Hacker News new | past | comments | ask | show | jobs | submit login
Test scores are not irrelevant (dynomight.net)
281 points by colinprince on Oct 26, 2022 | hide | past | favorite | 476 comments



I worked in Admissions at a large state school from 2005-2007. Our method for determining admissibility reminded me of the heat map chart in this article - ie GPA on one axis, SAT/ACT score on the another axis, trace your finger along both paths to see if the applicant was in or out.

I probably reviewed 20-30 applications a day. If high school transcripts were universally formatted, decisions could have been instantaneous but alas, we live in the real world and some human-in-the-loop normalizing had to be done over-and-over.

With all that reviewing, patterns emerged, namely that SAT and ACT scores strongly correlate to GPA. Now, I’m the kind of person that roots for the genius to overcome his grades and emerge a genius on the SAT/ACT. But in two years, it probably happened only twice. Before calculating a normalized GPA, I could look at the test scores and predict “admit” or “deny”.

While the author is correct to say “the irrelevance of test scores is greatly exaggerated”, in my experience, whether or not something is irrelevant has very little to do with what universities do.

I’d recommend only using test scores. Or, only go with GPAs. Only test scores is more efficient. Only GPAs looks better on press releases.


I can see the value of the GPA. When I was applying to Oregon State way back in the day (this would have been early 90s), my SAT score was really great. Probably top 10 in my school. My grades, on the other hand, were sub-2.00. I may have had a problem keeping myself interested in school. But I found tests extremely easy.

They rightfully told me to go to community college instead, because my GPA was abysmal. I'd have failed right out of OSU. I went to that community college and failed right out of it ;-). Then I went in the USAF, and maybe a year into that experience something clicked and suddenly I felt like an adult. I felt like my priorities realigned and I knew what I wanted and how to get there, and I could stick with it.

So when I left the military, I went back to the community college, got perfect grades, convinced the admissions gal at OSU to ignore all the HS & college grades from a few years earlier, and let me in. Graduated with my bachelor's in CS with excellent grades. Ultimately went on to get my masters, though that was years later.

Anyway, all of that to say ... my SAT score would have said "admit him" but my GPA was a more accurate assessment of my grit. I think both scores are useful, but don't give up on GPA.


Basically the story of my teens and early 20s except that my grades were good enough to get into CS at UIUC... where I promptly drank/smoked my way into an academic suspension. Switched to an easier school and easier major and failed out again which prompted me to enlist in the Navy. Graduated with highest honors from my boot camp class--it's fucking amazing what a little structure can do for people who need it. 16 years later I have a BS in economics, MBA, MSCS, and now trying to beat the final boss and get a PhD before I turn 40.


GPA is problematic because not all high schools have access to the elite courses that push a GPA beyond 4.0 and many top schools look for an admission barrier around 4.5 (some with exceptions for valedictorians only).


Another reason GPA is problematic: it's sexist. Male students face structural barriers when it comes to college acceptance, to the point where a woman is ~50% more likely to get a college admission. Male students tend to both score higher on average on the SATs and have a higher variance. Deemphasizing SATs in favor of GPA (which itself is a discriminatory measure-- teachers systemically show bias against boys when it comes to grades, particularly boys of color) is exacerbating the structural barriers boys face when it comes to attending university.


For what it's worth, I regularly interact with high school seniors and this topic has come up a few times. When I was in school, there was one teacher that was known to be sexist against boys. If you were regularly an A student, you'd be getting Cs. I was one of them. She taught English, which is probably the best subject to teach if you want to be subjective with your grades. Hell, there was another guy in class that kept getting the same percentage grade down to the second decimal point.

Anyways, I graduated from high school in 2006. At the same school now, there are apparently multiple/many teachers that are fully accepted by both the guys and girls as being sexist against boys. I doubt this trend is limited to my school and it's concerning as hell.


And yet people wonder why the MRA and MGTOW movements are gaining ground...


> Male students face structural barriers when it comes to college acceptance, to the point where a woman is ~50% more likely to get a college admission.

As far as college acceptance itself, the norm is for colleges to apply a lower admission standard for boys than they do for girls.

The structural issues are elsewhere, such as the emphasis on GPA and the related emphasis on grading for effort without caring whether the student knows the material.


>the norm is for colleges to apply a lower admission standard for boys than they do for girls.

Where are you getting this from? The data and all incentives going female>either>male would suggest otherwise, if it would suggest anything at all.


Schools that deliberately don't go out of their way to select for gender parity wind up with far more girls than boys. UNC Chapel Hill, for example, is 60% female. From what I understand, this does interesting things to their social dynamics.

Highly selective colleges that strive for parity must accomplish that by rejecting a few girls who would have gotten in on merit alone, and admitting a few boys who should have been the first ones out.

What incentives are you thinking of? What's the source of your data?


It's more nuanced than that. Engineering schools have mugh higher admission rates for women, sometimes over 2x, in the case of MIT: https://www.collegetransitions.com/blog/can-your-gender-give....


Please cite a source for gender discrimination in admissions. Seems like this would be a textbook Title IX violation.


Look at the incoming class numbers. Harvard/MIT/Princeton/Yale are all within 2% of 50/50.

I don't know how they do it legally, but ¯\_(ツ)_/¯


Ah, that makes sense. Since those are private schools they don't need to comply with Title IX.


Title IX applies to schools that accept federal funding such as Harvard, MIT etc. see for example https://studentlife.mit.edu/titleix


Title IX specifically does not apply to undergraduate admissions at private universities.

https://www.law.cornell.edu/uscode/text/20/1681

> in regard to admissions to educational institutions, this section shall apply only to institutions of vocational education, professional education, and graduate higher education, and to public institutions of undergraduate higher education


Ah, so even though the rest of title IX applies, admissions does not.


I wasn't aware they accepted funding. So it seems odd they are so evenly balanced.


Another possibility: men have higher variation in performance than women. Even if the male mean were lower, a greater standard deviation would swamp that effect at the high end.


> I wasn't aware they accepted funding.

Their researchers apply for grants. That is considered "funding". Enrolling students who have taken out student loans is also considered to be "goverment funding" for the school rather than the student.


"Enrolling students who have taken out student loans is also considered to be "goverment funding" for the school rather than the student."

Do you have a source for this?


It's well known. Colleges will admit to it in public, often saying things like "we need to have a lower standard for boys, or else no girls will want to attend our school". (The ideas being that (1) the goal of running a college is to admit more girls; and (2) girls don't want to attend all-girl schools, so you can get higher overall female admissions by admitting some unqualified boys.)

If you think the data suggest that colleges are overall giving admissions preferences to females in favor of males, you haven't looked at the data.


I am yet to see a single university admit to biasing males in admission. Most of them admit to biasing by race, so it shouldn't be hard to find an example.


If this is public, surely you have some actual sources to verify this?

>girls don't want to attend all-girl schools, so you can get higher overall female admissions by admitting some unqualified boys

This implies either unqualified boys are better attractors, or there not being enough qualified boys. I doubt either of those being true given tournament selection being omnipresent.

>you haven't looked at the data

The data most available shows admission rates to be fairly equal, with far more scholarships for girls in areas they lack presence than the other way around. The data on standards is practically invisible, and I'd be very skeptical of colleges publically admitting to sexism in today's age, let alone sexism in favor of boys. Openly doing so and not trying to change this is asking for a massive boycott.


>>you haven't looked at the data

> The data most available shows admission rates to be fairly equal, with far more scholarships for girls in areas they lack presence than the other way around. The data on standards is practically invisible

So you're saying... you haven't looked at the data, and that's why you're comfortable interpreting what it says.

Solid intellectual effort there.

> If this is public, surely you have some actual sources to verify this?

Indeed! You can find them yourself too, just look for any coverage of the issue over the last 20 years. I can't be responsible for everyone who wants to contradict stuff that's been known for decades. Get your own house in order.


I'm not the person you are arguing with. I'm only dropping in to say it's extremely frustrating seeing you claim data this, data that, without ever providing that data. The crux of the issue you are debating is the data you claim to have. Please either provide it or stop going in circles.


It might not be sexist; it might be that boys deserve lower GPAs because they tend to be worse at classwork.

I say that as a boy who was worse at classwork.


Classwork isn't some static thing; historically, boys had higher GPAs than women, and I suspect you can find some subfields of study today where boys consistently outperform women as far GPA goes. The choice of what subject matter goes into the curriculum that a GPA represents and how's it's taught and evaluated is a political choice. You could very easily make GPA equitable by removing one class that girls over perform in and replacing it with a class that boys over perform in.


The truly awkward question is that GPA is important because colleges give grades too, but maybe the classwork grading and tests are bad and not conducive not education overall.


I'd be interested in seeing if there's a gender effect on how predictive SAT scores are on college completion rate. A quick survey indicates that this isn't an area education researchers have thought is a good use of their time.


If it's as you say systemic that boys do worse at classwork, then continuing to use that (a la GPAs) as an admissions criteria _is_ sexist. It would be akin to having a pull-up competition determining admissions knowing full well boys perform better than girls.


But classwork is what you'll have to do in college anyway. It's not the admissions criteria that's the problem, it's the format of undergraduate higher education itself.

A pull-up competition is a perfectly valid test if the job/program requires exactly that kind of upper body strength.


The #1 difference between college and HS curriculum is that colleges have far far less classwork.


The opposite is true in my experience. I basically coasted through high school spending very little time on homework. I'd often spend hours, or even tens of hours, on single college assignments.


Far more courses in college rely on one or two exams over coursework. To be more precise, men tend to do better on tests of mastery of material over evidences of participation.


I view it similarly to military, or construction. It's not the biases' fault, as many as there are, that men are overrepresented in these domains. I'd urge to address systemic issues, but the profile of representation is not necessarily a symptom of such an issue.


(1) Many colleges are already practicing affirmative action for boys because many boys are not applying. Schools are glad to have girls but girls don't want to go to a school which is 70% girls so they try to admit more boys to make the social life more normal.

(2) There is a serious representation problem in primary education, particularly elementary schools. Both boys and girls benefit from having male teachers, boys particularly, since as it is they get the unambiguous message that school is an institution by and for women, one in which they don't have a place. It's bad enough that it shouldn't be thought of "we need to hire more men as elementary school teachers" but "we need to stop hiring women as elementary school teachers".


"then continuing to use that (a la GPAs) as an admissions criteria _is_ sexist."

Please define sexist/sexism. This use doesn't match with my definition.

The measure is an objective one, which is highly correlated to graduation rate and thus pertianate. It's not like this is used to discriminate.

Yes, you could have individual teachers showing bias, but it seems that the data in the article doesn't support this being impactful. I'm actually a little skeptical that bias is rampant given how overbearing many of the school policies have become and an inability to explain grading differences would be highly suspect. I'd expect people to get sued/fired left and right over this if it's truly widespread.


> I'd expect people to get sued/fired left and right over this if it's truly widespread.

The attribute of being labeled male is not a protected class (maybe legally, definitely de facto) in the same way that of being labeled female is.

To your broader point, disparate outcomes is ipso facto evidence of sexism. At least, that's what our society has settled on for other excluded identities, so we should do the same for men.


Literally, legally, Title IX protects males as well as females. Any pattern of arbitrary grading against male students and not female students would provide a basis.

"disparate outcomes is ipso facto evidence of sexism."

Lol, no. Again, please provide your definition of sexism, as this does not match the one in the dictionary.

There are plenty of things that are not sexist and have disparate outcomes.


> Any pattern of arbitrary grading against male students and not female students would provide a basis.

https://mitili.mit.edu/sites/default/files/project-documents...

This has been well studied and is a pervasive, consistent result. However, we don't see the EEOC launching lawsuits against or even studies investigating bias against boys. Can we really say that being male is a protected class if no one bothers to protect it?


That study is from France and doesn't mention the pervasive nature (in fact it shows some areas/groups don't have problems). Do you have one from the US? I'm particularly interested in one that has data to back up your "pervasive" claim.


American educational researchers are surprisingly uninterested in this topic; I managed to find recent studies on it out of France, Italy (from last week!)[0], the UK, Sweden, Denmark, and the Czech Republic, but none out of the US. In fact, I found more material in the US about the gendered bias in grading by students of teachers than by teachers of students(!!!).

I'll look more into this to see if I can dig out an American specific study.

[0] https://www.tandfonline.com/doi/full/10.1080/01425692.2022.2...


They're trying to determine how well you will do in courses in college. Surely how well you did in courses in high school is relevant. It is not sexist just because some groups tend to do better than others at a certain point in time. Girls have increased their academic success because they've historically not been encouraged to do so, but in recent decades they have. Girls in high school today are like 2nd (or 3rd) generation of women in the US to have been actually broadly encouraged to do well in school and obtain professional careers. There is an issue with boys feeling hopeless about their futures, and maybe this is a factor in why they as an aggregate have done worse in school relative to girls in recent years, but the solution isn't to lower standards.


SATs are predictive of college completion rates. Deemphasizing them in favor of another predictive metric (that women tend to do better on) exacerbates the disparate graduation rates that are adversely affecting men.

If we knew SATs are more predictive of college success for men than women, we'd even be able to simultaneously increase both representation of men in college and overall success rates. That's received a lot less study than other topics in vogue for educational researchers.


Is it then more or less sexist if colleges aim to balance gender ratios (as most do)?

Colleges often re-weight both GPAs and test scores based on sex, race, socioeconomic background.


It's also a maturity thing. Girls tend to be more mature than boys (emotionally and mentally) in their late teens, so their grades are a bit better. By the time everyone reaches their early 20's the differences have largely vanished.


We regularly use outcomes as evidence of discrimination, so I think it’s reasonable to strive to be consistent.

If we look at SAT discrepancies to identify groups harmed by structural issues in the test, we should also do the same with grades.


One solve for this is to require schools to anonymously publish their GPA percentiles along with some demographic info.

Bayesian analysis could then be applied to pull the signal out of the GPA data. GPAs can then be renormalized such that students are not incentivized to go to a HS that gives everyone As.


This points to the real problem with GPA. The institution assigning GPAs to students have a strong incentive to inflate GPAs. If colleges only looked at GPAs, high schools would inevitably give in to the temptation to give away perfect scores like candy, and students from schools retaining some shred of dignity would suffer for it. Standardized exams, for all their faults, are standardized, and don't have this problem.


> This points to the real problem with GPA. The institution assigning GPAs to students have a strong incentive to inflate GPAs.

I think it is even worse -- the inflation is not universal. At Cornell (undergrad), most engineering classes set the mean to B- or C+. You had to go 1SD higher to get to A- and near 2SD to get A+. Imagine how shocked I was to hear my friends at Princeton had a mean set to A-. Consider the pressure this puts on some students and not others, esp when you're all applying to the same graduate programs. Is a Cornell B+ worth more than a Princeton A-? Statistically yes, but in reality no one is harmonizing the distributions.

There is also the issue of fancy private prep schools which have 10 or 15 "Valedictorians" which many of us public school students were shocked to learn when we arrived at college. You'd meet multiple people claiming to be valedictorian from the same school's graduating class until you learn that private prep schools stretch the meaning of valedictorian.

Some reason life lessons: pay enough and you get to bend the rules.

Some reason life lessons: more elite, less pressure.


> If colleges only looked at GPAs, high schools would inevitably give in to the temptation to give away perfect scores like candy, and students from schools retaining some shred of dignity would suffer for it.

All three of these things are already happening.

For example, the Cal State University system has purely objective admissions for most campuses. That's good! But while they technically consider SAT scores, the last time I looked the difference in admissions points between a minimum SAT score and a maximum SAT score was roughly equal to the difference between a 2.0 GPA and a 3.0.


We don't have SAT in Canada and mostly rely on high school grades. Most universities/colleges adjust student GPAs on a per high school basis based on alumni performance in their first year courses.

I personally prefer the GPA system as it measures your performance over your final 2 years of high school rather than having your future determined by one test.


You can retake the test as many times as you want. Some schools average them, others will take the highest. The whole point of creating the test(s) in the first place were to deal with variability of GPA's between schools, so it's a bit useless to say "just adjust GPA's between schools instead".

In the US now we have federal school testing to try and normalize schools better. But before that many schools were graduate students that were illiterate (actually some still are, but I hope it's gotten better..). The US is 10x the population of Canada so image the difficulty for a liberal arts school in the northeast looking at a student from some rural 500-person school they've never heard of from the southwest.


Adjusting GPA between schools is one approach to deal with grade inflation between schools. SAT is just another approach with its own tradeoffs. I've simply listed why I prefer the former over the latter approach.

Canada's population is smaller but we also have many obscure rural schools. Non-prestigious universities don't have much data but they also don't have much choice in top applicants so it doesn't really matter. Over time, they can build up decently accurate profiles of high schools for majority of applicants.


They do a regression that includes both GPA and SAT. That is mathematically an adjustment of grades, where the adjustment is the SAT term. Then, as articles such as this one show, this correction correlates well with success at the university. So ultimately it becomes the same correction you want.


It would be interesting to see a study that compares SAT scores vs alumni performance to see which one has a better correlation with success to use for said adjustment


It's been many decades since I took the SAT (3x, at ages 11, 12, and 16), but I seem to recall that I had a choice of which schools to share my SAT results with, such that a school that only ever saw my age 16 SAT test couldn't possibly average in the other two scores.


It creates bad outcomes where students retake classes in the summers as a profession and game the system.

It also puts you at a disadvantage if you come from an immigrant area or poor area where your 90% will not count as highly as someone with 60% from the richest areas with a trackwork of success.

It really punishes anyone not rich and locks them out of elite universities and pushes most of the poor out of average universities.

A single test where everyone can study and pass seems more fair than basing it on things that can be impossible to change like where you grew up


Said test can also bias against the poor depending on how it's created. It can ask questions based on subtleties that only private schools/tutors have the time to cover or on life experiences that only upper middle class realistically have experienced.

Neither system is perfect but I think the purely SAT approach creates too much unnecessary stress for high school students when there are other proxies for determining future success.


> your future determined by one test.

There is nothing stopping you from taking the SATs multiple times. It's encouraged because you most likely will go up a few points the 2nd pass. I've taken it twice myself, 3 times if you include the PSAT. And I also took the ACT, only once.

GPAs you can only screw up once. If you fail even one class it's impossible to climb back to the top.


Tests are fair

Grades can be easily biased


I'd say the GPA is problematic because it's often a "weakest link" measure rather than focusing on strengths that a person can leverage for success. By that I mean, a GPA can easily be dragged down by a lack of interest in one or two mandatory subjects. A student can be a brilliant scientist who aces those classes but simply isn't very interested in world history or Catcher in the Rye or conjugating verbs in a foreign language. Or the opposite -- loves English and History and Arts but simply has no aptitude for geometry or physics. So these individuals graduate with 3.2 while others who are ok but not excellent at anything manage a 3.8.


Seriously, schools in Palo Alto or Stanford itself, producing GPA>4 only means that they’re cooking the numbers by giving some courses a greater than 4 average.

I think the correct thing is to require schools to publish their minimum and maximum grades, and then everyone can just scale that range as appropriate - because why should some kid going to Stanford get to report a 4.0 GPA when they’ve got low grades, but a kid at CSUEB has to report something lower because they’re a state school that can’t invent new grades?

From my personal experience, my GPA when I graduated was 8.7 or something, which at least makes it clear that I’m not on the same scale as you might normally think - whereas a 3.9 from Stanford might seem really good but is bolstered by courses with >4 scores. An extreme would be if you heard someone from my school talk about a 3.9GPA, which if IIRC would imply an average grade in the region of 60% (and yes you could in principle go negative, but that would get you on the “let’s talk about whether uni is right for you” list with the uni admin)


Is that a typo or did you actually have an 8.7 gpa? I don’t understand how that’d be possible


Which is entirely the point I was making - having anything based on GPA requires that the GPA have an agreed range, and a bunch of private universities in the US have decided to give students >4.0 GPAs - now if you see someone with a "4.5 GPA" you immediately know they're cooking the books, but a person can also have a 3.9 GPA with a bunch of bad grades countered by a few >4 grades.

My university's GPA was (IIRC) -2..9 or something where 9=A+, 8=A, 7=A-, etc but that wasn't used in any externally facing mechanism, because a GPA as a single score is not useful outside of the school administration. Single average GPAs aren't remotely robust enough to warrant any kind of external value, and they actively discourage any course experimentation because if you are wanting to do anything GPA gated you cannot risk anything that you don't already know you'll get a high grade in.


For sure. But weird I didn't know of any universities/colleges that gave above a 4.0 for GPAs (in the US) - thought that was more of a high school thing relating to Honors/AP classes.


Stanford and Harvard at least do this.


Scaling the ranges doesn't fix things for the students. At my high school, some electives only awarded grades up to 4 and others were offered in flavors that went up to 5 or 5.5. Every foreign language actually spoken by modern humans had the first year offered with grades up to 4 only, so optimal play was to take Latin or take Spanish 1 in summer school since that counts for qualifying for Spanish 2 but does not contribute to your GPA. Most kids took Latin because they had lives.


If you want insane you should look at the old sixth form certificate system in NZ - each school is allocated a set of grades to assign to kids based on the grades the prior year got from school certificate. As in say your school got 5 As, 5 Bs, 5 Cs, etc in school certificate exams (the nationwide standard tests from when you are 15), then your school would get 5 1s, 5 2s, etc (1 being the best). Then the school gets to choose how those available grades are distributed - so say the 5 As came from math exams, the school could allocate all the 1s to the English department, so the best grade you could get in 6th form cert outside of the English department would be a 2. I had a friend who was a super talented musician, but his highest possible grade in 6th form music was a 3. Despite getting essentially As across the board in school cert music his best 6th form cert grade for music was notionally a C+/B- or some such.

Given you could get into uni on a high enough 6th form cert grade this is obviously absurd (I think that it was something like your total score for your best 4 subjects had to be less than 6, so you can see how one course consuming 3 points even if you were absolutely perfect screws that)


> GPA is problematic because not all high schools have access to the elite courses that push a GPA beyond 4.0 and many top schools look for an admission barrier around 4.5

FYI, they rate according to the school's scale. This knowledge is shared.

You don't think they'd notice if every single candidate from specific schools never got higher than a 4.0? :)


There are thousands of high schools, and for each college, most applications come from a relatively small group (usually based on geography, while for elite schools, a set of top "feeders"). So to answer your question, no, most colleges are only familiar with the scale of a small percentage of schools.


They also weight schools differently. Shitty high schools + good scores let you avoid all of those cello lessons.


Also, grade inflation is real and not uniformly distributed.


These days to avoid this problem most schools normalize GPA's before looking at them for admission.


Admissions departments re-norm GPAs and many of the best high schools do not engage in grade weighting.


I was the same. Top fraction of a percent on SATs without studying but never turned in homework. Got Cs.

Got into state school in engineering, cruised through intro term, almost failed out second term when the weed-out started. Felt the same thing click into place over the summer: might as well try. No military experience required — 4.0 from then on.

I think the whole process is biased against male students because they mature later, and there is nothing difficult in high school and no real form of soft failure to wake you up.

We ought to make things harder, but support and include people when they fail the first time. We should normalize learning from failure.

So much of it is down to your family environment. My parents hardly existed as an environment.


> Anyway, all of that to say ... my SAT score would have said "admit him" but my GPA was a more accurate assessment of my grit. I think both scores are useful, but don't give up on GPA.

Do we care to measure grit? If someone can do the courses and pass the final exams, does it matter whether they did "sufficient grit" during the process?


Grit is exactly what's needed to pass college courses and final exams. It's also one of the most important skills in the real world. It's easy to have a stroke of luck or be a genius for a day on a single exam. It's much harder to persevere and learn unfamiliar material that you don't want to do for many months, culminating in a final exam.


Finals exams were pretty easy, you don't need grit to just show understanding of material, you just need to show up to lecture, listen, participate in class

I got A in all of my stats exams despite not doing homework, I still got a B in the class because homework is part of the grade

But why should everyone be forced to do homework if some people don't need the extra time waste?


Refusing to do assigned tasks despite knowing it will be a factor in your final assessment score is probably something valuable to measure as a factor in predicting your future success at tasks. I had some college courses where the homework was optional, but this is not that, this is courses where the homework is required, but not completed.


Because a job requires doing repetive work you already understand, so you shoul learn to do that. And because practice builds long term memory.


I prefer to replace "grit" with "programs on my calculator". Turns out the calculator is still better at stoichiometry than the chemistry student with a lot of grit.


But in that case did the chemistry student actually learn the chemistry, or just learned the tool? What's the goal for taking that class?


His anecdote clearly says that his low GPA correlated to his inability to finish the degree (by doing the courses and passing the final exams), which is a real criterion to consider by admissions, and that relates to grit.


I think if you take "grit" to mean "discipline", it goes a long way.

In my opinion, it's better think of school grades as a measure of efficacy instead of intelligence. Given a fixed amount of time, how correctly could you perform the task? Discipline, or "grit" as it has been called earlier in the thread, refers to one's propensity to allocate larger portions of time towards accomplishing a single task, especially in the face of adversity.

Hopefully this perspective makes it more clear how "grit" is useful.


TFA actually calls out that very high test score can be a a signal someone won't graduate.


But why take graduation as the target? Clearly it's not the school's mission in the world.

The schools here certainly do get bonus money from the government for every person who graduates.

On the other hand, companies like to employ people no matter what their degree - if they passed courses and did projects and learned useful skills in some school, that's definitely great, but graduation by itself isn't such a big thing.

Then there's the whole other aspect of school, getting to know people, learning social skills, organizing events, even activism. Just as an example lot of politicians come from student politics background. If everybody just "minds their business" and "doesn't have an opinion about politics" then dictators can just seize power.

So why should school intake optimize for graduation? We could get people who are really good at barely passing a large amount of narrow topic exams over a few years with methods like cramming in their room the previous night, but really bad at their job or other aspects of life.


When you start pulling that thread, the whole system of education metrics falls apart.

Including people's assumptions that seats in school should be prioritized for people already learned more than everyone else.


A very weak signal, about as strong as whether the person is male or female. And it's outliers in both directions that create this effect, not just high scores. This probably just indicates that the student was not a great match for the school and is likely to lose motivation before making it to graduation.


But how should they measure "I matured and now I have grit"?


Good question. In my case, I talked to the person in admissions who was making the decision and explained the situation and why I thought they should disregard the earlier grades.

But absent that conversation and direct evidence, GPA is a start. Assuming that the high school curriculum is at least passably similar to what they're going to expect from you in college (of course it isn't, exactly, college tends to be more heavily focused on tests than homework), a good GPA demonstrates that you will do the academic work required to succeed in class.


> I went back to the community college, got perfect grades

Presumably they had a separate GPA, or at least term-specific GPAs, from their second round of community college.


According to the University of California’s own internal study, there is a small group of students with terrible GPAs and good test scores. They tend to do well in college.

There used to be a set percentage of spots for students in that situation, but the admissions people decided that “Standardized tests are racist!” is a good twitter sound byte and eliminated the spots.

Of course, their own reports said that the “bad GPA, high test score” cohort were mostly from underprivileged areas (and disproportionately black or hispanic). That nuance didn’t make the 140 character cutoff, so that group is officially abandoned by UC now.

In other news, they’re now promoting a big push to make the student body composition match the general population.

However, there is only one over-represented ethnic group in the UC system. Guess which one?

Hint 1: Whites are under-represented.

Hint 2: They swore they’d never bring back UC’s old “asian quota”.

I honestly think the people running this state’s school system are malicious racists. Their results speak for themselves: California schools are currently ranked 43rd in the nation. They used to be top ten.


> According to the University of California’s own internal study, there is a small group of students with terrible GPAs and good test scores. They tend to do well in college.

This was me, fwiw. I’m so lucky that the engineering school I got into was enamored with my math score.


This was me as well; I should not have been allowed in, I took someone else's spot and I forever feel a tiny bit bad about that. I was not ready for college, and ultimately never graduated.

I'm plenty smart, I just needed an extra couple of years to mature before I was ready to take on the responsibility required for school, but by then my software career had already begun in earnest so I never returned.


The arguments against tests, from SAT to civil service exams is that effective tests correlate to IQ. IQ=racist, because we say so, so tests=bad.


[flagged]


> Tests are sexist, so we need to adjust for that to give men a boost

This is the first time I've heard this opinion and I would be very surprised to find out that it's what the HN hivemind thinks. I would be even more surprised to find that the same people who think this also think the other part about racism.


> Tests are sexist, so we need to adjust for that to give men a boost

I don't think I saw anyone claiming that. What was being written was that boys do a lot better on anonymized tests relative to girls than on GPA (which is inherently non-anonymous), so removing testing is actually exacerbating sexism in education.

This is probably the comment you're referring to: https://news.ycombinator.com/item?id=33339235


> IQ=racist, because we say so

The argument is a bit less stupid than that. IQ = racist because whites and Chinese have it while blacks, Puerto Ricans, and Mexican immigrants don't.

Is that reflected in different levels of group performance? Well, yes, but ignore that.


Obviously we currently don't ignore it while not studying it in details either, so we've come up with IQ=racist and stopped there.


I think you're closer to something than the other poster.

Let's give the assumption that certain racial/ethnic groups perform better on IQ tests as a whole than others.

If that were true, that would just be a fact. Yes. But why that is matters. And a lot of the why is just plain old racism. The groups who don't score well on IQ tests don't do so because they've been held back from all of the progress enjoyed by the rest of the world.

It's that progress that increases IQ scores. It's not that white people and Chinese people are smarter due to genetics or culture or whatever people use to spout. It's because they've been able to leverage the progress of the world to advance themselves. Black people, Puerto Ricans, and Mexicans, especially immigrants have not been allowed the same access to leverage it. You give them the same resources, the same chances, and over time, you'll see the same increases.

In fact, you do. Because those groups constantly get progress's hand-me-downs. Last year's progress. So they lag behind.


You jump from the UC to primary/secondary schools. They are not run by the same people.

Most of the decline is due to underfunding due to prop 13


> underfunding

This is a decades-old talking point that simply isn't true and hasn't been for years. Funding levels keep breaking records year after year.

"K-12 per-pupil funding [in 2022-23] totals $15,261 Proposition 98 General Fund—its highest level ever—and $20,855 per pupil when accounting for all funding sources." [0]

"Reflecting the changes to Proposition 98 funding levels noted above, total K-12 per-pupil expenditures from all sources are projected to be $18,837 in 2020-21 and $18,000 in 2021-22—the highest levels ever (K-12 Education Spending Per Pupil)." [1]

[0]: https://www.ebudget.ca.gov/2022-23/pdf/BudgetSummary/K-12Edu... (p. 3)

[1]: https://www.ebudget.ca.gov/2021-22/pdf/BudgetSummary/K-12Edu... (p. 4)


I don't know the specifics here, but it seems possible for something to both:

A) have received more money than ever before

B) not have enough money to function properly


OK, please see the chart at the top of page 5 of the 2021-22 budget PDF: https://www.ebudget.ca.gov/2021-22/pdf/BudgetSummary/K-12Edu... .

You can see that in 2012-2013 there was a big jump in funding from 2011-2012, going from $47.3B to $58.1B. In 2022-23 we're at $102B.

We're at roughly double funding levels from a decade ago, and the schools still don't have enough money to function properly? If so, this is alarming, and signals something is deeply wrong. We should investigate what is wrong and fix it rather than throwing ever more money into the black hole and hoping that will improve outcomes, despite a decade of evidence to the contrary.


How come everywhere else they get LESS money and have BETTER outcomes?

Hint: teacher salaries are a very tiny part of the expenses


NAEP scores have been pretty much flat since the 1970s while inflation adjusted spending per student tripled.

https://www.cato.org/policy-analysis/they-spend-what-real-co...


California spends about twice as much per student today (inflation adjusted) as it did before Prop 13 passed, so it doesn't track that Prop 13, or underfunding, is the cause of the poor performance.


All the 200+ student classes would certainly hurt them a lot though since that's a big chunk of some rankings.


I meant that primary/secondary funding per student had doubled. The UC and CSU systems have gotten less money from the state, but not because of Prop 13 which limited property tax rates. UC/CSU weren't and aren't receiving property tax revenue.


Prop 13 made it harder to raise all tax rates


That does not include capital expenditures


Actually California's state school system at the UC level is 1st in the country. And 1st in the world.


> However, there is only one over-represented ethnic group in the UC system. Guess which one?

Does the ethnic group that gets a quarter of Nobel prizes and was hated by Hitler not count or something?


Entrance exams are a common option in other countries, and they are a good way for students who are late bloomers to get into top schools.


I'm your outlier!

My ACT was 34 and I can't recall my SAT score but it was also 99th percentile. My GPA on the other hand was abysmal. I can't recall, but I wouldn't be surprised if it was well below 2.5.

Once I actually made it to college I was a pretty good student, though. Graduated with honors in 3.5 years.


The edge cases are more common at elite institutions, which are the typical admissions mania targets. At the top ~10 university I attended, there were plenty of people (myself included) falling in the poor HS GPA bucket (mostly due to extracurriculars) with perfect or near perfect ACT/SAT scores, among other outliers.


Short term one get can rid of test scores or GPA since they indeed seem to be correlated. But long term:

1. If we get rid of test scores, it would cause too much upward pressure on GPA which would make GPA less reliable. There are tens of thousands of high schools in the US. Even if the GPA distribution inside every high school is correct, how do you compare with GPAs from other high schools? And from experience I can predict that, once people realize that only the GPA is important, there will be outside influences (i.e. pay the teachers for good grades) that could cause even the distribution of GPA inside one school to not be correct.

2. If we get rid of GPA, students could just study for the test with obvious disastrous effects for their education.

We need both GPA and test scores to keep both of them in check.


28 on the ACT, no studying.

1.4 GPA.

I was eligible for the maximum academic grants for my state, just to be denied due to GPA because the string of high schools I went to were utter dystopian nightmares.

this tracks.


Another problem with relying on GPA is it is easily gamed. Parents and students can pressure teachers and administrations since so much is reliant on it. And a 4.0 at one school may not be the same at another school.

I understand some people "test poorly" but the truth is, if you actually know the material well, you'll test just fine in most cases. We can also coach students on how to approach testing. Other ideas are to break the test into various sessions with different admins to limit cheating potential and to take an average of the scores (or some other method) for each subject.

I guess my point is - make test results the metric to remove all bias. Change how we approach testing and preparation.


It doesn't remove all bias. It does remove a lot of bias and concentrates the remaining bias in a particular dimension, which probably does substantially improve the fairness of the system, but we need to be clear-eyed that it's not eradicating bias.


Your refusal to take responsibility for your part in that 1.4 GPA also tracks, heh...


of course, but that just illustrates that GPA is less about actual academic achievement, and more about playing a revolving game of teacher's pet at best, and at worst, being another tortured cog in a horrible system of petty social games called "high school"

giving any credence to GPA is upholding this hellish set of norms.

K-12 should be K-9. High school is glorified babysitting, with bullying and drugs being the only escape, for both parents, teachers, and students.


Academic achievement is entirely a measure of how well you're able to play the game, and has little to do with intellect until you reach the highest levels, and even then you can be a little less intelligent if you are very very good at the game (whereas you cannot be very intelligent and bad at the game).

So yeah, you weren't very good at the game. I wasn't either. Own that, and stop making excuses!


I think my HS GPA was... 2.9 or perhaps 3 (out of 4), yet I had a 29 on the ACT.

The ACT score was in the 80s before some scoring model change. It's probably changed multiple times since then, but I'd learned there was a model change in 1990(?) which makes comparisons between them a bit less useful. My siblings had slightly lower scores years later, but because the model changed, there's not a lot of comparison we could make.


Yes, nobody is dropping test scores because they are irrelevant -- although that overreach does exist in the press. They are dropping them because they add little after GPA is considered and may introduce costs, barriers and biases.

The author identifies the steel man version early on: "standardized test scores are nearly useless (at least after you know GPAs)" but then mostly ignores that parenthetical. This part that addresses it seems deliberately obtuse:

> If you care about the difference in accuracy between GPA and ACT, then you should care even more about the difference between (GPA only) and (GPA plus ACT). It’s incoherent to simultaneously claim that the GPA is better than the ACT and also that the ACT doesn’t add value.

Again, the version of this that reasonable people argue isn't that the ACT doesn't add any kind of value, it's that the amount and kind of value it adds (incrementally more predictive accuracy) aren't worth the costs and biases are introduced. If you myopically value accuracy, it doesn't make folks who want to balance accuracy with other goals incoherent.


What do you think about GPAs being non-standard? Test scores calibrate between schools - GPAs only kind of do.

I remember many classes that were taught by two separate teachers: one teacher consistently gave students higher grades. The other teachers students consistently did better on the AP test - much more strict with homework and test scores, but it ended up positively impacting the students.


Why did't the other teachers correct their grades at the end of the term?


My GPA was bottom 10ish at my high school. I graduated high school essentially because I already had most of the credits I needed to graduate from middle school. I got into college based on test scores alone (and I was surprised that this was possible). My GPA improved a bit early on in college and then improved much more around junior year once I started getting into harder classes. I ended up graduating with a math + CS double major in 4.5 years.

So, sometimes people are just bored. And there are a bunch of other factors here too - perhaps the biggest one being that "graduated within six years" is not a great indicator of success. Linus Torvalds did not graduate within six years but did build the first version of Linux at university.

GPA is a good indicator of how much work you're willing to put into what you're learning _now_, so it tracks well as a predictor of how much you care about GPA which is a pretty good predictor of whether you'll graduate in X years and probably not much else.


My high-school GPA was mediocre because I simply was unable to pay attention in class, or to study at home, or to turn in homework that wasn't dinged over and over for "careless errors". I was born in 1956. Given today's knowledge of things like ADHD-Inattentive, I'm pretty confident that I would have been diagnosed. Some drug would probably have been prescribed, and it probably would have helped my grades (although I'm not convinced it wouldn't have hurt me in other ways).

But for some reason, when I took my SATs, I found it to be fun. I got completely immersed for the period it took to take the test, and truly gave it all I had, with real focus. I got the best SAT scores in my school, which was shocking to some people.

And it turned out that that was far more predictive of my future, including my college GPA performance, than my high school grades were.


Interestingly, I was one of those "geniuses" that had a perfect ACT score but a horrible GPA (long story). I got rejected by every college I applied to except for the local state school, so that's where I ended up going. In the end I dropped out in my final semester and just went to work, but that's besides the point. I was more than qualified to academically excel at any university I wanted to attend, and I maintained a 3.9 GPA in college up until I dropped out.

The problem with using just GPA is that GPA can easily be manipulated by school authorities who dislike you, test scores cannot.


Grades and gpas are based on tests too, but on crappy tests without any psychometric validity, often correlated with race and broken in a bunch of ways which standardized tests are not. How many high school math teachers run irt or dif analysis on their tests? Meanwhile, standardized testing companies spend a lot of money and energy trying to remove race from their tests. It is as though smart people have been working really hard on this problem for many decades and found an optimal solution. The solution is standardized tests.


Sure, we can use only GPAs as long as all the applicants are graded on the same common platform. But today the GPA from one college is not exactly equivalent to the GPA from another one. The grading system as I heard is also not standard across US (leave alone international) with some instructors using absolute while others relative grading. So an A student from college X is not necessarily as good as a B student from college Y.


> With all that reviewing, patterns emerged, namely that SAT and ACT scores strongly correlate to GPA.

My school experience was that it's easy to cheat during the year and much harder to do so in exams, so often people with generally good grades would do at best average at the end.

I always had terrible grades(except for math), but they're not part of college admission scores in my country so I had no incentive to do anything about it.


I guess test scores are more efficient because GPAs are somewhat subjective and can be gamed to a degree. By both the school and the student.

Two schools with two different grading tolerances can produce students with similar GPAs but different SAT/ACT scores. As the SAT/ACT is a national standard, not a regional one.


Out of curiosity how did you normalize for say 3.0 GPA at a competitive admissions magnet school, and a 3.0 GPA at an underperforming high school? Did you correlate pervious students from that high school and their high school gpa and their graduating gpa?


Great question - the short answer is that our Admissions office did not attempt to normalize non-mathematical factors. At first, this surprised me. How often do we as students somehow get the impression that extracurriculars, magnet, Montessori, AP, IB, student council, <more trophies/achievements of your choice/> are somehow positively associated with college admissions?

From a non-Admissions perspective, accounting for these non-empirical differences seems wise and fair. However, they introduce subjectivity into the equation, which increases the likelihood of a human decision maker being unfair.

So what normalization did we do? What would you do? Knowing that no equation is perfect, you must start with a goal and find the least-bad equation.

The goal: Only admit students with a high likelihood of not failing out.

Why: Failing out has consequences for most US students who take out large loans and must repay immediately if they fail out.

The realization: High schools already provide a proxy for degrees of past failure - letter grades.

Assumption: Assume that high schools use their letter grading system as a method of normalizing and displaying how hard their courses are. A B+ should represent a certain level of achievement regardless of school. The school has the autonomy to use any letter system they choose.

Solution: Assign a point system to letter grades and calculate an average. Use only classes correlated to success at your university. This is the normalized GPA.

Surprise #1 - If your high school thinks a tougher-than-average letter grading system (eg a 93% is a B+) somehow makes students more attractive to the majority of the 7000 US colleges, they are wrong. And probably causing students to lose out on scholarships.

Surprise #2 - If you believe this system is unfair, you can aim the blame at technology. Applications to most universities (until recently I’m told) increase every year as digital applications make it easy to apply to 10 schools instead of 3. To handle the increase, colleges can hire 3x more admissions people, or find ways to speed up the process.

Surprise #3 - Colleges always want more applications than ever before. By dropping SAT/ACT as criteria, it removes a point of friction in the customer experience, and removes a step in the calculation effort. (If they could drop GPA and keep SAT/ACT, this would the ideal situation from a metric and efficiency perspective, but Marketing and the public would throw a fit.)


I have long wondered about this. Note that you also need to be able to normalize grades across students who went to the same school but had different teachers. I suspect the answer is that both normalizations are hard so admissions people don't bother.


They don't not bother. They just do it by gut feel without writing anything down about how they're doing it. Admissions counselors at elite high schools "network" with admissions offices at universities to benefit their students in this process. This is common knowledge in the industry to the point that foreign elite high schools, which have a harder time getting name-recognition, spend a pretty sum of money periodically sending their representatives on month-long US trips to go around, say, the top 30 US universities to market their brand.


This is true.

I took the ACT on a lark my sophomore year of high school. I scored a 29. I also attended one of the top high schools in the nation.

When I was registering for classes in college, my advisor kept trying to put me in Trigonometry or Pre-calculus (I can't remember). Because my ACT score in math said that's what I needed. I took the ACT while I was taking Geometry and Algebra II, so yeah, my math was probably weak. I tried explaining to them that my high school transcript shows I've taken up to Calculus. They wouldn't hear it. I refused to register for any sort of math below Calculus.

Eventually, they told me to go talk to the head of the math department. I told them fine. I went down there all prepared to make my case. As I started to make my case, he saw my transcript, said "Oh, <SCHOOL NAME REDACTED>. What class do you want to take?"

He didn't care about my transcript in the opposite way. I probably could have registered for 300 or 400 level math courses.


To be fair, some state schools do operate on just GPA. I recall my HS had that GPA/SAT scatterplot available for past applicants: UVA had a flat cutoff at ~4.1 weighted GPA, and the SAT literally didn't matter at all.


Well, I had pretty good SAT scores (2350+) and a very poor GPA (3.0/4.0).

Main reason was that I just spent all my time pursuing ECs and didn't have enough time to spend on grades.

Think you'll lose a lot of smart kids who spend time on ECs that way.


That was probably more or less my college story for grad school. Very good GREs but lousy GPA in college. (And only somewhat less my story in grad school before going to the next grad school :-))


I just posted a comment about my step son who had stellar test scores after we spent $1200+ on a private tutor. Did he get more qualified for college than the kids who had parents who couldn’t afford private tutors?


Yes he is more ready to be a clog in the machine. He passed the first test. He studied enough leetcode and it doesn't matter how he got there. Will the tutor be there in college and will he be able to absorb all the information college gives are bigger questions.


Doesn’t that argue that neither college test or the standard tech interviews are effective.


Excellent question. I’m of two minds here. As a high school student, I too struggled to get great test scores. I took the SAT four times, which finally resulted in a score that could get me scholarships.

Did I become smarter? - I sincerely doubt it.

Did I become less likely to fail out of college? …Seems unlikely, but I don’t know.

As an admissions rep, I understood that high scores, whether SAT or GPA, are certainly both proxies for a prediction: If I admit this student, are they going to fail out?

Unknown: It actually might be the case that students who get private tutoring are less likely to fail out because their parents can financially endure them struggling to repeat failed courses. I’d be interested to see a study on that.


Did your final score more accurately reflect your abilities? Probably. Can everyone increase their score by hundreds of points by repeated testing? Probably not.


Who knows? My first score in 10th grade without any specific SAT prep made me eligible for the only well known school I applied for my senior year - Georgia Tech. By the time I was doing specific prep for the SAT in 12th grade, that was only to get the award for the highest SAT score.

By the time I started prepping , I had already accepted a scholarship for a local college with the plan on doing joint enrollment after 3 years.

I didn’t do that either. I was tired of college after three years and just graduated the next year and started working. I didn’t really care about GT.


The underlying assumption in these anecdotes is that if everyone had equal access to test prep resources they would show an equal increase in performance. But the studies[1] don't bear this out. If someone is an outlier in terms of coachability, that probably means there were gaps in knowledge that could be filled quickly to have an outsized effect on score. But this seems relevant to suitability for college admission.

[1]https://slate.com/technology/2019/04/sat-prep-courses-do-the...


The same article states that even the minor change in scores:

> That means they ought to be irrelevant to college admissions officers. Briggs found otherwise, however. Analyzing a 2008 survey conducted by the National Association for College Admission Counseling, he noted that one-third of respondents described a jump from 750 to 770 on the math portion of the SAT as having a significant effect on a student’s chances of admissions


> to overcome his grades and emerge a genius on the SAT/ACT

Isn't it usually the opposite? Artificially high GPA from grade inflation and favoritism is "set straight" by a standardized test?


SAT scores seem quite unfair as a single indicator of performance. GPA at least provides a view of a much wider range over which performance may naturally fluctuate.


I had a 3.18 GPA and a 31 ACT, is that outlying enough?


From a state school’s Admissions perspective, where the unit test resolves to admit/deny, there is nothing particularly outlying about your scores from what I recall.

I see 31 and think, “admit”. I see 3.18, I think “probably admit”. I check your math scores (which is where a lot of people trip up), and see you got mostly Bs- heck, you’re on Hacker Mews, probably As, and I think, “admit”. Now I do the normalizing math and yeah, Admit. Congratulations!


Ooh okay okay try me.

I had a 30 on my ACT my sophomore year then dropped out of high school my junior year. I was getting F’s my junior year because I couldn’t stay awake and had maybe like a 2.3 GPA before that. Would have reduced drastically if I’d stayed in school.

It was only that high because some teachers were always super annoyed with me, and a lot of them would try to bend the system to give me a better grade but some just were like “homework is the grade sorry” even though I’d ace tests. Also my stepdad was doing meth so that caused issues.

So am I an actual rarity?


You might be an outlier!

But it sounds like you did not graduate high school or apply to college. If that’s correct, I would not have the context to know since the baseline of inclusion into my experience is people who graduate high school in some form - GED included, AND who provide test scores.

I’m really sorry to hear about your stepdad, wincy. I hope things are ok for you now.


Things are great for me now. Thanks! I’ve got a wife and two kids and just got promoted to a SWE team lead. Mostly self taught. College just wasn’t for me.


I replied to you in another subthread, but I'm curious how you would have judged me (I'm assuming "deny"), I had a 36 on the ACT (a perfect score) and a 1.97 GPA.


I like to think I wasn't "judging" anyone, but I know what you mean. Because we used both GPA and test scores, yes, you almost certainly would have been a "deny" with a 36 ACT and a 1.97 GPA.

If I had seen your file, it would have been the talk of the day. "Can you believe this file!" I probably would have talked to the Dean, maybe called you and given you some advice. That advice would have been something like:

"Listen. Don't tell anyone I said this, but, this school is more expensive than it needs to be. Heck, who says you need college at all? But if you really want to go here, you've got two options. 1. See if you can get into the local community college for a semester or two and then transfer your credits - we won't care about your HS grades anymore. 2. AND REALLY DON'T TELL YOUR PARENTS/TEACHERS I SAID THIS You could, in theory... just get your GED. That's a loop hole around us caring about your GPA."


I pretty much did option 1. I attended a community college and then transferred to the state school that was in my home town. A lot of folks did this anyway because it was much cheaper, you could do all your general ed at community college first and get an AA and then move to another school for your major studies.


Was there a secondary adjustment step to finalize admit/deny? Your heatmap-based method would almost certainly result in an unacceptable racial composition of the class.


I worked in the underrepresented student office, so I was keenly aware/curious about that possibility when I began. However, after reviewing thousands of applications, my experience, to my surprise, was that GPA to SAT/ACT scores remained highly correlated regardless of race. If memory serves (and it’s been awhile) the outliers who were more likely to have higher SAT/ACT than their GPA normally suggested, tended to come from predominantly non-white schools. I’ve seen so many reports to the contrary that this surprised me to no end, but that’s what I observed.


I see your point on the GPA/test correlation and that is good in that it reflects that grade inflation wasn’t a major problem while you worked there.

But my question about final admit/deny stands. If you look at the racial distribution of the upper performers of any standardized exam (your “heatmap tracing”), it is unlikely to me that a college would be fine with that class composition. Either there must be a secondary layer where your admit/deny’s are adjusted or your acceptance rate must be very high.


So one of the best standardized tests I’ve ever taken is the Subject GRE. When I was applying for PhD programmes I and a few of my colleagues took it. It was perfect - it tested both knowledge and critical thinking skills and experimental design. It was obscure so no material was available for specific training.

In the end the scores were so accurate in predicting eventual success - a decade later, the folks who got 96% plus are all either professors or deliberately chose not to, while the rest just took industry jobs (and in my opinion because they realized that’s best for them).

Here’s the kicker - every institution explicitly said they will NOT consider these scores as part of admission process. None of the people who got in the top percentiles (4 I know) made it to a top 10 institution in the US, while a bunch of others did because they had paper authorship in their undergrad. But paper authorship in undergrad has NO correlation with your actual scientific skills! It just meant you were connected and or a hustler. It’s sad that such an awesome test is deliberately ignored by these institutions.

The irony was that these tests were also not elitist. All you needed to do was thoroughly read Lodish and Lehninger and you’re good. We did study in a semi-premium institution in India, but by no means were we privileged by any special facilities or help (at least in the context of this test). The only barrier might be the test fee itself if anything.


Respectfully, I call bullshit. You can't quantify success in a PhD with a single variable. There are a billion ways research can go right or wrong -- irrespective of your personal pedigree. Your ideas might be too early or late for your community to grasp, maybe you appeal to the wrong audience, maybe you're unaware of an application of your research, maybe you don't have the right set of collaborators or need a perspective that, often times, emerges out of a lucky encounters with someone. I respect your experience but I don't want it to give people the wrong idea about research success...

> Paper authorship had NO correlation with research success

I think paper authorship demonstrates that you're willing to put in a non-trivial amount of work to persue a problem. That seems to be atleast one attractive skill in a PhD, wouldn't you agree?


Uhm, what exactly are you trying to get at? I said subject GRE is a very good measure of eventual success in academia, do you have a solid response to that or just a rambling tirade?

Paper authorship if the student is the first author shows grit and “gumption” I suppose? As if that’s what’s needed in academia at this moment (it’s important but not the main requirement). But almost no undergrad gets a first author paper. They get mentioned in the middle because they ran a bunch of sds gels. I wasn’t even interested in trying to become a professor and I got 10 papers before I finished my PhD, do you know how many I (or any of the folks I actually know who are now professors) had during our undergrad? Zero. And not for lack of trying. You know who actually got papers? The son of the department head.


The original comment, to me, reads more like "subject GRE is a definitive measure of eventual success in academia." I was arguing against the definitive part. Thanks for the clarification. Maybe it might be a good measure for your cohort, you, and people in similar situations.

> Almost no undergrad gets a first author paper

Maybe this is different in different fields but we have a lot of undergraduate first author papers in programming languages and machine learning. I mean -- through and through -- undergraduate students bringing up a topic, getting guidance from professors and senior PhD students, getting results by the end of the semester, and publishing results by next year. Even the people who end up "running the sds slides" either fall out by next year or end up working towards their own first author publications. I've always chalked this up to the experimental setup cost being very cheap in CS compared to in the "hard sciences" so most undergraduate students are already comfortable with all the tools they need to do research.

> I wasn't interested in trying to become a professor

I think this is precisely the variable that a standardized test cannot account for! I feel an "authentic" undergraduate research experience is successful if it helps students realize if research is right for them or not.

> ... papers ... the son of the department head...

I see where your frustration is stemming from. Sorry this was your first experience with undergraduate research.


I dunno, my snarky answer would probably be: Just as if they'd test for A, but you actually need to be good at B.

I've not been to any school in the US and I don't have a PhD, so all these acronyms like GPA only have a vague meaning to me. But as someone who kinda disliked school and didn't put much effort in it, I can see how school grades can be a bad proxy for university. I didn't even do that well in math in school, but I managed to go above and beyond with all my math classes for my CS degree. I'm not completely sure how the admission for PhDs work and if they look at your BSc/MSc tests... but again, I guess doing ANY work related to finishing a paper is probably closer to what you'll be doing later than e.g. your grades in database stuff when you'll be researching programming languages...


You can't quantify success in a PhD with a single variable.

Of course, but I'd bet the farm there's a very strong correlation!


I have 20+ years experience in academic research.

Connections and hustle are very strong predictors of success.

Test scores do indicate ability, and ability is helpful but far from sufficient.


Connections and hustle are stuff you can learn. At least enough to become a professor. Most of my nerd friends who became professors learned these during their PhDs. I didn’t say they became famous professors just folks doing decent research in different corners of academia. I’ll still get my house in the subject GRE as the single most important metric if I were to choose a grad student. Not that I’m in that game anymore of course.


Can you clarify for me, a non-native speaker, what you mean with hustle here? Energetic activities or fraud? (I guess the first, but I also heard of a lot of academic frauds)


>But paper authorship in undergrad has NO correlation with your actual scientific skills!

This seems like a bad take because the goal of PhDs in the US is to produce people who can publish a lot of papers, whether that's a good ideal or not. So having produced papers as an undergrad gives you a powerful signal that you are such a person.

>Here’s the kicker - every institution explicitly said they will NOT consider these scores as part of admission process.

I know in my field, Physics, the subject GRE was very important for admissions and outside of the recent changes due to Covid I believe every top program explicitly requires the Physics GRE for admissions and its not until you get to much lower ranked programs that it becomes optional or not used at all.


> a decade later, the folks who got 96% plus are all either professors or deliberately chose not to,

> All you needed to do was thoroughly read Lodish and Lehninger and you’re good.

How can these be true at the same time? If you can just read a book and get a good score, why don't everyone just get the good score?


Lodish (a then popular cell biology book) and Lehninger (a 1500 page biochem tome) are substantial books. 101 courses would cover the first 5-9 chapters of 25. Reading back to back and understanding all of it puts you at the top of the best students in biology in the world.

I was lucky, I genuinely loved reading them. I think I went through them both once every time I had a bad breakup lol.


It's clear they mean that the barrier to entry was available to anyone. That doesn't mean that anyone has the capacity to understand and apply the contents of the text.


> How can these be true at the same time? If you can just read a book and get a good score, why don't everyone just get the good score?

All you need to do to get 4.0 in Calculus 1, 2 and 3 and Real Analysis is work through Stewart, doing all the questions. Just because it’s simple to describe doesn’t mean it’s easy.


>while a bunch of others did because they had paper authorship in their undergrad. But paper authorship in undergrad has NO correlation with your actual scientific skills! It just meant you were connected and or a hustler

uh, the whole point of a phd is learning how to produce scientific knowledge usually in the form of a paper? I can't think of a better measure for that then if you have experience publishing papers because that means you've contributed to the production of some scientific knowledge? sure there's tons of variability there due to circumstance, but I think my point holds

and finding connections is bad? the scientific process is one of cooperation, teams building off the work of other teams. the lone polymath is a long dead myth.

and again, having hustle is bad? I think you'd want someone who can be self motivated enough to get involved with research going on at their university


Getting a paper in undergrad has absolutely nothing to do with your ability to be a good scientist or even a good PhD student. I’m talking about regular smart kids though, not people you could call a True Genius™. As I mentioned above, most undergrad paper authorship is peripheral where you just succeeded in establishing yourself in a lab long term and contributed to some research.

It’s great experience but I am saying again that my personal experience suggests this to be a worse indicator of success than a good standardized test.


>As I mentioned above, most undergrad paper authorship is peripheral where you just succeeded in establishing yourself in a lab long term and contributed to some research.

and what do you do in a research lab as an undergrad? you contribute to research of course. you plan, monitor, and execute experiments. you write up the results and their impacts, limitations, and related aventures for future work. you work with others in a research environment. you gather the nexessary background knowledge required to understand the field of research. in short, you do most every you would do in a PhD just with more guidance. you're not giving much reasoning beyond your "personal anecdote".


All of us did all of the above - no one finishes undergrad without extensive lab experience and expects to get into a good institution for a PhD directly. The main differentiator is whether you get a paper out of it. I worked every evening and summer all through my undergrad in multiple labs without ever coming close to a paper. This outcome is often out of the undergrads hands - except again if you’re connected or are a hustler just hunting a paper without trying to get a variety of experience.


How many PhD students have you mentored? Do you run a lab at a major University?


As I mentioned above, I chose not to go the academic route past my PhD. I don’t believe it’s a viable system and want no part in perpetuating it. Not that I didn’t have a choice (my professors actively encouraged me to do so, which they typically don’t to most of their students).


I feel like standardized tests are way more fair than GPA.

I personally struggled a lot in school due to health and other issues. I also hated doing homework which is BS work, and I always had high scores on my midterms and finals which I think matter way more.

Why does busywork play such a big role? It should be exempted. If I can get a 95+ on midterm and final I obviously understand the material. Why did I need to spend hours of my day doing BS exercises when I already understood the material? Nonsense.

So I get a B at best overall because I get a F on homework? What a joke. I really think school is a waste of time when it’s done that way. Learn stuff, move on. I wish that when I have kids I can afford to send them somewhere they can actually grow and thrive, not just sit down and do a bunch of homework problems.

(Just my $0.02)


> If I can get a 95+ on midterm and final I obviously understand the material.

That's assuming that these midterms/finals accurately and completely assess understanding of the material.

As a teacher, I need more assessment moments than one or two tests in a term to monitor, guide, and support students' learning processes, for all students in my class. This "busywork" gives me the extra input I need to adapt my teaching to my classes and students.

Of course, from my perspective, this isn't "busywork" because I try to create meaningful learning activities. However, from the perspective of some students, I can understand it feels like busywork. I try to differentiate, but that's not always possible, particularly if you're performing significantly above average.

The problem is that even if you are doing fine without the busywork, your classmates might not and need the training through busywork. Or they might need the structure and incentive to do something at all. Unfortunately, giving some students no work and other some work and other a lot of work will not go down well, particularly in larger heterogeneous groups. So we end up treating everyone the same, more or less. This can really suck for any student who's an outlier to the positive OR negative side.

Ideally, you'd get an education tailored to your needs, but in practice that's difficult to realize depending on where you live, your parents' background, etc. For example, if you're living in a smaller village with just a single high school, you don't have much choice. Similarly, if your parents don't understand what you need, they'll not go look for it. Or if your parents don't have the means to support your alternative schooling, bad luck. Etc.


A professor I had had a pretty good solution. Give homework, but only make it 5% of the score for the class. A couple of quizzes are 5% each, and the midterm + final are the remaining 85%. That way, if a student needs the structure, they get it, and on the other hand if a student is able to ace the tests without doing the busywork, they can still get their A or A+. The instructor also then gets lots of input points throughout the term.

Some students (particularly those who spent lots of time on the HW but didn't do well on the tests) found this unfair, but I think it teaches an important lesson: that you don't get credit for trying. Your performance actually matters.


I agree with many aspects of this; however, I do think learning how to grind through work is important. However, I remember being a child and a teenager trying to do the designated school and extracurriculars and then homework felt insane. I would often be doing work til 11. This does not feel like a way to inspire children or make them excited to learn. I loved school in many ways, but the homework caused so much stress when I think if I had more time to explore intellectual interests I think I would have more excited about many subjects that weren’t my favorite


Most real world work is grinding through boring work as well, and it’s worthwhile to build the discipline for it.


There are better ways to build discipline than spending a huge chunk of ages 5-18 doing otherwise pointless busy work.


But are there better ways to build obedience?


Completely agree with this. I used to skip most of my classes but would get As on all of my tests.

Now I struggle immensely on going through the every day grind of adult life.


Correlation or causation? Perhaps if you didn’t skip classes, you’d still struggle through adult life but also struggle through childhood.

Perhaps you’d just be more numb.


That's why I don't do "real work" anymore. You understand not everyone has a goal of being a salaried employee in life, right?


What do you spend your time on instead, and how do you support yourself? Unless there's some other definition for "real work" that still involves money but isn't some form of a job - be it self-employed, entrepreneurial, or otherwise.


I became rich and retired. I spend my time on myself


Perhaps it's worth sampling the population and seeing whether outcomes match. I'll go first. I didn't do this. My parents intentionally did my cursive writing homework for me so I could read and go out and explore the countryside.

I now run eng at a prop trading firm that you'll occasionally see in the FT.


It was the arbitrariness of academic work that got to me. Real world work tends to be clearly directed at a goal, and if you acheive the goal it doesn't tend to matter much how you did it. Academic assessments (poorly designed ones, which happens to be most of them in my experience) expect you to do things exactly in a prescribed manner. Essentially 50% of what you're doing isn't actually useful learning, it's domain-specific learning where the domain is "what my professor happens to care about".


I wish you were wrong, but you're not. Ability to work a BS system absolutely is a marker of success in society.


Well, I have the opposite opinion. First of all because tests don't remotely represent the nature of doing work. It's a momentary expenditure of energy, and all of rides on perfect performance that day. Luck, rote memorization, and some skill development.

So in my view, I read it as "Why should I have to do bs when I'm lucky and tend to do well on tests?"

In my cs classes, I did terribly on tests, and got 100% on all the work, so I failed. Why? Well I have chronic sleep issues and miserable short term memory that's gotten worse over time, and the mid-term was 40%. I also just don't really care about the outcome, because it doesn't represent my ability to do any of the tasks in the real world. Oh, I was too slow in writing out the exact java syntax for a generic implementation of a priority queue implemented with an array? Tests aren't inherently virtuous. You have to believe in them to do well, and you have to feel adrenaline in the process. Usually has nothing to do with incremental progress, measurable skill, or anything valuable really. I can now recall various key bits of key events from the Byzantine empire, but did I do well on the test? No, I was a bit tired that day.

Both should be balanced, and relevant. The tests and the homework should meaningfully test whether or not I can do what they intend to indicate. Not "He couldn't sit there for 3 straight hours from 6pm to 9pm on one night of the year writing out java in pencil in a dead silent room.


WTF is lucky? You can hit all of the questions you know in one test. You can't be "lucky" on every test, that's a ridiculous claim


Lucky in the sense that if 50% of your whole grade rides on one block of time on one day, you better hope to have a really good day. If you sit down and realize "sheesh, I have no mental clarity at all today", you're kinda fucked.


That's why there are midterms, if it's only 25% of your grade, getting a bad grade on one doesn't mean you can't pass the class

I would prefer having quizzes on each chapter as well, so every few weeks you review the material


The flip side of the fairness argument is that test scores are just one time point (often just a couple of hours, maybe several hours at most, of one day) in one format with no flexibility for unusual circumstances or anything. GPA, in contrast, is reflective of repeated things over a long time period, with lots of variables accounted for.

There's issues with GPAs too, like lack of standardization, but they both have their issues with fairness.


It seems to be the case that GPAs are more important with scholarships than admissions, where boards are looking to find any edge to be able to differentiate students.


If this were the case then couldn't we just make high school diplomas, heck even university degrees, a series of midterm and final tests? Why have lectures at all when people can come in knowing the material online? What if you could start "school" in one week, test nonstop for the next month, and have your degree in CS assuming passing marks at the end?


The implication of the “test scores don’t matter” argument is that we have the wrong people staffing virtually every credential-based position. There’s lots of kids with 4.0 GPAs from high school at directional state university getting education degrees. Test scores is what distinguishes those kids from the ones at Stanford. And Stanford is where everyone from Big Tech to Big Banks to Big Law to Big Consulting goes to hire. Who is working at Pfizer or Astra Zeneca doing drug development? It’s the people with high test scores. Every single Supreme Court justice in the last half century got there based on test scores. (The nominee whose test scores led her to a law school ranked outside the top 50, Harriet Miers, has to withdraw because of that.) Warren Buffett got into Wharton based on test scores. Charlie Munger got into Caltech based on his scores on the army’s intelligence test. Bill Gates got an 1590 on the SAT. Even the professors saying test scores don’t matter got where they are because of test scores.

Maybe that’s all true. But it’s quite a remarkable claim that would turn the world as we know it upside down.


Let's steelman that a bit.

If you structured society such that only the top 0.5% of households by income could send their kids to college, you'd find that a lot of your successful entrepreneurs, judges, scientists, etc. went to college and had a high income growing up.

But I don't know that proposing something else, like standardized testing, as an alternative to this would be quite a remarkable claim. You're just saying that the scoring function produces a lot of false negatives.


In your hypothetical, institutions would quickly realize that a bunch of non-college graduates were just as smart as college graduates. Which is actually the public perception in places where jobs are filled based on social status and connections rather than objective metrics.

What I’m saying that the current scoring function appears to match up pretty well with our intuition about who is smart enough to work on Wall Street or at Facebook. At least, these profit-seeking institutions don’t seem to think it would be better for them to shift their hiring from Stanford to kids with 4.1s and lower SATs who went to UCSB instead.


> In your hypothetical, institutions would quickly realize that a bunch of non-college graduates were just as smart as college graduates.

Which would cause those institutions to move away from income as a means of deciding who gets in; much like in institutions now are moving away from standardized testing as a means of deciding who gets in.

> At least, these profit-seeking institutions don’t seem to think it would be better for them to shift their hiring from Stanford to kids with 4.1s and lower SATs who went to UCSB instead.

There's two assumptions here that you may want to reexamine:

1. Profit selling institutions are purely rational at hiring, and only hire based on merit. As in, if you're hired to work at a Wall Street firm, it's not because you went to Harvard; it's because you're smart. Which is very optimistic, to say the least.

2. That people are intrinsically smart or not, and the institution you go to doesn't change that. If test scores didn't correlate at all with merit, you'd still want to hire from Stanford over UCSB; not because the SAT scores are higher at Stanford but because Stanford is a better institution for educating people. So even if the people who go in are the same caliber, the people who come out are different.

To be clear, I'm not disagreeing that test scores can measure aptitude. I'm disagreeing with the line of reasoning where we have the following premises:

1. We live in a society where resources are allocated by performance in standardized testing.

2. Having more resources increases the likelihood you're successful.

3. People who are successful have high standardized test scores.

And conclude that standardized test scores are an accurate predictor of future success.


So a test score opens doors. It’s more of a statement about how our society and institutions are structured. That implication says nothing about the quality and acumen of a person once they step through those doors. Test scores do not determine qualities and acumen, only that you meet a standard (acumen and standard is not always one-to-one). All those you named got the opportunity to demonstrate their qualities and acumen because a test score opened a door for them to do so. So I don’t believe test scores have ever mattered to society or institutions other than that: opening a door and not a guarantee of any sort of success once you step through it.


I think what you're pointing out is the point of the "test scores don't matter" argument actually.


I don't think test scores are a predictor of success, but I do think they are one of the best metrics we have, and actually help poor kids to "make it" to good schools. Rich kids might have access to better teachers, but Math is the same for everyone, a poor kids using materials freely available in a public library can do very well in exams, allowing him to get into an elite school, otherwise only reserved for the elites.


There's no shortage of students at state universities with elite test scores.


Test scores are accurate at predicting your ability to succeed at further testing, I don't know if this is all that controversial.

What people are getting up in arms about is the validity of testing as a whole regarding real life outcomes, and this is an effective argument as it is damn hard to grade people by any metric that doesn't get gamified to death and resulting in people wasting a ton of time on playing the game.


I can't believe I had to scroll this far down to find this comment! People are just assuming that positively correlating testing with further testing somehow implies that the way we test is indicative of future work performance or even your quality as a human (whatever that means). They may be correlated, but a person's ability to sit exams moves the needle much more significantly. More so for universities — I've taught many students who graduate with great scores that I would never ever hire.

Using the same methods to measure wildly diverse people is a tragedy that causes individuals who could otherwise contribute to society in very meaningful ways to have fewer doors open to them.


A student who scores highly on a Calculus exam may be generally good at Calculus, or maybe they know a couple exam-specific algorithms to succeed and got a little lucky besides.

A student who fails a Calculus exam is nearly guaranteed to be poor at Calculus.

To say exams have no predictive power is to ignore the obvious.


Student A that passes a Calculus exam may be be more likely to know Calculus than student B that fails but so what?

Passing a Calculus test doesn't necessarily predict if the student can apply their knowledge to production to generate new ideas and widgets which make our lives better, it just predicts that Student A can follow rules and learn something _when required by an authority_.

I think this sentiment is similar to how people feel about Leet Code interview questions. So what if an engineer can invert a binary tree if they can't formulate new, useful ideas and convince others to work on them?

Perhaps student A is more likely be create new, better widgets than student B but I think academicia/LC cannot test this directly, which is what people are getting at when they criticize tests.


>Passing a Calculus test doesn't necessarily predict if the student can apply their knowledge to production to generate new ideas and widgets which make our lives better

It actually does though, because if he fails, then you can be completely certain that he can't.


I know this isn't true because of my own experience.

I've always been bad at math, I failed precalculus multiple times in both highschool and college, I never got past my first year of calculus in college, I wasn't able to pass descrete maths, and I never took an algorithms class but, never the less, I am applying statistics to novel problems at my day job and people are lining up to work with me and my career growth is accelerating faster than those around me who certainly got better marks than I did.

How is this possible? I'm not entirely sure, however, it indicates that academic tests don't actually measure ones ability to apply themselves and their knowledge to real problems that people actually have outside of an academic setting (i.e the real world), which I believe is the primary point in the "anti-testing" narrative.


But do you actually calculus and are the problems you attack actually difficult?

Failing something multiple times in the past also doesn't mean that you don't know it now.


No, you can't. You can only possibly be completely certain that the student is not good at calculus tests.


If you can't do calculus in a stressful situation, you can't do it, anymore than you can understand french if you can't understand it when people speak it at full speed.


Huh? How often is calculus ever done in stressful situations outside of tests? Seriously this seems like a bad example...

I feel like you're making my point here. I can speak french and be spoken to, order off a menu, etc. but I can't understand when two native speakers are conversing. Are you saying I don't understand french?


When the deadline for your papers are approaching, when your competitors are closing in and about to scoop you, when you're using calculus in order to make financial trading decisions, etcetera.

Everything is time constrained.


> Everything is time constrained

Yeah but not to the same degree or in the same way. Having to sit in a room with no sources for an hour and shit out some test answers is not really same as having weeks to apply yourself with access to appropriate resources, references, and guidance from more senior coworkers.


The way I see it, you should be able to do calculus without pen or paper, so certainly without references or guidance.

If you want to be creative the first step is to put the problem in your head so that you can actually ponder it. If you have to pour over references and look things up, then you're not in the game any more than you can play chess if you have to look up the rules as you play.


This doesn't align with my experience writing software and applying maths to engineering problems. I am working a lot with statistics in my current role but if you asked me to take a stats test, I would almost certainly fail without references. Does this mean I am bad a stats? Perhaps? Does it matter? No.

> If you want to be creative the first step is to put the problem in your head so that you can actually ponder it. If you have to pour over references and look things up, then you're not in the game any more than you can play chess if you have to look up the rules as you play.

It seems like you are confusing the problems with the solutions. If you can understand the problem well enough, the solution will become obvious, even if you never took the math class--at least that's how it works for me. I do best when I have an actual real-world problem to solve, not a conceptual one created by an instructor.


So you have all the Laplace transforms memorized just in case some internet person asks you how to model a wheel going up a ramp at the origin?

Everyone on my engineering course did this question, but my guess is zero of us could solve it right now from memory. A fair portion would be able to solve it with some references available.

But you're making the other guy's point for him here. It's a waste of your mind to have test-taking skills memorized.


No, it's completely against my philosophy to memorize things just to remember; and things like Laplace transforms you look up in 'Beta: mathematics handbook' which you're typically allowed to bring to tests in engineering and in applied physics, at least here in Sweden.

What I mean is that if you know calculus you can do it fluently, like a language. You don't remember the solution, you calculate it, preferably in your head.

Edit: Though, back at my university tests took five hours, and I understand one or two hours is normal in America, and I could imagine that those shorter tests mean that there's more memorization and less sensible study.


This is not clear to me, on one hand you say it's acceptable to look things up, on the other hand you should be able to calculate things in your head?


It's acceptable to look things up that don't matter.

The way I see it, you should have the machinery in your head, so that you can derive things if pressed, and you should have the ability to work largely in your head and actually ponder things, but you shouldn't necessarily memorize formulas.

For example, I haven't even memorized the formula for the quadratic equation, I just complete the square in my head every time.

I think it's important to end up knowing them and to end up being able to calculate with paper and pencil to some degree and because calculus is very basic it's something one should be able to do in that way.

You should be able to imagine a right angle triangle sitting on the hypotenuse of another right angled triangle and figure out the x and y coordinates of the corner and how to get the sum of angles formulas from that, or to in general imagine things from calculus, like how the product rule can move what what is differentiated from one function to another etcetera.


I've never worked in a company where we were time constrained or stressed. I could always tell my boss "it's not going to be done by the deadline" and he'd have to come up with a way to do plan B.


Yes, the deadlines come from customers and the competition, not from bosses.


Then it's not necessarily a "deadline", but a suggestion.

Most of the time it's better to do something the right way than to artificially hurry developers to hack something together that will bite you later in the form of bugs


Yes, although in this I was thinking less about software and the like as such and more about things like pricing derivatives contract that somebody wants and things like that, and of course, then you must do the right computation.


Again, though, if creating new, better widgets requires calculus, student A may not have what it takes, but student B almost certainly can't do it.

Some jobs require a creativity, a spark, a something that you can't test for. But they also require some solid skills, and those skills are testable.


> Again, though, if creating new, better widgets requires calculus, student A may not have what it takes, but student B almost certainly can't do it.

I'm not so sure about this. What if student A was just better suited to learn Calculus when they were tested while student B bored by the class and teacher and stopped going/caring but, if given the chance to apply themselves in a non-academic setting working on real problems that affect people's lives, student B would learn Calculus twice as fast as student A and deliver a better widget in less time?

There are many more scenarios here that I could list but I think the primary point is that the context around the students, their lives, and the testing environment matters a lot and this is something that cannot easily be accounted for using numerical test scores

> But they also require some solid skills, and those skills are testable

Maybe badly. Leet Code is supposed to do this but is famously criticized. Probably better than nothing I suppose


Environment matters

Nobody says that tests are perfect

They are just the best thing we managed to find

Go ahead and design something as fair as tests that takes env/context as input too.


Go ahead and figure out this method for determining who is genius and who is not

But that method will need to be able to predict future events that shape the person

Tests dont say that you cannot be successful, it just says how prepared you are, now.


Not sure why you are being flippant.

I am trying to point out that the methods we use to "determine who is a genius and who is not" are deeply flawed and rudimentary, letting people fall through the cracks and preventing people with different learning styles for succeeding.

Like I said in another comment, I am terrible at testing and failed my way through school but I am still able to have a successful career as a software engineer applying maths to engineering problems that other people are excited about and want to work on. The reasons for this are complex because I am complex, like all people, but not like numerical test scores.

> But that method will need to be able to predict future events that shape the person

I think individuals are capable of qualitatively assessing a person's ability and trajectory given enough information, time, and context. Humans may come to the wrong conclusions a lot of the time but so do academic tests...


Because while tests arent perfect, then they are the best available method that is transparent/fair at scale

If you want the change, then propose an improvement.

________

>I think individuals are capable of qualitatively assessing a person's ability and trajectory given enough information, time, and context. Humans may come to the wrong conclusions a lot of the time but so do academic tests...

Ive heard countless stories like teacher telling student that he will be not even close to being successful due to not giving a damn about school and yet they manage to do fine

So no, i dont think people of school are capable of this at scale

You would need smart and open minded and up to date people to at least try to do that reasonably


> If you want the change, then propose an improvement.

I don't need to have a solution to have valid criticisms of an existing system. I am just sharing my experience with testing, how it has affected my life, and why it didn't work for me.

That said, I do think we can improve our ability to assess (not test) people by:

* Putting less emphasis/focus on tests and more focus on long form project work and interpersonal relationships with peers and instructors * provide more flexibility to students when testing them (like access to resources/notes, more time, allow them to ask questions etc).

> You would need smart and open minded and up to date people to at least try to do that reasonably

I would certainly hope that the people teaching the next generation were smart, open minded and up to date. Testing doesn't control this anyway, a bad teacher will be unfair and have bad curriculum regardless if they test their students formally or not.


>I don't need to have a solution to have valid criticisms of an existing system.

I agree.

>I would certainly hope that the people teaching the next generation were smart, open minded and up to date. Testing doesn't control this anyway, a bad teacher will be unfair and have bad curriculum regardless if they test their students formally or not.

Hmm, I originally didnt think about those people as a teachers, but as if they were some kind of examiners.

But thats fine, they could be teachers too

Yet, I do believe that the bias would be too big :(

At every level of eduction except higher teachers were visibly biased towards students

Some benefited from it, some had harder time due to that.


I remember doing so well in calculus until the 3D graphing unit.

I think I ended up with a D or worse on that one units test. My friend was like "oh bro you just add a pi to everything". This isn't thanksgiving and it's more complicated than that!


Barring fraud, standardized tests are not gameable, unlike GPA, references, etc. You can study and improve your score, of course (just like you can study and improve GPA) but that's not "gameing" in the subversive way you mean it. Their ungameability is inconvenient for people that are looking to obfuscate the real admission criteria, which is why there is enormous effort to downplay their predictive performance.


You can absolutely "game" tests.

They're taught as "test techniques." And there's plenty of tutors who teach it.

Process of elimination, read the questions then passage, common question structure, are all examples.

These are "skills" but they don't really apply outside of testing.

Hell I went into standardized tests with common question solvers programmed in my calculator. If that's not gaming I don't know what is.


Most of these techniques are picked up with a few mock tests and they only matter on the margins - people that obsess about these things tend to care a lot about rankings rather than absolute scores. They make a 1550 SAT into a perfect 1600. They do not take a 1000 into 1600.


Absolutely there are limits.

But there are limits to GPA gaming and references gaming too

I'm not trying to say tests are useless or anything, but they're definitely gameable.

Test-taking is a skill that many students learn that's not applicable outside of school really


> there's plenty of tutors who teach it

There are also $30 books that teach it and youtube videos that give it away for free.


Sure, but you have to know to do it. The game isn't hard but you have to know to play.

I don't think it makes tests useless -- almost everything is gameable.

But claiming that tests are ungamable objective measures of knowledge/ability is not true


I've only gone to school in Europe (Norway, specifically) - so I'm not sure how one can game their GPA. I'd say that a solid 80% of my degree was graded/scored on written or oral exam. Rest were home exams, or a mix of both - with 10% being the Masters thesis alone.

But then again, we don't have any standardized tests here. Not for college admission, or afterwards. The grades you get, are what you have - and everything is judged by those grades.


College graduation rate is a real world outcome, which is author is referring to. You don't have to be particularly good at testing to graduate. Cs get degrees. A lot of courses are project or paper-based. If all else fails, you can take GPA boosting classes.

Of course, the far left will claim that "real world outcomes" are also systemically biased when they don't align with their priors.


> Test scores are accurate at predicting your ability to succeed at further testing

Very well put. The whole question of "do test scores predict college graduation" is easily conflated with "do test scores predict success/ability". In reality, college is a big set of tests. It would be interesting to compare with other metrics.


My take is that standardized tests are a road to social mobility. People can bullshit their way to higher grades but there are some rich kids who can't test their way out of a paper bag and the one way we can stop them from laundering their parents accomplishments is standardized testing.


Except you can buy your way into tutoring too, and pay for test accommodations, and so forth and so on. It might not amount to a huge effect, but if you're on the boundary area, it can make the difference. Someone might also argue that although bullshitting with grades makes a difference, it usually isn't going to turn a D into an A either (speaking as a university professor, with many colleagues at different universities, in those situations it's usually a matter of being pressured to turn an F into a D).


> you can buy your way into tutoring

You still have to listen to and understand what the tutor says, though. And it's not as if the tutor is saying anything that's not published in books or on YouTube videos - the only real benefit a private tutor could even theoretically have is that he could potentially find a few different ways to explain a concept until the student understood it.


>the only real benefit a private tutor could even theoretically have is that he could potentially find a few different ways to explain a concept until the student understood it.

Yes that's one benefit. Also with the focus on one student the tutor can gear lessons to best suit their learning style and current knowledge.

Really anyone who can afford a tutor would benefit from one.


studying for standardized tests doesnt work.


Except you can buy your way into tutoring too, and pay for test accommodations, and so forth and so on.

I don't buy this. It's like saying someone has a remarkable physique or remarkable fitness only because they used PEDs when it actually requires a significant amount of effort and discipline regardless of what's in the gas.

Certainly targeted studying, tutoring, test accommodations, etc. can help. But the difference is at the margins.


It certainly helps to have some familiarity with standardized tests, particularly to take the PSAT and then take the SAT multiple times. Some people "freeze up" in that environment or they don't understand the strategy of "I've eliminated 3 out of 5 options, I should guess one of the other two".

Test prep helps the average person test at their level, but there are some people who can at best make excuses like "I must have skipped a line when I was filling out the form" and for those people the one thing that helps is hiring a ringer to take the test for them.

School on the other hand offers vastly greater options for people to turn financial and social capital into false achievements in terms of chiseling for grades, expensive sports and other activities, etc.


You would be surprised at the number of kids who have tutoring... and still can't test their way out of a paper bag


It's the B to B+ and B+ to A chiselers that bug me the most.


I see it the opposite. You can hire people to prep you for the test and pass it while having done zilch during HS; whereas bs'ing your way to higher grades over a period of several years is harder.


Well, no, because schools have a profound interest in having those students parents pay them for a top notch education.


How does that stop them in any capacity from living off their parent's wealth?


I don't think that's the point the parent post is making


Cute. Disclosure, I make the data (transcripts, test scores) flow to decision makers as part of my day job at a Big 10 University. We just went "test optional".

I view "test optional" as a signaling exercise. In brief, how closely can you read the instructions and infer what you are not being told? The key is "test optional". If you sit for the test, you get more than just admissions, as merit-based financial awards are unlocked. The University Admissions does not require a test, but many of the donor contracts that fund scholarships do require those tests.


> I view "test optional" as a signaling exercise. In brief, how closely can you read the instructions and infer what you are not being told?

Sorry, what does that mean - who is the "you" in that sentence, and what are they inferring?

It sounds like what you mean is that that your school went test optional (ostensibly to help level the playing field), but those who are lucky or privileged enough to have the secret decoder ring can decipher the instructions and find the key to optimizing their chances of admission. If so (as someone who grew up with immigrant parents who were educated but definitely would not have had the awareness to navigate this), that seems really fucked up...


The English language has a plural "you". The word "you", in this case, means "anyone who applies for admission". I said nothing about increasing or decreasing your chances of admission.


It could also have easily meant "you" as in the person evaluating the application, "you" pedantic ass.

If it doesn't change you chance of admission then what is the applicant supposed to be infering from your cryptic instructions?


Interesting, and good to know! But wouldn't you say that students are better off not submitting their scores if they scored low on the tests? Only given a sufficiently high enough score does it become a "nothing to lose" situation.

Unless the issue is students just don't want to take the test to begin with.


I'm saying that you can be admitted without test scores. You leave money on the table when negotiating the bill when you do not take the test. Opting not to take the test could be an optimum strategy if you already know that the offered awards will not apply to you.


I see zero discussion of test scores and people with learning disabilities. Those who cannot be properly accommodated will be unfairly represented and often have to put up with a lot more barriers (such as frequently messed up accommodations on test day leading to anxiety). Usually these accommodations help normalize the specific testing scenario and people with disabilities do not need as drastic accommodations in real job setups. The tests are designed for the average test taker but have lots of hiccups for those that deviate from the average.

It’s a difficult problem to solve, but it’s unfair if a student gets easily distracted or anxious when taking a test around others in a time constrained setting. Those types of external factors generally shouldn’t be represented in a student’s test results yet often are. If tests are to be a good indicator for all students, I think accommodations have a ways to go.

If tests are indeed a generally good indicator of a student’s success and are continued to be used in such a manner, I feel it’s important that the experience is fair for all types of test takers. Many US tests such as the CPA exam do a great job of handling accommodations. However, many others (like the PSAT) have a lot of room for improvement.


> I see zero discussion of test scores and people with learning disabilities. Those who cannot be properly accommodated will be unfairly represented…

I have long wondered how these interact / can support people who are a couple of sigma out from what we have decided is the “norm”. I can see how additional test time can help someone with anxiety or distraction issues, which can help someone with other strengths shine.

But how can this work outside the academy? A trial lawyer has only so much time to work on the case before it goes to trial or before some response to a filing is required. An engineer building a rocket still has to get something designed before assembly. Arent such time accommodations are hard to implement in the “real world”?

There are nice counter examples to my question. For example dyslexic people struggle with all sorts of cases, yet technology can help many of them (e.g. new kinds of fonts), and not just current students but people already in the work force.

I can also see the school accommodation helping someone who is separately working on their anxiety issues — though mental health cost support is quite poor, at least in the US.

But at their root: do these accommodations help or do they provide an unrealistic hope to the student?

NB: I want everyone to have a fair shake, and am a fan of ADA accommodation and affirmative action and related diversity efforts. I ask this question within that context.


... so here's the thing (speaking as someone who has taken many of these tests, been in admissions committees, and researches these tests). Take learning disabilities, and now expand it out to consider the similar factors, but with things like cultural background, life circumstances, and so forth and so on. Learning disabilities are not the same as those other things -- I don't want to equate them -- but they share some of the same issues with the test being normed on a certain standard population, and once you get outside of that, weird things happen.

Those density gradient plots in the linked blog post are interesting and useful to think about, but they're kinda hiding the fact that the vast majority of the data going into them is based on test-takers with a certain standard background, characteristics, and so forth. That in turn shapes the contours of what is error in those plots and analyses. The problem isn't necessarily that the tests are useless per se, it's that it's hard to interpret them in a way that accommodates people in nonstandard situations, or even accommodates the idiosyncracies of differences between people in general, differences that wouldn't matter in the real world.

The errors of using ACT-only, GPA-only, and so forth in the post is pretty interesting, but the author is missing the fact that that table has never really been the point of contention. The point of contention is whether that table's patterns apply equally across divisions of gender, race, SES background, age, disability status, and so forth and so on. They raise the point that the improvement in error from switching from ACT to GPA is comparable to (or better than) adding ACT to GPA, but isn't the real question whether doing so increases certain types of "predictable error variance", in the sense that you could predict the residual from things like SES, race, and so forth and so on?

I think GPA is seen as more acceptable than ACT because it's exchangeable in an important sense with the criterion being predicted. That is, if you want to know college GPA, maybe secondary school GPA is a little flawed, but at least it's ostensibly similar in terms of what it actually is. I think people have a sense that, say, you aren't using standardized tests as the criterion for college graduation, so why use it for prediction? Why have college and college GPA at all? Why not just let people take standardized tests and skip the whole degree program thing? There's reasonable arguments for doing that, but also reasonable arguments for not doing it, and many of those are the same arguments for and against using the test for admissions.


Tangent - I think that my particular disability (ADHD) helped me with my test scores. I was terrible at studying, but on the tests I would hyper-focus and do extremely well.

On the other hand, on the GRE, the writing portion was last and I could not bring myself to care. I scored in the 50th percentile, which for someone who regularly scores in the 99th percentile on standardized tests (including ones with writing portions), is pretty bad.


This whole concept of measuring SAT/ACT scores or GPAs and using them to "predict" graduation rates is bonkers.

Sounds a lot like the precogs in Philip K. Dick's Minority Report.

Several decades ago, I had a high-school History teacher. We had an elite private high-school in the same city as I lived (India has both private and public schools). They had an extremely rigorous selection process, including infamous interviews with parents for admission to kindergarten. Our history teacher once remarked, "They take horses and send them out as horses. We take horses, asses and everything else and send them out as horses".

The SAT/ACT/GPA filter seems to be focused on selecting horses and proudly declaring all horses graduated.


> They take horses and send them out as horses.

Correct. Although there are huge benefits of a gate-kept horse-only environment. Competing against each other, the horses become more excellent. This is why the parents are paying the big bucks and putting up with the process. It is about gaining access to a horse-only environment.

> We take horses, asses and everything else and send them out as horses.

This is just envy at the easier job of the private school teachers, not indicative of their actual transformational capabilities of the public school teacher. The best that these teachers can do (which is no small feat) is to manage the classroom such that the donkeys and the horses do not fight and the horses can stay motivated despite having fewer peers.


Test scores have already been irrelevant for certain segments for over a decade, speaking as someone of East Asian ethnicity. Personal anecdote: 3.7 GPA in average high school with 8 APs (some of which were self-study), 2400 SAT, 36 ACT, lots of state-level top 3 finishes in multiple academic competitions, varsity track, musical instruments, first generation immigrant, etc.

Ended up going to what I thought was my safety school and rejected/waitlisted at every reach school I applied to. I sometimes wish colleges mandated name/ethnicity-blind application reviews - not to sound ungrateful, but I’m still incredulous to this day that the best I could do was a public state school (still top 50 admittedly).

It ended up being a great experience in that it was a forcing function and made me realize most of these rules and expectations around admissions were meaningless. But I probably could have saved years in high school exploring things that mattered rather than optimizing for a college application process that didn’t ultimately end up feeling very fair.


Glad it worked out for you. Asian college applicants have it rough these days.

This is the President of Stanford speaking out of both sides of his mouth:

2022-10-12: "Stanford apologizes for admissions limits on Jewish students in the 1950s and pledges action on steps to enhance Jewish life on campus". Stanford President Marc Tessier-Lavigne apologized on behalf of the university and pledged action on recommendations in a task force report confirming Stanford limited the admission of Jewish students in the 1950s.

https://news.stanford.edu/report/2022/10/12/task-force-repor...

2022-08-02: Amicus brief filed in support of Harvard and University of North Carolina “A diverse student body enriches the academic experience for all Stanford students,” said Stanford President Marc Tessier-Lavigne. “Considering race as one part of holistic review helps us foster a diverse campus community, one in which all students have the opportunity to learn from each other’s experiences and to think critically about their own views and preconceptions. It also means that Stanford graduates from all backgrounds can go on to bring their unique insights to leadership roles in the government, business, and nonprofit sectors.”

https://news.stanford.edu/report/2022/08/02/amicus-brief-fil...

Guess which race/ethnicity Stanford is limiting now?


> pledges action on steps to enhance Jewish life on campus

It seems Jewish life is already thriving on campus - despite being just ~2% of US population, they are 7/11% of Stanford's undergraduate/graduate students [1]. To put that into perspective, and excluding from analysis the 10% international students, they are 3.9x over-represented among non-international undergraduates, relative to their US population.

In contrast, white (including Jewish) undergraduates are only 32.6% of undergraduates, but 57.8% of US population [3]. Meaning non-Jewish whites are 25.6% of undergraduates, making them 0.49x under-represented.

So despite a Jewish person having a 7.9x greater chance of getting into Stanford than a non-Jewish white, Stanford's president is apologizing that this disparity isn't even greater.

[1] https://www.hillel.org/college-guide/list/record/stanford-un...

[2] https://www.collegefactual.com/colleges/stanford-university/...

[3] https://en.wikipedia.org/wiki/Us_demographics


I've feel I've reached a strange line of reasoning here where the OP complained about asian discrimination of college admissions, but the most recent reply seems to imply that asians are also over represented as well (~22% of Stanford, but 6% of US pop).

Any time I see this come up it always feels like asian students are being used a cudgel in order increase admissions for white students.


Discrimination and over-representation could both happen right?

Group A is 5% of the overall population, but 15% of the best admissions to a college, and the college ends up being 10% from group A.


Fine, but then the reasoning against Jewish students doesn't stand. How can anyone, except for the Stanford admissions office, deduce whether or not Jewish students are over or under represented?


What does "over represented" mean?


In my post, I specifically defined it relative to US population. But what you are implying is correct - in general, "over-represented" is subjective. Had I used only US population with SAT score above some threshold, the numbers would have differed.

In fact, it's past the time time to address your question, and start using the term consistently. Stanford themselves boast how "underrepresented minorities" are the fastest growing group [1], without defining the term. Presumably they mean relative to population, but somehow they don't count whites among that group, despite them being, at Stanford, both under-represented and a minority.

We would all greatly benefit if, instead of being so coy, they would openly state: "Yes, there are fewer non-Jewish whites than what would be expected from population numbers, but that is because they're just not smart enough."

After all, how can the conversation advance, if the participants are dishonest, hiding or ignoring data as is convenient?

[1] https://news.stanford.edu/report/2021/12/08/underrepresented...


There's literally a definition of under-represented groups in that article. What are they hiding? The whole point is that raw intelligence on some absolute scale (however measured) is not the only factor in admissions.


Their definition states which groups they consider under-represented, but not the criterion for under-representation. That this criterion is total US population (and not e.g. population with SAT score worthy of Stanford) is implied, not explicit.

The implication may be clear to you and I, but the post I replied to was literally asking what over-represented means (what, not who). Presumably they would ask the same of that Stanford article.


Dude, he is replying to a comment saying "over-represented among non-international undergraduates, relative to their US population.". I think it doesn't take a Ph.D. in linguistics to comprehend what everyone is saying here: some racial subgroup is represented at x% in total population but at y%>x% in the population under consideration.

You can disagree with it, as I do (where I don't expect every non-racial sub-population to show proportions of race corresponding to the total population) but how can you not comprehend what he's saying? Like, you don't understand it? Come on, postulate a few explanations in English for what it could possibly mean.


I continue to be amazed at the notion that anyone could walk the grounds of an elite university and conclude they discriminate against Asians. Why not focus on institutions where Asians are actually under-represented? Government, executive suites, Hollywood, sports, music, arts?


err... because Asians typically have to score higher/better on GPAs, exams, and such to get in than other ethnic groups. Yes, Asians are over-represented in US universities if comparing to the population of the US but are doing so despite the discrimination.

Also - grouping all Asian groups is racist in itself. It's implying that Koreans, Chinese, Japanese, Filipinos, Thai, Vietnamese are the same, which is patently false.


Don't forget your demographic adjustments! The ~5% of Fortune 500 Board seats held by Asians corresponds roughly to the share of the population that's Asians who are citizens, fluent in English, and in the age bracket for that kind of position. As to "Hollywood" and "art"--let white kids waste their lives on such pursuits.


> As to "Hollywood" and "art"--let white kids waste their lives on such pursuits.

This sentence comes out as both racist and elitist at the same time.


Music and art (as careers) are elitist. Those careers didn’t exist in my dad’s third world village.


The HackerNews classic


Just because you see a lot of Asians at Stanford doesn't mean that they aren't anti-asian. There should be even more of them given just how much better the average asian student is compared to the average student.


Wow, so what does it take to get into the top schools as an Asian?

My (Asian) kids are going to face the same challenges when they apply for college. When the time comes, I'm tempted to ease off the whole tiger parent thing, let them have a more well-rounded high school experience, and go to a public state school. It's what I did. Had to spend a few years clawing my way into a FAANG compared to graduates from fancier schools, but in the end it worked out.


IMHO, let your kids be kids, don't push them too hard. Life gives you lots of chances if you have a support structure - my family was full of fuckups but our parents were there for us and we all eventually made it through with good careers.

The main thing I want for my kids (they're still young - 7 & under) is to find things they're passionate about - things that makes them want to work hard and overcome a challenge. That could be school, sports, music, games, or whatever.

Of course, I'm just a random internet stranger, why would anyone listen to me :)


> so what does it take to get into the top schools as an Asian?

The comments below on how to get into an elite university (like Harvard or Stanford) have a fairly high hit rate when all of the boxes are checked. This is a fairly detailed and actionable list, but I am sure I left some important parts out (esp. regarding edge cases and corner cases). Feel free to ask follow-up questions if you have any.

My comments/suggestions:

-Strong grades and strong test scores will get a student into most state schools. This is a very solid baseline goal, imho. These same grades and test scores will be table stakes at elite schools -- more will be required in order to be admitted.

- Make sure your kids are actually interested in going to an elite school and making the extra effort that entails. If they are not, then steer them towards state schools. The comments below assume that they want to go to an elite school and are willing to make the extra effort.

- Take a challenging curriculum and get As with maybe a B or two -- the goal is to be top 5% of the graduating class, ideally the top 1%. Note that Bs (more than two) can be overlooked very easily if the rest of the application is strong. Note that the grade criterion can absolutely suck if your kids go to a high school that thinks that AP/honors classes just means that the students should be assigned more busy work. The busy work will take time away for them to do something that is actually interesting (see below).

- Learn how to write well. This will serve the student well in the application, HS classes, college classes, and life in general. Note that students will probably be rated on their writing skills by their references, and they will need to rate extremely highly (e.g., top 1% ever for this teacher).

- Do solid on the SAT. Contrary to popular belief, you do not need a perfect score. That said, a strong score is table stakes. Specifics below.

- My main advice for folks is to prepare for the SAT, mainly by taking practice tests with recent old tests at the same time and a similar location that the student will be taking the real test. The SAT is a marathon, and most people perform below their potential due to fatigue. Getting used to the long slog and the pacing of taking the test helps most people a lot. They will also be able to find gaps in their test-taking skills, and these are usually fairly easy to address when not under time pressure to do so.

- Math -- I would definitely aim for a perfect score in the Math (since it is fairly basic), but a slightly less than perfect score in Math is fine. Engineering people and folks who apply to MIT tend to have perfect or one tick below perfect.

- Verbal -- Most people try to game this a month or two before the test. Big mistake. My suggestion is to read well-written newspapers like NYT, WaPo, and WSJ as well as "high brow" magazines like The New Yorker, The Atlantic, and/or National Review. They should make note of every word that they don't know, and they would do well to learn them. This is easy to do over three or four years and tougher to do over one or two months. Bonus points if you, as a parent, discusses these articles with them so that they are reading critically rather than just casually. The text structures are the same types of (edited and abbreviated) text structures that they will see on the SAT.

- More than grades and SAT scores, the student should be able to impress one or two teachers that they are in the top 1% of students that the teacher has ever seen. Why? This is a common question on the reference letter form. I strongly encourage the student to check out the reference letter forms early in their high school career to see the manner and extent to which they need to impress the people who write them references. The student can do this by showing interest in one or more areas and really going deep into the subject beyond what is in the high school material -- basically, have passion for the topic.

- Also note that there are schools and teachers who know understand how the admissions game is played. The two things that these folks do that help a lot are: 1) to round up their rating of the student (e.g., if top 1%, maybe rate them "best I've ever taught"), and 2) to give detailed examples of outstanding work the student has done. The recommendations give credence to whatever it is that the student writes in their application. Picking who writes recommendation letters is a very important decision that I think many applicants put very little thought into. Not only do these teachers/people need to know you well, but they need to be able to write a good recommendation letter as well (this is a rare talent, imho, unless you attend an elite university feeder school).

- Related to grades/subjects, I would personally encourage students who can do it to enroll in a joint hs/jc program. There will be much less busy work, and one or more of the professors will know how to write a good recommendation.

- Try to figure out one or more areas in which they can do something notable at a national or international level. There is a wide scope of what this can look like, but this is where most people have a substantial gap in their application. This area can be social/leadership (probably the "easiest"), sports, arts, or academics (probably the toughest). I buried this relatively deep in the list assuming most people won't read this far, but this is the one thing that really sets aside exceptional applicants from threshold applicants who have great grades and test scores.

- Examples of "notable" (social/leadership): Start a successful business (ideally in an interesting area), start a non-profit that does meaningful work, develop a community program that makes meaningful change (esp. in under-served communities), develop some ongoing development aid project abroad (often seen in church missions), amazing Eagle Scout service projects. Note that all of these are starting or leading a group to new levels of success. Just participating is not enough (anyone can do that).

- Examples of "notable" (athletics): Be a recruited athlete, be an athlete that is not recruited but would be competitive on the university varsity team and expresses an interest to play at the university, win a state championship, be chosen as "all state" in your state (preferably first team), be on an Olympic team or on the short list development squad, win a regional (or wider) event in an "Olympic sport" (e.g., figure skating, ice skating, diving, swimming, etc. that might not be covered in HS sports). On a personal level, I would probably encourage folks to look at less common sports that can be excelled in at a young age -- competitive shooting, obscure martial arts (Japanese sword drawing, anyone?), noodling, etc.

- Examples of "notable" (arts): Be a published author, have an exhibit in a gallery, win a national/international competition (this might be easier than it seems if you're clever), be a national/international touring artist. Note that this can be combined with social/leadership by doing something like creating a highly regarded arts competition, creating a successful touring music group, etc.

- Examples of "notable" (academics): high placement in a nationwide math contest, be selected for the international science fair (a series of competitions), win a Westinghouse competition, publish a research article (as primary or a significant contributor). Note that this is the toughest category to stand out in, imho. The people who apply to elite schools are absolutely crushing this category. "Placed 3rd in state with HS trivia team", while commendable, won't really look that impressive or unusual when compared to a Westinghouse winner.

- With regards to the "do something notable" point made above, this something should be referred to in one or more of the student's references, otherwise it looks like it is completely made up and will probably be discounted. Supporting documentation helps (e.g., awards, newspaper articles, published papers, etc.), and be sure to send in this documentation as "supporting materials". If your kid is doing something special, I hope that they reach out to the press to have their special thing covered -- local news outlets love stuff like this, and it looks great in an application.

- Note that there are weird edge cases in admissions. As an example, UPenn is a private school, but they have to accept a certain number of PA residents per year. While there are extremely strong students from PA at UPenn, there are also some students who are relatively weak (and it shows). Also at UPenn, each of the four schools has an independent applicant pool, so the Wharton pool (9% admission rate) is different than the Nursing pool (25% admission rate). Not to pick on UPenn, but I happen to know details about these two examples.

- When applying to a school, the applicant should have a good reason for going to that school. "Because it is #1" or whatever doesn't cut it. The access to resources at elite schools is amazing, and the applicant should show that they have at least considered how they might utilize some of those resources.


(part 2, since first reply was too long)

A few additional notes:

- Note that many people who attend elite schools almost stumble into checking off all of the boxes above without realizing it. It's amazing how some people just randomly do the right things that pave the path to admission into an elite school, while others who try so hard to get into an elite school either do the wrong things (like overly focus on grades and SATs because someone told them to) or have someone else in the process (e.g., a teacher or guidance counselor) fail them.

- Some people think that what happens at elite schools is what makes their graduates so successful. This is true at the margins. Specifically, the access to specific networks of people that control high value parts of the nation and economy is exceptional, and people who avail themselves of this feature of elite schools will see an outsized return on their attendance at an elite school (secret: many/most don't avail themselves). All that said, many of the people who have crazy success after attending an elite school would also have crazy success after attending any school because they and their families are very well connected. Note that state schools also have these sorts of networks, but they are typically limited to the state or regional area in terms of scope.

- IMHO, the at-school opportunities available to a great student at a big state school (something that I think is relatively easy to be) is probably comparable to the opportunities available to a slightly above average student at an elite school (harder to be due to the admission's barrier). The reason is that most state school students are not ambitious and overly curious and extremely talented, so the best students stand out to the professors at state schools while they are relatively common at elite schools. I think the only potential "loss" of going to a state school is if a student would be one of the best both at the state school as well as being the best at an elite school (this bar is almost unbelievably high).

- Most people think that there is discrimination against Asians because they know or have heard of some Asian who got a perfect or near perfect SAT score and a perfect GPA (i.e., the "standard strong" applicant that is a dime a dozen at elite schools) and did not get into Harvard or Stanford or wherever. As I hope the above post has shown, that result is not surprising if their grades and SAT scores are the most compelling part of their application. Harvard's entering class this year is ~25% Asian (most being Asian-American). The threshold Asian folks who didn't get in probably look a lot like the threshold White folks who didn't get in -- they were very solid but didn't stand out or had one or more red flags in their application. As such, the goal is to be a strong admit applicant rather than be a threshold applicant and hope for the best.

- All of the above seems like a lot of work, and I suppose it is. That said, I think that most of it is something that ambitious and motivated people will be doing anyway just because they want to, and the rest is just polishing the edges (e.g., making sure to get good references).

- Lastly, Cal Newport has written some amazing books about "being a superstar" and "standing out without burning out". I recommend giving his books a read.

I hope this helps.


> Some people think that what happens at elite schools is what makes their graduates so successful...

Thanks for this comment. I think this is really important. I see so many families make huge sacrifices to try to get their kids into top colleges when they really dont seem like the type of kids to be elite class people. I grew up in a country where this didn't really exist so am a bit baffled by the whole game. I'm happy to take it easy and get a regular college, but worry my kids might miss out on some pot of gold.


> but worry my kids might miss out on some pot of gold

The default assumption should be that if your kids enter an elite school as $SES kids, then they will leave the same (SES = socioeconomic status). So a middle class kid will likely leave Harvard still a middle class kid due to their connections, way of perceiving the world, way of interacting with the world, etc. There usually is no pot of gold.

The dream of social climbing is often more of a rug pull than many middle class people realize since it doesn’t just end with school. Some folks with classic hard-working working class or middle class values will go to an elite school, do well, and then get a top job in IB, consulting, law school/law firm, or whatever, and they think they have made it. But then their career path just flattens after a few years of very grindy work. If the social climber did not make a concerted effort to develop a network of higher class friends (which is not easy if you’re not already in it, imho), then partner or whatever becomes much tougher when job performance is basically based on if they can make it rain or not via their social networks.

There are exceptions, typically in academics, in specializing in a topic or group of people that are foreign to northeast corridor people (e.g., oil people in Texas or car people in Detroit), or in certain organizations (e.g., Goldman was/is known for facilitating social climbing).

An example of one guy I know who made a big jump in SES (sort of)…

He came from a family of educators in small town Michigan, but had a very working class mentality (pro-union, kids did manual labor, etc.). The guy was obsessive about being rich from a young age — the type of guy who actually wrote himself a check for one million dollars dated for his 30th birthday and was hell bent on cashing it.

Long story short, he managed to work his way into a PE firm as a partner in his mid 30s after a decade of decent-but-not-great pay grindy work. His speciality was suppliers to Detroit auto manufactures. Note that he knew nothing about cars and the auto industry at the start, but he knew how to talk to Detroit people without sounding like a NYC shyster.

Made a bunch of money (mid 8 figures), closed down the fund, and retired in his early 40s in the Hamptons.

Note the path though… super grindy, didn’t really make it early, relied on his prior SES, a lot of social climbing, and a bit of luck to set himself apart and make it “big”.

This is not a path I would recommend to my kids unless they are hellbent on taking it like the example above.

The experiences I see of those who tried to social climb is that a lot of them hit mid-life and wondered where their high-trajectory career went. That “high-trajectory” was a class-based illusion, imho, and they didn’t do what it would take to make that illusion become real.

Apologies if this sounds overly cynical, but I think it’s important to debunk some of the myths about what makes the elite schools seem so impressive.

To end on a more positive note, I will say that elite schools often help folks reach the top end of their SES range, especially for upper middle class folks. So kids who come from families with wealth in the 10m range might end up with wealth in the 50m range, with a few making a bigger jump to lower upper class. This may seem like a big jump, but it is not functionally since that level of extra money largely just buys QoL improvements rather than access life changing levers.

There was a guy on HN who wrote a really good social class guide and what affordances each class gave you. I will try to find it and link it here later.

Here is a link:

https://indiepf.com/michael-o-churchs-theory-of-3-class-ladd...

Michael O’Church had the original idea. I can’t quickly find a link to his original article in the internet archive (way back machine), but I think it’s still there.


I just got around to checking my HN replies. Thank you so much for writing this out. I'm at least a decade+ from having to put these plans into motion, but I'll refer back to it from time to time.


You can grind leetcode for a few months and get a FAANG role with a boot camp degree. Don't bother tiger parenting.


Something similar happened to me. I had a perfect high school GPA, 1540/1600 SAT. Applied to several schools, all of which I was above average score wise

Only got into flagship state schools that were well below my average SAT

The one thing I've learned from this (a valuable lesson): never, ever do the self identification. You'll always get fucked. Self identifying is always a mistake


If you don’t self identify you just get binned as white. Putting down mixed race is the standard workaround. It’s vague enough that you’re probably not lying and it avoids being labeled white.


I don’t understand this. Does race or does it not exist? If it does not, you should be able to put down any race when asked, since how can they prove it? If you said you’re black and walk in but clearly your skin tone is fairer than expected, do they have a color palette they hold up to your face and say “nope, sorry you’re too light to be black, or is there a generic test they force you to perform?


> do they have a color palette they hold up to your face and say “nope, sorry you’re too light to be black, or is there a generic test they force you to perform?

It’s more of a sniff test that is only performed at elite schools or for privileged positions (affirmative action and/or scholarships).

I knew a woman who identified as black who was lily-white. It turns out her grandfather (in the south was black), and the locally scandalous out-of-wedlock interracial relationship that produced one of her parents was a big part of her identity and influenced what she studied. She ended up at an elite school.

On the other hand, if someone puts a race on the application that doesn’t really match how they look and doesn’t have much in the way of explanation in their application, then I think that interviewers at elite schools would (at a minimum) be curious. This is especially true if the race selected would likely get them some sort of preferred admissions status.

With regards to Asians in particular, many (most?) Asians can reasonably put down multi-racial since there has been much racial mixing within Asian over the past few millennia. Whether that jibes in a US elite college application is a different issue.

At most schools, unless race confers someone a special status like affirmative action or scholarship eligibility, then no one cares. If someone got in specifically due to an inaccurate claim, then they run the risk of being expelled or even having their degree revoked for unethical behavior.


>then they run the risk of being expelled or even having their degree revoked for unethical behavior.

Have either of these actually happened? Someone put an 'innacurate' representation of their 'race' down and that happened? I'm trying to imagine how you could prove it. For instance, some theories hold that humans emigrated from sub-saharan Africa and thus most/all (even Americans) would be technically 'African-American' by some definition of the word.

It's my understanding that these forms request your self identified race thus there would be no wrong answer, you merely need to self identify for the second it takes to check the box.


> Have either of these actually happened?

Yes, people have been expelled and had their degrees from elite schools revoked for unethical behavior (usually academic misconduct).

Has it happened for misrepresenting their race on their application? I have no idea, since I don't keep up with such things. I would imagine that this situation is handled much more subtly and much earlier in the process (e.g., by noting that race doesn't match appearance in the interview and/or content in application). Said another way, if someone is able to con their way into an elite school via lying about their race, my guess is that they dodged the most likely filter (the admissions process).

That said, there is a blurb in almost all applications that says that folks can have their admission rescinded etc etc if their application is not true to the best of their knowledge.

There is also a code of conduct that says something similar regarding ethical behavior.

I've seen some whacky stories about these clauses being invoked, so anything is possible. Need an example, do a search for "the water buffalo incident" that happened at one of these elite schools (ostensibly for racial harassment under the code of conduct).


> With regards to Asians in particular, many (most?) Asians can reasonably put down multi-racial since there has been much racial mixing within Asian over the past few millennia. Whether that jibes in a US elite college application is a different issue.

By that exact same reasoning, couldn't literally anyone else do this? If we start blurring the definition, why does it conveniently stop there?


> couldn't literally anyone else do this? If we start blurring the definition, why does it conveniently stop there?

Sure.

I don't think anyone is getting dinged or getting special affirmative action consideration for listing themselves as unspecified multi-racial.

If someone checks a box that gets them special affirmative action consideration, and it makes up no part of their application, and it doesn't pass the sniff test at the alumni interview, then a note will probably be made that race, appearances, and application do not appear consistent, and they will proceed from there.

The admissions officers are looking for a compelling narrative to admit the applicant. It's up to the applicant to provide that compelling narrative.

If I were advising someone who is multi-racial with an affirmative action twist while not really looking the part, I would recommend that they make a note of it in their application. Note that being part Black while appearing very White usually has some impact on shaping the person's identity.

For reference, I knew one person at an elite school who would introduce herself in icebreakers as "My name is $NAME, and I am Black"... and her skin was extremely pale white. She had a Black grandfather who had a scandalous affair with a White woman (when and where that was not apparently acceptable) to produce one of her parents. She told us stories about how that impacted her identity. I imagine she may have written a little bit about it in her one or more of her admissions essays.


> If someone got in specifically due to an inaccurate claim, then they run the risk of being expelled or even having their degree revoked for unethical behavior.

This is actually what I’m asking. How do they measure your race? If you make the claim someone is driving under the influence, you have a defined blood alcohol level threshold and a way to test that. How do they do the same for race?


> How do they measure your race?

Great question.

My guess is that they would only take action if you did something like brag on social media about hacking the admissions system by lying.

Stranger things have happened...


Would they really? Was Elizabeth Warren punished by Harvard or UPenn for unethical behavior?


Apparently, she does have some Native American ancestry many many generations back, like a lot of Americans


Well, if that is the case, then it would probably be hard to actually accuse the person of behaving unethically.

I'm extremely pasty white, blond hair, blue eyes, Norwegian last name. But 23 and me said that I have 0.1% North African ancestry, or 1 ancestor from ~1600 AD..so African American it is!


> Would they really? Was Elizabeth Warren punished by Harvard or UPenn for unethical behavior?

Good question.

Just because they can doesn't mean they will.

As I said elsewhere, I think that the most likely filter for race bamboozlers is in the application process when the stated race and the combination of application, references, and alumni interview does not tell a coherent story.

Regarding Warren specifically, I will honestly say that I don't think race mattered very much for her. I know it's a meme in certain circles, but she applied in a very different era, one in which admissions standards were much lower. Being capable and ambitious was probably enough regardless of her race. I could be very wrong about her specific case, but I am certain that the Harvard and Penn of that era was nothing like the Harvard and Penn of today in terms of competitiveness of admissions.


Imo we should just ban the practice entirely. The supreme court's about to do it for universities, hopefully employers are next


According to the statistics other is slightly worse than white but much better than asian


> if you don’t self identify you just get binned as white.

Still better than being binned as Asian


I had the nearly the exact same stats and situation. Even the mf colorado school of mines rejected me


It makes me really angry hearing Harvard defend their racist admission policy, especially when you know it was designed to limit the number of Jews admitted.


> Personal anecdote: 3.7 GPA in average high school with 8 APs (some of which were self-study), 2400 SAT, 36 ACT, lots of state-level top 3 finishes in multiple academic competitions, varsity track, musical instruments,

I love how the GPA requires a decimal place of precision, yet musical achievement requires nothing more than listing the plural of "instrument."


I was shocked they didn't divulge their 400m time.

We all care about different things, I guess


My first thought is that maybe they didn't like your college essays. The funny part is that college essays are the most obviously biased towards the privileged. Poor people can't afford touring colleges, so their "why this college" essays can't be as specific. Meanwhile, the rich can pay someone to ghost-write their essays.


> Poor people can't afford touring colleges, so their "why this college" essays can't be as specific.

The "why this college" essay can trivially be addressed by looking through the university website.

Most applicants don't even do this.

A simple formula for this essay is:

- $APPLICANT has a demonstrated interest in $TOPIC(S).

- $ELITESCHOOL has $RESOURCES about $TOPIC(S) that $APPLICANT would like to explore.

This can obviously be iterated, expanded, and refined, but that's the basic formula for knocking it out of the park on that sort of essay.

Touring colleges has little to no impact on this essay.

Ghost-written essays also come across as shallow unless the parents also pay off teachers to write recommendation letters that corroborate the BS. It happens, but it's not common.


> Most applicants don't even do this.

> A simple formula for this essay is:

Sounds like a perfect job for a talented ghost writer. The problem with college websites is that only maybe 5% of is actually useful for essay writing. The rest is meaningless generic fluff. College tours allow you to experience some quirks of the school that you can sprinkle in to make your essay more interesting.

>Ghost-written essays also come across as shallow unless the parents also pay off teachers to write recommendation letters that corroborate the BS

You're comparing strong student with a modest background to a weak student from a privledged background. For students of equal ability, privledged ones have much better access to resources for writing a good essay. As far a recommendations go, that's also highly biased towards elite private schools. They have much smaller class sizes, so students are able to form much closer relationships with their teachers and teachers can write more personalized rec letters.

I suspect that you're greatly overestimating your BS-detecting abilities, which makes sense because you'd otherwise be hopelessly paranoid and cynical. My wife used to work part-time editing essays at a large college-consulting service, and it definitely helps. For example, simply by writing a lot of essays, you gain an intuition about how much you're able to BS without sounding fake. On their own, students are much more likely to overshoot or undershoot.

As far as my personal experience goes, my approach towards college essays was to write raw, avant-garde essays that obviously signal my personality. My writing skills weren't strong enough to make this work, but I could rely on the adults around me, a few of whom were alum at prestigious universities to refine the drafts into something that was actually coherent. I definitely couldn't have done this if I were from a poorer background.


Wow! East Asian here who didn’t do anywhere near as well as you with test scores (2200 SAT, no competition wins) but still transferred into a top 20 school. Can I ask when you applied for college? I’m class of 2012.


> but still transferred into a top 20 school

Columbia? A public school in a state that mandates acceptance of Community College graduates with good enough grades?


I wonder if a court would allow Asian teenagers to change their last name…


As someone who applied to college too long ago to remember anything - do you self-provide your demographics? Or do they try to guess what race/ethnicity you are based on your name?


They ask you to explicitly mention in the application. You can choose not to disclose, but then they'll try to guess it themselves.


You should vote your interests. The end goal here is a system of race-based patronage. Instead of just taking an objective test like an equal, your kids' futures will depend on breaking out sob stories to tickle the fancy of white gatekeepers. Not just for college. For jobs. For promotions. The sky's the limit.

Prop 16 failed, but that won't stop them. Asians at Meta and Alphabet look to your left and look to your right. One of you will be gone when corporate American implements the same policies as Harvard (which cuts Asian numbers in half). And who is standing in the way of that future? Republican judges.


There is another solution to this: inflation. As inflation eats away at income, assets and savings, true merit & productivity will be the only thing that matters. Look at big tech earnings today and the reaction in the stock market. Big tech can increase their EPS any time they want by shedding 50% of their workforce.

We're no where near such draconian decisions. But if we get there, the decisions will be based entirely on the financial contribution of an employee/team/project/department to the company's bottom line. I don't believe tribal/race/group affiliation will matter.


What was your college essay about?


There is something missing from your story, but I can’t exactly figure out what it is.

Some things that may have gone wrong:

- Maybe your GPA was too low for your high school. If it put you outside of the top 10% or so of your class, then that hampers your chances substantially.

- Maybe your references did not know how to write good reference letters for elite schools. I think this limits a lot of people, since tepid reference letters from people who don’t know better can kill an application.

- Conversely, maybe you went to an elite university feeder high school. If so, maybe you didn’t compare well against other folks who did get into elite schools other than test scores.

- You applied based on academic prowess, and maybe your academic prowess was not enough for elite schools. Elite schools typically want national or international levels of success in something (especially academics, which is hyper-competitive), so maybe your state competition results combined with a relatively low GPA (for elite schools) didn’t really match up to the other 100-200 or so people at each school who got in on the back of academics.

- Maybe you didn’t have much/any significant community involvement and/or leadership experience. This may seem soft, but it makes applicants stand out, imho.

Fwiw, I don’t think your ethnicity played as much of a role as you think it did. Those killer test scores you have are basically table stakes at elite schools (i.e., you need a lot of something else to get in), but they are a meal ticket at large state schools. Many/most people think that test scores and above average grades alone should be enough for elite school admission, but this hasn’t been true for many, many decades.

For reference, “top 50 state school” is something like University of Georgia or Ohio State University, both the type of school that will not slow down someone who would have fit in at Harvard or Stanford, imho.

> But I probably could have saved years in high school exploring things that mattered rather than optimizing for a college application process that didn’t ultimately end up feeling very fair.

I guessing that you optimized incorrectly.

I’m curious about where you got your information on how to optimize for elite school admissions. Most people will say something like “have great test scores and great grades”, and they would be wrong for elite schools but very accurate for most state schools.

I don’t want to say that elite school admissions are perfect (or even good), but people who have never been able to see the processes from the inside (e.g., actual applicant pools) usually make very unreasonable criticisms of the processes.

Fwiw, every elite school tells applicants exactly what they want in their recruiting materials. It’s just that different people interpret those materials in different (usually inaccurate) ways.

All that said, I’m glad it worked out for you.


Yeah, in hindsight fair to say a little bit of all of the above. I agree with a lot of what you mentioned.

At the same time - if I'd gotten 300-400 points less on the SAT, I likely still would have ended up at the same school. So in that vein my gut says the test scores were close to irrelevant. Or maybe not, I guess I wouldn't know.

Separately I think it's valuable to evaluate this through another lens. It sounds like you have some experience with the admissions process, and something I've been curious about for the longest time is this: if I had been in a slightly different segment, e.g. first-generation immigrant of non-East-Asian ethnicity, how would that have affected my chances?

Is it that test scores mattered less for me in this particular case, or is it that there's generally a higher bar because of competition from peers with similar East Asian backgrounds? In both cases it feels like test scores matter less overall (even if paradoxically the bar is generally higher!).

---

> For reference, “top 50 state school” is something like University of Georgia or Ohio State University, both the type of school that will not slow down someone who would have fit in at Harvard or Stanford, imho.

This was the only piece I felt differently on. There's a significant advantage to attending a top school - the alumni network and a generally stronger and more well-connected student body for starters. Going to a state school didn't necessarily prevent me from finding success later in life, but I definitely took the long way around.


Let's be real, Stanford could fill it's entire freshman class with kids who got perfect GPAs and perfect SAT scores. That's not the deciding factor. You mentioned doing well in State level competitions but the elite schools of the US, Ivy+ etc, are the best schools on the planet. It's not surprising that you didn't get in given that fact. Their freshman classes are full of State Champions, National Champions, Olympiad Winners, Intel Science Fair winners, Published Authors, kids that have given concerts at Carnegie Hall, etc. I think you may have misjudged the level of competition. It's not 1940 when a perfect SAT got you into any school you wanted.

>This was the only piece I felt differently on. There's a significant advantage to attending a top school - the alumni network and a generally stronger and more well-connected student body for starters. Going to a state school didn't necessarily prevent me from finding success later in life, but I definitely took the long way around.

People have studied this and found that students that were accepted into elite colleges but ended up going to lower ranked schools had equivalent levels of achievement after graduation. So it turns out graduating from Harvard isn't as important as you think. Unless you want to go into IB or Big4 Consulting.


I halfheartedly agree.

Getting into a top 5 school would have definitely been a surprise, but I also doubt the top 20 or even top 50 schools have equal levels of talent.

> People have studied this and found that students that were accepted into elite colleges but ended up going to lower ranked schools had equivalent levels of achievement after graduation. So it turns out graduating from Harvard isn't as important as you think. Unless you want to go into IB or Big4 Consulting.

This tracks but as mentioned elsewhere it's probably a longer journey all around. There very much is an in-crowd of connection pooling when you're an alumni of an elite school, from my experience of going to some meetups with heavy membership from them.


>Getting into a top 5 school would have definitely been a surprise, but I also doubt the top 20 or even top 50 schools have equal levels of talent.

I tend to think that that outside of like the top tiers of Harvard, Yale, Princeton and MIT/Caltech that the top 20 have student bodies that are basically interchangeable and it's a coin flip whether someone gets into UPenn but not Columbia or UChicago but not Vanderbilt.

>This tracks but as mentioned elsewhere it's probably a longer journey all around. There very much is an in-crowd of connection pooling when you're an alumni of an elite school, from my experience of going to some meetups with heavy membership from them.

You can short circuit this by doing a graduate degree. MBA from HBS is at least as good if not better than an AB from the College. But like I said outside of a few industries those connections aren't as valuable as you may think. Is getting a job in a VC firm easier as a Harvard/Stanford Alum? Sure, but least in theory though getting startup funding from a VC as a Harvard vs Michigan alum should be roughly equivalent in difficulty(lest we break the illusion of meritocracy in the tech world).


I concur with all points, and I envy your brevity.


> It sounds like you have some experience with the admissions process

I do. Not as an admissions officer, but in a variety of advisory capacities to admissions offices over the decades.

> if I'd gotten 300-400 points less on the SAT, I likely still would have ended up at the same school

Maybe. It depends on the state school. I certainly hope that your scores and (maybe) grades got you access to some sort of honors or scholarship program that a significantly lower score would not have. If you were not in one of these types of programs, I am curious about why that is the case.

> if I had been in a slightly different segment, e.g. first-generation immigrant of non-East-Asian ethnicity, how would that have affected my chances?

My default answer is that it would not have affected your chances at all unless someone in your application process was racist. This would most likely have happened with someone who wrote a reference letter rather than someone affiliated with the universities.

The exception would be if your different ethnicity put you into an affirmative action category (specifically Black, Latine, or indigenous people). You might get in based purely on academics, but it seems like you would still be relatively average or weak in other areas. There would be a lot of other factors to consider (e.g., socioeconomic status, school/school district, geographic area, etc.).

While a few of the affirmative action admits are merely "standard strong" (i.e., not admit and not close because so many people are similarly strong while not standing out) who get in due to their race, many/most tend to be shine in other areas (esp. community leadership activities). Affirmative action basically gives some of these folks a slightly lower curve on academic rating, but their academics are (usually) good enough, and the other parts of their application made them shine. There is plenty to debate about whether that is a good policy or not, but I think that this is a reasonable policy if a specific kind of racial diversity is an institutional goal (as it seems to be). Almost all of the folks who get into elite schools are extremely impressive in some factor that the school values.

> Is it that test scores mattered less for me in this particular case, or is it that there's generally a higher bar because of competition from peers with similar East Asian backgrounds?

Without knowing more about your specific situation, my guess is that you focused on the wrong things -- specifically SAT scores over almost everything else.

As I have said elsewhere, no "strong admit" East Asian candidates are being denied admissions to elite schools because of a quota. Zero. None.

The folks who don't get admitted who feel jilted are typically "standard strong" (good grades, great SATs, nothing exceptional, will never get in) and the threshold applicants who just miss an offer because the whole package doesn't quite add up to enough (rare, but it happens). That said, lots of White people fall into these categories as well, so I don't think that it is specifically an East Asian phenomenon.

The issue of discrimination comes up because the average SAT scores of Asians is higher than that of other races. Some people think that is a sign of racial discrimination. This can be seen in articles that claim that Asian students need ### more SAT points to get into $ELITESCHOOL (prime example of misuse of statistics around the theme of correlation does not indicate causation). Personally, I think that this just points to a focus on test prep in the East Asian community that is consistently greater than other racial communities (by a lot).

> There's a significant advantage to attending a top school - the alumni network and a generally stronger and more well-connected student body

This is true... if the students avail themselves of this network. Most folks who are not already part of that network rarely do. I have found that there are relatively few people who effectively utilize this network during school or after school who would not have had similar access via another school and/or their already-existing family connections.

I call this the dirty little secret of the Ivies -- the Ivies largely don't make people successful, they just accept a large chunk of people who are already going to be successful and let them go to school there. There are a very small percentage of people of modest background who go to elite schools and then becoming someone exceptional via the school network, but that is extremely rare (for a number of interesting reasons, imho).

Note that state schools have similarly strong alumni networks, but the scope of these networks is usually state or regional rather than national or international.

The real exception in my eyes is if someone has relatively narrow ambitions to be an investment banker, a consultant, a supreme court justice, etc. (i.e., some line of work that almost exclusively consists of elite school grads). Yes, elite schools help with this tremendously. That said, I imagine that the elite folks in GA and OH society (which I listed as top 50 state school examples) are more likely than not graduates of UGA and OSU, and certain roles like state-level politicians, judges, business magnates, etc. are disproportionately from these schools.

Feel free to ask more questions. I feel like the misinformation on this topic is abundant, and I like to dispel as many of the myths as I can (at least based on my experiences).

I made very detailed reply to a concerned Asian parent who replied to you:

https://news.ycombinator.com/item?id=33341897

Please add to that line of the thread as well if you find it interesting.


> There are a very small percentage of people of modest background who go to elite schools and then becoming someone exceptional via the school network, but that is extremely rare (for a number of interesting reasons, imho).

Would you mind expounding on these interesting reasons? I've heard a theory that it's because the truly powerful networks are established at prep schools such as Andover and Exeter, so the vaunted power of the elite university network is really an extension of the prep school network.


> My default answer is that it would not have affected your chances at all unless someone in your application process was racist. This would most likely have happened with someone who wrote a reference letter rather than someone affiliated with the universities.

Of course there was. An admissions officer saw it and they have to stop the school getting too Asian.


> An admissions officer saw it and they have to stop the school getting too Asian.

Where do you believe that this specific situation has actually happened this century?

I am fairly certain that this has not happened at any elite school in the recent past.

The worst actual evidence I’ve seen is of an admissions officer referring to a “standard strong” applicant as something slightly less flattering like “a familiar profile” Asian applicant.

Note that “standard strong” basically means great grades and scores and little else of significant note. This title defaults to non-admit.

Below is a link to a click-bait article that discusses the topic. Note how almost all of these “inappropriate” statements fall into the category of least charitable possible interpretations of the rater’s comments rather than the explicit malice that you seem to suggest is happening in admissions offices.

https://www.buzzfeednews.com/article/mollyhensleyclancy/asia...


These are people who know that anything written down will eventually be used in a lawsuit. They’re not going to use slurs. I bet they wouldn’t even have any problem with their children marrying an Asian. But it’s their job to keep the Asian percentage down below what it would be if the student was White, and very far below what it would be if they were Black, among students with otherwise similar credentials. If that doesn’t qualify as anti-Asian racism what is it?

The reason the holistic admissions system was made up was to avoid having too many Jews. Now it’s used to prevent too many Asians.

Plus ça change, plus c’est la même chose


Let me tell two stories:

I graduated with the highest SAT scores in my school and the second highest in my city back in the early 90s. I’m not bragging, it was a relatively small town in the south with not great school systems and a total graduating class of 1500 across all of the schools.

What are the chances that happen because I was so smart or because I had a mother who was not only a high school math teacher but also spent years volunteering to teach SAT prep classes?

Second story: my step son didn’t have the ACT scores to get into the college he wanted to go to. Four months later and with the help of a private tutor who was a teacher (not in his school) that we paid $100/hour for 12 hours, he suddenly raised his scores more than enough. Did he gain aptitude in 12 weeks or did he learn test taking techniques?


Yeah, I hear this, but in 1995 I just bought an SAT prep book for $12 and spent a couple months of Saturdays practicing and did very well. My family lived well below the poverty line. SAT prep is not confined to the privileged.


Selecting for people who have some intelligence and actually give a fuck is actually a great way to cull the herd.


> What are the chances that happen because I was so smart or because I had a mother who was not only a high school math teacher but also spent years volunteering to teach SAT prep classes?

It sounds like you are saying that good parenting shouldn't have any effect of kids success in life.

Look, we know that poor parenting results in kids with poor outlooks, but penalising good parenting by taking away the effect of good parenting is probably not a good way to go about helping those kids with poor parenting.


I’m saying that the test isn’t indicative of “aptitude”. Well neither are GPAs though. It’s not hard getting a 3.7 GPA when the teachers know that if they make the coursework challenging or don’t give plenty of make up chances that either most of the students will fail (in the small town I grew up in) or that the Karens will complain to the principal that this one teacher will keep her little Timmy from getting into whatever prestigious school she wants him to get in (in the case of the “good schools” my step sons went to)


> I’m saying that the test isn’t indicative of “aptitude”.

Maybe it isn't, maybe you're correct, but the practical effect of ignoring tests is to penalise good parenting, instead of helping those kids with aptitude but no opportunity.

We need a really good reason for why kids that worked hard at the tests and demonstrated that they were willing to study (whether payed for and/or pushed for it by the parents) should be rejected in favour of those who never demonstrated this.


> Maybe it isn't, maybe you're correct, but the practical effect of ignoring tests is to penalise good parenting

“Good parenting” == “being able to pay for private tutors for little Timmy”?

Just like a “good software engineer” == “someone willing to ‘grind LeetCode’ for six months?

(And just in case someone replies that I’m just upset because I couldn’t “get into a FAANG”, yes I work for one now. No I didn’t do the LeetCode grind or the DS&A interview. Yes, I’m in an IDE coding most days doing the same type of work I’ve been doing for almost three decades)


Didn't you claim that your mom helped you?

Are you saying that ignoring tests won't have any effect on the majority of good performers who got that way because of parental attention and involvement?

Ignoring the tests just teaches those kids who worked hard that they shouldn't have.

The message it sends to all kids is terrible: don't bother with learning because you'll be chosen, or not, based on your demographic.


There is a big difference between “learning” and “preparing for a specific test.”

Just like there is a big difference between “being a good software engineer” and “practicing reversing a binary tree on the whiteboard while juggling bowling balls while riding a unicycle on a tightrope”


> There is a big difference between “learning” and “preparing for a specific test.”

Yes, but the PRACTICAL EFFECT of doing away with tests would be to disadvantage the hard workers.


Don’t you think a mother able to teach those classes would be in general a better test taker than the average person, and so her children likely would be too?

And nobody is denying that test scores can be raised. That’s likely a fixed amount so just amounts to constant noise. I think you’re as likely to find a way to consistently raise a 600 SAT to 1600 as you are trying to train any couch potato into an Olympic sprinter.


So let’s say my mom had been a “good test taker”. But she didn’t spend anytime with me. Was always too busy to help me when I had a question, etc. Do you think I would have done as well?

What are the chances that I would have chosen to be software developer coming from that same small town if my parents hadn’t bought me a fully tricked out Apple //e in 1986?

Even when I was in college in 1992, most of the CS students didn’t have a computer home. By then I had a $4000 LCII set up with a $1000 laser printer and a $300 SoftPC application to run x86 compilers. My parents weren’t “rich”. My mom was a teacher and my dad was a factory worker.


Why is it that the anti test score argument almost always tries to chalk it up to hiring tutors? They consistently try to make it sound like people "buy" higher test scores with tutoring in order to discredit tests like the SAT


There is research on this. I believe it was Thomas Espenshade, social scientist at Princeton, who had data on test prep access, modeled along with SES, etc. Test prep is worth about 40 points on the SAT. Not trivial, but not a game changer.

It's true that people who are otherwise well off are more likely to have test prep, and those people's test scores are often several standard deviations higher than low-income students from bad schools without test prep. But the counterfactual of (same person) +/- test prep = about 40 SAT points.


Great & useful article. We should have more of these for “popular papers”.


Agreed! Although, I admittedly didn’t read the post carefully enough to evaluate how accurate its analysis of the original paper. But all else being equal, we’d live in a better world if papers that get a lot of popular attention are subjected to this scrutiny with an eye towards explaining what works/ doesn’t work for a general audience


My favorite point:

"2. It might be that test scores don’t predict college graduation rates[...] is provably false."


It'a sunken cost effect at paly, little more than that. Oh, and parents pressure, least we forget the role of family. You can test my claim empirically by asking yourself: "are all college graduates I met more intelligent than me?". Admission test measure how well you can prepare for admission tests, nothing more. There are too many confounding factors and co-factors that play in the career of a relatively fresh semi-adult aged 17-18, that reducing it to a single number is just moronic or bad-faith utilitaristic escape route. Universities only care that you graduate so they can make money, therefore they use the simplest KPI to measure that. Let's not forget Universities are staffed with people and are a business.


It's unlikely to be a sunken cost fallacy.

You're technically right with "Admission test measure how well you can prepare for admission tests, nothing more.".

It's definitely not a silver bullet and doesn't assure success after admission. nonetheless, the ability to prepare for these tests is strongly correlated with success along this path.

If they can't be arsed to prepare for them they're unlikely to bother with studying either, making their admissions pointless too.

And if they attempted to prepare for the test but still failed... Then they're unlikely to be able to learn in the University setting either and should follow up on other avenues.


> The full dataset isn’t available, but since we have the number of students in each ACT / GPA bin above, we can create a “pseudo” dataset, with just a small loss of precision.

Big oof. This is an example of terminal data science poisoning. The data in the paper already proved the point of the article! Using it to create some chart gore where you reproduce the same plot in uncanny valley levels of resolution doesn't add to the argument. This is using very sophisticated techniques to achieve what you could by simply blurring an image of the first chart.


It is not that they are irrelevant, it is that they are ableist, selecting for a narrow skill and ability combination, and creating a barrier for anyone else who lacks it, for whatever reason.


“C’s get degrees”. Measuring college graduation rate is a vanity metric


If anything over indexing on solely on test scores will filter out a large % of psycho/socialpaths who gets by on personality scores.


Did I miss the correlation across school systems? Everything in the study was from Chicago public schools.

My older brother switched from a B student in a private high school to the public high school and slept his ways to A's


Every metric has problem, but experience in a university suggests that, if one has to use a single metric to rank students, by far the best choice (the only choice really) is average grade.


I think that when people say test scores are irrelevant, they don't mean irrelevant for studies but irrelevant as an indicator of job performance.


> Meta-analyzing the relationship between grades and job performance.

Roth, P. L., BeVier, C. A., Switzer, F. S. III, & Schippmann, J. S. (1996). Meta-analyzing the relationship between grades and job performance. Journal of Applied Psychology, 81(5), 548–556. https://doi.org/10.1037/0021-9010.81.5.548 Employers and academics have differing views on the value of grades for predicting job performance. Employers often believe grades are useful predictors, and they make hiring decisions that are based on them. Many academics believe that grades have little predictive validity. Past meta-analyses of the grades–performance relationship have suffered either from small sample sizes or the inability to correct observed correlations for research artifacts. This study demonstrated the observed correlation between grades and job performance was .16. Correction for research artifacts increased the correlation to the .30s. Several factors were found to moderate the relationship. The most powerful factors were the year of research publication and the time between graduation and performance measurement.


Maybe a hot take but I think school pedigree is irrelevant.

I went to a top university in my country after transferring in after not getting in from highschool. Turns out myself and many others would have got a lot of value from going to a less prestigious university. Ultimately most of us end up working at the same select companies. But idk I am from Canada so maybe the intensity of selection is not the same.

We put far too much value in prediction of success and not enough value in actual success. Seems like the academic version of keeping up with the Jones's


Test scores are the most objective metric available. And thus I really consider them fairest. Or we could just have entirely random lottery.


I think the author is missing some really key points here. Namely that GPA and test scores are highly correlated. I'd go so far as to say that there is a confounding variable that predicts both. I'd even say there are several upstream predictors, but I think we all know that (or I thought we did). Every single graph they show demonstrates this yet they keep missing it. The top comment right now (by ajwinn[0]) notes that they used a heatmap like the one in the article for determining admission. This is great because it builds multiple pathways to gain admission. That's honestly how we should treat variables like these: rely on either.

As to explain the confusion in the models, well when you have confounding variables like this it is easy to reduce one variable's weight nearly to zero and place the rest in the other variable. So if I'm right (might not be) I think you could also show that the inverse is true, that grades aren't a good predictor and test scores are. The breakdown suggests that this is true too tbh.

But I also disagree with some of the analysis and techniques here. It's not always easy to use models that have continuous(ish) variables (GPA/ACT/income) and discrete ones (race/gender). I bring this up because there's some hints that they aren't a stats expert. For example there's the line

> The “likely-to-graduate” score becomes a probability after a sigmoid transformation.

Which just isn't true. A Softmax function[1] creates a probability distribution but a Sigmoid[2] doesn't. A Sigmoid bounds the output to [0, 1] (or [-1,1]) but being bound on that range isn't the only requirement to a probability distribution. A sigmoid does not guarantee that the sum of the elements is 1 (cumulative density), but a softmax does. In fact, it almost never will. (I mention this in my ML class a lot so make sure you all don't make this mistake too! I'll be really annoying if you do)

There's some more errors and novice approaches here and so I just want to note this because we shouldn't just take the results and conclusions as is when the modeling is bad. But I still think he makes some good (and bad) points and those are definitely up for discussion.

[0] https://news.ycombinator.com/item?id=33338809

[1] https://en.wikipedia.org/wiki/Softmax_function

[2] https://en.wikipedia.org/wiki/Sigmoid_function


In my opinion discriminating against people given their past test scores is unfair. People can improve and having scored bad (which might well be a consequence of hard life circumstances at a specific moment of their life) in the past should not be allowed to ruin their future. So I believe nobody should be allowed to see one's previous test scores and judge them this way. Whenever it is legitimately necessary to assess a person's proficiency a new test should be taken.


Who's saying that test scores are uncorrelated with college graduation, which is largely dependent on passing further tests?



It's a strategy to game the rankings.

Factors for college rankings often include acceptance rate (lower is better), and average SAT/ACT scores. By not requiring SAT/ACT scores, you do two things at once:

1. you allow more people to apply, especially those who did not take the test or received a poor score, who you can simply reject to bring down your acceptance rate

2. by making it optional, only the applicants who think their score would be an advantage (i.e. those with high scores) would send them, which brings up your "average test scores".


I don't know how this works in practice tbh.

We have some evidence from MIT that it is a bad idea to drop testing.

https://mitadmissions.org/blogs/entry/we-are-reinstating-our...


There's certainly a strong whiff of "is this the right question" to the whole discussion. School and school-adjacent results predict school results. Whoo.

Tests are just leetcode for students, after all.


The post starts with four quotes on that theme.


Does the same hold true of medical professionals? ie MCAT scores correlate with med school and/or practice performance?


It's been some time since I read this work, but at one point there was a situation where medical schools in a state were temporarily (like a year or two) prevented from using MCATs in admission decisions due to a judge's court order. Most applicants still had MCATs due to other schools, though.

Some researchers collected data before, during, and after this time, and followed-up on the physicians later.

My fuzzy memory (which might be incorrect) is that MCATs were predictive of GPA in the first two years of medical school like 0.20 on a Pearson correlation metric, and it dropped from there as you went out increasingly further in time and included things like ratings on clinical rotations. So it was predictive, but fairly weakly so and seemed less predictive as time went forward.


I’d gladly hire on test scores alone.


I think there’s an issue on how these data are examined.

There is a huge amount of data backing up the claim that the school attended influences the SAT, etc.

There is a huge amount of evidence backing up the claim that SAT/ACT aren’t 100% predictive of later academic success.

But “not being 100% predictive” does not mean “is not predictive”.

The core problem the removal of SAT, etc is attempting (imo hamfistedly) to mitigate is the following chain:

(1) Student A comes from affluent family, in an affluent region.

(2) Student A goes to a private school with better teachers, etc, or because it’s an affluent region the public schools have better teachers, etc

(3) Student A’s family can afford additional SAT, etc tutors

Now imagine a somehow identical Student B, but from a poor family in a poor region. Before any other factors become involved, they don’t get the benefits of 2 or 3.

We would expect that on average 2&3 result in higher scores, and so a higher acceptance rate, and a higher rate of scholarships for Student A vs Student B. Again this is on average.

The TLDR is given two kids who have identical mental ability, the one from the richer family will, on average, have better scores. So schools say let’s drop the scores entirely, but that is also clearly nonsense: Even in this exact scenario Student A will do better in uni or what have you, because part of why they got a better score is that they have been taught and learned more - IE they absolutely do know more and are more academically proficient than Student B by this point. Moreover we know (because of yet more studies), Student A will tend to continue to do better than Student B once they’re at school, even once B “catches up”, because we know that people with fewer external stressors perform better at everything.

So even if we discard the score based entry to unis, we’ll see a strong correlation between original score and subsequent academic success.

So while I understand the desire to remove the “accident of birth” factor that benefits the lucky and penalizes the unlucky, simply removing the test requirements is fairly questionable as an approach.

This is also before asking about Students C and D who are the peers of students A and B - surely the better scoring of A and C (or B and D) is who you should prefer, but now you’re discarding theoretically ideal and unbiased data in favour of…?


SAT / ACT are basically IQ tests. Grades are far more fuzzy but generally show an ability to stick with something over the long term and execute on projects.

I think both are important to being successful. Ignoring one for the other is probably bad. That said I prefer IQ over grades.


I think a major counter point is how these tests can be improved upon individually without much improvement over other subjects. The first time I took the SAT, I went in totally blind and got 1320, a few months later, with a bit more practice with pre existing precalc, I was able to get 1490. I also had many classmate report similar trends of their score increasing as the number of retakes increases.


This has been studied. There is a retest effect on most "cognitive performance" tests: scores increase on the first and second re-test, then level off. (i.e. the test scores for the 3rd and 4th time you take the test end up about the same)


The term of art is practice effect.

https://www.explorepsychology.com/practice-effects/


It measures your ability to perform on a new subject after studying for it, plus your willingness to study. The fact that you can study for it and improve your score (but not entirely - raw abilities still matter) is a feature not a bug.


But SAT is the aptitude test, and recognizing the importance of aptitude and working on that aspect would benefit the candidate greatly. You are supposed to be able to work and improve urself and benefit from it. A person is not an static object. What happened to `bring the best out of oneself`?


I'm not sure you're making the point you think you are. As I recall the SAT many years ago, there were definitely question forms like X:Y as A:? which rely on vocabulary to some degree but may also be hard (or at least take some extra time) if the form is unfamiliar. Also even a surface-level understanding of some areas of math you realize you sort of didn't really get can be useful. So with that fairly modest knowledge and refresh, you got what's a possibly useful but also relatively modest uptick.


Don't discount the bump from understanding the pacing and difficulty of the exam. You might have 25 questions in 30 minutes for a math section. Understanding that which sections are easy and which are hard allows a student who has seen the exam before to allocate their time more efficiently.


Sure. And even taking the test once under real live conditions is probably beneficial to complement a bit of study. Just as really giving a presentation to even a few people almost certainly beats just mentally walking through it in your head.

Obviously you can try to create a real simulation but it's probably not the same thing.


The increase from 1320 to 1490 is pretty significant when it comes to college rankings. Although 1490 isn't anything stellar, it's the top 10 or 5 percent at my school. My main point was just the fact that, by paying collegeboard more money for more retakes, your score will increase. The SAT doesn't seem like a great indicator of anything besides one's ability to perform on the SAT


There’s a range you can improve on. If you are an idiot you don’t have a chance of improving that much, but mommy and daddy can buy you a 4.0 at a private school. Can’t bribe your way to a 1600 SAT anywhere near as easily as the private school game can be played. And this is another reason why I like SATs over grades - the poor can get a leg up over the rich.


Yeah, let's base it all on a test mom and dad pay for after all the prep classes they paid for. High school is free most places, friend.


I'll take a standardized test over arbitrary grades any day. Grades are far, far easier to game in so many ways (switch to an easier major!) than the SAT. Just ask anyone gunning for med school or law school - they will opt for the easier class over the harder one if it means their GPA has less risk of dropping.


> Grades are far more fuzzy but generally show an ability to stick with something over the long term and execute on projects.

What grades show is how well the student presented what the teacher wanted returned. Generally, but not always, students that are more intelligent than their teachers and may even have a better grasp of the subject don't receive the highest grades. Grades are about obedience, not learning. The "best" students are children of teachers.


> That said I prefer IQ over grades.

Why would any rational individual take an IQ test? IQ cannot be changed and whatever outcome doesn't change your course of action, being the fact that you have to work hard.

Low IQ: Gotta work hard

High IQ: Gotta work hard


> Why would any rational individual take an IQ test? IQ cannot be changed and whatever outcome doesn't change your course of action

From what I can tell, most professions have a rough IQ floor. If you're above the floor, more IQ can always help, but other traits like conscientiousness and agreeableness tend to matter more. But if you're substantially below the floor, there's almost no chance that hard work will help you bridge the gap.


Have you taken one?

It certainly felt like the kind of test that could be improved with practice. Not as easily as the SAT, sure, but certain patterns seemed to appear, in a way I bet I could prep for.

I haven't tested that hypothesis, but if anyone wants to fund me to take a bunch over the next year... let me know ;)

(Why did I take it in the first place? Curiosity. Maybe also a bit of vanity. "Rationality" seems like a false idol to focus purely on in general, for decisions.)


I've taken at least 5 over the course of my life. Possibly a 6th that I didn't know about (possibly age 5 or 6?), but I know of taking multiple between ages 9 through mid 30s.

Regardless of the test, the percentile numbers didn't change.

I've read people claim on and off for years that you can "study" for them and "improve" your scores, but that's not been my experience. Any variations you get might be slight, but you're extremely unlikely to get any statistical difference... unless....

I had a test at age 9 for a specific 'gifted' program starting in our school district. 2 months later, we moved districts, and I was given the exact same test again. I aced it, because I'd asked my parents some of the questions I didn't understand the first time. I 'confessed'(?) to the test giver in the second district that I'd already had these exact same questions 3 months earlier, which they noted. I'd already 'qualified' after the first test, so again, I don't think there was any significantly different outcome (I 'qualified' after the second test in the new district as well).


IQ scores do change over time, largely as a function of curiosity and conscientiousness. You could be a genius toddler but if you fail to learn, your IQ continually diminish relative to your peers


IQ has a map/territory problem as well. The theory is that you have some underlying IQ, and the test is a tool to determine what your IQ is. It is well established that practice can improve your score on an IQ test. However, that does not imply an improvement in your underlying IQ.


Close. Improving your score on an IQ test does improve your IQ. Your IQ is literally that score. What it does not improve is g, your underlying general intelligence.


You could change your life/career plans based on the results. Whether to go to college or not, whether to pick a job that pays well immediately or one with more potential for long-term advancement, that sounds like the kind of decision where knowing your IQ would help.


No. Whatever your IQ you should aim high and let nature put you in your place.

you cannot pre-emptly put yourself in your place, that’s the best route for depression and depression sucks whether you have a high, normal or low IQ


So much for being rational.


I lost points on the SAT for mixing up 2 pi r and pi r^2 - left out the 2 for the former. (That's the one that stands out as a duh, I knew better, one - I forget what else I missed.) There were also a LOT of analogies. I've read that those have been removed, I dunno how the math section is, but both of those were much more "knowing things" than innate intelligence. Grades tend to require both "knowing things" and longer-term discipline; I could see where that would be advantageous.

Grades have a disadvantage of A's being way too common, though - no resolution at the top of the scale.


SAT/ACT are basically poor IQ tests, because study for them results in heavily loading onto memory aspect of IQ but unloading on some others somewhat.

Ideally, you would want an IQ test and the 'Big Five' personality traits. The latter though can be annoying to get.

I have seen some firms get very good results by basically using a disguised IQ test, there are some firms that basically compose the tests, give them to 1000s of people and perform the factor analysis to extract IQ from the test, then contract the the test out for hiring departments. Some of those tests are pretty creative in that most people wouldn't know they were taking an IQ test.


Before people state strong assumptions one way or the other, have there ever been any studies that measured the relationship between test scores and IQ? If so, what did they say?


It’s an IQ test. Not a culture fair one, like Raven’s Progressive Matrices, but it’s as much an IQ test as coarse, quick and dirty one like Wonderlic is.

> Scholastic assessment or g? The relationship between the scholastic assessment test and general cognitive ability

> There is little evidence showing the relationship between the Scholastic Assessment Test (SAT) and g (general intelligence). This research established the relationship between SAT and g, as well as the appropriateness of the SAT as a measure of g, and examined the SAT as a premorbid measure of intelligence. In Study 1, we used the National Longitudinal Survey of Youth 1979. Measures of g were extracted from the Armed Services Vocational Aptitude Battery and correlated with SAT scores of 917 participants. The resulting correlation was .82 (.86 corrected for nonlinearity). Study 2 investigated the correlation between revised and recentered SAT scores and scores on the Raven's Advanced Progressive Matrices among 104 undergraduates. The resulting correlation was .483 (.72 corrected for restricted range). These studies indicate that the SAT is mainly a test of g. We provide equations for converting SAT scores to estimated IQs; such conversion could be useful for estimating premorbid IQ or conducting individual difference research with college students.

https://journals.sagepub.com/doi/abs/10.1111/j.0956-7976.200...


Yes. If you look you will find published literature on the subject. The plots plotting SAT/ACT and IQ will have an obvious trend but be very noisy. Hence why they are poor IQ tests, you can have a decent guess at someone's IQ based on their test results but it's not ideal.


The whole point of IQ tests at the very beginning was to predict how well you can do academically. Technically every test is an IQ test.


"That said I prefer IQ over grades." The evil in this is: Your IQ doesn't change. So, have the the children take an IQ test at age 10, then just setup their college admissions right there. Deny kids that don't have the right IQ. No need to bother with GPAs or "trying".


I mean, it's the logic I used to decide not to pursue a career in professional basketball as someone of below average height. Do you think encouraging short people to pursue a career in basketball is productive in any way?

If you're 10 years old and have a below average IQ, don't bother aiming for a PHD at MIT in atrophysics.


Very much disagree. Lazy genius is useless. Hard working, low IQ gets stuff done. Obviously hard working genius is ideal. But without the ability to focus on longer term goals, which grades is an indicator of more so than test scores, your high IQ means nothing.


I would argue it very, very much matters what you're trying to do. If you're harvesting apples, then sure. Hard working beats IQ. Hell, IQ is a handicap, not a plus.

If you're doing research, you have the other side of the spectrum, IQ beats everything, and hard work can actually prevent you from finding solutions.


Research is lots of hard work. There’s a minimum level of intelligence needed, but I’d rather have an acceptable intelligence+hard worker as a collaborator than a genius+slacker.

There aren’t many Archimedes out there and most research is the result of lots of labor.


How do you define "acceptable intelligence"? By their ability to better a lazy genius? I don't really see a non-tautological way to define "acceptable intelligence"


What I meant is that intelligence is required to complete the required education, design studies, carry them out, etc.

It’s some level but not genius.


This sort of idea ... I get it, but it runs stuck on what exactly "acceptable intelligence" is. That would be the sort of intelligence 99% or more of the population doesn't have.

Also at some level "I'd rather have" becomes pointless. Choices are usually 1 individual, or waiting 6 months to a year. I don't know if we're in a bubble but I've seen the same in biochemistry research and computer research: academics take everyone they possibly can, because master or above and they will get an unbeatable offer from a huge company very soon.


Good thing SAT scores also measure hard work, planning, and follow up. You can improve your scores significantly with the hard work of a few hours and checking out a sample test or prep book at your local library or school


SAT/ACT measure how good you are at following rules and how much time/money you have to go to tutors and study 1000 page textbooks outlining specific strategies to maximize your score.


Most of the evidence I see suggests that working through a test book at your local library goes a long way towards erasing most of the disadvantage associated with going into the test cold. Yes, there's some time involved but I'm pretty sure declining returns set in fairly quickly. There are clearly advantages associated with recognizing the type of questions that will often be on the test. But a lot beyond that is actually knowing the vocabulary and the various geometry, etc. material.

Now, if you basically don't know or have trouble understanding the material? I wonder to what degree some tutoring and even a fair bit of time are going to make up for years of poor school outcomes (for whatever reason). Some specific strategies are 1.) pretty straightforward--if you can quickly narrow options down, maybe guess and move on--and 2.) won't move the needle a lot.


Really? My SAT scores were average for admit to the college I attended. I never had a tutor, never looked at test strategy guides, and took the test only once, and my parents were lower middle class. The people I know who got 800s (perfect SAT scores in those days) did not game the system - 5 minutes talking to them would reveal that they were very, very smart individuals.

Want to do well on the SATs?

1. take all the math classes your high school offers

2. pay attention in those classes

3. read a lot of adult-level books (I read tons of scifi)

and you'll do fine without needing any gamesmanship.


BTW, I'd advise caution with gaming your way into a tough college, rather than doing the work. It will assume a solid foundation in math, and you'll still have to work hard. Having to struggle, then transfer out after a year, is not doing yourself a favor.

Me, I struggled and came pretty close to flunking out.


Do you have any sources for this? SAT scores aren't highly correlated with household income, there is something but it's far from everything.

In my opinion and experience of the results my peers got, these kinds of standardized tests were really pretty good. The tests and studying methods themselves didn't necessarily mimic what you would do in a real job or even a university degree and sure seemed stupid and annoying at the time when you had to do the damned things... But stepping back and thinking about it, for the most part the people who were intelligent and capable of a motivated and disciplined approach to academics and fields that typically require degrees did actually tend to score better than people who were lacking in those things.

Yes there was the occasional very intelligent person who wasn't disciplined or motivated for whatever reason who did poorly. They weren't or wouldn't have been happy or done well in college either though, for the most part. And there were a few uninspiring intellects who got extremely high marks withe the sheer grind, although I think that's an impressive attribute itself and probably very useful for a lot of fields. There certainly were not hordes of Einsteins and Newtons just slipping through the cracks because they didn't test well.

So I don't know. There is this "common wisdom" that tests are terrible, which is an attractive idea. I just don't know if that's really true. Maybe they're actually pretty good for a large proportion of people.


Anything you can prep for - grades included - is going to trend towards being largely influenced by "who has the most time and money."

I don't have any good ideas to get away from that.


> who has the most time and money

Or what matters most - who is willing to put in the effort.

You already have 6 hours a day in high school. Don't need more time than that. As for money, that's not needed either. Your high school math books will do. There's also the public library, the Kahn Academy, youtube, etc.


Having a more centered economic distribution amongst children is a good one. It is non-sensical and anti-meritocratic to have a large enough skew amongst child starting positions as to render meritocratic testing irrelevant.

It's possible our declining academic performance is simply another artifact of rising inequality.


Yes, let's solve the problem of skewed student assessment by solving inequality. Your solution is far less tractable than the problem!


Or is it? we generally know how to solve inequality - we don't know how to solve student assessment. Higher taxes and rising inflation/interest rates reduce inequality.


Compared to grades it is certainly a magnitude less dependent on how much time/money you have, how many tutors you have, and how good you are at reading textbooks.


Tutoring and extensive exam-specific prep can boost scores. They are not necessary for getting a high score.


Which is a good predictor for succeeding in college, including college success.


So if a student who has to work 30 hours a week to support his family scores worse than a student who has no job and visits a tutor to study an SAT prep book outlining bizarre strategies about deciding when to not answer a question to score higher, that means the latter student should be preferred for college admissions?


IQ tests show the ability to sit down and do pointless tasks which an authority figure has told you are meaningful and you must guess what that authority figure wants you to do.

Extremely valuable skills for white collar drones.

Completely meaningless as a measure of actual intelligence though.


IQ is the most reliable and most valid measure of general intelligence we have been able to come up with. It is good enough for many clinical purposes, eg assessing brain damage, and also just so happens to correlate very strongly with just about every "life outcome" you can think of. Does it measure everything imaginable about human intelligence? No. But it measures what we actually know about very well. To call it completely meaningless is absurd.


You should reread what I said.


Be specific.


The ability to do boring arbitrary tasks is much more strongly correlated with the measures that IQ proponents cite as proof it's a meaningful measure than with whatever they mean by intelligence.

We could get the same results by forcing people to complete white noise patterns and seeing how long it takes them to quit.


That's an interesting idea, and while I doubt that it is true it seems like something you could determine experimentally. Has anyone done that or are you just asserting it?


I’m not sure what IQ tests you mean but that doesn’t sound like any I’ve used.

The intent is to gauge general mental ability, not rule following.

Take spatial in tests [0]. They aren’t about rule following so much as problem solving and lateral thinking.

Also iq tests are standardized and normalized. They have flaws, certainly, and are no way meaningless. It’s the best that we have until we get something better and there’s lots of correlations between iq and positive life outcomes. [1]

[0] https://en.wikipedia.org/wiki/Spatial_intelligence_(psycholo... [1] “ IQ correlates positively with family income, socioeconomic status, school and occupational performance, military training assignments, law-abidingness, healthful habits, illness, and morality. In contrast, IQ is negatively correlated with welfare, psychopathology, crime, inattentiveness, boredom, delinquency, and poverty.” https://blogs.scientificamerican.com/beautiful-minds/iq-and-...


>The intent is to gauge general mental ability, not rule following.

General mental ability as defined by sitting down and doing an arbitrary boring task, viz. following a rule that you don't see much point in but someone important told you to do.

The ability to do well on those tasks correlates much more strongly with:

>family income, socioeconomic status, school and occupational performance, military training assignments, law-abidingness, healthful habits, illness, and morality. In contrast, IQ is negatively correlated with welfare, psychopathology, crime, inattentiveness, boredom, delinquency, and poverty

than whatever intelligence is.

In short: IQ is a great tool to measure your ability to be a white collar drone.


The task isn’t arbitrary. It’s designed to measure mental ability. And it’s not boring as it can be quite short. And the rules should be quite easy to understand as they involve things like “which shapes fit together.”

The test is not designed to be confusing or hard to understand.

I expect IQ is also correlated with being a white collar drone (pretty awesome) but it’s also correlated with whatever intelligence is, as imperfectly as we’ve defined it.


The tasks which can be reduced to finite integer sequences are absolutely arbitrary.

E.g. you're given a finite set of pictures and expected to guess what the next picture is from a few possibilities. The answer is that it's all of them since any finite sequence has an infinite number of generating functions. The only way to get it right is to try and read the mind of whoever wrote the test and guess what they meant. An invaluable skill to have in the white collar workforce when guessing what your boss actually wants from you. But white collar empathy is not intelligence.

From the IQ test listed here: https://www.mensa.org/public/mensa-iq-challenge literally all of them are reducible to questions about integer sequences.


> short: IQ is a great tool to measure your ability to be a white collar drone.

But why so snarky?


I’m not a fan of ad hominem but the most reasonable guess to me is that someone did poorly on IQ tests and is working to justify why they are bad. Or they have a super high score and are trying to ultimately justify their argument with “well I scored 179 and even I think they are bad.”

I don’t think we’ll have closure.


> Test scores are highly predictive

Any study in this area that doesn't control for socio-economic factors is near useless, because it mainly predicts that children of successful (ie rich) people will likely grow up to be successful (ie rich). This is almost tautological.


You seem to be completely unfamiliar with the research in the field. It is well known that controlling for socioeconomic factors does not substantially impact predictive validity of standardized tests. See for example

https://journals.sagepub.com/doi/abs/10.1177/095679761243873...

> Scores on the SAT (a test widely used in the admissions process in the United States), secondary school grades, college grades, and SES measures from 143,606 students at 110 colleges and universities were examined, and results of these analyses were compared with results obtained using a 41-school data set including scores from the prior version of the SAT and using University of California data from prior research on the role of SES. In all the data sets, the SAT showed incremental validity over secondary school grades in predicting subsequent academic performance, and this incremental relationship was not substantially affected by controlling for SES.

Or this

https://eric.ed.gov/?id=ED562860

> (a) SES is related to SAT scores (r = 0.42 among the population of SAT takers), (b) SAT scores are predictive of freshman grades (r = 0.47 corrected for school-specific range restriction), and (c) statistically controlling for SES reduces the estimated SAT-grade correlation from r = 0.47 to r = 0.44.

There are more studies on this, with rather consistent results. In short, poor students do about as well as rich students when matched on SAT score.


Why would one expect socioeconomic factors to do one or both of:

a) influence grades in primary and secondary school

b) influence standardized test scores

but not also do:

c) influence college performance (esp. freshman year when you're least removed from those same effects in HS)?


Not sure what your point is. My point is that controlling for SES is useless in context of the predictive validity of SAT. Regardless of the degree in which socioeconomic factor casually influence test scores and educational outcomes, you get no substantial gain from learning them above what you learn from test scores alone, as long as the goal is to select students who are most likely to learn and succeed at your university.

This means that admission officers who want to learn about SES of candidates in addition to learning their test scores usually either are looking for something else the student’s aptitude for learning and ability to succeed, or are completely unfamiliar with state of the research (this is sadly quite common, and this is often willful ignorance too).


They certainly do all three.

This doesn't mean that SAT scores can't still be similarly predictive across socioeconomic gradients. Both of these things can be true without any inherent conflict.


Sure, it just means it's a bit of a pointless debate...

Questions that are much more interesting to me are:

* should any level of the education system attempt to correct for underlying socioeconomic factors?

* if so, how? (I would love more discussion about this because it both seems like an extremely difficult problem and because otherwise we waste all our time and energy on non-effective things like removing standardized tests from admissions requirements.)

In the software engineering interview world, for instance, a whole lot of active discussion is spent around interviewing methods. Let's make that bigger! Let's figure out how to get the people who never even make it to the interview stage today a better opportunity.

Otherwise it's like taking two plants, reducing the amount of sunlight one gets to 50% of what the other one gets for several years, then moving them both to a new environment, waiting four more years, and saying "huh, the one that was taller before we moved them 4 years ago is still taller, guess that's a good measurement to use to pick plants, since even controlling for that, the taller ones from the light-limited cohort are also ended up taller than their cohort-mates after the last 4 years!"


> Questions that are much more interesting to me are

Sure, but it's still worthwhile to point out factually inaccurate articles and deceptive papers. If they want to argue for how the education system should correct socioeconomic factors, they should do so honestly, not by lying to us that tests are useless.


What debate?

The test scores are useful. That's the point. There's nothing to debate.


I am actually familiar with some of the research actually, but thank you for your condescendence.

I didn't mean to imply that the SAT or grades are not predictive of future performance, only that lower socio-economic statuses will correlate to lower SAT grades in the first place, ie your (a) proposal in the second source.

From this same source:

> The SAT-grade relationship is not an artifact of common influences of SES on both test scores and grades

To substantiate my initial comment, test scores are highly predictive yes, but as you just linked, SES is highly predictive of test scores (and this is only for people who get to the damn test). This is the tautological part. High incomes make people test well which lead to jobs with high incomes. But since this correlation is not 100% but merely very high, we can pat ourselves on the back and say hey it's accurate!

Articles that defend the accuracy of the system tend to implicitly defend its legitimacy, which I think is a much more important question than "is SAT indicative of future performance" when considering whether the SAT is useful.


> I didn't mean to imply that the SAT or grades are not predictive of future performance, only that lower socio-economic statuses will correlate to lower SAT grades in the first place, ie your (a) proposal in the second source.

Why then talk about “controlling for SES”? The correlation SES/educational ability very much exists, but why would one want control for SES? The goal is, allegedly, to test scholastic aptitude, so that we can identify students that are likely to succeed at school. If SES correlates with this ability, by controlling for this, all you’ll do is control out the effect.

> High incomes make people test well which lead to jobs with high incomes.

No, there is little evidence that high incomes makes people test well. Here, and in the entire thread, you are committing what is called a “sociologist’s fallacy”. This is from Meehl, P. E. “Nuisance variables and the ex post facto design.”:

> While every sophomore learns that a statistical correlation does not inform us as to the nature of the causality at work (although, except for sampling errors, it does presumably show some kind of causal relation latent to the covariation observed), there has arisen a widespread misconception that we can somehow, in advance, sort nuisance-variables into a class which occurs only at the input side. This is, of course, almost never the case. The usual tendency, found widely among sociologists and quite frequently among psychologists (particularly among those of strong environmentalist persuasion) is to assume sub silentio that there is a set of demographic-type variables, such as social class, domicile, education, and the like, that always operate as nuisance variables to obscure true relationships, and that function primarily as exclusively on the input side from the standpoint of causal analysis. This automatic assumption is often quite unjustified. Example: We study the relationship between some biological or social input variable, such as ethnic or religious background, upon a psychological output variable, such as IQ or achievement. We find that Protestants differ from Catholics or that Whites differ from Blacks. But we find further that the ethnic or religious groups differ in socio-economic class. We conclude, as an immediate inference and almost as a matter of course, that we have to ‘control’ for the socio-economic class variable, in order to find out what is the ‘true’ relationship between the ethnic or religious variable and the psychological output variable. But of course no such immediate inference is defensible, since on certain alternative hypotheses, such as a heavily genetic view of the determiners of social class, the result of such a ‘control’ is to bring about a spurious reduction of unknown magnitude in what is actually a valid difference

This was published over 50 years ago, but nevertheless, this mistake is done over and over in sociological research and discussions.

> Articles that defend the accuracy of the system tend to implicitly defend its legitimacy, which I think is a much more important question (…)

I actually have no idea what you mean by “legitimacy” here.


Your comment would be equally or more effective without this sentence:

>You seem to be completely unfamiliar with the research in the field.


I disagree. When a comment is sufficiently wrong, it's useful to underscore that.


If a user wishes to underscore that a comment he is replying to is wrong, then he should do just that. There is no need for the user to attack the person making the comment.


Calling someone "completely unfamiliar" is not much of an attack. I don't think it's any worse than saying the comment is "completely wrong", but it gets the point across slightly better.


I was being charitable here. The less charitable alternative is that the person I was responding to was familiar with the literature, but nevertheless engaged in willful deception.


Since acknowledging your evidence will inhibit my ability to blame the white supremacist cis-hetereonormative ableist patriarchy for my problems I will instead respond with an anecdote about how I know a lesbian indian in a wheel chair that bombed the SATs and is now president of the united states.


well, maybe, just maybe, that's because the socio-economic status correlates positively with cognitive abilities and the kids somehow inherit the traits of their parents?


Meh. Dumb rich kids end up with positive college and career outcomes regardless of their test scores. Let’s not suggest that wealth correlates to ‘better traits’.


That's funny I grew up as one of six children and my mother was a school teacher. Guess I have to go back to my HS and tell them they need to lower my scores because I wasn't rich enough.


You should instead chide them for not emphasizing the fact statistics measures trends, not absolutes.


Methodology classes didn't excel either.




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

Search: