Hacker News new | past | comments | ask | show | jobs | submit login

In principle, SAT scores enable people like me, who were very very good at such tests, to break into elite schools.

In fact, test prep works well enough to enable a shockingly high percentage of legacies and donor brats in elite schools.

Ending SAT use in admissions might enable a more balanced approach to admissions, but it does not broaden access to elite schools. They could increase class sizes, or they could get rid of legacies and donor admissions.




Standardized tests are designed to be as non-gameable as possible. They're definitely less gameable than the other admissions criteria still used in elite colleges: admissions essays, playing obscure sports, fancy extracurricular experiences, etc. Most studies find SAT prep confers very limited benefits to the average person.

Having an essay essentially written for you by an expensive consultant with contacts in admissions departments is undetectable and widely practiced. But when it comes to standardized tests, many wealthy families get so desperate for good scores for their kids that they engage in criminal conspiracy (recall the Varsity Blues scandal in 2019).

It's extremely hard for me to see how, when you remove the least gameable aspect of admissions, that it gives the rich less of an advantage. I'm almost certain it gives them more of an advantage instead.


The SAT is way more gameable than the College Board would have you believe. Most coaching is not very effective, but there is a group that is very effective. It’s almost by design that they aren’t very popular.

That said it is less gameable than most of the rest of the process. Family wealth and income are probably less gameable, but many schools are need blind.


Are you saying there is some particular training program that increases scores much more than the 30-90 points claimed by various studies?


Yes. And I’m going to be coy and not tell you much about it. So you’re right to be skeptical. In a few years I’ll speak more freely on it.


Skeptical is putting it mildly. I'd give you better chances of having a secret proof of the Riemann Hypothesis than producing miracle educational results at scale (even if it is just on a standardized test).


They are being so vague that’s it’s impossible for us to know, but they didn’t say “at scale”. You’re certainly correct that there isn’t some secret sauce that we could just apply to everyone. But it wouldn’t surprise me if there was a service that costs (high) 5 or 6 figures per year that could substantially increase test scores.

Though I would imagine such a service would have to take place over months/years and not days/weeks, and that kind of blurs the lines of whether it even counts as test prep at that point.


I’d be just as skeptical. It’s something that would require strong evidence and I’m providing you none.

I’ll be around on HN in five years. Ask me then.


If it's real, you might as well tell everyone and I'll hear about it that way - I am certain you'll be very popular.


To be clear, the plan is to not just tell you. :-)


Can you link published research indicating that you can increase SAT scores through training? The research I have read says the opposite.


14% of Harvard undergrads are legacies. It is exceedingly unlikely that SATs help a higher percentage of deserving but otherwise overlooked applicants.

My point is, there is a way to have a far higher impact on opening seats for kids who should be getting an elite education than the question of whether SATs help or harm attaining that goal. They're not really serious until they don't preference legacies.


> 14% of Harvard undergrads are legacies. It is exceedingly unlikely that SATs help a higher percentage of deserving but otherwise overlooked applicants.

I don't follow. Are you implying that the first sentence is evidence for the second? Do you think legacies are getting in via high standardized test scores? I don't think that's true. If standardized test scores were a major determinant of admission, these schools would be 2-3x more East Asian than they currently are. And virtually none of those students would be legacies. The Harvard admissions case showed all this with evidence from Harvard's own admissions data.

Additionally, the article cites a study saying that SATs do in fact help "deserving but otherwise overlooked" applicants.

> My point is, there is a way to have a far higher impact on opening seats for kids who should be getting an elite education than the question of whether SATs help or harm attaining that goal. They're not really serious until they don't preference legacies.

I agree with you that elite schools do not explicitly prioritize the talented poor. My impression is they prioritize (1) the powerful, (2) those who possess or can afford expensive class signals, (3) those who will donate money, (4) racial balance, (5) high achievers. Standardized tests allow students who can't get in via Categories 1-4 to sneak in via Category 5. Even these students tend to be of higher socioeconomic class, but to a lesser extent than Categories 1-4.

In my opinion, the only way to do better than standardized tests would be (standardized tests) - (all the non-academic mechanisms that privilege the rich) + (explicit priority for the talented poor). But these second and third terms will never happen, so standardized tests are better than nothing.

It is hard for me to understand those who support standardized tests being discarded along the way to a more fair admissions process. As you point out, there are much more obvious ways to achieve that first.


All dropping standardized test does is grant them the ability to discriminate.

There are going to be special cases and examples of unfairness in any system, and so you can find cases where a standardized test resulted in an unfairness, but the alternative is only 100x worse.

The alternative is not "Without the burden of these inflexible uncaring mechanistic tests we can serve each precious flower better and leave no one out."

That is only the sales pitch.

The alternative is actually nothing but "Now we can discriminate."


I sometimes wonder if rich people pay tutors to test prep their stupid children simply because they don't want their kids knowing they only got into a good school because of corruption.


There’s no amount of test prep that will get a “stupid child” to a 1500+ on the SAT.


To get into Jesuit High in New Orleans, you need to have parents with money, no criminal record, and a pulse. The average student ACT score there is 31. That's the 95th percentile nationally. Those kids get great educations and have involved parents, but they aren't averaging 31s due to merit unless you're counting their parents' wallets too.


That looks to be about the equivalent of 1400 in the SAT. An ACT of 31 is more aligned to Tulane’s admitted class than Harvard’s. (Tulane is a top 50 university, so still “a good school” relative to the entire population of 18 year old high school grads.)


I’ve taught at Tulane and I promise you some of these 31 ACT kids can’t add fractions or understand how exponents work, leave aside integrals.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

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

Search: