Hacker News Reader

Comment by doctorpangloss | Hacker News Reader

how many elected leaders are in STEM? would high SAT scores and grades exclude many US presidents and congressmen? (yes) winning elections seems kind of important to me. so if you were just like, selecting for leadership - and many of our leaders are brilliant people, just not in the sense of being good at taking tests - would that be good or bad? or... what is your real opinion? what are you actually mad about?

obviously the UC system should give spots to the kids who will use those spots the best. but it is very hard to define what "using spots the best" means.

View on HN

We certainly don't want them to fail out, which is what is happening. Berkeley reported 10% failure rate in the intro CS course and 35% in the pre-intro course. https://www.dailycal.org/news/campus/academics/failing-grade...

> not in the sense of being good at taking tests

The trick to doing well on the SATs is to pay attention in class.

That's not how it plays out in practice. There is overwhelming evidence that students who otherwise excel academically score fairly mediocre SAT scores on their first attempt and then jump substantially after weeks of targeted practice and/or tutoring, even though they didn't learn anything new in the classroom.

If attention in class were all it took then that improvement couldn't happen. What changed was familiarity with the test, not classroom focus.

I'm not aware that there is any method to dramatically increase SAT scores (and neither IQ test scores). could you point me to your sources?

Straight from the College Board themselves:

https://research.collegeboard.org/media/pdf/osp-technical-re...

That report also references another study showing that each hour of tutoring was associated with an increase of 2.34 points on the SAT score which unfortunately is behind a paywall:

https://www.researchgate.net/publication/282492223_Preparing...

I guess times have changed. When I took the SAT, I did zero prep work. Nobody else I knew did prep work, either.

Paul Graham recently posted SAT advice along the form of "when you finish the test and have more time, go back over the test and check for mistakes."

I was kinda astonished at this advice, isn't it obvious? A strategy I also employed was to do the easy problems first, so I don't miss a question that would have been easy. Apparently this has to be explained to people?

I suppose prep work would be fine for the students who didn't pay attention in class.

Yes, if you took your SAT among a cohort of people where none of you practiced for the SAT, then what you're saying holds true.

That's not really the case anymore. Top tier students nowadays prepare for the SAT, they don't go into it blind and haven't done so for the better part of 20 years.

I didn't get a perfect score, but it was good enough to get what I wanted.

Well, then what if someone with a better score got the job you were going for, even though they are not smarter, just prepared more for the test?

I'd get the jobs I wanted by bringing samples of my work to the interview.

An interview ... with scores like that?!

The trick is to prepare specifically for SATs.

One would say the tests (and job interviews) should have been designed with the original intent of testing candidates AS IS, i.e. preparing specifically for such tests should have been considered as cheating... But at some point it turned into prep gymnastics, and measuring how desperate the candidates are.

Once it is a forced ranking over all the candidates - it becomes a test of preparation and execution not of underlying skill.

In the US or China?

Reply on news.ycombinator.com