Hacker News Reader

Comment by someguydave | Hacker News Reader

Yes there is a way, you put kids in test prep and see if they do better. They generally do not improve much.

The SAT is an IQ test, and IQ is mostly inherited from your parents.

There is no way the SAT is a purely IQ test, when i didn't get a hot score on my first SAT, i practiced for 1 week straight and substantially improved my scores (like 300 points range). The way i liken this to is if a MLB pitcher throws the same pitch at the same velocity and i know this ahead of time, i could practice for a year and when the time comes i could hit that ball. Does that mean i'm a major league level batter? Heck no, it is the same way with the SAT's, Leetcode and other tests, they measure how much you want it more than raw IQ.

Yes there is a way, you put kids in test prep and see if they do better.

For this to be a real test, you'll need parents of twins willing to send only one kid to test prep..and not send the other one..you know..for science. Doubt there is a parent who would do that.

> For this to be a real test, you'll need parents of twins willing to send only one kid to test prep..and not send the other one

What's the issue with testing someone without prep, then giving them a week to study up, and then testing them again? People take SAT multiple times, you know, and it is a brand new set of questions every time.

I did that myself (took it 3 times total), and I can say this: taking it without prep and then taking it later with just a week of studying on my own for about an hour per day made a massive difference. Even before I got my results back on the second try, I already felt i did much better, as I was second guessing much less and felt more confident about my answers. And the final scores reflected that really well.

However, studying beyond that (which I did for my third attempt, months later) didn't help much and just marginally changed my scores by +-20 in different categories (out of 700 total max in each of the three categories, which is how it was scored at the time). I didnt feel any more prepared either. Judging by my acquaintances, it was a similar pattern. Massive boost in scores on the second try, marginal or non-existent in scores on attempts beyond the 2nd.

I don't think the SAT is an IQ test. I'm sure IQ helps but parents clearly have a pivotal role - the single biggest external factor to explain variation in score perhaps. Surely the biggest factor in SAT scores is how much you the student _care_ about the SAT score. Why do you care? Maybe because your parents care, maybe because your parents don't care? At the end of the day parents are crucial to the success of their children positively and negatively.

This paper found a 0.82 correlation between general mental ability g and SAT scores

https://archive.org/details/ScholasticAssessmentOrG/page/n1/...

That’s relatively low considering what kind of test it is. Elementary school sees a 0.7 correlation between IQ and academic performance.

.82 is an insanely high correlation in any kind of social science

A benchmark of everything across all of social sciences is extremely broad when comparing two different tests of mental performance designed to correlate with other tests of mental performance.

It’s like being shocked that resting heart rate and VO2 Max happen to correlate quite well relative to other health indicators. Yes they are testing different things, but not that different.

> The SAT is an IQ test

In that case I weep for the future, because I’ve met plenty of dumbass people who managed to get a high enough SAT score to get in school.

I also weep for people that extrapolate too much meaning from an IQ test.

(I'm not talking about you).

So you are saying that Asians have higher intelligence, and not just a little bit higher, but significantly higher than other race?

of the ones present in Michigan, yes.

The Asians present in North America are not a random sample. A large portion of them came here specifically to work in STEM or other fields requiring high levels of intelligence and education. It's very probable that the ones here are, on average, significantly more intelligent than both the average American and the average Asian in Asia.

For that matter, the Asians you see in studies of the average test scores in somewhere like “China” aren’t representative of the educational ability of the average rice paddy farmer, but of more urbanized populations.

I feel this must at least be a factor.

Is this supposed to be a "Gotcha"?

Using that logic, everything is an IQ test :-)

Extensive test prep can do a significant amount to increase SAT scores, but the test is designed to minimize it.

People do see over 100 point increase ~20 hours of prep, and extensive as in 3+months can push that to 200 though with diminishing returns at the top of the curve. Which shouldn’t be surprising as increasing speed, even just from not needing to read instruction, provides an advantage for most people.

Reply on news.ycombinator.com