

Maybe somebody else can restate the point I'm trying to make in a form that will make more sense to you.

Ironically, yummyfajitas definition of "intinsic" to mean "not teachers, curriculum, nor facilities" would thus include life circumstance.

For example, the kid's quality of life outside of school. Which is demonstrably improved by helping poor kids with rent. Which I believe he is trying to argue against.

Is it clear to anyone else that 'yummyfajitas is implying that the kinds of students who attend poorly-performing schools are genetically incapable of academic performance?

I may be misreading. I'd like a second opinion. I'd rather not assume people I'm arguing with hold revolting core beliefs. This would not be, like, the first time I'd jumped the gun on something like this: I myself hold a (probably faulty) core belief that the median HN participant is likely to host retrograde beliefs about race, which is unfair of me too.

I'm genuinely curious why you are asking third parties to put words into my mouth. Why not just ask me?

If you want a summary, I don't share your religious faith that God ensured an identical statistical distribution of genes across all subpopulations when he created man. But Kalium's argument is not dependent on this - there are more things in the world than simply teachers, curriculum, facilities and genetics. For example, poor people might be disproportionately bad parents (perhaps they exhibit many bad behaviors which also make them poor) and this might cause their children to be disruptive in school.

That's a concrete theory that's directly compatible with your secular faith in statistically identical genetic distributions, but still supports Kalium's claim.

My question still stands.

If you do _not_ believe humans possess "identical statistical distribution of [academic performance] genes across all subpopulations" why hide such insight? It'd be interesting to hear where in the gnome the "academic performance gene" sits.

And have you also "discovered" a correlation between this academic performance gene and melanin content? Or gender too? Surely you don't want to deprive us of such important "discoveries".

Reply on news.ycombinator.com