Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

That's true, but at the same time it's a standardized test so anyone can replicate those prep courses. There's tons of prep books with practice questions, tips, etc.

It's a matter of awareness and taking it seriously (obviously a simplification, but hope you get my point).



I mean, it's not really true. People spend a lot of money on test prep, but this isn't even a case where the evidence hasn't come in yet. We know perfectly well that test prep has negligible influence on SAT scores.


Those studies were mostly funded and influenced by the college board, and they needed that to be true, so that’s what they found. But there is prep and there is prep!, they are different with different influences. Studying the former (say a few weekend long prep course) to make conclusions about the latter (intensive cram school for a couple of years) is criminal.

I’ve seen an SAT cram school in China, and yes...it will influence your scores significantly because they pull out all of the stops.


> Those studies were mostly funded and influenced by the college board, and they needed that to be true, so that’s what they found.

The SAT draws a huge amount of overtly hostile attention. So does SAT prep. But these highly motivated SAT-hostile researchers haven't been able to produce the results they need.


That simply isn’t true, eg see https://www.washingtonpost.com/news/answer-sheet/wp/2017/05/...

The College Board can’t really have it both ways. And of course: https://www.reuters.com/investigates/special-report/college-...

Which is just one example of pulling out all the stops.


Yes, no one disputes that memorizing an answer key can give you a higher score. But by the same token, no one believes that the influence of cheating is of interest when studying potential gains from "test prep".


That isn’t just it though: they have access to old tests, they have access to test takers who memorize questions, given the importance and popularity of any test, these options will arise and people with means will have access to them.

There is no avoiding Goodhart’s law.


You can make a case for a much weaker form of Goodhart's Law, that with the SAT as a well-known target, SAT scores are less informative than they would be otherwise.

But "there is no avoiding Goodhart’s law" is obviously wrong if you interpret Goodhart's Law as it's actually phrased; the SAT is doing a fine job of avoiding it today. The SAT hasn't ceased to be a good measure. It's still very informative.




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

Search: