Sufficiently prestigious schools are drowning in perfect applicants. You don’t even need to admit any rich slackers. Just let some of the qualified substitute money for luck.

> Sufficiently prestigious schools are drowning in perfect applicants.

This is by design. There exist tests that can distinguish between the merely very good and the outstanding. Elite US universities go out of their ways to not use them, precisely so they can admit the merely very good legacies.

I suspect the schools are trying to curate their student body so it doesn't turn into a freak show, which would itself reduce prestige. It would be like a zoo that contains exactly one species of animal. This is the dilemma. Taking the top 1000 scorers on some particularly difficult exam, and putting them all together in one place, would be a freak show. Not because they are freaks per se, but because any monoculture of an extreme trait will turn itself into a freak show.

Also pure test based admission ends up being something around 75% asian, 15% white, and 10% others. See Stuyvesant for example[1]. For some reason Harvard feels like being 75-80% Asian would be bad for their brand so they got sued for discrimination. But Harvard is Harvard and it's hard to win a lawsuit against an institution that is that well connected.

[1] https://en.wikipedia.org/wiki/Stuyvesant_High_School#Demogra...

As someone who went to one of those “pure test based” NYC schools: this is not the testament that you think it is.

The number of Black and Hispanic students at Stuyvesant and Bronx Science was an order of magnitude higher in the 1970s; the current demographics are a consequence of paid preparation being a reliable way to score highly on the SHSAT.

It’s amazing that people don’t realise this. I always think back to the study that showed the amount of books in a household was a strong predictor for IQ. Background matters a whole lot, but high performance people don’t want to hear it, because they want to feel like it’s all their own credit.

Nope. Just like money, there’s also an intellectual family bank that gets inherited. And it’s not genes, it’s access to knowledge which is purchased by capital.

It can be even more basic. Just having your own quiet space makes a huge difference to study outcomes. Someone living in a tiny apartment with a shared bedroom, the TV next door, and constant street noise is going to have a much harder time concentrating than someone with a private bedroom at the top of a detached house.

Of course genes play a part, but eugenics is mostly economic and political discrimination, not objective assessment of absolute potential.

>the amount of books in a household was a strong predictor for IQ. Background matters a whole lot...

That doesn't follow at all. IQ is correlated with parental IQ, due to genetics. Equally plausible, if not more so, than "books cause high IQ" is "high IQ causes books". Smart people a) like to read, and b) have smart children. I don't think it's likely that kids are boosting their IQ scores by cracking open their parent's book collection.

Then you’re ignoring the whole study for Intellectual convenience.

Well, you didn't provide a link to the study, and people mistaking correlation for causation is a frequent thing even in science (replication crisis).

The hypothesis "smart people are likely to buy more books, and smart people are likely to have smart children" sounds quite plausible to me. Does it seem implausible to you?

The best way to convince me that books directly increase IQ is to donate a ton of books to kids who need it, and measure the gains. (As a side effect, you would probably win a Nobel price if that worked.) My experience with charity says that actually many people are willing to donate books to charities; charities refuse them, because they already have enough books and the kids are not reading them anyway.

There are many ways how poverty can hurt intellectual development. This is not one of them.

There are many people who want to get rid of old books for kids, because the kids grew up and don't need them anymore; if you asked them to donate the books to you, you would be doing them a favor -- saving them from a dilemma between the bad feeling of throwing the book away, and the wasted space at home. If any NGO would bring a truck and say "please give us the books for kids you no longer need", the truck would be full in an afternoon. Then... according to your theory, the truck could go a few blocks further and turn hundred poor kids into Einsteins. I would be so happy if that worked! Unfortunately, it does not.

Was that study done before everyone had an internet connection?

All studies ever have been done before everyone had an internet connection.

So is household book count even a valid metric any more, since quite a lot of the world has more information at their fingertips than ever before?

Book count being a predictor isn't the same thing as book count being the cause. It is likely that instead, book count is an easy to measure attribute of a household which has many other correlative factors that contribute to success.

Perfect students are not “freaks”.

On the other hand, a Mensa gathering (not a member, myself) is one of the few situations I've been in that was hard to characterize as anything other than freakshow on earth. Shows that you can maximize a single dimension so much that you are left with people incapable of communication.

Similarly, an organization of students maximizing academic performance has a certain vibe that definitely doesn't feel "perfect". It doesn't even seem like that useful of a skill, but I'd have to know the purpose of a uni to comment further.

Mensa is for those who are insecure about their intelligence and like the idea that they can easily rank themselves above others with a number designed to check if you're mentally handicapped. I'm not surprised they can't communicate.

I always thought to myself, what if I take the test and pass it? Then what?

If I fail, that's easy. I lick my wounds and go home.

But if I pass? The shape of the distribution -- basically slicing off one end of a bell curve -- means that in all likelihood, I would be at the lower end of the IQ scale within the group, even if at the upper end within the overall population.

So I'd be taking a test and paying money for the privilege of being the dumbest person in the room. It hardly seemed attractive to me.

If you're talking about the SAT and ACT, I would challenge the idea that they can accurately distinguish between the good and outstanding. At most they provide a rough indication. I know this because I got a much higher score than I deserved to on the ACT because you can read the test pretty easily if you put yourself in the shoes of the people making it (I got a 31 overall and a near-perfect 35 on the science portion without studing)

I'm specifically not talking about the SAT and ACT since they're widely used by elite universities. Those tests top out way too early and can't reliably distinguish between the top 1% and the top 0.1% (The fraction of perfect scorers is much smaller than 1% but the standard error of measurement is way too large). The exams I have in mind are those like the science olympiads and the Cambridge STEP. STEP in particular is basically used for this purpose (olympiads are a slightly different beast).

Are you trying to get students that can pass specific kinds of tests or students that will be successful? The two aren't strongly correlated. I'd argue successful builds more prestige for the university. And let's be real, everyone knows these tests are purposefully tricky. In the real world people don't give you multiple choices and try to trick you into picking the wrong answer and then laugh at you when you picked the one try tried to get you to pick (politics aside). In classes, unless your professor is an asshole, they aren't trying to trick you. Success depends on a lot of factors well beyond what these tests measure. This is why SAT and ACT don't measure success or even academic success (as in they don't strongly correlate [0][1] with success. This source should decent correlation with first year grades but not second).

I'm not sure why one would even think an academic test could determine success. One of the biggest lies many of us have been told is that if you're just smart enough you'll be successful. Sorry, but that's not how it works.

[0] https://edpolicyinca.org/publications/predicting-college-suc...

[1] https://www.latimes.com/california/story/2019-12-22/grades-v...

At the extreme end of the spectrum with quant hedge funds this actually becomes somewhat true. They expect strong academic understanding of math and make a killing by hiring those who do.

I have some useful background here that leads me to strongly disagree with you. I did well at these international science olympiads in high school. To do that I studied for the particular shape of the exam questions. I became friends with many other people who were high performers at these olympiads. Later on I got a PhD from a prestigious institution and I am now a scientist at another one. Throughout this time, I was very involved in the pedagogical work at these institutions.

One very important piece of insight I gained throughout that time is that after you use some exam to get the top 25% (ish) percent of students, the exam performance really does not matter: it does not correlate with scientific output or originality of work. If anything, I had to unlearn some of the skills that made me good at olympiads because they were severely limiting my creativity. At an olympiad you know there is a solution, while true research problems might be unsolvable and need to be approached differently.

TL;DR: Exams (selective or not, hard or not) are great at giving you the top 20% of students, but they are inherently terrible at telling you who among the top 20% will be a productive scientist or engineer.

Your anecdotal evidence is contradicted by wider evidence.

In particular the famous Benbow study [1] that established that even among the top percentile of the population there are massive differences: the top quarter percentile (99.75-99.99) was 2-3 times more likely to have authored academic research later in life, and about 1.5x more likely to have attended postgrad education, to have gone for a STEM degree, to have gone for a PhD than the bottom quarter of the top percent (99.00-99.24).

[1] https://www.gwern.net/docs/iq/smpy/1992-benbow.pdf

To update on the sibling comment I posted, after reading this study, I think it is pretty flawed. They indeed have an interesting population they have chosen to test, but to claim that they are the 1% top performers is incredibly misleading. The other 99% of their peers were never tested in the same fashion, which leads to obvious sampling and bias problems. This special population was treated differently from the other 99% of students from the very start of the study, so there is no surprise that new internal dynamics will appear among them, simply by virtue of being observed and treated as a separate group.

To start testing the validity of the claims in this paper a study needs to be performed where the strength of effect has to be considered, when comparing a similar one-percenter group and a larger ten-percenter group. My prediction is that the strength of effect would show only negligible differences, confirming my hunch that the special treatment[1] is what created the new subdivision in this new group.

More reading material on this topic would certainly be interesting, if you have anything in mind.

[1]: The special treatment in this paper being telling a kid "you are only a 7th grader, but you are as smart as a high schooler".

This is interesting, thanks for sharing it! It would take me some time to read it, but given that you seem the have already looked into this, I was wondering whether you have suggestion for a wider set of studies, a meta-study maybe.

I am asking, because if I have to choose between my empirical observations (anecdotal as they are, given it is n=1 observers), and a single piece of research without followups, I do feel justified to stick to what I have seen myself. But I am open to be convinced otherwise.

In the US the appropriate exams would likely be the qualifiers for the International Olympiads.

It those were used for admissions, I wouldn’t be surprised if there was a huge market for answers (from somewhere in the production chain), stand-in test takers, etc.

> I got a much higher score than I deserved to on the ACT because you can read the test pretty easily if you put yourself in the shoes of the people making it

I don't understand why that is some secretly accessible knowledge that is only available to some. FWIW, the ACT always seemed more game-y to me what with its hard time crunches and stuff.

No, there are other ways to separate the pack. A perfect score on those tests isn't rare enough, plus you're still going to see a lot of exceptional candidates who got a few wrong.

They aren't being used, so they aren't the existing standardized tests, which are being used.

But it's not that hard to simply extend the difficulty gradient higher. Stick some math olympiad-type problems on there, for instance.

Further, the only thing that the SAT claims to be able to predict is a student’s first semester performance, and wasn’t found to be a better predictor than other factors on other measures.

Of course I can’t find a reference for that right now, but it certainly doesn’t predict college graduation rates best.

https://www.forbes.com/sites/prestoncooper2/2018/06/11/what-...

Useless. They topped out the SAT buckets at 1100.

A test that splits the top .1% of test takers vs. the top 1% likely isn't the best indicator of future career performance ( the source of university prestige ), or necessarily academic performance.

Ok, a test that splits between the top 0.01% and the top 1%. One student goes to national competitions, the other is merely a decent student at your local high school. The SAT lacks the granularity to distinguish between them.

My point was that there is likely diminishing likelihood that the a test with such granularity would return a meaningful score. One would need to include a both a wider and deeper set of subject matter that still conforms to the context of the test. Eventually such a test would pick out the best studiers of generalized aptitude tests. While studying for a test is useful in academia - it's not "The" skill.

It's much simpler to have a candidate include notable achievements on their CV including competitions and after-school programs. However, even here you'll find factors that don't correlate to future performance e.g. those with the means and desire to min-max a CV for college admittance ( A very important life skill ) may not be the best leaders, academics, or students.

While both of the above skills will be correlated to future performance there are almost certainly diminishing returns from selecting the top .1% according to a single dimension.

Curiously the list of fields medalists does not appear to contain a mathlete at present. Providing at least one datapoint that performance in math competitions may not be correlated to future (Academic) mathematics achievement.

I just included that to point out that there is a tangible difference at the high end, it’s not just “everyone getting 800 on the math SAT is similar in their mathematical ability”. Whether competition performance correlates to mathematical achievement in college later in life is unclear, although I do want to point out that many do end up becoming professors and such. Many of the rest don’t show up on the “math radar” because they major in computer science instead and get jobs in the industry.

Throughout my academic career, I've met quite a few brilliant mathematicians with no interest in such competitions. And quite a few "goes to competitions" folks with little interest in higher pursuits. Placing too much faith in any measure is fraught

I’m not saying it’s the only measure we should have. But in the context of the chain this was in I was feeling a sort of “does it really matter if we can separate these people” and I just wanted to illustrate that in actual competitions that discriminate to this level you get a meaningful difference out of it. Now, whether that is an effective indicator of success…I’m not completely sure. A lot of the math competition winners I know are doing startups and HFT, so there’s that.

> A lot of the math competition winners I know are doing startups and HFT, so there’s that.

That's in line with my experience, too. Selecting olympiads is a preference for highly competitive people. In math, at least, collaboration is much more important.

Most of 21st century Fields medal winners are IMO medalists, so that seems to contradict your hypothesis. Terence Tao, Venkatesh, Miryam were all famously young and/or good contestants.

It doesn't have to be the best, merely better than what is used now. And what is used now is some obfuscated proxies of socioeconomic status.

That’s because future career is mostly influenced by how much money your parents have.

I think admitting legacies is morally and ethically wrong but the idea that what we need is another test even harder than the SATs to distinguish the "outstanding" is misguided.

The difference between kid who gets a 1550 and a 1600 on the SAT is so marginal that calling one "outstanding" and one "merely very good" is just silly.

Is that difference the extra $20,000 parents spent on tutoring? Or is that difference actually correlated with outcomes in and after college? Doubtful.

I think what the parent comment is saying is that neither the 1550 or 1600 is "outstanding" and the SAT is a flawed exam that doesn't really get you the "genius" level kids

And my point is that no test exists which will determine who are the "genius" kids.

There are tens of thousands of adults all over the world who went to Harvard, Princeton, Yale, Stanford etc. that then went on to live completely mediocre middle class lives.

There's absolutely nothing wrong with that but my point is that how you do on a test at 16-17 years old does not make you a "genius", nor does scoring in the 99.999% ensure that a person will achieve academic or career outcomes superior than the person who scored 99.9%

Elite schools don't use them at least partially so they can take the same tests as every other school.

> There exist tests that can distinguish between the merely very good and the outstanding.

What tests are these?

Yes does outstanding mean a high depth of knowledge in one area, does it mean a high breadth of knowledge in many different areas? Does it mean a high potential to learn (memory/metacognition skills)? Does it mean comfort with high levels of abstraction? All of these would require different tests.

Or what even is "success?" Isn't that what garners prestige for the school? Do these tests determine your soft skills? Your ability to network? Handle different stressed environments? How you can manage people?

There are certainly some rich slacker/less talented people at prestigious schools, though I agree with the main thrust of your argument.

That is an absolutely wonderful use of the word cache.

You're looking for cachet ("the state of being respected or admired; prestige") not cache ("hidden away for future use").

Also not to be confused with Caché (apparently pronounced like cachet):

https://en.wikipedia.org/wiki/InterSystems_Cach%C3%A9

I will be the picky one: Caché, cache, cachette and cachet have a different pronunciation.

Cachet and cachait do have the same though. Same for caché and cacher.

> Caché, cachet

I'll be honest I have no idea what the difference is between these two.

Ones French, and is the past tense of the verb cacher, “to hide”

No, I meant the pronunciation!

Haha it’s should be the same - but maybe that’s because I learned French in Canada.

It's different, though maybe native English speakers will struggle because there is no equivalent of "é".

"et", "è", "ai" are all pronounced the same. Similar to the "e" in "best".

For "é" and "er" I struggle to find a definite English equivalent sound. Sometimes "a" is close, but it depends on the accent of the speaker. Maybe the best example is just the last "é" when pronouncing "Beyoncé" or "fiancée"

> "e" in "best".

But I pronounce chachet nothing like the e in best. It seems much more similar to a french "é"

This is how I pronounce it:

https://www.google.com/search?q=cachet+pronunciation&oq=cach...

Then you don't pronounce it well :D The link you gave me looks like someone with a strong English accent trying to pronounce French (which make sense for an AI)

No, it's someone speaking the anglicised pronunciation which is correct in English just like for the other 1000s of borrowed French words in English.

Though you'll be happy to know that where I grew up in Texas, everyone tried to pronounce "croissant" the French way when (and only when) ordering one at a deli counter. But we weren't far from Paris, Texas.

I did not say the audio clip pronunciation is wrong. I said it was pronounced with an English accent, which bounces a bit the vowels, though the sound of "è" is correct.

What I said is incorrect, is the parent poster saying it sounds closer to "é" than the sound of "best".

If you can't tell the difference, too bad for you, but that doesn't make you right.

"è" (or "et" in French) is pronounced /ɛ/. Wikipedia will confirm that this is indeed the same sound as in "best" (/ˈbɛst/). https://en.m.wiktionary.org/wiki/best

"é" is not the same sound. It's pronounced /e/ like in "fiancée" (/fiˈɑnseɪ/) https://en.m.wiktionary.org/wiki/fianc%C3%A9e

> anglicised pronunciation which is correct in English

All the links I have posted are for the anglicised version of the French words.

> I said it was pronounced with an English accent,

Cachet is an english word in this context, so how could the English pronunciation be incorrect?

> "è" (or "et" in French) is pronounced /ɛ/. Wikipedia will confirm that this is indeed the same sound as in "best"

Yep.

> "é" is not the same sound. It's pronounced /e/ like in "fiancée"

Also yep. Nobody is disputing you on those.

The problem is when you're saying that the english pronunciation for an english word is incorrect." It can't be. Cachet, in this context, is an english word and this is the english pronunciation.

The link is right, it's pronounced "cash, eh?". Which is basically the same way you'd pronounced "caché" in French.

> Which is basically the same way you'd pronounced "caché" in French

Just... No.

I don't get it, what is wrong with you people for trying to teach a French how a French word is pronounced?!

"é" is pronounced /e/ https://en.m.wikipedia.org/wiki/Close-mid_front_unrounded_vo...

"è" is pronounced /ɛ/ https://en.m.wikipedia.org/wiki/Open-mid_front_unrounded_vow...

The Wikipedia page contains audio samples which are accurate.

"cachet" (phonetically "cachè") is definitely pronounced differently than "caché".

The difference may seem subtle and inexistant for English speakers, but for native French speakers it's definitely different sounds...

I don't understand what is the point of arguing when a native speaker + Wikipedia tell you that you're wrong.

> "cachet" (phonetically "cachè")

Do you understand? This is the statement that we're saying is incorrect. The english word cachet is not pronounced phonetically "cachè" in english. There is another word, cachet, in french. It is not the same word. It does not mean the same thing. It has a common etymology, but that doesn't make you the authority on its pronunciation.

You are trying to argue with an American how an English word since the 1600s is pronounced!

Alright, I concede.

I was looking at the English wiktionnary but in the French section of the word which mentions /ka.ʃɛ/. The English section indeed uses /ˈkæʃ.eɪ/.

My apologies, you were right.

lol, I speak French. Québecois French, but still.

I'm not telling you how it's pronounced in French, I'm suggesting how I pronounce it in English using French as a reference. I don't really think an accent aigu sounds anything like an accent grave.

I don't pronounce "cachet" as "cachè" but rather "caché".

Indeed, I got confused reading the English pronunciation on wiktionnary.

You are right it's pronounced like "caché" in English. My apologies.

Nothing to apologize for! That's what I love about HN, we're all here to learn. Leave the apologizing to us Canadians ;)

This is "the right mix".